1
00:00:00,000 --> 00:00:03,479
[MUSIC PLAYING]

2
00:00:03,479 --> 00:00:15,583


3
00:00:15,583 --> 00:00:16,500
IAN: All right, folks.

4
00:00:16,500 --> 00:00:20,260
Welcome back to week
six, where we're going

5
00:00:20,260 --> 00:00:22,760
to talk a little bit about exposure.

6
00:00:22,760 --> 00:00:28,940
So to this date we've talked a lot about
composition, lensing, storytelling,

7
00:00:28,940 --> 00:00:33,920
without focusing too much on
the technical details of how

8
00:00:33,920 --> 00:00:38,180
to actually make an image, and
match it to our intentions.

9
00:00:38,180 --> 00:00:40,700
So we'll dive a little
bit more in that today.

10
00:00:40,700 --> 00:00:43,730
And I think it's important
to define exposure.

11
00:00:43,730 --> 00:00:46,550
So the idea of exposure
is that we want to render

12
00:00:46,550 --> 00:00:52,100
a scene in a specific way using
our camera controls to interpret

13
00:00:52,100 --> 00:00:55,100
the amount of light in a given scene.

14
00:00:55,100 --> 00:00:56,800
So what does that really mean?

15
00:00:56,800 --> 00:00:58,883
Well we're going to dive
deep into this, and we're

16
00:00:58,883 --> 00:01:01,175
going to look at each of the
different camera controls.

17
00:01:01,175 --> 00:01:03,590
How they affect the image,
and how we can utilize

18
00:01:03,590 --> 00:01:07,010
them to make different exposures.

19
00:01:07,010 --> 00:01:10,880
So what's our goal
with exposing an image?

20
00:01:10,880 --> 00:01:13,910
It's really to capture
that intentional image.

21
00:01:13,910 --> 00:01:17,540
It's to make a decision about
how we want an image to look

22
00:01:17,540 --> 00:01:20,325
before we press the shutter button.

23
00:01:20,325 --> 00:01:23,450
When you have a camera that's on auto
mode and you press the shutter button

24
00:01:23,450 --> 00:01:28,190
you're basically giving up all the
decision making to a small machine.

25
00:01:28,190 --> 00:01:31,250
And they're sophisticated, sure,
but they're not intelligent.

26
00:01:31,250 --> 00:01:33,620
And our goal is for us
as intelligent operators

27
00:01:33,620 --> 00:01:36,620
to make the decisions for
the camera, or at least

28
00:01:36,620 --> 00:01:40,430
override decisions that may be poor.

29
00:01:40,430 --> 00:01:43,220
So I think what we
really need to understand

30
00:01:43,220 --> 00:01:46,940
is what exactly is it
that we're trying to do?

31
00:01:46,940 --> 00:01:50,830
And so in any given scene
there's some amount of light.

32
00:01:50,830 --> 00:01:54,200
And we can use absolute measurements
to tell how much that is.

33
00:01:54,200 --> 00:01:59,240
It could be 150 foot candles of light,
but I don't know what that means.

34
00:01:59,240 --> 00:02:03,950
I sort of conceptually understand that,
but it doesn't help me take a camera

35
00:02:03,950 --> 00:02:05,262
and make an image.

36
00:02:05,262 --> 00:02:06,470
I have to do a bunch of math.

37
00:02:06,470 --> 00:02:08,990
It gets very confusing.

38
00:02:08,990 --> 00:02:11,030
There's easier ways.

39
00:02:11,030 --> 00:02:16,250
So what we really need is
a relative amount of light,

40
00:02:16,250 --> 00:02:17,680
or at least the scale--

41
00:02:17,680 --> 00:02:22,160
a relative scale that we can use
to adjust our camera settings.

42
00:02:22,160 --> 00:02:24,770
So for us in this class when
we're talking about photography,

43
00:02:24,770 --> 00:02:29,300
and then later cinematography, we're
going to use the concept of f-stops.

44
00:02:29,300 --> 00:02:34,460
And all an f-stop is is a
doubling or halving unit.

45
00:02:34,460 --> 00:02:37,310


46
00:02:37,310 --> 00:02:41,090
So if you have a camera
that has some sensitivity,

47
00:02:41,090 --> 00:02:44,750
and you double the sensitivity
of it, that's one stop.

48
00:02:44,750 --> 00:02:47,300
If you have a camera that has
some light coming into it,

49
00:02:47,300 --> 00:02:51,680
and you cut the amount of light coming
into it by one half, that's one stop.

50
00:02:51,680 --> 00:02:53,720
They're going in opposite
directions, but it's

51
00:02:53,720 --> 00:02:55,310
an equivalent unit of one stop.

52
00:02:55,310 --> 00:02:59,780


53
00:02:59,780 --> 00:03:02,390
So just briefly to get
a sense of how this

54
00:03:02,390 --> 00:03:06,350
might look when we look at an image we
have here an image from Mount Auburn

55
00:03:06,350 --> 00:03:11,450
Cemetery that's at exposure, roughly.

56
00:03:11,450 --> 00:03:15,350
And I made this exposure using
a DSLR, and the light meter

57
00:03:15,350 --> 00:03:20,090
that was in the camera told me that
if I set the camera settings to this

58
00:03:20,090 --> 00:03:21,770
it'll be at exposure.

59
00:03:21,770 --> 00:03:23,660
And I think it did a pretty good job.

60
00:03:23,660 --> 00:03:26,360
There's no real complaints there.

61
00:03:26,360 --> 00:03:31,190
So in an effort to sort of investigate
how much difference a stop might be

62
00:03:31,190 --> 00:03:33,380
I opened up one stop.

63
00:03:33,380 --> 00:03:39,380
So I allowed double the amount
of light to strike the sensor.

64
00:03:39,380 --> 00:03:41,480
And you can see that
it gets much brighter.

65
00:03:41,480 --> 00:03:47,330
If we go backwards we have some dark
shadow detail in here, neutral gray,

66
00:03:47,330 --> 00:03:50,930
some bright white, and
everything feels naturalistic.

67
00:03:50,930 --> 00:03:53,990
We start to brighten up
that neutral gray begins

68
00:03:53,990 --> 00:03:58,160
to push towards the lighter gray,
this definitely begins to clip,

69
00:03:58,160 --> 00:04:01,460
these trees feel a little bit brighter.

70
00:04:01,460 --> 00:04:04,730
Not completely unnatural,
but getting there.

71
00:04:04,730 --> 00:04:08,392
So now if we allow four times
as much light in, or two stops--

72
00:04:08,392 --> 00:04:10,850
remember that you're doubling
each time you open up a stop.

73
00:04:10,850 --> 00:04:13,220
So you double, and
then you double again.

74
00:04:13,220 --> 00:04:15,760
Now this image really
starts to fall apart.

75
00:04:15,760 --> 00:04:19,220
And this is the classic overexposure
that maybe you've accidentally done,

76
00:04:19,220 --> 00:04:21,529
or you've been struggling
with your camera controls.

77
00:04:21,529 --> 00:04:26,210
And you end up here where you have
clipping elements, there's no detail,

78
00:04:26,210 --> 00:04:28,130
there's no actual shadow detail.

79
00:04:28,130 --> 00:04:29,570
No dark tones in this at all.

80
00:04:29,570 --> 00:04:31,790
It's all light grays.

81
00:04:31,790 --> 00:04:34,340
And if we go even further
it completely falls apart.

82
00:04:34,340 --> 00:04:36,285
And we could keep opening this--

83
00:04:36,285 --> 00:04:38,660
allowing more light to hit
the camera and eventually we'd

84
00:04:38,660 --> 00:04:41,940
end up with a solid white image.

85
00:04:41,940 --> 00:04:45,920
So, OK, that's overexposure.

86
00:04:45,920 --> 00:04:48,550
But when we look at this image again--

87
00:04:48,550 --> 00:04:49,160
add exposure.

88
00:04:49,160 --> 00:04:51,310
This is the same image
from the beginning.

89
00:04:51,310 --> 00:04:54,500
And we go the other
way, I've now reduced

90
00:04:54,500 --> 00:04:59,840
the amount of light entering into
the camera by a half, or one stop.

91
00:04:59,840 --> 00:05:03,070
And you can see that all the
tonalities get depressed.

92
00:05:03,070 --> 00:05:05,612
They're getting pushed
down into the shadow areas.

93
00:05:05,612 --> 00:05:08,570
There's more detail in that white
snow, which was much brighter before,

94
00:05:08,570 --> 00:05:10,920
but now it's sort of a shade of gray.

95
00:05:10,920 --> 00:05:14,330
And if we keep going two
stops so four times less light

96
00:05:14,330 --> 00:05:18,960
is reaching in the sensor you now
see like beautiful detail in here,

97
00:05:18,960 --> 00:05:24,030
but everything else is turning into
a muddy, crushed, shadowy darkness.

98
00:05:24,030 --> 00:05:29,040
And if we keep going it
becomes almost unreadable.

99
00:05:29,040 --> 00:05:31,800
So all that is is to say
that we have these tools that

100
00:05:31,800 --> 00:05:35,520
allow us to increase and
decrease the amount of light,

101
00:05:35,520 --> 00:05:40,740
and we do so by measuring it
in doubling or having units.

102
00:05:40,740 --> 00:05:42,600
So how do we measure light?

103
00:05:42,600 --> 00:05:48,120
Well, we have a brief video online that
we posted earlier this week, on Friday,

104
00:05:48,120 --> 00:05:50,790
on light meters, which if
you've all watched, great.

105
00:05:50,790 --> 00:05:53,405
And if not we'll talk a little
bit about as a refresher.

106
00:05:53,405 --> 00:05:54,780
But there's a few different ways.

107
00:05:54,780 --> 00:05:58,140
You can use a handheld light
meter like this to measure light,

108
00:05:58,140 --> 00:06:00,570
but I think more often
than not what we'll use

109
00:06:00,570 --> 00:06:05,280
is the internal light
meter of the camera.

110
00:06:05,280 --> 00:06:08,730
And so this will measure the amount
of light that's striking the subject,

111
00:06:08,730 --> 00:06:11,280
and reflecting back through
the lens onto some sort

112
00:06:11,280 --> 00:06:13,620
of light sensitive material.

113
00:06:13,620 --> 00:06:17,370
And it will say if you set your
camera to these values you'll

114
00:06:17,370 --> 00:06:21,570
get a decent exposure.

115
00:06:21,570 --> 00:06:24,270
So they can be hand-held
or internal to a camera,

116
00:06:24,270 --> 00:06:26,400
but all light meters
are calibrated to expose

117
00:06:26,400 --> 00:06:31,260
for an idea called 18% reflective gray.

118
00:06:31,260 --> 00:06:35,250
And they do so in different ways.

119
00:06:35,250 --> 00:06:37,500
So briefly, before we look
at a couple different ways

120
00:06:37,500 --> 00:06:41,310
they do that, 18% reflective
gray is often sort

121
00:06:41,310 --> 00:06:44,220
of colloquially termed middle gray.

122
00:06:44,220 --> 00:06:49,585
When exposed properly it forms the
middle value between absolute light

123
00:06:49,585 --> 00:06:50,960
and absolute black in your image.

124
00:06:50,960 --> 00:06:54,270
It's right in the middle.

125
00:06:54,270 --> 00:06:58,020
So it's a very handy
reference tone for us.

126
00:06:58,020 --> 00:07:00,120
And it's usually found on a gray card--

127
00:07:00,120 --> 00:07:05,190
something like this, that reflects
back 18% of the light that strikes it.

128
00:07:05,190 --> 00:07:07,830


129
00:07:07,830 --> 00:07:09,330
Cool.

130
00:07:09,330 --> 00:07:16,920
So before we get here let
us quickly fire this guy up.

131
00:07:16,920 --> 00:07:19,150
So this is the output
of our camera right now.

132
00:07:19,150 --> 00:07:21,810


133
00:07:21,810 --> 00:07:24,600
If we take a middle gray
card and we place it

134
00:07:24,600 --> 00:07:27,580
in front of the camera
well illuminated--

135
00:07:27,580 --> 00:07:29,610
we'll fill the frame with it.

136
00:07:29,610 --> 00:07:33,150
We can see that in the middle
of this there's this histogram,

137
00:07:33,150 --> 00:07:38,970
and sort of all of the image data
is centered right in the middle.

138
00:07:38,970 --> 00:07:43,590
If we set it for this
white background you

139
00:07:43,590 --> 00:07:46,410
can see that it's sort of drifted
back to the middle as well.

140
00:07:46,410 --> 00:07:48,368
And that's sort of
interesting because we would

141
00:07:48,368 --> 00:07:51,660
expect this tone to reflect more light.

142
00:07:51,660 --> 00:07:55,050
So there is a bit of a trick going
on with reflective light meters

143
00:07:55,050 --> 00:07:57,360
that we need to pay attention to.

144
00:07:57,360 --> 00:08:02,100
Reflected light meters always
assume that every single element

145
00:08:02,100 --> 00:08:06,330
that they're metering is middle gray.

146
00:08:06,330 --> 00:08:08,820
So there's a problem when you
meter something that is not

147
00:08:08,820 --> 00:08:12,030
middle gray like this white background.

148
00:08:12,030 --> 00:08:19,890
And so if we actually are
to open up this camera

149
00:08:19,890 --> 00:08:22,070
and allow a little bit more light in--

150
00:08:22,070 --> 00:08:24,640
what am I doing here?

151
00:08:24,640 --> 00:08:25,570
There we go.

152
00:08:25,570 --> 00:08:28,210
I has auto ISO on.

153
00:08:28,210 --> 00:08:29,280
So there's the shadow.

154
00:08:29,280 --> 00:08:31,420
Let's get this shadow
out of here so we can--

155
00:08:31,420 --> 00:08:34,819
so now all of a sudden we brighten the
image up by allowing more light in,

156
00:08:34,819 --> 00:08:40,679
and this tone is rendering
normally or as we would wish it to.

157
00:08:40,679 --> 00:08:45,990
So we can also do the same
thing if we have a black object.

158
00:08:45,990 --> 00:08:48,930
You can see that the camera
using the reflective meter

159
00:08:48,930 --> 00:08:51,570
set this value to add
exposure, which is lining up

160
00:08:51,570 --> 00:08:54,922
that small cursor with
the caret in that scale.

161
00:08:54,922 --> 00:08:56,380
You can see that this is not black.

162
00:08:56,380 --> 00:08:58,560
It's middle gray.

163
00:08:58,560 --> 00:09:03,360
It's actually rendering this incorrectly
because what it's expecting to see

164
00:09:03,360 --> 00:09:05,250
is this middle gray card.

165
00:09:05,250 --> 00:09:12,390
So what we need to do
in order to offset that

166
00:09:12,390 --> 00:09:16,440
is actually underexpose
what the camera is reading.

167
00:09:16,440 --> 00:09:19,510
So now when we check this the camera
is like you're over three stops

168
00:09:19,510 --> 00:09:22,980
underexposed, but this
starts to look correct.

169
00:09:22,980 --> 00:09:24,870
So we talked a lot about
this-- thanks, Ben.

170
00:09:24,870 --> 00:09:30,750
[INAUDIBLE]-- in the light meter video.

171
00:09:30,750 --> 00:09:35,400
And we'll come back to this, but that
is to say that in any given scene

172
00:09:35,400 --> 00:09:42,330
the exposure suggested by a
camera is made to be 90% accurate.

173
00:09:42,330 --> 00:09:46,960
Most scenes have a mix of light values,
dark values, and values in between.

174
00:09:46,960 --> 00:09:50,130
So if you sort of assume that that
all melds down to about middle gray

175
00:09:50,130 --> 00:09:53,220
you can suggest an exposure.

176
00:09:53,220 --> 00:09:55,220
And the meter will suggest
an exposure, and then

177
00:09:55,220 --> 00:09:58,080
when you expose for that
you'll be pretty accurate.

178
00:09:58,080 --> 00:10:02,690
But so this scene here is
pretty much all highlights.

179
00:10:02,690 --> 00:10:05,340
And my camera said if you
set it to these settings

180
00:10:05,340 --> 00:10:09,180
you'll get a properly
exposed image, but it's not.

181
00:10:09,180 --> 00:10:11,790
It's too dark.

182
00:10:11,790 --> 00:10:13,050
So if we open up one stop.

183
00:10:13,050 --> 00:10:14,310
Well, we're getting closer.

184
00:10:14,310 --> 00:10:16,460
Snow starting to look like snow.

185
00:10:16,460 --> 00:10:21,500
And if we open up one more stop
maybe we're a little bit too far.

186
00:10:21,500 --> 00:10:24,810
It's getting a little bit bright
over in this area over here,

187
00:10:24,810 --> 00:10:28,380
but it's rendering as
a very bright white.

188
00:10:28,380 --> 00:10:30,980
And sort of side by side
you can see it here.

189
00:10:30,980 --> 00:10:36,052
The metered exposure versus compensating
by increasing our exposure one stop.

190
00:10:36,052 --> 00:10:39,260
So this is what I mean to say when I
say that we need to be a little bit more

191
00:10:39,260 --> 00:10:40,910
intelligent than the machines.

192
00:10:40,910 --> 00:10:42,560
They're very sophisticated.

193
00:10:42,560 --> 00:10:46,130
They can measure all kinds of things,
but they're still sort of locked in

194
00:10:46,130 --> 00:10:48,310
and reference at specific values.

195
00:10:48,310 --> 00:10:50,060
So we need to understand
what those are so

196
00:10:50,060 --> 00:10:53,360
that we can compensate, and
sort of use our intention

197
00:10:53,360 --> 00:10:58,040
to override the camera when it's
making, essentially, a dumb decision.

198
00:10:58,040 --> 00:10:59,200
Yeah?

199
00:10:59,200 --> 00:11:01,158
AUDIENCE: If middle gray
is meant to be halfway

200
00:11:01,158 --> 00:11:04,625
between pure black and pure white
how come it's 18% and not 50?

201
00:11:04,625 --> 00:11:06,995
IAN: So it has to do
with the sensitivity of--

202
00:11:06,995 --> 00:11:12,320
or the way that human eyes render
light, and because it's logarithmic.

203
00:11:12,320 --> 00:11:13,940
So it's a power function.

204
00:11:13,940 --> 00:11:17,210
So 50% is not quite
halfway on a linear scale

205
00:11:17,210 --> 00:11:23,120
when it gets transformed,
I think, but essentially it

206
00:11:23,120 --> 00:11:26,002
is because if you have say--

207
00:11:26,002 --> 00:11:26,960
this is a good example.

208
00:11:26,960 --> 00:11:29,090
If you have one light
bulb, and you turn it on,

209
00:11:29,090 --> 00:11:31,535
and you add another
light bulb it's almost

210
00:11:31,535 --> 00:11:33,410
like you've doubled the
amount of brightness,

211
00:11:33,410 --> 00:11:36,202
and it's a very obvious difference,
but if you have 100 light bulbs

212
00:11:36,202 --> 00:11:41,510
and you turn one more light bulb on
it's such a tiny incremental difference.

213
00:11:41,510 --> 00:11:45,020
So we're dealing with light
on this power function scale

214
00:11:45,020 --> 00:11:48,380
where we're doubling and we're
halving, and so the value that gets us

215
00:11:48,380 --> 00:11:50,717
to that middle gray is
actually 18% and not 50%

216
00:11:50,717 --> 00:11:52,050
because it's not a linear scale.

217
00:11:52,050 --> 00:11:54,590


218
00:11:54,590 --> 00:11:57,540
Which there's plenty of math, and
I certainly won't do it justice,

219
00:11:57,540 --> 00:12:02,440
but we can dive down that
rabbit hole another time.

220
00:12:02,440 --> 00:12:04,990
So we looked a little bit
about what exposure is doing

221
00:12:04,990 --> 00:12:07,900
and how we might fool
ourselves with the light meter,

222
00:12:07,900 --> 00:12:12,200
but what are the actual settings that
I was just changing on the camera?

223
00:12:12,200 --> 00:12:15,780
I was pushing and pulling
at some buttons and dials,

224
00:12:15,780 --> 00:12:19,840
and I was clicking some things, and the
image was changing, but what was it?

225
00:12:19,840 --> 00:12:21,830
So the three main camera controls.

226
00:12:21,830 --> 00:12:27,340
The first one is ISO, which is the
sensitivity of a sensor to light,

227
00:12:27,340 --> 00:12:32,320
or a film stock, or any kind of medium
that we're using to capture light.

228
00:12:32,320 --> 00:12:34,180
The second is shutter speed.

229
00:12:34,180 --> 00:12:40,270
How long do we let light strike
that sensitive medium for?

230
00:12:40,270 --> 00:12:44,170
In the case of DSLRs it's
how long the shutter is open,

231
00:12:44,170 --> 00:12:48,760
but maybe there are other
mechanisms in play on other cameras.

232
00:12:48,760 --> 00:12:50,050
And the final one is aperture.

233
00:12:50,050 --> 00:12:53,730
So in every lens there's a
diaphragm that opens and closes.

234
00:12:53,730 --> 00:12:56,860
And depending on how open that
is or how small that is it

235
00:12:56,860 --> 00:12:59,650
lets in more or less light.

236
00:12:59,650 --> 00:13:03,740
That's a control that
we have at our disposal.

237
00:13:03,740 --> 00:13:06,730
And so by using all
three of these elements

238
00:13:06,730 --> 00:13:13,150
we can control how much light
reaches the photographic sensor,

239
00:13:13,150 --> 00:13:17,390
and how sensitive that sensor
is to light more generally.

240
00:13:17,390 --> 00:13:20,380
So let's dive in a little to each
of these categories because there

241
00:13:20,380 --> 00:13:25,910
are different artifacts that happen when
we change and control each one of them.

242
00:13:25,910 --> 00:13:27,790
So first one, ISO.

243
00:13:27,790 --> 00:13:30,970
So it, again, measures the
sensitivity of a medium to light.

244
00:13:30,970 --> 00:13:32,738
In digital cameras it's the sensor.

245
00:13:32,738 --> 00:13:35,530
In old film cameras it's actually
a type of stock, like film stock,

246
00:13:35,530 --> 00:13:38,800
and they each have different ISO values.

247
00:13:38,800 --> 00:13:41,530
And the sensitivity
doubles and halves, which

248
00:13:41,530 --> 00:13:47,350
is that unit of a stop, when the
ISO value is doubled or halved.

249
00:13:47,350 --> 00:13:51,132
And in this little scale down below
there's some common ISO values.

250
00:13:51,132 --> 00:13:53,590
A few of them you'll notice
are bigger than the other ones.

251
00:13:53,590 --> 00:13:58,090
Those are-- maybe you can
call them major stops of ISO.

252
00:13:58,090 --> 00:14:00,265
When you bought film
stock back in the day

253
00:14:00,265 --> 00:14:03,160
it generally did not come in
these third stop increments.

254
00:14:03,160 --> 00:14:08,050
It came as 100, or you'd
get 200, or 400, or 800.

255
00:14:08,050 --> 00:14:11,050
So through that they've
become our major stops,

256
00:14:11,050 --> 00:14:13,960
and there's ISO values in
between that digital cameras can

257
00:14:13,960 --> 00:14:19,450
replicate at third stop intervals
or maybe half stop intervals.

258
00:14:19,450 --> 00:14:22,220
But every time you double the number--

259
00:14:22,220 --> 00:14:28,250
so if you go from 100 to 200
it's now twice as sensitive.

260
00:14:28,250 --> 00:14:33,095
And if you half the number going from
800 to 400 it's half as sensitive.

261
00:14:33,095 --> 00:14:34,720
And it works even with the third stops.

262
00:14:34,720 --> 00:14:36,790
So say we had our camera
at 320 and we double it

263
00:14:36,790 --> 00:14:42,160
to 640 that's doubling the
ISO value, that's one stop.

264
00:14:42,160 --> 00:14:47,020
We've increased the exposure
value of the camera by one stop.

265
00:14:47,020 --> 00:14:51,100
DAN: And by default most DSLRs that
you buy, and mirrorless as well,

266
00:14:51,100 --> 00:14:53,470
will increment by thirds of a stop.

267
00:14:53,470 --> 00:14:56,530
So you can set it to half
stop or only to do full stops,

268
00:14:56,530 --> 00:14:59,530
but by default your camera
as you kind of click up--

269
00:14:59,530 --> 00:15:01,270
and this is all the exposure values--

270
00:15:01,270 --> 00:15:02,558
is a third of the stop.

271
00:15:02,558 --> 00:15:03,100
IAN: Exactly.

272
00:15:03,100 --> 00:15:07,717
So when we were shooting
on film predominantly

273
00:15:07,717 --> 00:15:09,550
you would load a camera
with a roll of film,

274
00:15:09,550 --> 00:15:11,740
and you would have one
ISO at your disposal

275
00:15:11,740 --> 00:15:13,910
until you finished that roll of film.

276
00:15:13,910 --> 00:15:15,910
So the flexibility of
digital cameras is sort of

277
00:15:15,910 --> 00:15:17,798
amazing because in any given moment--

278
00:15:17,798 --> 00:15:19,840
I had it on auto ISO by
accident, and it was just

279
00:15:19,840 --> 00:15:25,660
sort of changing ISO to every stop
in between for any different shot

280
00:15:25,660 --> 00:15:28,225
that we want to make.

281
00:15:28,225 --> 00:15:30,100
So let's do a little
exercise 'cause I really

282
00:15:30,100 --> 00:15:34,750
want to drive home this idea
of the relationality of stops.

283
00:15:34,750 --> 00:15:41,980
So if we go from 200 ISO to 800 ISO
how many stops difference is that?

284
00:15:41,980 --> 00:15:44,498
And is it more or less sensitive?

285
00:15:44,498 --> 00:15:47,290
This is a good time for the internet
to chime in if you're on there

286
00:15:47,290 --> 00:15:49,240
and want to join us.

287
00:15:49,240 --> 00:15:50,420
Yeah?

288
00:15:50,420 --> 00:15:52,915
AUDIENCE: I would say more sensitive.

289
00:15:52,915 --> 00:15:55,040
IAN: It's more sensitive,
but how much more?

290
00:15:55,040 --> 00:15:56,498
AUDIENCE: Two stops more sensitive.

291
00:15:56,498 --> 00:16:00,638
IAN: Two stops because we to start at
200, and we go to 400, and then to 800.

292
00:16:00,638 --> 00:16:01,930
So we've doubled, then doubled.

293
00:16:01,930 --> 00:16:03,850
That's two stops difference.

294
00:16:03,850 --> 00:16:04,900
Absolutely.

295
00:16:04,900 --> 00:16:05,800
More sensitive.

296
00:16:05,800 --> 00:16:07,000
Great.

297
00:16:07,000 --> 00:16:08,220
How about this one?

298
00:16:08,220 --> 00:16:11,200
Ooh, so much math.

299
00:16:11,200 --> 00:16:14,390
All the arithmetic.

300
00:16:14,390 --> 00:16:15,230
1,600 50.

301
00:16:15,230 --> 00:16:16,545
How about someone from Zoom?

302
00:16:16,545 --> 00:16:17,420
Anyone feeling brave?

303
00:16:17,420 --> 00:16:21,290


304
00:16:21,290 --> 00:16:22,970
DAN: Carla says five stops.

305
00:16:22,970 --> 00:16:23,760
IAN: Five stops.

306
00:16:23,760 --> 00:16:24,260
Yep.

307
00:16:24,260 --> 00:16:24,980
Less sensitive.

308
00:16:24,980 --> 00:16:25,640
Exactly.

309
00:16:25,640 --> 00:16:28,190
Very good.

310
00:16:28,190 --> 00:16:29,835
How about 320 to 400?

311
00:16:29,835 --> 00:16:30,960
This is a little bit weird.

312
00:16:30,960 --> 00:16:33,780


313
00:16:33,780 --> 00:16:36,998
I was talking about it halving
in doubling, being a whole stop.

314
00:16:36,998 --> 00:16:38,290
What do we think this might be?

315
00:16:38,290 --> 00:16:41,020


316
00:16:41,020 --> 00:16:42,090
One third up.

317
00:16:42,090 --> 00:16:42,710
There we go.

318
00:16:42,710 --> 00:16:44,350
Nice.

319
00:16:44,350 --> 00:16:48,430
And then finally one where we have a
weird incremental stop that's maybe not

320
00:16:48,430 --> 00:16:54,698
one of those major stops, 500 to 1,000.

321
00:16:54,698 --> 00:16:55,740
Is it doubling or having?

322
00:16:55,740 --> 00:16:56,615
Right, it's doubling.

323
00:16:56,615 --> 00:17:00,600
It's going 500 to 1,000 It's
one more stop more sensitive.

324
00:17:00,600 --> 00:17:04,200
This feels very basic in the
moment, but at the end of the day

325
00:17:04,200 --> 00:17:07,950
having a firm grasp that when
you double and half you're

326
00:17:07,950 --> 00:17:11,700
moving things in a stop increment it's
helpful because every other camera

327
00:17:11,700 --> 00:17:16,020
control also moves in stop increments.

328
00:17:16,020 --> 00:17:17,920
Can be controlled in stop increments.

329
00:17:17,920 --> 00:17:20,910
So if you add a stop here you can
take a stop away somewhere else.

330
00:17:20,910 --> 00:17:24,060


331
00:17:24,060 --> 00:17:27,210
So because there is no
free lunch in anything

332
00:17:27,210 --> 00:17:30,570
that we do and there are
nothing but trade-offs

333
00:17:30,570 --> 00:17:36,180
there are artifacts that are
associated with different ISO values.

334
00:17:36,180 --> 00:17:40,710
So increasing the sensitivity
of a medium introduces noise.

335
00:17:40,710 --> 00:17:43,050
Does anyone have a sense of
what ISO noise looks like

336
00:17:43,050 --> 00:17:44,818
or what noise is more generally?

337
00:17:44,818 --> 00:17:48,900


338
00:17:48,900 --> 00:17:49,400
Yeah.

339
00:17:49,400 --> 00:17:52,220
So in the old days of film
it was actually grain.

340
00:17:52,220 --> 00:17:56,810
You'd see particles that were actually--
the molecular structure of the film was

341
00:17:56,810 --> 00:18:00,260
bigger so that it was more sensitive
to light because there was more surface

342
00:18:00,260 --> 00:18:01,142
area.

343
00:18:01,142 --> 00:18:02,600
In electronical-- or electronical--

344
00:18:02,600 --> 00:18:03,100
[CHUCKLES]

345
00:18:03,100 --> 00:18:09,140
In electronic systems it's
actually just random data

346
00:18:09,140 --> 00:18:16,400
that gets introduced into our image.

347
00:18:16,400 --> 00:18:21,140
And so the more
sensitive ISO values tend

348
00:18:21,140 --> 00:18:24,410
to have much higher levels of noise.

349
00:18:24,410 --> 00:18:26,750
And lower ISO values have less noise.

350
00:18:26,750 --> 00:18:29,600


351
00:18:29,600 --> 00:18:33,200
It's much more apparent in
the shadows of an image.

352
00:18:33,200 --> 00:18:35,090
Why do you think that might be true?

353
00:18:35,090 --> 00:18:41,690


354
00:18:41,690 --> 00:18:42,190
Yes?

355
00:18:42,190 --> 00:18:44,140
AUDIENCE: There's probably
less data in the shadows.

356
00:18:44,140 --> 00:18:44,640
IAN: Yeah.

357
00:18:44,640 --> 00:18:46,360
So there's less data in the shadows.

358
00:18:46,360 --> 00:18:50,920
There's no information to
overwrite a noise value.

359
00:18:50,920 --> 00:18:55,442
So you can see even low
value noise in shadows

360
00:18:55,442 --> 00:18:57,400
because there's very
little data there to begin

361
00:18:57,400 --> 00:19:00,610
with whereas if you have a lot
of-- like a bright highlight

362
00:19:00,610 --> 00:19:05,120
it may overwrite sort of a
medium level value of noise.

363
00:19:05,120 --> 00:19:09,580
The first place you're going to
see noise is always in the shadows.

364
00:19:09,580 --> 00:19:12,460
So this is actually
from a cinema camera.

365
00:19:12,460 --> 00:19:15,790
This is still of our
friend Dan Armendariz.

366
00:19:15,790 --> 00:19:21,820
And here we can see him sitting in
this magnificent room at 400 ISO.

367
00:19:21,820 --> 00:19:23,940
I don't see too much
wrong with this image.

368
00:19:23,940 --> 00:19:25,590
I don't see a lot of noise anywhere.

369
00:19:25,590 --> 00:19:29,780
It sort of seems fairly
clear and crisp to me.

370
00:19:29,780 --> 00:19:36,540
At 3,200, well, I can sort of start
to see a little bit of degradation.

371
00:19:36,540 --> 00:19:41,620
If we go back and then forward
there's some something there.

372
00:19:41,620 --> 00:19:45,500
Still not crazy-- it doesn't
seem from this distance.

373
00:19:45,500 --> 00:19:49,600
But then if we go to
12,800 that looks wild.

374
00:19:49,600 --> 00:19:52,060
There's just so much
textural element to it.

375
00:19:52,060 --> 00:19:53,550
So much noise.

376
00:19:53,550 --> 00:19:56,240
And so it's a little
hard to see at that scale

377
00:19:56,240 --> 00:20:00,820
so if we zoom in we have
Dan looking handsome.

378
00:20:00,820 --> 00:20:03,890
Dan looking, ooh, a little horse.

379
00:20:03,890 --> 00:20:04,390
And whoa.

380
00:20:04,390 --> 00:20:07,350


381
00:20:07,350 --> 00:20:10,618
And you definitely can see
it down in the shadow areas.

382
00:20:10,618 --> 00:20:12,410
And I didn't mean to
make fun of Dan there.

383
00:20:12,410 --> 00:20:14,980
He's a handsome man.

384
00:20:14,980 --> 00:20:18,130
DAN: Ian, you we're talking about
the ISO levels like at 400--

385
00:20:18,130 --> 00:20:18,630
IAN: Mm-hm.

386
00:20:18,630 --> 00:20:20,672
DAN: I forget-- something
in the middle, and then

387
00:20:20,672 --> 00:20:22,910
1,200, but is every camera the same?

388
00:20:22,910 --> 00:20:23,410
IAN: No.

389
00:20:23,410 --> 00:20:25,640
So every camera is not the same.

390
00:20:25,640 --> 00:20:31,810
The range of ISO values that you have
available to you is hardware dependent.

391
00:20:31,810 --> 00:20:38,200
So you'll find that certain cameras
can go from 50 to maybe 32,000.

392
00:20:38,200 --> 00:20:40,090
Some of them have ludicrous numbers.

393
00:20:40,090 --> 00:20:43,240
Like the A7S, I think, is
like 120,000 or something.

394
00:20:43,240 --> 00:20:46,060
It can see in the dark.

395
00:20:46,060 --> 00:20:47,980
Whereas a film camera
might be locked into one

396
00:20:47,980 --> 00:20:51,310
ISO because you put
100 speed film in it.

397
00:20:51,310 --> 00:20:56,050
Or maybe your camera only
has ISOs from 50 to 3,200.

398
00:20:56,050 --> 00:20:58,300
And maybe it only has half
stop increments in between.

399
00:20:58,300 --> 00:20:59,890
It doesn't have third stops.

400
00:20:59,890 --> 00:21:04,150
So the range of ISOs that you have
available to you is hardware specific,

401
00:21:04,150 --> 00:21:08,650
but the relationship is the
same regardless of hardware.

402
00:21:08,650 --> 00:21:10,840
Doubling or having
increases or decreases

403
00:21:10,840 --> 00:21:14,900
the sensitivity of the
medium by half or double.

404
00:21:14,900 --> 00:21:16,260
DAN: And my question also--

405
00:21:16,260 --> 00:21:20,830
like 400 on one camera, is it the
same as 400 on another camera?

406
00:21:20,830 --> 00:21:26,980
IAN: I would say no because
at the end of the day

407
00:21:26,980 --> 00:21:32,980
the way the hardware is interpreting
the electronic signals is

408
00:21:32,980 --> 00:21:34,840
different for every single camera.

409
00:21:34,840 --> 00:21:38,997
And so you may have one camera and
say, oh, there's no noise at 400 ISO,

410
00:21:38,997 --> 00:21:42,080
and then use a different camera, and
you may find that in a specific image

411
00:21:42,080 --> 00:21:45,670
there is a lot of noise at 400 ISO.

412
00:21:45,670 --> 00:21:52,690
And so there is definite variability in
the quality of the electronic circuits

413
00:21:52,690 --> 00:21:54,370
that are in digital cameras.

414
00:21:54,370 --> 00:21:56,790
And so obviously higher
tier cameras are going

415
00:21:56,790 --> 00:22:02,610
to do better at more ISO
values than lower tier cameras.

416
00:22:02,610 --> 00:22:03,628
So, yeah.

417
00:22:03,628 --> 00:22:05,420
DAN: So sensitivity-wise
they are the same,

418
00:22:05,420 --> 00:22:08,230
but there are definitely trade
offs between different models

419
00:22:08,230 --> 00:22:10,280
and different sensors as far as quality?

420
00:22:10,280 --> 00:22:10,780
IAN: Right.

421
00:22:10,780 --> 00:22:11,280
Exactly.

422
00:22:11,280 --> 00:22:14,710
So has the same sensitivity,
but you may actually

423
00:22:14,710 --> 00:22:17,110
end up with more noise,
more artifacts, and it's

424
00:22:17,110 --> 00:22:19,540
because of the quality of the camera.

425
00:22:19,540 --> 00:22:21,670
That's a good way to say it.

426
00:22:21,670 --> 00:22:25,990
So why would you ever accept more noise?

427
00:22:25,990 --> 00:22:27,970
Why would you why would
you increase the noise?

428
00:22:27,970 --> 00:22:31,360
If it looks so bad why would you do it?

429
00:22:31,360 --> 00:22:33,280
Well, it's a really pragmatic decision.

430
00:22:33,280 --> 00:22:36,850
There sometimes is
just not enough light.

431
00:22:36,850 --> 00:22:39,580
It's nighttime, it's dark, it's dusk.

432
00:22:39,580 --> 00:22:41,680
There's just not a lot
of light and so you

433
00:22:41,680 --> 00:22:45,010
need to boost up the
sensitivity of the sensor

434
00:22:45,010 --> 00:22:48,580
to even render any kind of image.

435
00:22:48,580 --> 00:22:51,010
So there's a very practical side of it.

436
00:22:51,010 --> 00:22:53,380
Or maybe you're going
for aesthetic effect.

437
00:22:53,380 --> 00:22:54,670
To mimic grainy footage.

438
00:22:54,670 --> 00:22:59,050
Maybe you're trying to mimic some
surveillance footage or something

439
00:22:59,050 --> 00:23:01,270
like that, or you want to go back to--

440
00:23:01,270 --> 00:23:04,450
a lot of street photographers
used to shoot 3,200 speed film

441
00:23:04,450 --> 00:23:07,780
and it always had really
heavy grain in it.

442
00:23:07,780 --> 00:23:10,270
And maybe you like that
kind of look and style

443
00:23:10,270 --> 00:23:12,970
and you want to add a
little bit of more noise

444
00:23:12,970 --> 00:23:18,430
into your image to mimic that
for a textural effect, perhaps.

445
00:23:18,430 --> 00:23:20,230
Here's a picture of a ghost I took.

446
00:23:20,230 --> 00:23:22,310
I was ghost hunting last night.

447
00:23:22,310 --> 00:23:23,634
Do you see it?

448
00:23:23,634 --> 00:23:24,910
AUDIENCE: [INAUDIBLE]

449
00:23:24,910 --> 00:23:25,017
IAN: All right.

450
00:23:25,017 --> 00:23:27,059
Well, you guys will have
to investigate it later.

451
00:23:27,059 --> 00:23:30,010


452
00:23:30,010 --> 00:23:31,150
So shutter speed.

453
00:23:31,150 --> 00:23:32,350
So that's ISO.

454
00:23:32,350 --> 00:23:35,170
It's the sensitivity of the sensor.

455
00:23:35,170 --> 00:23:37,450
Increasing the sensitivity
introduces noise.

456
00:23:37,450 --> 00:23:40,690
Lowering the sensitivity
sort of minimizes noise,

457
00:23:40,690 --> 00:23:45,440
but it also requires more light
so there's a trade off there.

458
00:23:45,440 --> 00:23:47,990
So shutter speed.

459
00:23:47,990 --> 00:23:52,280
So shutter speed is the amount of
time the sensor is exposed to light.

460
00:23:52,280 --> 00:23:55,130
There is a shutter in here.

461
00:23:55,130 --> 00:23:57,410
It opens for some period of time.

462
00:23:57,410 --> 00:24:01,040
The sensor is struck by
light and then it closes.

463
00:24:01,040 --> 00:24:03,410
We measure it in fractions
of a second, though you

464
00:24:03,410 --> 00:24:05,960
may see them written on
cameras as integer values.

465
00:24:05,960 --> 00:24:07,970
This is to save space.

466
00:24:07,970 --> 00:24:10,640
On those tiny knobs
you may see like 1,000.

467
00:24:10,640 --> 00:24:16,400
That's actually one over
1,000, 1/1000 of a second.

468
00:24:16,400 --> 00:24:19,950
And listed there are the
major stops or shutter speeds.

469
00:24:19,950 --> 00:24:25,070
This is what you would find on most
old school 35 millimeter film cameras.

470
00:24:25,070 --> 00:24:27,920
Starting at a second and
going up to about 1/1000.

471
00:24:27,920 --> 00:24:29,780
So, again, hardware specific.

472
00:24:29,780 --> 00:24:32,960
Your camera may have more
shutter speeds available

473
00:24:32,960 --> 00:24:37,010
to you, and in different increments.

474
00:24:37,010 --> 00:24:39,650
Or it may have less.

475
00:24:39,650 --> 00:24:44,150
But the important thing to take away
is that doubling the length of time

476
00:24:44,150 --> 00:24:47,690
doubles the amount of light
that can strike the sensor.

477
00:24:47,690 --> 00:24:50,660
Halving the length of time
halves the amount of light

478
00:24:50,660 --> 00:24:51,980
that can strike the sensor.

479
00:24:51,980 --> 00:24:55,550
So, again, we have this stop interval.

480
00:24:55,550 --> 00:25:01,610
So you could imagine that if I have to
decrease the sensitivity of my sensor

481
00:25:01,610 --> 00:25:05,720
maybe I can open the
shutter for twice as long.

482
00:25:05,720 --> 00:25:08,240
So go down one stop here,
and up one stop here.

483
00:25:08,240 --> 00:25:12,996
And it's the same amount of light just
manifesting a little bit differently.

484
00:25:12,996 --> 00:25:16,810


485
00:25:16,810 --> 00:25:19,930
So we're gonna count stops again.

486
00:25:19,930 --> 00:25:22,330
So a 1/60 to a 1/15.

487
00:25:22,330 --> 00:25:23,410
DAN: 1/60 of second.

488
00:25:23,410 --> 00:25:24,730
IAN: 1/60 of a second.

489
00:25:24,730 --> 00:25:25,300
I apologize.

490
00:25:25,300 --> 00:25:27,820
I shorthand a lot of this,
but I should be more specific.

491
00:25:27,820 --> 00:25:32,140
So 1/60 of a second to 1/15 of a second.

492
00:25:32,140 --> 00:25:34,000
So it can be a little counter intuitive.

493
00:25:34,000 --> 00:25:38,348
1/60 of a pizza is more or
less than 1/15 of a pizza?

494
00:25:38,348 --> 00:25:40,681
That's how I have to do it
my head to be frank with you.

495
00:25:40,681 --> 00:25:42,094
[CHUCKLES]

496
00:25:42,094 --> 00:25:45,255


497
00:25:45,255 --> 00:25:45,880
How many stops?

498
00:25:45,880 --> 00:25:46,380
Yes?

499
00:25:46,380 --> 00:25:47,230
AUDIENCE: Two stops.

500
00:25:47,230 --> 00:25:48,310
IAN: Two stops.

501
00:25:48,310 --> 00:25:52,150
So it goes from 1/60
to 1/30, 1/30 to 1/15.

502
00:25:52,150 --> 00:25:56,230
Two stops more sensitive, or
more light striking the sensor.

503
00:25:56,230 --> 00:26:02,360
1/1000 to 1/30?

504
00:26:02,360 --> 00:26:04,900
So much counting.

505
00:26:04,900 --> 00:26:06,746
DAN: Alex says five stop.

506
00:26:06,746 --> 00:26:07,580
IAN: Bam.

507
00:26:07,580 --> 00:26:08,130
Five stops.

508
00:26:08,130 --> 00:26:12,960


509
00:26:12,960 --> 00:26:13,750
And that's wrong.

510
00:26:13,750 --> 00:26:15,365
That should say more sensitive.

511
00:26:15,365 --> 00:26:17,947


512
00:26:17,947 --> 00:26:18,530
That's a typo.

513
00:26:18,530 --> 00:26:19,820
I apologize.

514
00:26:19,820 --> 00:26:23,182
Because we're going from 1/1000
of a second to 1/30 of a second

515
00:26:23,182 --> 00:26:24,140
that's much more light.

516
00:26:24,140 --> 00:26:25,932
That should read five
stops more sensitive.

517
00:26:25,932 --> 00:26:29,210
I'll fix that before
the end of the lecture.

518
00:26:29,210 --> 00:26:30,920
So 1/500 to 1/400.

519
00:26:30,920 --> 00:26:31,880
This is odd.

520
00:26:31,880 --> 00:26:34,400
That doesn't feel like
a doubling or halving.

521
00:26:34,400 --> 00:26:36,056
What might that be?

522
00:26:36,056 --> 00:26:37,550
AUDIENCE: A third, maybe?

523
00:26:37,550 --> 00:26:39,010
IAN: A third.

524
00:26:39,010 --> 00:26:41,540
Yep, a third more sensitive.

525
00:26:41,540 --> 00:26:44,750
And from 1/180 to 1/90?

526
00:26:44,750 --> 00:26:50,130
Those are odd numbers that weren't
on our list, but it's still half.

527
00:26:50,130 --> 00:26:51,550
Yep.

528
00:26:51,550 --> 00:26:52,050
Exactly.

529
00:26:52,050 --> 00:26:53,120
So it's one stop more.

530
00:26:53,120 --> 00:26:55,860


531
00:26:55,860 --> 00:27:00,150
So like all things that
we've talked about there

532
00:27:00,150 --> 00:27:03,120
is a trade-off with shutter speed.

533
00:27:03,120 --> 00:27:07,770
You can't just open up
your camera for as long

534
00:27:07,770 --> 00:27:11,430
as you want and still expect
to render a crisp image.

535
00:27:11,430 --> 00:27:14,520
So this is an image of a
dam that Dan took, and you

536
00:27:14,520 --> 00:27:16,515
can see that the water is frozen.

537
00:27:16,515 --> 00:27:19,350


538
00:27:19,350 --> 00:27:21,750
It was falling down, and it's
literally frozen in midair.

539
00:27:21,750 --> 00:27:24,150
So this must be a very
short shutter speed

540
00:27:24,150 --> 00:27:28,680
in order to freeze motion like this.

541
00:27:28,680 --> 00:27:31,290
The blink of an eye.

542
00:27:31,290 --> 00:27:36,000
Here's the same image with
the same amount of light.

543
00:27:36,000 --> 00:27:39,600
So technically these
exposures are equivalent,

544
00:27:39,600 --> 00:27:42,040
but they look very different.

545
00:27:42,040 --> 00:27:47,310
In this one there's time for the water
to fall all the way down and create

546
00:27:47,310 --> 00:27:52,120
this sort of streaking
effect in the image.

547
00:27:52,120 --> 00:27:54,780
So this is a slower shutter speed.

548
00:27:54,780 --> 00:27:55,590
What is it?

549
00:27:55,590 --> 00:27:56,710
Half a second.

550
00:27:56,710 --> 00:27:59,653
So the shutter is open
for half a second,

551
00:27:59,653 --> 00:28:01,320
whereas in the previous one what was it?

552
00:28:01,320 --> 00:28:05,410
1/4000 of a second.

553
00:28:05,410 --> 00:28:06,660
That's beyond fathoming to me.

554
00:28:06,660 --> 00:28:09,320
It's a little too fast.

555
00:28:09,320 --> 00:28:11,370
And so this image
takes a second to load,

556
00:28:11,370 --> 00:28:15,060
but you can see here a
composite image that Dan's made

557
00:28:15,060 --> 00:28:18,930
that starts at very slow shutter speeds,
and moves towards very fast shutter

558
00:28:18,930 --> 00:28:21,322
speeds.

559
00:28:21,322 --> 00:28:23,280
And if we actually put
the shutter speeds there

560
00:28:23,280 --> 00:28:28,620
you can see incrementally
where motion begins

561
00:28:28,620 --> 00:28:31,980
to look how we might
perceive it normally

562
00:28:31,980 --> 00:28:35,040
into being frozen
instantaneously or drifting

563
00:28:35,040 --> 00:28:39,240
into this fairy-like wishy-ness.

564
00:28:39,240 --> 00:28:42,750
DAN: This kind of raises the question
of how fast is fast enough too?

565
00:28:42,750 --> 00:28:45,540
So it really depends on your subject.

566
00:28:45,540 --> 00:28:51,570
For sports maybe one 1/250 of a second
is fast enough to freeze motion,

567
00:28:51,570 --> 00:28:55,020
but obviously to freeze waterfall you
need to crank it up even further, so.

568
00:28:55,020 --> 00:28:55,770
IAN: Yeah.

569
00:28:55,770 --> 00:28:56,562
DAN: So it depends.

570
00:28:56,562 --> 00:28:58,410
IAN: And also perception
too because when

571
00:28:58,410 --> 00:29:03,660
I think about when I turn my tap on, and
what I see I see something over here.

572
00:29:03,660 --> 00:29:07,140
I can't make out individual
drops unless I'm tracking them.

573
00:29:07,140 --> 00:29:09,720
It's gushing, but it
doesn't look like this.

574
00:29:09,720 --> 00:29:12,313
This is sort of very ethereal.

575
00:29:12,313 --> 00:29:14,730
And it definitely doesn't look
frozen in time like the one

576
00:29:14,730 --> 00:29:19,740
before so the speed
of the object matters.

577
00:29:19,740 --> 00:29:22,770
And sort of our perception
of that speed as humans

578
00:29:22,770 --> 00:29:25,579
defines how what it
looks naturalistically.

579
00:29:25,579 --> 00:29:29,072


580
00:29:29,072 --> 00:29:33,563
AUDIENCE: That's to say human eyes
are sort of equivalent to some f-stop,

581
00:29:33,563 --> 00:29:36,560
would you say?

582
00:29:36,560 --> 00:29:37,490
IAN: I don't know.

583
00:29:37,490 --> 00:29:38,890
Maybe--

584
00:29:38,890 --> 00:29:41,920
DAN: There are arguments
online that yes, but--

585
00:29:41,920 --> 00:29:42,920
IAN: Oh, I'm going back.

586
00:29:42,920 --> 00:29:46,290
DAN: I think It varies very much
by person to person as well.

587
00:29:46,290 --> 00:29:46,790
IAN: Yeah.

588
00:29:46,790 --> 00:29:52,730
And I think the argument is you
have to be looking at a fixed point.

589
00:29:52,730 --> 00:29:55,610
And everything-- when
humans track objects--

590
00:29:55,610 --> 00:29:59,630
fast moving objects to achieve
more clarity than maybe

591
00:29:59,630 --> 00:30:02,540
if something passes right
in front of your vision

592
00:30:02,540 --> 00:30:04,670
while you look at a fixed point.

593
00:30:04,670 --> 00:30:07,430
So it's a really hard
experiment to perform

594
00:30:07,430 --> 00:30:09,740
because people eyes are always moving.

595
00:30:09,740 --> 00:30:13,280
And if you move something in front
of them they just will track it.

596
00:30:13,280 --> 00:30:14,540
And you sort of get--

597
00:30:14,540 --> 00:30:17,450
it's all skewed data I would say.

598
00:30:17,450 --> 00:30:20,210
But there probably is
some sort of upper bound

599
00:30:20,210 --> 00:30:23,810
of what we can resolve with our eyes.

600
00:30:23,810 --> 00:30:27,110
Just as fast as they refresh.

601
00:30:27,110 --> 00:30:29,570
So this is last night's snowy night.

602
00:30:29,570 --> 00:30:32,210
Snow is just streaking
down through these wires.

603
00:30:32,210 --> 00:30:34,693


604
00:30:34,693 --> 00:30:35,985
Here's another one with a tree.

605
00:30:35,985 --> 00:30:38,940


606
00:30:38,940 --> 00:30:42,530
DAN: Just to go back one second
to the previous conversation.

607
00:30:42,530 --> 00:30:45,470
I think it also largely
comes down to something

608
00:30:45,470 --> 00:30:49,490
known as frame rate, which is how
many images you see in sequence,

609
00:30:49,490 --> 00:30:53,700
and what human perception is to believe
something is actually in motion.

610
00:30:53,700 --> 00:30:56,780
And so I think that is a
bigger part than shutter speed,

611
00:30:56,780 --> 00:31:00,230
for example, as far as what
feels natural and feels normal.

612
00:31:00,230 --> 00:31:03,380
And we're going to punt to
the next lecture in here

613
00:31:03,380 --> 00:31:05,025
on that when we talk about video.

614
00:31:05,025 --> 00:31:05,525
IAN: Yeah.

615
00:31:05,525 --> 00:31:08,680
It's sort of the idea of
persistence of vision.

616
00:31:08,680 --> 00:31:10,490
Video is not moving images.

617
00:31:10,490 --> 00:31:13,370
It's a sequence of still
images, and we perceive them

618
00:31:13,370 --> 00:31:16,490
as moving because they happen
fast enough that we don't notice.

619
00:31:16,490 --> 00:31:19,490
So there is some sort of
threshold where that happens.

620
00:31:19,490 --> 00:31:23,110
I'm just not convinced
I know what it is.

621
00:31:23,110 --> 00:31:25,940
So to sort of jump
back to that idea of we

622
00:31:25,940 --> 00:31:30,050
can actually track objects to
increase the perceptive focus of this,

623
00:31:30,050 --> 00:31:34,430
this is a car that's driving
by, but by panning with the car

624
00:31:34,430 --> 00:31:39,500
I was able to capture this in focus
while the rest of it is out of focus

625
00:31:39,500 --> 00:31:41,360
at a much slower shutter speed.

626
00:31:41,360 --> 00:31:44,100
If I hadn't moved the camera at
all everything would be blurred.

627
00:31:44,100 --> 00:31:45,940
It would just be a large streak.

628
00:31:45,940 --> 00:31:47,510
DAN: And I think we should
be careful with what you say

629
00:31:47,510 --> 00:31:49,340
is in focus versus what's not in focus.

630
00:31:49,340 --> 00:31:52,160
This is actually motion blur,
which is different from something

631
00:31:52,160 --> 00:31:53,270
being in soft focus.

632
00:31:53,270 --> 00:31:54,290
IAN: That is true.

633
00:31:54,290 --> 00:31:56,720
A very important distinction.

634
00:31:56,720 --> 00:31:59,120
So this is actually--

635
00:31:59,120 --> 00:32:02,930
as it moves across the sensor
it ends up looking blurry,

636
00:32:02,930 --> 00:32:05,060
which is the idea of that
motion blur for sure.

637
00:32:05,060 --> 00:32:08,400


638
00:32:08,400 --> 00:32:14,430
So it does lead us to this question
that if shutter speeds get low enough

639
00:32:14,430 --> 00:32:18,580
you can actually introduce shake or
movement just from your own human body.

640
00:32:18,580 --> 00:32:19,860
This camera's on a tripod.

641
00:32:19,860 --> 00:32:23,020
It's very stable, but if I hold
something I'm always moving.

642
00:32:23,020 --> 00:32:24,080
Always.

643
00:32:24,080 --> 00:32:25,680
Try as hard as I might.

644
00:32:25,680 --> 00:32:28,950
So the rule of thumb is to minimize
hand-held camera shake set your shutter

645
00:32:28,950 --> 00:32:32,490
speed to faster than one
over your focal length.

646
00:32:32,490 --> 00:32:35,910
So it seems confusing, but I have
a 50 millimeter lens on here.

647
00:32:35,910 --> 00:32:38,970
In order for there not to be any
camera shake when I hand hold it

648
00:32:38,970 --> 00:32:43,560
I should be shooting it
faster than 1/50 of a second.

649
00:32:43,560 --> 00:32:47,760
If I put a 70 millimeter lens on here
or a 100 millimeter lens then maybe

650
00:32:47,760 --> 00:32:53,550
I want to shoot at 1/70 or 1/100.

651
00:32:53,550 --> 00:32:57,155
So, especially-- this comes into play
a little bit more too with zoom lenses

652
00:32:57,155 --> 00:33:00,030
where you might be zooming in and
out and changing your focal length,

653
00:33:00,030 --> 00:33:02,790
and not sort of paying attention
to what shutter speed you're at.

654
00:33:02,790 --> 00:33:09,180
And you're say at 1/70 of second and
you're shooting at 70 millimeters,

655
00:33:09,180 --> 00:33:11,640
but then you sort of snap
into 200 all of a sudden

656
00:33:11,640 --> 00:33:17,010
you're going to introduce
camera shake into the image.

657
00:33:17,010 --> 00:33:21,090
A little bit of motion blur that
reduces the crispness of your images.

658
00:33:21,090 --> 00:33:24,530


659
00:33:24,530 --> 00:33:26,130
So aperture.

660
00:33:26,130 --> 00:33:31,010
The final of our three is the
size of the opening in a lens.

661
00:33:31,010 --> 00:33:36,950
And we have a little short
video to tease you with this

662
00:33:36,950 --> 00:33:38,300
if we can just watch like this.

663
00:33:38,300 --> 00:33:40,970


664
00:33:40,970 --> 00:33:45,583
So as the lens gets more and more
open what do you notice happening?

665
00:33:45,583 --> 00:33:47,000
AUDIENCE: The numbers get smaller.

666
00:33:47,000 --> 00:33:48,380
IAN: Yeah, the numbers get smaller.

667
00:33:48,380 --> 00:33:48,950
This is odd.

668
00:33:48,950 --> 00:33:51,770


669
00:33:51,770 --> 00:33:53,690
Interesting.

670
00:33:53,690 --> 00:33:58,740
So aperture actually refers to the size
of the diaphragm opening in a lens.

671
00:33:58,740 --> 00:34:01,490
And it's a fractional relationship
between the size of the opening

672
00:34:01,490 --> 00:34:02,615
and the length of the lens.

673
00:34:02,615 --> 00:34:07,970
So it, again, is
written in the integer--

674
00:34:07,970 --> 00:34:13,258
as integers or decimals on camera
bodies and lenses and things like that,

675
00:34:13,258 --> 00:34:15,050
but it's actually like
a fractional amount.

676
00:34:15,050 --> 00:34:18,239
So it's 1/2 or 1/2.8.

677
00:34:18,239 --> 00:34:21,949
So 1/22 is smaller than one half.

678
00:34:21,949 --> 00:34:24,620
And that's why the
diameter of the opening

679
00:34:24,620 --> 00:34:28,097
gets smaller the larger the number gets.

680
00:34:28,097 --> 00:34:30,139
It's a little frustrating,
and counter intuitive,

681
00:34:30,139 --> 00:34:32,540
and can be confusing, if
you're not used to this,

682
00:34:32,540 --> 00:34:36,290
but with practice I
promise you'll grab it.

683
00:34:36,290 --> 00:34:42,860
So the major f-stops are
listed below up to 22.

684
00:34:42,860 --> 00:34:48,889
You can have smaller apertures
like f/32, or 45, or 64.

685
00:34:48,889 --> 00:34:50,389
And there are increments in between.

686
00:34:50,389 --> 00:34:51,949
There are third stop
increments in between.

687
00:34:51,949 --> 00:34:54,996
You may find someone that shoots
something at f/9 or something like that

688
00:34:54,996 --> 00:34:56,913
and you're like, well,
that's not listed here,

689
00:34:56,913 --> 00:35:00,410
but so there's increments in between.

690
00:35:00,410 --> 00:35:03,860
And just to drive home the point
that the smaller the f number

691
00:35:03,860 --> 00:35:05,720
is the larger the opening.

692
00:35:05,720 --> 00:35:10,850
So over here we're in like 2.8,
and over there we're at f/22.

693
00:35:10,850 --> 00:35:14,120
And so when you refer to
the size of an aperture

694
00:35:14,120 --> 00:35:17,240
you might just say F
whatever the number value.

695
00:35:17,240 --> 00:35:22,400
So I'm at f/2, or I'm at
f/16, or f/8 to denote

696
00:35:22,400 --> 00:35:25,160
that you're talking about
the size of the aperture,

697
00:35:25,160 --> 00:35:30,254
and not some other numerical value
associated with camera exposure.

698
00:35:30,254 --> 00:35:33,200


699
00:35:33,200 --> 00:35:36,770
So artifacts of aperture.

700
00:35:36,770 --> 00:35:40,580
Again, there are
trade-offs with everything.

701
00:35:40,580 --> 00:35:44,900
So changing the aperture directly
affects the depth of field of an image.

702
00:35:44,900 --> 00:35:47,660
We have not talked about
depth of field except sort

703
00:35:47,660 --> 00:35:50,120
of in glancing blows in critique.

704
00:35:50,120 --> 00:35:56,180
And depth of field is defined as the
amount of an image in apparent focus.

705
00:35:56,180 --> 00:35:59,960
So in reality with the
way optics are there's

706
00:35:59,960 --> 00:36:03,920
only one single plane of
critical focus in an image

707
00:36:03,920 --> 00:36:08,540
that runs perpendicular
to the lens axis.

708
00:36:08,540 --> 00:36:11,027
And it's set at some
distance from the lens.

709
00:36:11,027 --> 00:36:12,860
And if you look at your
lens you'll probably

710
00:36:12,860 --> 00:36:15,860
see that there's feet and
meter indicators on there.

711
00:36:15,860 --> 00:36:17,848
And when you adjust
those to a witness mark

712
00:36:17,848 --> 00:36:20,015
that's how far away that
critical plane of focus is.

713
00:36:20,015 --> 00:36:22,940


714
00:36:22,940 --> 00:36:27,590
But, before we get there,
that's sort of obviously not

715
00:36:27,590 --> 00:36:28,870
how we experience photographs.

716
00:36:28,870 --> 00:36:30,620
We often look at
photographs where there's

717
00:36:30,620 --> 00:36:36,500
more things in focus on the
z-axis than a single plane.

718
00:36:36,500 --> 00:36:40,370
So there's some artifact
that's happening

719
00:36:40,370 --> 00:36:43,460
when cameras make images that
allow us to have more in focus

720
00:36:43,460 --> 00:36:46,850
than just a single plane, and
that is what depth of field is.

721
00:36:46,850 --> 00:36:52,502
How much distance on the
z-axis is in apparent focus.

722
00:36:52,502 --> 00:36:54,460
And we'll come all the
way back around to depth

723
00:36:54,460 --> 00:36:56,210
of field in just a few
minutes, but that's

724
00:36:56,210 --> 00:36:59,320
the concept that we're talking about,
and aperture directly affects it.

725
00:36:59,320 --> 00:37:00,100
Oops.

726
00:37:00,100 --> 00:37:01,650
Forward.

727
00:37:01,650 --> 00:37:04,995
So here we have an image
of a young man in--

728
00:37:04,995 --> 00:37:06,000
DAN: Handsome young man.

729
00:37:06,000 --> 00:37:06,260
IAN: What?

730
00:37:06,260 --> 00:37:07,343
DAN: A handsome young man.

731
00:37:07,343 --> 00:37:11,390
IAN: Handsome young man in a swing.

732
00:37:11,390 --> 00:37:16,303
And I used Dan's son as my stand in.

733
00:37:16,303 --> 00:37:18,845
But you'll notice that the back
of the image is out of focus.

734
00:37:18,845 --> 00:37:21,805


735
00:37:21,805 --> 00:37:22,680
And so we're at what?

736
00:37:22,680 --> 00:37:24,620
At f/1.4.

737
00:37:24,620 --> 00:37:27,230
So in this image-- which
is a really large opening--

738
00:37:27,230 --> 00:37:29,360
really large aperture opening.

739
00:37:29,360 --> 00:37:35,140
If we go one more image we're now at
11 so the aperture has gotten smaller.

740
00:37:35,140 --> 00:37:37,640
And we've compensated for that
with other exposure controls,

741
00:37:37,640 --> 00:37:40,858
but now you can see that all
of this background is in focus.

742
00:37:40,858 --> 00:37:43,400
I love that he sort of looks
over his shoulder at that point.

743
00:37:43,400 --> 00:37:45,610
What's that?

744
00:37:45,610 --> 00:37:48,890
And so side by side you can see that
these images while the exposure is

745
00:37:48,890 --> 00:37:53,390
the same, the brightness
of the light values

746
00:37:53,390 --> 00:37:58,540
and the darkness of the dark values
is the same, they look very different.

747
00:37:58,540 --> 00:38:02,600


748
00:38:02,600 --> 00:38:06,090
And so you can use depth of
field as a creative tool.

749
00:38:06,090 --> 00:38:09,900
And it's often used to separate
people from environments

750
00:38:09,900 --> 00:38:15,090
to make things more intimate, or to
show how expansive an environment might

751
00:38:15,090 --> 00:38:19,510
be if you go the other way.

752
00:38:19,510 --> 00:38:20,880
DAN: Can I--

753
00:38:20,880 --> 00:38:21,380
IAN: Yeah.

754
00:38:21,380 --> 00:38:24,100
DAN: [? Can ?] I give an easy
way to remember the f-stops?

755
00:38:24,100 --> 00:38:24,600
IAN: Yeah.

756
00:38:24,600 --> 00:38:26,310
DAN: Can you set a drawing for me?

757
00:38:26,310 --> 00:38:27,120
IAN: Yeah, I can.

758
00:38:27,120 --> 00:38:30,418
DAN: So I to this day I have
trouble remembering these numbers,

759
00:38:30,418 --> 00:38:32,460
just like-- you know, you
get to know the majors,

760
00:38:32,460 --> 00:38:36,420
but the easiest way for me to
visualize this as you're going along

761
00:38:36,420 --> 00:38:39,383
is start with one and 1.4.

762
00:38:39,383 --> 00:38:42,300
And so the nice thing now is you can
just keep doubling along the way.

763
00:38:42,300 --> 00:38:44,520
So one you double to two.

764
00:38:44,520 --> 00:38:47,238
1.4 doubles to 2.8.

765
00:38:47,238 --> 00:38:49,530
Two doubles to four.

766
00:38:49,530 --> 00:38:52,140
2.8 doubles to 5.6.

767
00:38:52,140 --> 00:38:54,330
Four doubles to eight.

768
00:38:54,330 --> 00:38:56,760
5.6 doubles to approximately 11.

769
00:38:56,760 --> 00:38:58,070
We round here.

770
00:38:58,070 --> 00:39:00,606
Eight doubles to 16.

771
00:39:00,606 --> 00:39:02,610
11 doubles to 22.

772
00:39:02,610 --> 00:39:07,440
And that covers most lenses that
you'll pick up and operate with.

773
00:39:07,440 --> 00:39:12,092
So if you're trying to remember this
scale just remember one and 1.4,

774
00:39:12,092 --> 00:39:13,050
and then keep doubling.

775
00:39:13,050 --> 00:39:17,695


776
00:39:17,695 --> 00:39:18,570
IAN: Quick and dirty.

777
00:39:18,570 --> 00:39:22,380


778
00:39:22,380 --> 00:39:25,020
So we talked-- oops.

779
00:39:25,020 --> 00:39:26,970
So I was alluding to--

780
00:39:26,970 --> 00:39:28,815
these have the same--

781
00:39:28,815 --> 00:39:31,440
they're allowing the same amount
of light to strike the sensor.

782
00:39:31,440 --> 00:39:34,890
They're just using different
settings in order to do this.

783
00:39:34,890 --> 00:39:37,020
So that means that
there's some sort of idea

784
00:39:37,020 --> 00:39:40,140
that you can have different
exposure settings that

785
00:39:40,140 --> 00:39:44,520
yield equivalently exposed
images with different artifacts.

786
00:39:44,520 --> 00:39:48,210
And so as a photographer,
or an image maker,

787
00:39:48,210 --> 00:39:53,370
you have to make decisions about which
artifacts you want and which ones you

788
00:39:53,370 --> 00:39:54,360
don't.

789
00:39:54,360 --> 00:39:59,340
And it comes back to that idea that we
were talking about so much of intention

790
00:39:59,340 --> 00:40:03,090
and supporting your narrative story.

791
00:40:03,090 --> 00:40:07,520
'Cause this image over here is
very much about this young boy.

792
00:40:07,520 --> 00:40:11,700
And this image is actually about
this boy in a larger environment.

793
00:40:11,700 --> 00:40:15,420
And by making the decision to
have narrow depth of field we're

794
00:40:15,420 --> 00:40:18,270
focused in with the child, but
to have expensive depth of field

795
00:40:18,270 --> 00:40:24,060
we're sort of looking at the child
in relationship to the space.

796
00:40:24,060 --> 00:40:25,560
So exposure equivalencies.

797
00:40:25,560 --> 00:40:29,760
So we can expose the same
scene with different settings,

798
00:40:29,760 --> 00:40:33,810
and yield an image that is at exposure.

799
00:40:33,810 --> 00:40:36,930
But how does the image change?

800
00:40:36,930 --> 00:40:44,710
So to come back to this image
again, this is exposed at ISO 100.

801
00:40:44,710 --> 00:40:45,990
So a lot of noise?

802
00:40:45,990 --> 00:40:47,710
Not a lot of noise in that do you think?

803
00:40:47,710 --> 00:40:48,510
AUDIENCE: Low.

804
00:40:48,510 --> 00:40:50,310
IAN: So low noise probably.

805
00:40:50,310 --> 00:40:53,460
It's at f/5.6.

806
00:40:53,460 --> 00:40:57,630
So is that a lot of depth of
field or a little depth of field?

807
00:40:57,630 --> 00:40:58,520
Shallow or narrow?

808
00:40:58,520 --> 00:40:59,520
AUDIENCE: In the middle.

809
00:40:59,520 --> 00:41:02,190
IAN: It's in the
middle-ish we could say.

810
00:41:02,190 --> 00:41:05,880
And it's exposed at 1/100 of a second.

811
00:41:05,880 --> 00:41:10,740
So when we think back to the image of
the dam with all the sequential shutter

812
00:41:10,740 --> 00:41:15,870
speeds on that it's not super
fast right to freeze anything

813
00:41:15,870 --> 00:41:19,650
like falling water or anything
moving really fast, but fast enough

814
00:41:19,650 --> 00:41:21,240
to freeze most things.

815
00:41:21,240 --> 00:41:24,030
And it's also not slow enough to
allow significant motion blur.

816
00:41:24,030 --> 00:41:27,600


817
00:41:27,600 --> 00:41:31,980
So here's the same image, and
we've changed a couple of things.

818
00:41:31,980 --> 00:41:34,200
We're now shooting at ISO 400.

819
00:41:34,200 --> 00:41:37,872
So what have we done
compared to the last image?

820
00:41:37,872 --> 00:41:40,500
AUDIENCE: Doubled the ISO?

821
00:41:40,500 --> 00:41:44,150
IAN: We were at 100 before
so we went 100 to 200.

822
00:41:44,150 --> 00:41:45,180
200 to 400.

823
00:41:45,180 --> 00:41:50,220
So we increased the
sensitivity by two stops,

824
00:41:50,220 --> 00:41:54,150
but we also opened up the
aperture by several stops,

825
00:41:54,150 --> 00:41:58,860
and then shrank the shutter
speed by a fair amount.

826
00:41:58,860 --> 00:42:04,710
But what you'll see is that f/1.4 is
a really big opening and has a very

827
00:42:04,710 --> 00:42:06,670
shallow depth of field.

828
00:42:06,670 --> 00:42:08,180
And you can begin to see that.

829
00:42:08,180 --> 00:42:10,910
If we go backwards you
can see detail here.

830
00:42:10,910 --> 00:42:13,020
Crispness.

831
00:42:13,020 --> 00:42:17,590
We go forward and it's totally blurry.

832
00:42:17,590 --> 00:42:19,590
And we know that that's
probably not motion blur

833
00:42:19,590 --> 00:42:24,707
because we're shooting at 1/6400 of
a second, which is incredibly fast.

834
00:42:24,707 --> 00:42:26,790
It's fast enough to freeze
water and most anything

835
00:42:26,790 --> 00:42:28,540
that we would deal
with in our daily life.

836
00:42:28,540 --> 00:42:32,580


837
00:42:32,580 --> 00:42:36,690
So here's another version of this image.

838
00:42:36,690 --> 00:42:39,000
We're still shooting at a
very high shutter speed.

839
00:42:39,000 --> 00:42:41,010
We're shooting at a deeper--

840
00:42:41,010 --> 00:42:44,470
or a smaller aperture, which
gives us deeper depth of field,

841
00:42:44,470 --> 00:42:49,410
but we're shooting at this
really wildly high ISO, which

842
00:42:49,410 --> 00:42:52,043
introduces a lot of noise.

843
00:42:52,043 --> 00:42:54,960
And it's difficult to see, but we'll
zoom in on this in just a second.

844
00:42:54,960 --> 00:42:59,490


845
00:42:59,490 --> 00:43:02,060
So again, just to go the other way.

846
00:43:02,060 --> 00:43:03,870
So we have f/1.4.

847
00:43:03,870 --> 00:43:07,460
And this is 1/8000 of a second, which--
there's not a lot moving in here,

848
00:43:07,460 --> 00:43:09,750
but there's relatively no motion blur.

849
00:43:09,750 --> 00:43:10,870
DAN: There would be nothing
moving even if there were.

850
00:43:10,870 --> 00:43:11,700
IAN: Yeah.

851
00:43:11,700 --> 00:43:14,700
Exactly, but it wouldn't
be moving if it was.

852
00:43:14,700 --> 00:43:19,110
And a sort of relatively benign ISO 640.

853
00:43:19,110 --> 00:43:20,220
And again f/22.

854
00:43:20,220 --> 00:43:23,460
And we notice that the difference
between f/1.4 if you look at that

855
00:43:23,460 --> 00:43:25,870
gravestone in the very
far background there.

856
00:43:25,870 --> 00:43:28,980
This guy here.

857
00:43:28,980 --> 00:43:29,790
All of a sudden--

858
00:43:29,790 --> 00:43:31,090
oops, I went the wrong way--

859
00:43:31,090 --> 00:43:35,980
at f/22 it's sort of crisp,
whereas before it was out of focus.

860
00:43:35,980 --> 00:43:39,360
DAN: And, Ian, you said a
moment ago a benign ISO here,

861
00:43:39,360 --> 00:43:42,840
but I don't know that you defined
what is an acceptable ISO range?

862
00:43:42,840 --> 00:43:45,990
Like we talked about the trade
of a high ISO introducing noise,

863
00:43:45,990 --> 00:43:48,635
and a low ISO having less
noise, but like in your-- maybe

864
00:43:48,635 --> 00:43:50,010
this is an experiential question.

865
00:43:50,010 --> 00:43:50,820
In your experience--

866
00:43:50,820 --> 00:43:51,270
IAN: Yeah

867
00:43:51,270 --> 00:43:52,740
DAN: --like what would
your target range be

868
00:43:52,740 --> 00:43:53,850
if you're going to go
out and shoot something

869
00:43:53,850 --> 00:43:57,450
and wanted to keep as little noise as
possible while giving yourself a range,

870
00:43:57,450 --> 00:43:58,590
what would you operate in?

871
00:43:58,590 --> 00:44:03,360
IAN: So I tend to shoot
between 100 and 400.

872
00:44:03,360 --> 00:44:10,250
And I think that may be just habit
from shooting 35 millimeter film

873
00:44:10,250 --> 00:44:13,820
where I would buy it at 100
speed, or 200 speed, or 400 speed,

874
00:44:13,820 --> 00:44:17,880
but mostly 100 and 400.

875
00:44:17,880 --> 00:44:21,300
But I think on any given
shot I'm willing to push up

876
00:44:21,300 --> 00:44:26,090
to like 800, maybe 1,000, and once
I get past that it just starts to--

877
00:44:26,090 --> 00:44:29,280
I need to really want
to have the grain there

878
00:44:29,280 --> 00:44:32,430
because it gets hard to get rid of.

879
00:44:32,430 --> 00:44:36,970
So I would say experientially,
yeah, I shoot around 400.

880
00:44:36,970 --> 00:44:40,080
400 to 800 is sort of what I shoot
because it's sensitive enough

881
00:44:40,080 --> 00:44:44,640
that I can be in a reasonably dark
situation and capture what I want,

882
00:44:44,640 --> 00:44:47,130
but it's also just not
introducing that much noise

883
00:44:47,130 --> 00:44:50,210
that there's a real problem when I
go in and look at the images later.

884
00:44:50,210 --> 00:44:50,710
Yeah?

885
00:44:50,710 --> 00:44:52,943
AUDIENCE: Are you shooting
full frame or crop?

886
00:44:52,943 --> 00:44:53,610
IAN: Full frame.

887
00:44:53,610 --> 00:44:54,420
Well, it depends.

888
00:44:54,420 --> 00:44:56,160
It depends which camera I'm using.

889
00:44:56,160 --> 00:45:00,990
So I'll shoot full frame, which
means a 35 millimeter size sensor,

890
00:45:00,990 --> 00:45:04,200
but I own cameras that have
smaller sensors than that.

891
00:45:04,200 --> 00:45:11,070
And so the-- yeah, and so
and so it really depends

892
00:45:11,070 --> 00:45:12,570
on what piece of hardware I'm using.

893
00:45:12,570 --> 00:45:18,210
Again, all of these values and
things are dependent on the hardware

894
00:45:18,210 --> 00:45:19,010
that you have.

895
00:45:19,010 --> 00:45:21,848
DAN: Yeah, but I think Ralph's
question is actually interesting

896
00:45:21,848 --> 00:45:22,890
asking which size sensor.

897
00:45:22,890 --> 00:45:28,650
Is there a performance difference with
a bigger sensor versus a smaller sensor?

898
00:45:28,650 --> 00:45:32,310
IAN: Yeah, there is because you
can have larger photo sites.

899
00:45:32,310 --> 00:45:37,680
So because you have a larger sensor the
photosites that are sensitive to light

900
00:45:37,680 --> 00:45:41,700
can be larger which means
that they are effectively

901
00:45:41,700 --> 00:45:46,560
better at higher sensitivities than
sensors that have smaller photosites.

902
00:45:46,560 --> 00:45:50,840
So a full frame or a
larger sensor is going

903
00:45:50,840 --> 00:45:58,320
to have better quality in
lower light than something

904
00:45:58,320 --> 00:45:59,850
that has a smaller photo sensor.

905
00:45:59,850 --> 00:46:00,862
DAN: With a higher ISO.

906
00:46:00,862 --> 00:46:02,070
IAN: Yeah, with a higher ISO.

907
00:46:02,070 --> 00:46:04,710
I think, all things
being equal, if you say

908
00:46:04,710 --> 00:46:08,220
have like a small micro 4/3
mirrorless camera at ISO 400

909
00:46:08,220 --> 00:46:10,860
they're probably indistinguishable.

910
00:46:10,860 --> 00:46:15,400
Maybe if you really, really, really
dig into the image you can find it.

911
00:46:15,400 --> 00:46:19,170
But if you're shooting it say 3,200,
or 6,400, or something like that,

912
00:46:19,170 --> 00:46:21,487
having a larger photosite,
which means a bigger sensor,

913
00:46:21,487 --> 00:46:23,070
is going to be more beneficial to you.

914
00:46:23,070 --> 00:46:27,010
That's where you're going to find that
little bit of edge that it gives you.

915
00:46:27,010 --> 00:46:27,510
For sure.

916
00:46:27,510 --> 00:46:29,623
It's a good question.

917
00:46:29,623 --> 00:46:31,290
DAN: And then I just want to highlight--

918
00:46:31,290 --> 00:46:35,140
Alec said that there's
vignetting on the lens.

919
00:46:35,140 --> 00:46:37,648
It's very noticeable between 1.4 f/11.

920
00:46:37,648 --> 00:46:38,940
IAN: Yeah, there definitely is.

921
00:46:38,940 --> 00:46:39,732
AUDIENCE: There is.

922
00:46:39,732 --> 00:46:43,580
IAN: Yeah, like you can see it here.

923
00:46:43,580 --> 00:46:46,620
At 22 all the corners
are bright, and at 1.4

924
00:46:46,620 --> 00:46:49,825
for there's this serious vignetting.

925
00:46:49,825 --> 00:46:52,200
And it almost feels like it's
softening too a little bit.

926
00:46:52,200 --> 00:46:55,798


927
00:46:55,798 --> 00:46:57,840
And what that is is that
the coverage of the lens

928
00:46:57,840 --> 00:47:03,120
is sort of just getting a little
bit too small for the sensor size

929
00:47:03,120 --> 00:47:10,200
and it's just-- not quite enough light
is reaching the sensor at that time.

930
00:47:10,200 --> 00:47:12,900
So I'm going the wrong way.

931
00:47:12,900 --> 00:47:14,530
That's why they keep going.

932
00:47:14,530 --> 00:47:18,150
So if we jump in and look at both.

933
00:47:18,150 --> 00:47:22,000
This is the first image, which was
sort of medium values of everything.

934
00:47:22,000 --> 00:47:24,840
It's very vanilla.

935
00:47:24,840 --> 00:47:29,460
This is 1.4 and you can
see my focus was just off.

936
00:47:29,460 --> 00:47:30,960
AUDIENCE: It's hard to focus at 1.4.

937
00:47:30,960 --> 00:47:31,877
IAN: It's really hard.

938
00:47:31,877 --> 00:47:34,050
And I sort of left this
in here as an illustration

939
00:47:34,050 --> 00:47:39,750
that you might think you have
something in focus when you're

940
00:47:39,750 --> 00:47:44,460
looking at it with a
smaller aperture, but then

941
00:47:44,460 --> 00:47:47,820
if you open up it may not quite be--

942
00:47:47,820 --> 00:47:53,920
you're plane of focus might not be
exactly where you expected it to be.

943
00:47:53,920 --> 00:47:55,020
And this is very--

944
00:47:55,020 --> 00:47:56,070
like this happens a lot.

945
00:47:56,070 --> 00:47:57,900
Like it's not far off really.

946
00:47:57,900 --> 00:48:01,470
It's on this bush here somewhere,
but it's just too far forward.

947
00:48:01,470 --> 00:48:03,280
Like I made a mistake.

948
00:48:03,280 --> 00:48:06,390
DAN: And it's really unfortunate when
you're shooting with a person and--

949
00:48:06,390 --> 00:48:06,810
IAN: Yeah.

950
00:48:06,810 --> 00:48:08,820
DAN: The thing you want when you're
shooting a person is for their eyes

951
00:48:08,820 --> 00:48:12,028
to be in focus because that's like the
first thing your eyes typically go to.

952
00:48:12,028 --> 00:48:15,220
And when you perceive an image in
focus the eyes are typically in focus.

953
00:48:15,220 --> 00:48:19,740
So if you notice that the focus is
like just on the end of someone's nose,

954
00:48:19,740 --> 00:48:23,033
but their eyes are not in focus it's
definitely a moment to kind of-- it's

955
00:48:23,033 --> 00:48:25,450
good to double check if you're
taking a picture of people,

956
00:48:25,450 --> 00:48:27,690
I guess, to punch in
digitally on your screen

957
00:48:27,690 --> 00:48:31,120
and make sure that the
image is actually sharp.

958
00:48:31,120 --> 00:48:31,620
So

959
00:48:31,620 --> 00:48:33,420
IAN: This is that image
with a really high ISO.

960
00:48:33,420 --> 00:48:35,462
And I think it was a little
bit difficult to see,

961
00:48:35,462 --> 00:48:37,590
but in when we zoom
in now you really can

962
00:48:37,590 --> 00:48:42,807
see the noise just sort of
introduced and all of the texture.

963
00:48:42,807 --> 00:48:45,390
And you'll almost see-- like
it's almost a little bit brighter

964
00:48:45,390 --> 00:48:51,757
too because there's so much
added bright data in the shadows.

965
00:48:51,757 --> 00:48:54,090
Lift them up a little bit
because there's so much of it.

966
00:48:54,090 --> 00:48:56,760


967
00:48:56,760 --> 00:48:59,540
So, again, 1.4 again was my bad focus.

968
00:48:59,540 --> 00:49:00,170
So frustrating.

969
00:49:00,170 --> 00:49:01,730
AUDIENCE: It happens to all of us.

970
00:49:01,730 --> 00:49:02,360
IAN: Right.

971
00:49:02,360 --> 00:49:04,670
But then if we go to f/22--

972
00:49:04,670 --> 00:49:08,040
and this is what tricked me because I
think I actually shot this one first

973
00:49:08,040 --> 00:49:09,457
and I was like, ah, it's in focus.

974
00:49:09,457 --> 00:49:10,220
Looks great.

975
00:49:10,220 --> 00:49:12,435
You can see that this
background is in focus.

976
00:49:12,435 --> 00:49:14,060
And this foreground object is in focus.

977
00:49:14,060 --> 00:49:18,950
So it has a large depth of field.

978
00:49:18,950 --> 00:49:21,590
So this should visually
illustrate that we

979
00:49:21,590 --> 00:49:27,020
have the ability to have different
exposure settings for the same exposure

980
00:49:27,020 --> 00:49:28,910
value.

981
00:49:28,910 --> 00:49:32,930
So if we think about a
scene in this way there

982
00:49:32,930 --> 00:49:37,790
is some sort of amount of available
light that we want to sort of render.

983
00:49:37,790 --> 00:49:44,190
And we have three controls
at our disposal to do that.

984
00:49:44,190 --> 00:49:49,010
And by increasing or decreasing
one we can increase or decrease

985
00:49:49,010 --> 00:49:52,820
the overall sensitivity of the
camera to some amount of light.

986
00:49:52,820 --> 00:49:56,720
Or we can increase one and
decrease another to not change

987
00:49:56,720 --> 00:49:59,960
the sensitivity of the camera, but
the change the way that image looks.

988
00:49:59,960 --> 00:50:02,960


989
00:50:02,960 --> 00:50:13,280
So if we think about this-- if we assume
that are base exposure that we metered

990
00:50:13,280 --> 00:50:23,000
returned some values like f/5.6, 400
ISO at 1/60 a second, which is, again,

991
00:50:23,000 --> 00:50:24,650
not that grainy.

992
00:50:24,650 --> 00:50:25,820
It's not that fast.

993
00:50:25,820 --> 00:50:27,570
Not that slow.

994
00:50:27,570 --> 00:50:29,730
There's not a lot of
room for motion blur.

995
00:50:29,730 --> 00:50:30,980
Someone walking would be fine.

996
00:50:30,980 --> 00:50:32,390
A car might be out of--

997
00:50:32,390 --> 00:50:35,210
a moving car might have
a little blur to it.

998
00:50:35,210 --> 00:50:36,920
Someone standing still would be fine.

999
00:50:36,920 --> 00:50:39,720
And the depth of field
is in the middle range.

1000
00:50:39,720 --> 00:50:42,470
We haven't figured out quite what
that means, how much that means,

1001
00:50:42,470 --> 00:50:47,070
but we know it's in the
middle compared to 1.4 or 22.

1002
00:50:47,070 --> 00:50:48,500
So let's make some decisions.

1003
00:50:48,500 --> 00:50:52,280
Let's change the way this image looks.

1004
00:50:52,280 --> 00:50:56,197
And what I've done is I've
filled in two of the three blanks

1005
00:50:56,197 --> 00:50:57,155
with different numbers.

1006
00:50:57,155 --> 00:51:00,030


1007
00:51:00,030 --> 00:51:07,490
So we should at this point be able
to calculate what the empty is.

1008
00:51:07,490 --> 00:51:10,490
So we'll just walk through
the very first one.

1009
00:51:10,490 --> 00:51:15,930
So before we were at 5.6 and then
we closed down one stop to f/8

1010
00:51:15,930 --> 00:51:20,810
so we allow less light in by one stop.

1011
00:51:20,810 --> 00:51:21,990
Here we're at 400.

1012
00:51:21,990 --> 00:51:23,870
We're still at 400.

1013
00:51:23,870 --> 00:51:27,650
So we've allowed one stop less light in.

1014
00:51:27,650 --> 00:51:29,600
So now we need to compensate for that.

1015
00:51:29,600 --> 00:51:32,180
And we're going to compensate
for that in this empty square.

1016
00:51:32,180 --> 00:51:36,140
So we need to allow one stop more
light in to make them equivalent.

1017
00:51:36,140 --> 00:51:37,850
Does it make sense?

1018
00:51:37,850 --> 00:51:41,860
So what would the value
be in that square?

1019
00:51:41,860 --> 00:51:52,515


1020
00:51:52,515 --> 00:51:53,015
Yeah?

1021
00:51:53,015 --> 00:51:54,505
AUDIENCE: [INAUDIBLE]

1022
00:51:54,505 --> 00:51:55,630
DAN: I've got the internet.

1023
00:51:55,630 --> 00:51:55,950
You're right here.

1024
00:51:55,950 --> 00:51:56,310
Go ahead.

1025
00:51:56,310 --> 00:51:57,170
AUDIENCE: Yeah, but [INAUDIBLE].

1026
00:51:57,170 --> 00:51:58,564
I haven't heard them all day.

1027
00:51:58,564 --> 00:52:02,160
DAN: Miles says should we
go up to 1/30 of a second?

1028
00:52:02,160 --> 00:52:03,060
IAN: We should.

1029
00:52:03,060 --> 00:52:03,900
Well done.

1030
00:52:03,900 --> 00:52:04,950
That's awesome.

1031
00:52:04,950 --> 00:52:08,760
So we reduce the amount of light by
one stop by closing the aperture down.

1032
00:52:08,760 --> 00:52:11,880
So we had to increase the amount
of time that the shutter speed

1033
00:52:11,880 --> 00:52:16,560
was open by one stop.

1034
00:52:16,560 --> 00:52:21,322
This isn't the only decision we could
have made, but it is the decision here.

1035
00:52:21,322 --> 00:52:23,630
AUDIENCE: For the
proper exposure, right?

1036
00:52:23,630 --> 00:52:28,140
IAN: Yeah, to maintain our exposure.

1037
00:52:28,140 --> 00:52:31,260
The same exposure that we had given
by these yellow numbers up here.

1038
00:52:31,260 --> 00:52:36,480


1039
00:52:36,480 --> 00:52:40,240
So now we have a
different set of numbers.

1040
00:52:40,240 --> 00:52:42,870
And again, we're still going to
reference the yellow numbers.

1041
00:52:42,870 --> 00:52:45,270
Don't worry about the bottom one.

1042
00:52:45,270 --> 00:52:50,820
So we open up the
camera from 5.6 to 1.4.

1043
00:52:50,820 --> 00:52:52,340
DAN: Can you go back to my drawing?

1044
00:52:52,340 --> 00:52:53,170
IAN: Yeah.

1045
00:52:53,170 --> 00:52:55,650
DAN: To go this far I
still have to visualize it.

1046
00:52:55,650 --> 00:52:57,920
So we're going from 1.4--

1047
00:52:57,920 --> 00:52:59,200
IAN: Well, we started at 5.6.

1048
00:52:59,200 --> 00:53:00,660
DAN: Right, sorry.

1049
00:53:00,660 --> 00:53:03,050
IAN: Was the original
exposure value, and we're

1050
00:53:03,050 --> 00:53:05,670
going to open up the aperture to 1.4.

1051
00:53:05,670 --> 00:53:11,520
DAN: 1, 2, 3.

1052
00:53:11,520 --> 00:53:12,020
IAN: 4.

1053
00:53:12,020 --> 00:53:13,437
DAN: I won't say the final number.

1054
00:53:13,437 --> 00:53:14,700
Somebody else can say it.

1055
00:53:14,700 --> 00:53:20,190
IAN: So we've added four stops more
light by adjusting the aperture value.

1056
00:53:20,190 --> 00:53:24,000
We still haven't adjusted the
ISO value so that's the same.

1057
00:53:24,000 --> 00:53:27,610
So we now just have to compensate
for four stops more light.

1058
00:53:27,610 --> 00:53:30,366


1059
00:53:30,366 --> 00:53:31,930
Got a few answers from online.

1060
00:53:31,930 --> 00:53:34,230
AUDIENCE: Do you need
more light or less light?

1061
00:53:34,230 --> 00:53:35,670
IAN: Well, so you tell me.

1062
00:53:35,670 --> 00:53:39,838
We started at 5.6 and
we're going to 1.4.

1063
00:53:39,838 --> 00:53:40,776
AUDIENCE: OK.

1064
00:53:40,776 --> 00:53:41,714
We need to add more.

1065
00:53:41,714 --> 00:53:43,090
OK, yeah, I'm good.

1066
00:53:43,090 --> 00:53:43,590
I'm good.

1067
00:53:43,590 --> 00:53:44,160
IAN: Right.

1068
00:53:44,160 --> 00:53:46,080
So the smaller numbers
are larger openings,

1069
00:53:46,080 --> 00:53:52,680
which is sort of counter intuitive,
but we've gone five stops more open.

1070
00:53:52,680 --> 00:53:54,963
So where would the shutter speed go to?

1071
00:53:54,963 --> 00:53:56,130
You said you have an answer?

1072
00:53:56,130 --> 00:53:57,420
DAN: I have a few answers from online.

1073
00:53:57,420 --> 00:53:57,920
IAN: OK.

1074
00:53:57,920 --> 00:53:59,190
DAN: So I have 1/240.

1075
00:53:59,190 --> 00:54:01,410
I've got 1/1000.

1076
00:54:01,410 --> 00:54:03,660
1/960.

1077
00:54:03,660 --> 00:54:04,202
IAN: OK.

1078
00:54:04,202 --> 00:54:06,660
So we'll start with the lowest
and we'll go all the way up.

1079
00:54:06,660 --> 00:54:07,860
So 1/240.

1080
00:54:07,860 --> 00:54:14,130
So we were at 1/60 and we want to
reduce the amount by four stops, right?

1081
00:54:14,130 --> 00:54:15,510
Is that what we decided?

1082
00:54:15,510 --> 00:54:19,160
So we go 160 to 120.

1083
00:54:19,160 --> 00:54:22,020
1/120 to 1/250, which
doesn't match perfectly,

1084
00:54:22,020 --> 00:54:23,740
but that's sort of the way it is.

1085
00:54:23,740 --> 00:54:29,850
1/250 to 1/500, and 1/500 500 to 1/1000.

1086
00:54:29,850 --> 00:54:32,370
That's our four stops.

1087
00:54:32,370 --> 00:54:36,197
So we end up at 1/1000 of a second.

1088
00:54:36,197 --> 00:54:39,030
DAN: And this is a good moment to
highlight the rounding that we do.

1089
00:54:39,030 --> 00:54:40,363
The kind of fudging of the math.

1090
00:54:40,363 --> 00:54:43,980
It's the same thing from 5.6
to f/11 when we make that jump.

1091
00:54:43,980 --> 00:54:46,980
It's not quite, but it's
easier to talk in whole numbers

1092
00:54:46,980 --> 00:54:49,800
than it is to remember
1/960 of a second.

1093
00:54:49,800 --> 00:54:50,740
IAN: Right, exactly.

1094
00:54:50,740 --> 00:54:53,700
And so you will find as you
investigate some of these concepts

1095
00:54:53,700 --> 00:54:56,850
that some of the numbers
are scales on the cameras.

1096
00:54:56,850 --> 00:55:00,478
Maybe you do have that fudge
factor when you go from 5.6 to 11,

1097
00:55:00,478 --> 00:55:01,770
which doesn't quite make sense.

1098
00:55:01,770 --> 00:55:05,830
Which should be 11.2 I think.

1099
00:55:05,830 --> 00:55:08,340
All right, so let's do this one here.

1100
00:55:08,340 --> 00:55:11,250


1101
00:55:11,250 --> 00:55:14,340
So we're at 5.6, and we open up to 2.8.

1102
00:55:14,340 --> 00:55:15,984
How many stops?

1103
00:55:15,984 --> 00:55:17,460
AUDIENCE: Two stops?

1104
00:55:17,460 --> 00:55:19,460
IAN: 5.6, four, 2.8.

1105
00:55:19,460 --> 00:55:21,090
Yeah, two stops.

1106
00:55:21,090 --> 00:55:23,610
1/60 to 1/30.

1107
00:55:23,610 --> 00:55:25,710
Opening or closing?

1108
00:55:25,710 --> 00:55:28,293
More sensitive or less sensitive?

1109
00:55:28,293 --> 00:55:29,770
AUDIENCE: More.

1110
00:55:29,770 --> 00:55:31,170
IAN: More sensitive.

1111
00:55:31,170 --> 00:55:32,980
By how many stops?

1112
00:55:32,980 --> 00:55:33,610
AUDIENCE: One.

1113
00:55:33,610 --> 00:55:33,740
IAN: One.

1114
00:55:33,740 --> 00:55:34,532
So we add one more.

1115
00:55:34,532 --> 00:55:36,228
So that's three stops difference.

1116
00:55:36,228 --> 00:55:39,690


1117
00:55:39,690 --> 00:55:45,480
So now if we were at ISO 400 and
we've added three stops of light

1118
00:55:45,480 --> 00:55:47,551
how much less sensitive
do we need to make that?

1119
00:55:47,551 --> 00:55:52,635


1120
00:55:52,635 --> 00:55:53,760
There's so much arithmetic.

1121
00:55:53,760 --> 00:55:55,670
It's annoying, I know.

1122
00:55:55,670 --> 00:55:58,606
But I'll make you really good at it.

1123
00:55:58,606 --> 00:55:59,565
AUDIENCE: Three stops.

1124
00:55:59,565 --> 00:56:01,440
IAN: Yeah, so we need
to go down three stops.

1125
00:56:01,440 --> 00:56:03,423
So what is the numerical value for that?

1126
00:56:03,423 --> 00:56:06,810
AUDIENCE: [INAUDIBLE]

1127
00:56:06,810 --> 00:56:07,830
IAN: OK, perfect.

1128
00:56:07,830 --> 00:56:09,870
That's exactly right.

1129
00:56:09,870 --> 00:56:10,492
Great.

1130
00:56:10,492 --> 00:56:11,640
AUDIENCE: [INAUDIBLE]

1131
00:56:11,640 --> 00:56:13,050
IAN: Great.

1132
00:56:13,050 --> 00:56:16,827
All right, so our last
one we've doubled the ISO.

1133
00:56:16,827 --> 00:56:18,660
We've done something
with the shutter speed.

1134
00:56:18,660 --> 00:56:22,620
How many stops differences is
the shutter speed from 1/60?

1135
00:56:22,620 --> 00:56:26,710


1136
00:56:26,710 --> 00:56:28,990
Oh, I still have to do it my head.

1137
00:56:28,990 --> 00:56:29,754
1/30.

1138
00:56:29,754 --> 00:56:31,950
1/15.

1139
00:56:31,950 --> 00:56:37,830
So you go from 1/60 to 1/30,
1/30 to 1/15, 1/15 to 1/8.

1140
00:56:37,830 --> 00:56:42,450
That doesn't make any sense,
but that's what the number is.

1141
00:56:42,450 --> 00:56:46,890
So it's three stops more light.

1142
00:56:46,890 --> 00:56:49,650
And we've also doubled the ISO.

1143
00:56:49,650 --> 00:56:51,930
So that's one stop more sensitivity.

1144
00:56:51,930 --> 00:56:54,480
So that's a total of four stops.

1145
00:56:54,480 --> 00:56:57,363
So if we take the aperture we need
to close down four stops from 5.6

1146
00:56:57,363 --> 00:56:58,530
so we just go the other way.

1147
00:56:58,530 --> 00:57:04,230
We go 5.6 to eight, eight
to 11, 11 to 16, 16 to 22--

1148
00:57:04,230 --> 00:57:04,730
f/22.

1149
00:57:04,730 --> 00:57:09,420


1150
00:57:09,420 --> 00:57:11,460
And finally this last one.

1151
00:57:11,460 --> 00:57:13,250
I don't even understand this one.

1152
00:57:13,250 --> 00:57:14,790
We'll go 400 to 320.

1153
00:57:14,790 --> 00:57:17,670
That is not halving nor doubling.

1154
00:57:17,670 --> 00:57:20,017
What do we do?

1155
00:57:20,017 --> 00:57:21,350
AUDIENCE: It's a third isn't it?

1156
00:57:21,350 --> 00:57:22,410
IAN: Yeah.

1157
00:57:22,410 --> 00:57:23,490
Let's do the rough math.

1158
00:57:23,490 --> 00:57:24,310
It's about a third.

1159
00:57:24,310 --> 00:57:25,560
So I've got a third of a stop.

1160
00:57:25,560 --> 00:57:26,670
I don't know how to do
that with my fingers

1161
00:57:26,670 --> 00:57:28,020
so I'll just keep it in my head.

1162
00:57:28,020 --> 00:57:31,270
Now we go from 1/60 to 1/100.

1163
00:57:31,270 --> 00:57:33,360
Is that a full stop?

1164
00:57:33,360 --> 00:57:37,110
Full stop would be 1/120.

1165
00:57:37,110 --> 00:57:38,355
So it's 2/3.

1166
00:57:38,355 --> 00:57:39,480
So now we have 2/3 and 1/3.

1167
00:57:39,480 --> 00:57:40,657
Ah, one stop.

1168
00:57:40,657 --> 00:57:42,240
AUDIENCE: [INAUDIBLE] can handle that.

1169
00:57:42,240 --> 00:57:43,350
IAN: Perfect.

1170
00:57:43,350 --> 00:57:46,372
So we just have to open up to f/4.

1171
00:57:46,372 --> 00:57:47,580
So it can get a little funky.

1172
00:57:47,580 --> 00:57:51,360
And it can get-- like you can start to
sort of move and shake things around,

1173
00:57:51,360 --> 00:57:55,260
but you literally can
make different images.

1174
00:57:55,260 --> 00:57:58,800
So what does this image look?

1175
00:57:58,800 --> 00:57:59,574
This middle one?

1176
00:57:59,574 --> 00:58:02,460
1.4, 400, 1/1000 of a second?

1177
00:58:02,460 --> 00:58:03,960
AUDIENCE: There would be a vignette.

1178
00:58:03,960 --> 00:58:05,340
IAN: Yeah.

1179
00:58:05,340 --> 00:58:08,790
So we noticed that when it was at
1.4 with this lens on this camera

1180
00:58:08,790 --> 00:58:12,072
there was seriously vignetting so
if we use this same system again

1181
00:58:12,072 --> 00:58:14,280
we'll get serious vignetting,
but what's the artifact

1182
00:58:14,280 --> 00:58:15,270
that we'll really care about?

1183
00:58:15,270 --> 00:58:17,228
AUDIENCE: [INAUDIBLE] a
shallow depth of field.

1184
00:58:17,228 --> 00:58:19,380
IAN: It'll have very
shallow depth of field.

1185
00:58:19,380 --> 00:58:20,700
What about things that
are moving in that?

1186
00:58:20,700 --> 00:58:21,867
Say there's some cars in it.

1187
00:58:21,867 --> 00:58:25,024


1188
00:58:25,024 --> 00:58:26,440
AUDIENCE: They'll be frozen.

1189
00:58:26,440 --> 00:58:28,023
IAN: Yeah, they're going to be frozen.

1190
00:58:28,023 --> 00:58:28,880
1/1000 of a second.

1191
00:58:28,880 --> 00:58:31,430
That's pretty fast.

1192
00:58:31,430 --> 00:58:33,130
That's 1,000 pictures in one second.

1193
00:58:33,130 --> 00:58:36,790
That's actually pretty ludicrously
fast if I think about it.

1194
00:58:36,790 --> 00:58:41,380
So very shallow depth of
field, but no motion blur.

1195
00:58:41,380 --> 00:58:43,370
So blur because of depth of field--

1196
00:58:43,370 --> 00:58:48,790
or lack of focus because of the depth
a field, but what about this one?

1197
00:58:48,790 --> 00:58:53,860
F/22 at 800 at 1/8 of a second?

1198
00:58:53,860 --> 00:59:03,090


1199
00:59:03,090 --> 00:59:03,590
Yeah?

1200
00:59:03,590 --> 00:59:06,790
AUDIENCE: You'll have wide
depth-- very wide depth of field.

1201
00:59:06,790 --> 00:59:07,308
IAN: Mm-hm.

1202
00:59:07,308 --> 00:59:09,350
AUDIENCE: And potentially,
depending on the lens,

1203
00:59:09,350 --> 00:59:11,686
it's gonna have like a
bit of shake or blur.

1204
00:59:11,686 --> 00:59:12,667
IAN: Mm-hm.

1205
00:59:12,667 --> 00:59:15,710
AUDIENCE: At 800 I'm
assuming not too bad.

1206
00:59:15,710 --> 00:59:16,700
Not too much grain.

1207
00:59:16,700 --> 00:59:18,920
IAN: Yeah, but more than 400 for sure.

1208
00:59:18,920 --> 00:59:21,080
So large depth of field
and then some motion blur.

1209
00:59:21,080 --> 00:59:22,220
1/8 of a second.

1210
00:59:22,220 --> 00:59:24,230
I can do a lot in 1/8 of a second.

1211
00:59:24,230 --> 00:59:26,080
Like I moved.

1212
00:59:26,080 --> 00:59:28,640
A lot of dancing in 1/8 of a second.

1213
00:59:28,640 --> 00:59:30,800
So that is to say that
we can actually speculate

1214
00:59:30,800 --> 00:59:34,213
at what these images
look like abstractly just

1215
00:59:34,213 --> 00:59:35,630
by looking at the camera settings.

1216
00:59:35,630 --> 00:59:39,020
We have an understanding of how
to previsualize what a scene will

1217
00:59:39,020 --> 00:59:40,852
look like at given settings.

1218
00:59:40,852 --> 00:59:44,060
So when you're out photographing you
can look at a scene and you can be like,

1219
00:59:44,060 --> 00:59:46,453
oh, so there's cars moving,
there's some water flowing,

1220
00:59:46,453 --> 00:59:49,370
I know if I set a low shutter speed
I'm going to get interesting blur.

1221
00:59:49,370 --> 00:59:51,210
Cool, I want to try that.

1222
00:59:51,210 --> 00:59:54,532
And you know that you can
then decrease the shutter

1223
00:59:54,532 --> 00:59:56,240
speed so that you end
up with more light,

1224
00:59:56,240 --> 00:59:58,180
and then you maybe have to stop down.

1225
00:59:58,180 --> 00:59:59,510
It's going to increase
your depth of field,

1226
00:59:59,510 --> 01:00:00,927
which may be an interesting image.

1227
01:00:00,927 --> 01:00:03,620


1228
01:00:03,620 --> 01:00:06,550
Now I'm going backwards again.

1229
01:00:06,550 --> 01:00:14,090
So briefly, what do you think the
camera settings were for this image?

1230
01:00:14,090 --> 01:00:17,520


1231
01:00:17,520 --> 01:00:21,740
And we're going to talk abstractly
without sort of like digging

1232
01:00:21,740 --> 01:00:22,580
into the image file.

1233
01:00:22,580 --> 01:00:24,788
We don't know for sure, but
we can take some guesses.

1234
01:00:24,788 --> 01:00:27,702


1235
01:00:27,702 --> 01:00:29,670
AUDIENCE: F/11.

1236
01:00:29,670 --> 01:00:32,820
IAN: OK, so you're saying f/11,
but why are you saying that?

1237
01:00:32,820 --> 01:00:34,870
AUDIENCE: Well, a lot's in focus.

1238
01:00:34,870 --> 01:00:36,370
IAN: So there's a large depth field.

1239
01:00:36,370 --> 01:00:38,120
That's our first
indicator, so we're going

1240
01:00:38,120 --> 01:00:40,050
to say that it's probably
a small aperture.

1241
01:00:40,050 --> 01:00:41,340
F/11, maybe higher.

1242
01:00:41,340 --> 01:00:42,180
Who knows?

1243
01:00:42,180 --> 01:00:46,402
But like a small aperture for sure.

1244
01:00:46,402 --> 01:00:47,610
What about the shutter speed?

1245
01:00:47,610 --> 01:00:50,574


1246
01:00:50,574 --> 01:00:52,550
AUDIENCE: Relatively slow.

1247
01:00:52,550 --> 01:00:53,480
IAN: Relatively slow.

1248
01:00:53,480 --> 01:00:55,106
Why do you say that?

1249
01:00:55,106 --> 01:00:58,850
AUDIENCE: [INAUDIBLE] shutter speed,
there would be a shadow on that,

1250
01:00:58,850 --> 01:01:01,190
I guess.

1251
01:01:01,190 --> 01:01:04,130
IAN: Well, we don't we don't know
how much light was at the scene.

1252
01:01:04,130 --> 01:01:08,270
We don't know, but what we can
do is look at the artifacts.

1253
01:01:08,270 --> 01:01:11,990
And the artifact of shutter
speed is is there motion blur?

1254
01:01:11,990 --> 01:01:12,620
Is there not?

1255
01:01:12,620 --> 01:01:15,120
Is there motion blur on certain
things and not other things?

1256
01:01:15,120 --> 01:01:18,710
Because that gives us an
indication of how fast it is.

1257
01:01:18,710 --> 01:01:21,910
AUDIENCE: It doesn't look like there's
much motion [INAUDIBLE] at all.

1258
01:01:21,910 --> 01:01:24,740
IAN: Yeah, so I think--
you know, where there's--

1259
01:01:24,740 --> 01:01:28,778
the water is rippling in some wind
and that seems relatively crisp.

1260
01:01:28,778 --> 01:01:31,820
Like maybe there's a little bit of
motion blur there, but it's not a lot.

1261
01:01:31,820 --> 01:01:33,653
It's not sort of
distracting, and it doesn't

1262
01:01:33,653 --> 01:01:39,580
look like the water in the dam picture,
which was super smooth and flowing.

1263
01:01:39,580 --> 01:01:41,830
AUDIENCE: Carla says 600 ISO.

1264
01:01:41,830 --> 01:01:42,800
IAN: 600 ISO.

1265
01:01:42,800 --> 01:01:43,612
AUDIENCE: What?

1266
01:01:43,612 --> 01:01:45,410
IAN: Yeah, why?

1267
01:01:45,410 --> 01:01:46,090
So why?

1268
01:01:46,090 --> 01:01:48,350
AUDIENCE: [INAUDIBLE] be more.

1269
01:01:48,350 --> 01:01:49,850
IAN: What is the reasoning for that?

1270
01:01:49,850 --> 01:01:54,056
The logic behind that assessment?

1271
01:01:54,056 --> 01:01:55,544
AUDIENCE: Aperture's closed down.

1272
01:01:55,544 --> 01:02:00,020


1273
01:02:00,020 --> 01:02:04,430
IAN: I don't see a lot of grain so if
that is the idea-- like, yeah, there's

1274
01:02:04,430 --> 01:02:06,440
not a lot of noise.

1275
01:02:06,440 --> 01:02:10,280
And so I think we could say
that this scene probably

1276
01:02:10,280 --> 01:02:12,060
has a fair amount of light in it.

1277
01:02:12,060 --> 01:02:16,760
It's a sunlit vista.

1278
01:02:16,760 --> 01:02:22,190
And there's a reasonably high shutter
speed, a large depth of field,

1279
01:02:22,190 --> 01:02:24,077
and some relatively low ISO.

1280
01:02:24,077 --> 01:02:24,702
AUDIENCE: Yeah.

1281
01:02:24,702 --> 01:02:26,905
IAN: Right?

1282
01:02:26,905 --> 01:02:28,130
And maybe-- yes?

1283
01:02:28,130 --> 01:02:33,740
AUDIENCE: Isn't another artifact of
a low ISO good color reproduction?

1284
01:02:33,740 --> 01:02:38,700
'Cause like when I saw the picture at
the beginning with-- what's his name?

1285
01:02:38,700 --> 01:02:40,408
I don't know if his
name was Dan as well.

1286
01:02:40,408 --> 01:02:41,616
Was his name Dan [INAUDIBLE]?

1287
01:02:41,616 --> 01:02:42,150
IAN: Mm-hm.

1288
01:02:42,150 --> 01:02:45,120
AUDIENCE: So like at 400
the colors looked OK,

1289
01:02:45,120 --> 01:02:49,620
but when you went to 1,200,
800, it was blotchy and colors

1290
01:02:49,620 --> 01:02:51,740
that weren't there started to exist.

1291
01:02:51,740 --> 01:02:54,050
IAN: Yeah, so in that sense
because noise is random

1292
01:02:54,050 --> 01:02:55,758
and it's not just
brightness values, it's

1293
01:02:55,758 --> 01:02:58,690
also color values, that you
can get random red introduced

1294
01:02:58,690 --> 01:03:03,688
at a pixel or a bunch of
other pixels and so the color

1295
01:03:03,688 --> 01:03:06,230
loses fidelity with the introduction
of noise in the same way

1296
01:03:06,230 --> 01:03:08,545
that your exposure loses fidelity.

1297
01:03:08,545 --> 01:03:11,420
And you notice in the graveyard
image where it seemed almost brighter

1298
01:03:11,420 --> 01:03:13,712
because there was so much
noise in the shadow it easily

1299
01:03:13,712 --> 01:03:16,670
could have been like seemed more
colorful or the colors seemed

1300
01:03:16,670 --> 01:03:22,640
off and messed up because of like
random color data forming as noise.

1301
01:03:22,640 --> 01:03:24,125
So, yes, in that sense absolutely.

1302
01:03:24,125 --> 01:03:26,750
DAN: Yeah, and with the stretching
of the information basically

1303
01:03:26,750 --> 01:03:30,530
you're sensor's just like stretching the
information that it is able to collect.

1304
01:03:30,530 --> 01:03:31,670
As we saw there was a lot--

1305
01:03:31,670 --> 01:03:34,090
when the noise comes up in
the shadows that can also

1306
01:03:34,090 --> 01:03:35,840
lead to the perception
of lack of contrast

1307
01:03:35,840 --> 01:03:39,890
as well because there's just like
more noise across the darks that

1308
01:03:39,890 --> 01:03:41,930
seem to raise them
more than they actually

1309
01:03:41,930 --> 01:03:45,440
are so there's better reproduction at
a lower ISO because of that as well.

1310
01:03:45,440 --> 01:03:48,257


1311
01:03:48,257 --> 01:03:49,840
IAN: So what's going on in this image?

1312
01:03:49,840 --> 01:03:53,312


1313
01:03:53,312 --> 01:03:53,960
Hm?

1314
01:03:53,960 --> 01:03:56,320
AUDIENCE: Looks like a California.

1315
01:03:56,320 --> 01:03:57,677
IAN: Maybe.

1316
01:03:57,677 --> 01:03:59,010
Are they cormorants or pelicans?

1317
01:03:59,010 --> 01:04:02,500
I don't know.

1318
01:04:02,500 --> 01:04:04,570
High ISO or low ISO?

1319
01:04:04,570 --> 01:04:06,490
We'll start there.

1320
01:04:06,490 --> 01:04:07,990
DAN: Getting a lot of slow shutters.

1321
01:04:07,990 --> 01:04:08,660
Long shutter--

1322
01:04:08,660 --> 01:04:09,490
IAN: Yeah.

1323
01:04:09,490 --> 01:04:13,510
I think the most the most sort of
dramatic thing about this image

1324
01:04:13,510 --> 01:04:16,180
is the slow shutter speed
that allows the water to blur.

1325
01:04:16,180 --> 01:04:18,850


1326
01:04:18,850 --> 01:04:22,508
And the depth of field is--

1327
01:04:22,508 --> 01:04:23,050
I don't know.

1328
01:04:23,050 --> 01:04:25,008
I mean there's some
definition in the water out

1329
01:04:25,008 --> 01:04:31,480
in the far horizon and the rock so maybe
it's pretty large, but it's, you know--

1330
01:04:31,480 --> 01:04:33,100
and I don't notice a lot of noise so.

1331
01:04:33,100 --> 01:04:35,100
But the most dramatic
feature of this is someone

1332
01:04:35,100 --> 01:04:39,692
is utilizing a slow shutter speed for
an interesting compositional effect.

1333
01:04:39,692 --> 01:04:42,185
AUDIENCE: And I would say it's
low noise because when you

1334
01:04:42,185 --> 01:04:44,452
have the sun that's a powerful source.

1335
01:04:44,452 --> 01:04:45,160
IAN: Yeah, right.

1336
01:04:45,160 --> 01:04:46,630
So they may not need
a lot of sensitivity,

1337
01:04:46,630 --> 01:04:48,130
but we don't know what time of year?

1338
01:04:48,130 --> 01:04:49,420
What time of day?

1339
01:04:49,420 --> 01:04:54,478
This could be evening or early morning
so there's a lot of room for variation.

1340
01:04:54,478 --> 01:04:56,770
Our error bars are large with
that's that sort of idea,

1341
01:04:56,770 --> 01:05:00,320
but yep the sun is an incredibly
strong source of illumination

1342
01:05:00,320 --> 01:05:04,210
so probably not something
really, really high.

1343
01:05:04,210 --> 01:05:07,090
And we also don't see
any artifacts of that.

1344
01:05:07,090 --> 01:05:10,078
I don't see a lot of noise in the image.

1345
01:05:10,078 --> 01:05:10,870
How about this one?

1346
01:05:10,870 --> 01:05:23,020


1347
01:05:23,020 --> 01:05:24,980
AUDIENCE: Very small f-stop.

1348
01:05:24,980 --> 01:05:25,810
IAN: Yeah.

1349
01:05:25,810 --> 01:05:29,050
So a very small f-stop.

1350
01:05:29,050 --> 01:05:32,680
Well, actually I think we should
say a small aperture opening, which

1351
01:05:32,680 --> 01:05:33,940
is a large f-stop number.

1352
01:05:33,940 --> 01:05:36,317


1353
01:05:36,317 --> 01:05:39,400
Just to be specific because it gets
so confusing otherwise if you're like,

1354
01:05:39,400 --> 01:05:41,358
oh, it's a large f-stop
and you're like, but is

1355
01:05:41,358 --> 01:05:44,010
that the big number or the big opening?

1356
01:05:44,010 --> 01:05:44,510
So.

1357
01:05:44,510 --> 01:05:45,590
AUDIENCE: Small number, big opening.

1358
01:05:45,590 --> 01:05:46,410
IAN: Yeah, yeah.

1359
01:05:46,410 --> 01:05:51,370
No, but just-- it takes
muscle memory and practice.

1360
01:05:51,370 --> 01:05:55,600
So this is really about sort of diving
into a very small section of an image

1361
01:05:55,600 --> 01:05:59,610
and letting the out of focus play is
sort of a graphic element around that.

1362
01:05:59,610 --> 01:06:01,360
There doesn't appear
to be a lot of noise,

1363
01:06:01,360 --> 01:06:03,527
and there's not a lot to
suggest anything either way

1364
01:06:03,527 --> 01:06:05,530
about the shutter speed.

1365
01:06:05,530 --> 01:06:09,490
It's some sort of neutral value, but
they made a very conscious choice

1366
01:06:09,490 --> 01:06:14,200
to shoot at a very large aperture
to get a shallow depth of field.

1367
01:06:14,200 --> 01:06:18,190


1368
01:06:18,190 --> 01:06:20,845
Whoa.

1369
01:06:20,845 --> 01:06:21,831
What's this one?

1370
01:06:21,831 --> 01:06:25,510


1371
01:06:25,510 --> 01:06:27,260
DAN: Fast shutter
speed, [INAUDIBLE] says.

1372
01:06:27,260 --> 01:06:29,050
IAN: Yep, exactly.

1373
01:06:29,050 --> 01:06:30,880
And not much to suggest--

1374
01:06:30,880 --> 01:06:33,172
I mean, the depth the field
feels pretty reasonable.

1375
01:06:33,172 --> 01:06:35,380
I can see things in the
background that are in focus.

1376
01:06:35,380 --> 01:06:38,405


1377
01:06:38,405 --> 01:06:44,430
The water is completely
stopped midair for sure.

1378
01:06:44,430 --> 01:06:45,680
AUDIENCE: It's a big aperture.

1379
01:06:45,680 --> 01:06:46,597
IAN: Yeah, yeah, yeah.

1380
01:06:46,597 --> 01:06:48,940
Absolutely.

1381
01:06:48,940 --> 01:06:49,440
Whoa.

1382
01:06:49,440 --> 01:06:51,412
God.

1383
01:06:51,412 --> 01:06:52,398
What about this one?

1384
01:06:52,398 --> 01:07:01,280


1385
01:07:01,280 --> 01:07:02,310
AUDIENCE: High ISO.

1386
01:07:02,310 --> 01:07:02,927
IAN: Yeah.

1387
01:07:02,927 --> 01:07:03,510
I would agree.

1388
01:07:03,510 --> 01:07:04,128
Absolutely.

1389
01:07:04,128 --> 01:07:04,920
That's a good read.

1390
01:07:04,920 --> 01:07:07,800
AUDIENCE: [INAUDIBLE]. fast ISO.

1391
01:07:07,800 --> 01:07:08,670
IAN: Yeah.

1392
01:07:08,670 --> 01:07:09,870
What about the shutter speed though too?

1393
01:07:09,870 --> 01:07:11,037
This is sort of interesting.

1394
01:07:11,037 --> 01:07:11,874
AUDIENCE: Yeah.

1395
01:07:11,874 --> 01:07:14,268
I would say fast 'cause look,
they're frozen in the air.

1396
01:07:14,268 --> 01:07:16,560
IAN: Yeah, they're frozen in
midair so it's fast enough

1397
01:07:16,560 --> 01:07:19,800
to stop someone in mid leap.

1398
01:07:19,800 --> 01:07:22,340
But also they see in
order to do that they

1399
01:07:22,340 --> 01:07:26,070
needed to boost up the ISO to make it
more sensitive because it was allowing

1400
01:07:26,070 --> 01:07:28,777
such little light in for the time.

1401
01:07:28,777 --> 01:07:31,520
AUDIENCE: Yeah, and there's a
lot in focus in the picture.

1402
01:07:31,520 --> 01:07:32,142
IAN: Yeah.

1403
01:07:32,142 --> 01:07:34,958
AUDIENCE: How did they get this picture?

1404
01:07:34,958 --> 01:07:36,000
IAN: By boosting the ISO.

1405
01:07:36,000 --> 01:07:40,620
And I think you could-- like
the level of grain in the image,

1406
01:07:40,620 --> 01:07:42,633
or noise in the image,
is really apparent.

1407
01:07:42,633 --> 01:07:43,220
AUDIENCE: Yes.

1408
01:07:43,220 --> 01:07:43,720
IAN: Yeah?

1409
01:07:43,720 --> 01:07:45,470
AUDIENCE: One thing
on this image as well,

1410
01:07:45,470 --> 01:07:48,580
like you mentioned about color
reproduction and higher ISO.

1411
01:07:48,580 --> 01:07:49,080
IAN: Mm-hm.

1412
01:07:49,080 --> 01:07:52,160
AUDIENCE: I think this image
might be made black and white

1413
01:07:52,160 --> 01:07:55,110
and that's one kind of
thing people sometimes

1414
01:07:55,110 --> 01:07:57,180
do when they bump up their ISOs.

1415
01:07:57,180 --> 01:08:00,900
When you have bad color you just
make it black and white [INAUDIBLE]..

1416
01:08:00,900 --> 01:08:03,990
IAN: Yeah, because the color
gets really mushy and noisy,

1417
01:08:03,990 --> 01:08:06,060
and then the black and
white begins to feel

1418
01:08:06,060 --> 01:08:08,610
like a textural
compositional element rather

1419
01:08:08,610 --> 01:08:11,108
than a degradation of some other image.

1420
01:08:11,108 --> 01:08:12,150
That's a very good point.

1421
01:08:12,150 --> 01:08:14,935


1422
01:08:14,935 --> 01:08:15,765
Nope.

1423
01:08:15,765 --> 01:08:18,010
Backwards again.

1424
01:08:18,010 --> 01:08:22,330
So let's take a five minute break
at 7:04, and then we'll come back

1425
01:08:22,330 --> 01:08:24,330
and we'll talk a little
bit about depth of field

1426
01:08:24,330 --> 01:08:27,479
because I've sort of been saying
oh there's large depth of field

1427
01:08:27,479 --> 01:08:30,370
and shallow depth of field,
but we don't know what it is.

1428
01:08:30,370 --> 01:08:33,802
So let's demystify this when we
return in just a couple of minutes.

1429
01:08:33,802 --> 01:08:34,510
All right, folks.

1430
01:08:34,510 --> 01:08:36,760
Welcome back.

1431
01:08:36,760 --> 01:08:40,500
So We're gonna dive right in and
start talking a little bit about depth

1432
01:08:40,500 --> 01:08:46,859
of field as a creative tool, and
also how it works from a functional

1433
01:08:46,859 --> 01:08:49,050
standpoint with your camera.

1434
01:08:49,050 --> 01:08:54,210
So depth of field is the amount of
any image that's in apparent focus.

1435
01:08:54,210 --> 01:08:58,410
We've seen some images where there's
only-- a tiny part of the frame

1436
01:08:58,410 --> 01:09:00,779
is in focus and everything
else is sort of blurring out

1437
01:09:00,779 --> 01:09:03,180
in the foreground and the background.

1438
01:09:03,180 --> 01:09:07,050
And we've seen images where
everything from the foreground

1439
01:09:07,050 --> 01:09:12,330
object to the vast horizon
in the distance is in focus.

1440
01:09:12,330 --> 01:09:17,580
So we need to be careful because
apparent focus is the linchpin of this.

1441
01:09:17,580 --> 01:09:21,240
There is only one plane
of focus in an image,

1442
01:09:21,240 --> 01:09:25,290
and it's set when you adjust
the focal ring on your lens

1443
01:09:25,290 --> 01:09:29,130
and choose some distance
at that witness mark.

1444
01:09:29,130 --> 01:09:35,979
The rest of everything that you perceive
as in focus is apparently sharp.

1445
01:09:35,979 --> 01:09:42,300
It's sharp enough that our human eyes
don't notice that it's out of focus.

1446
01:09:42,300 --> 01:09:45,840
So there is a threshold
where at a certain point

1447
01:09:45,840 --> 01:09:48,630
our eyes do notice
that it's out of focus.

1448
01:09:48,630 --> 01:09:54,470
And this threshold is much
lower on smaller images.

1449
01:09:54,470 --> 01:09:56,220
And as you blow things
up, and I think you

1450
01:09:56,220 --> 01:09:58,420
may have experienced this,
where you take a picture

1451
01:09:58,420 --> 01:10:00,928
and it feels like it's sharp,
like my gravestone image.

1452
01:10:00,928 --> 01:10:02,970
And then I blow it up to
a big size and I'm like,

1453
01:10:02,970 --> 01:10:04,575
oh, it's actually out of focus.

1454
01:10:04,575 --> 01:10:05,450
AUDIENCE: [INAUDIBLE]

1455
01:10:05,450 --> 01:10:08,910
IAN: Yeah, on an 85 inch TV.

1456
01:10:08,910 --> 01:10:15,630
So there is this relationship to how
that apparent level of focus breaking

1457
01:10:15,630 --> 01:10:19,710
down the larger you enlarge an image.

1458
01:10:19,710 --> 01:10:22,170
So the bigger print that
you're going to make,

1459
01:10:22,170 --> 01:10:24,990
the larger you're going to project
something, the more important

1460
01:10:24,990 --> 01:10:28,380
it is to make sure that you nail
that critical plane of focus.

1461
01:10:28,380 --> 01:10:31,380
And to Dan's point we often,
when we're doing portraiture,

1462
01:10:31,380 --> 01:10:33,540
put that right through the eyes.

1463
01:10:33,540 --> 01:10:35,820
So that we know that this
person's perfectly in focus,

1464
01:10:35,820 --> 01:10:37,860
and that as the depth
of field grows we'll get

1465
01:10:37,860 --> 01:10:41,460
a wonderful image of their whole face.

1466
01:10:41,460 --> 01:10:43,410
So what does this look
like more generally?

1467
01:10:43,410 --> 01:10:46,050
So this is a funky
diagram that I made where

1468
01:10:46,050 --> 01:10:50,460
you have this sensor plane that's
inside of the camera, you have a lens,

1469
01:10:50,460 --> 01:10:56,610
and then you have your plane of
focus out here set to some distance.

1470
01:10:56,610 --> 01:10:59,820
So one of the interesting
things about depth of field

1471
01:10:59,820 --> 01:11:05,430
is that it's not perfectly 50/50
surrounding that plane of focus.

1472
01:11:05,430 --> 01:11:09,030
So the amount that's apparently in
focus is 1/3 third in front of it

1473
01:11:09,030 --> 01:11:11,190
and 2/3 behind it.

1474
01:11:11,190 --> 01:11:15,090
So there's actually a little more
behind that plane of focus that

1475
01:11:15,090 --> 01:11:19,050
appears in focus than in front of it.

1476
01:11:19,050 --> 01:11:22,530
And this is a handy
trick if you're really

1477
01:11:22,530 --> 01:11:25,680
shallow, you have a really
shallow depth of field,

1478
01:11:25,680 --> 01:11:29,850
you can actually cheat that plane
of focus a little bit forward

1479
01:11:29,850 --> 01:11:33,060
to make use of that extra
space that's behind it.

1480
01:11:33,060 --> 01:11:36,790


1481
01:11:36,790 --> 01:11:37,290
Nope.

1482
01:11:37,290 --> 01:11:39,860
I keep-- do I keep going backwards?

1483
01:11:39,860 --> 01:11:42,550
So we described depth of
field as being deep when

1484
01:11:42,550 --> 01:11:45,710
there's much of an
image in apparent focus,

1485
01:11:45,710 --> 01:11:49,730
and conversely when only a
small area is in apparent focus

1486
01:11:49,730 --> 01:11:53,420
we describe that as being shallow.

1487
01:11:53,420 --> 01:11:57,680
The three factors that control this
are our aperture, focal length,

1488
01:11:57,680 --> 01:12:00,210
and the focusing distance of the lens.

1489
01:12:00,210 --> 01:12:05,510
So the only exposure control that
effects depth of field is aperture,

1490
01:12:05,510 --> 01:12:09,860
but other elements effect
depth of field as well.

1491
01:12:09,860 --> 01:12:14,300
And that means the length of the
lens, whether you're at a wide,

1492
01:12:14,300 --> 01:12:17,060
normal, or tele.

1493
01:12:17,060 --> 01:12:20,480
And how closely the lens is focused.

1494
01:12:20,480 --> 01:12:22,650
Whether you're focused
at a subject really,

1495
01:12:22,650 --> 01:12:25,670
really close in front of the
lens or much, much further away.

1496
01:12:25,670 --> 01:12:28,430


1497
01:12:28,430 --> 01:12:32,270
So to sort of talk a little
bit more concretely about this.

1498
01:12:32,270 --> 01:12:38,150
So the smaller the opening the deeper
the depth of field for aperture.

1499
01:12:38,150 --> 01:12:45,092
So f/22 will make a deeper depth the
field in the same image than f/1.4.

1500
01:12:45,092 --> 01:12:48,050
And we saw some examples of that with
the cemetery shot where there's--

1501
01:12:48,050 --> 01:12:52,400
the background gravestone was out
of focus, but then it was in focus.

1502
01:12:52,400 --> 01:12:56,060
The larger the opening the
shallower the depth the field.

1503
01:12:56,060 --> 01:12:57,500
And there's this little basic--

1504
01:12:57,500 --> 01:12:59,750
major stops down there.

1505
01:12:59,750 --> 01:13:03,980
Shallow are going this way towards
more open and the smaller F number.

1506
01:13:03,980 --> 01:13:08,390
And deeper going towards the larger
F number but smaller opening.

1507
01:13:08,390 --> 01:13:11,210
So your rule of thumb to
help you make sense of this

1508
01:13:11,210 --> 01:13:14,930
is that doubling the aperture
doubles the depth field.

1509
01:13:14,930 --> 01:13:16,070
Yes?

1510
01:13:16,070 --> 01:13:17,870
AUDIENCE: [INAUDIBLE]

1511
01:13:17,870 --> 01:13:23,600
IAN: So it's a little bit helpful that
if you don't have enough depth of field

1512
01:13:23,600 --> 01:13:27,108
you can double the amount of it
just by opening up one stop--

1513
01:13:27,108 --> 01:13:28,150
or closing down one stop.

1514
01:13:28,150 --> 01:13:29,180
I'm sorry.

1515
01:13:29,180 --> 01:13:32,210
Closing down one stop and
adjusting one of the other exposure

1516
01:13:32,210 --> 01:13:34,460
controls to give you a little
bit more depth of field.

1517
01:13:34,460 --> 01:13:37,640


1518
01:13:37,640 --> 01:13:42,500
So a big thank you to Andrew Markham
for sitting in for us on this.

1519
01:13:42,500 --> 01:13:45,680
But what we have here
is a wide angle lens

1520
01:13:45,680 --> 01:13:51,150
set to f/16 with some amount
of lighting in a space.

1521
01:13:51,150 --> 01:13:58,670
And what I want to look at is not just
our subject, but the area behind them.

1522
01:13:58,670 --> 01:14:00,648
And watch what happens
as we play this video.

1523
01:14:00,648 --> 01:14:07,490


1524
01:14:07,490 --> 01:14:10,850
So what we're actually
doing in this moment

1525
01:14:10,850 --> 01:14:16,250
is opening up the aperture at
the same time as we decrease

1526
01:14:16,250 --> 01:14:18,800
the amount of light in the space.

1527
01:14:18,800 --> 01:14:24,110
So we're actually dimming the
lights as we open up the aperture.

1528
01:14:24,110 --> 01:14:30,440
And what you see is in the
starting frame when we're at f/16

1529
01:14:30,440 --> 01:14:32,930
all this appears crisp.

1530
01:14:32,930 --> 01:14:38,495
Maybe a little blurry, but much
crisper than this, for certain.

1531
01:14:38,495 --> 01:14:41,450


1532
01:14:41,450 --> 01:14:43,610
So this is a wide angle lens.

1533
01:14:43,610 --> 01:14:45,740
Do you remember what focal
length we were at, Dan?

1534
01:14:45,740 --> 01:14:48,890


1535
01:14:48,890 --> 01:14:50,570
25, let's say.

1536
01:14:50,570 --> 01:14:51,444
Maybe.

1537
01:14:51,444 --> 01:14:53,380
DAN: There was crop factor there too.

1538
01:14:53,380 --> 01:14:54,950
IAN: Yeah, yeah, yeah.

1539
01:14:54,950 --> 01:14:56,690
But it definitely is a wide angle lens.

1540
01:14:56,690 --> 01:15:01,550
And then if we look at this same
situation with a normal lens

1541
01:15:01,550 --> 01:15:03,980
we'll see something interesting happen.

1542
01:15:03,980 --> 01:15:07,760
Remember that I said that focal
length increases depth of field.

1543
01:15:07,760 --> 01:15:09,860
So here we are with a normal lens.

1544
01:15:09,860 --> 01:15:12,190
The angle of view has shrank a bit.

1545
01:15:12,190 --> 01:15:13,190
The camera hasn't moved.

1546
01:15:13,190 --> 01:15:16,300
We've just tightened
up the lens a little.

1547
01:15:16,300 --> 01:15:17,817
Remember from Dan's lecture.

1548
01:15:17,817 --> 01:15:20,150
And we'll do the same thing
again where we start at f/16

1549
01:15:20,150 --> 01:15:21,753
and we'll open up the aperture.

1550
01:15:21,753 --> 01:15:24,170
And we'll actually just sort
of dim the lights in concert.

1551
01:15:24,170 --> 01:15:34,960


1552
01:15:34,960 --> 01:15:41,540
And if we look at this one, again, you
see that the background is in focus.

1553
01:15:41,540 --> 01:15:44,060
And then it's out of focus,
but is it more out of focus

1554
01:15:44,060 --> 01:15:45,866
than the previous one?

1555
01:15:45,866 --> 01:15:47,180
AUDIENCE: Yes.

1556
01:15:47,180 --> 01:15:48,260
IAN: OK.

1557
01:15:48,260 --> 01:15:49,025
Yeah.

1558
01:15:49,025 --> 01:15:50,150
That's sort of interesting.

1559
01:15:50,150 --> 01:15:53,000


1560
01:15:53,000 --> 01:15:54,890
So now we'll move to a telephoto lens.

1561
01:15:54,890 --> 01:15:58,070
Again, the angle of view
cropping in-- or not cropping in,

1562
01:15:58,070 --> 01:16:00,767
but we'll run this again.

1563
01:16:00,767 --> 01:16:08,400


1564
01:16:08,400 --> 01:16:10,900
And now it's really soft.

1565
01:16:10,900 --> 01:16:13,943
Incredibly soft.

1566
01:16:13,943 --> 01:16:15,610
And this is the difference between them.

1567
01:16:15,610 --> 01:16:19,440
And I would say that this starting
softness out here is not quite as crisp

1568
01:16:19,440 --> 01:16:25,240
as it was in the earlier two iterations
of this, but that is much much softer.

1569
01:16:25,240 --> 01:16:29,500


1570
01:16:29,500 --> 01:16:34,910
So that's adjusting the aperture
in three different situations.

1571
01:16:34,910 --> 01:16:38,440
One on a wide angle lens,
one on a normal lens, and one

1572
01:16:38,440 --> 01:16:41,290
on a telephoto lens.

1573
01:16:41,290 --> 01:16:46,630
And by opening up the aperture,
making the opening larger,

1574
01:16:46,630 --> 01:16:51,390
we shrink the depth the field
in all three of those cases.

1575
01:16:51,390 --> 01:16:56,665
It's interesting to see how it's
different for wide, normal, and tele.

1576
01:16:56,665 --> 01:16:58,540
And we'll come back to
that in just a second.

1577
01:16:58,540 --> 01:17:02,380


1578
01:17:02,380 --> 01:17:06,190
So one of the other factors
that effects depth of field

1579
01:17:06,190 --> 01:17:07,690
is the distance to the subject.

1580
01:17:07,690 --> 01:17:11,145
And what I mean by that is
where your lens is focused.

1581
01:17:11,145 --> 01:17:14,020
Generally I'm just going to assume
that we're focused on our subject,

1582
01:17:14,020 --> 01:17:14,730
but perhaps not.

1583
01:17:14,730 --> 01:17:20,140
But the distance that your lens is
focused at, whether it's near or far

1584
01:17:20,140 --> 01:17:25,090
changes how deep or shallow
the depth of field is.

1585
01:17:25,090 --> 01:17:29,950
The closer that critical plane
of focus to the film plane

1586
01:17:29,950 --> 01:17:32,230
the shallower that depth of field is.

1587
01:17:32,230 --> 01:17:36,550
So the closer our subject is to us
the shallower the depth of field

1588
01:17:36,550 --> 01:17:38,920
will be versus when it's further away.

1589
01:17:38,920 --> 01:17:41,260
It's the inverse.

1590
01:17:41,260 --> 01:17:44,260
And a rule of thumb for this is
doubling the distance quadruples

1591
01:17:44,260 --> 01:17:47,420
the depth of field so
there's some power law here.

1592
01:17:47,420 --> 01:17:47,920
Yeah.

1593
01:17:47,920 --> 01:17:48,670
AUDIENCE: [INAUDIBLE].

1594
01:17:48,670 --> 01:17:49,170
IAN: Right?

1595
01:17:49,170 --> 01:17:53,977
Or halving the distance
cuts it by four, so--

1596
01:17:53,977 --> 01:17:55,060
DAN: It's dramatic though.

1597
01:17:55,060 --> 01:17:55,560
IAN: It is.

1598
01:17:55,560 --> 01:17:56,090
It is.

1599
01:17:56,090 --> 01:17:58,590
DAN: And I was shooting some
macro photography this weekend,

1600
01:17:58,590 --> 01:18:05,745
and I was at an aperture that was
5.6 six and f/8 and it was so thin.

1601
01:18:05,745 --> 01:18:07,870
You think of that depth
field as being fairly deep,

1602
01:18:07,870 --> 01:18:09,100
but it really matters how close you are.

1603
01:18:09,100 --> 01:18:11,475
With the macro lens you're
right up against your subject,

1604
01:18:11,475 --> 01:18:13,435
and so this really is exaggerated there.

1605
01:18:13,435 --> 01:18:15,310
IAN: Right, and I think
maybe even that image

1606
01:18:15,310 --> 01:18:18,520
that we looked at that was the
green plant with that really

1607
01:18:18,520 --> 01:18:21,520
narrow depth of field might
have been on a macro lens.

1608
01:18:21,520 --> 01:18:25,890


1609
01:18:25,890 --> 01:18:27,760
Let me just go--

1610
01:18:27,760 --> 01:18:29,660
what did I do?

1611
01:18:29,660 --> 01:18:32,470
I went backwards again.

1612
01:18:32,470 --> 01:18:37,720
So this is a wide angle lens
where we're focused at four feet.

1613
01:18:37,720 --> 01:18:39,055
Look at the background.

1614
01:18:39,055 --> 01:18:40,180
There's a lot out of focus.

1615
01:18:40,180 --> 01:18:43,030
We have our little rubber ducky there.

1616
01:18:43,030 --> 01:18:44,530
And as we play this video--

1617
01:18:44,530 --> 01:18:47,216


1618
01:18:47,216 --> 01:18:47,716
perhaps.

1619
01:18:47,716 --> 01:18:48,216
No.

1620
01:18:48,216 --> 01:18:53,490


1621
01:18:53,490 --> 01:18:57,740
And we moved the camera
further away from our subject

1622
01:18:57,740 --> 01:19:01,340
racking with our subject, so
keeping the focus on our subject.

1623
01:19:01,340 --> 01:19:05,570
Increasing that distance you can
see that the focus in the background

1624
01:19:05,570 --> 01:19:08,480
begins to change.

1625
01:19:08,480 --> 01:19:11,990
That by increasing the distance
the objects behind our subject

1626
01:19:11,990 --> 01:19:13,000
come into crisper focus.

1627
01:19:13,000 --> 01:19:22,400


1628
01:19:22,400 --> 01:19:23,150
DAN: It's so slow.

1629
01:19:23,150 --> 01:19:23,800
It's wonderful.

1630
01:19:23,800 --> 01:19:24,200
It's like--

1631
01:19:24,200 --> 01:19:25,834
AUDIENCE: Can you play it again for me?

1632
01:19:25,834 --> 01:19:26,334
IAN: Sure.

1633
01:19:26,334 --> 01:19:54,680


1634
01:19:54,680 --> 01:19:56,730
So if you think back to
that duck originally it

1635
01:19:56,730 --> 01:20:00,625
was much softer than it is now.

1636
01:20:00,625 --> 01:20:02,920
AUDIENCE: So you're
changing focal length.

1637
01:20:02,920 --> 01:20:05,020
Are you doing it literally
moving the camera,

1638
01:20:05,020 --> 01:20:06,620
but keeping the focus on the subject?

1639
01:20:06,620 --> 01:20:07,370
IAN: Yeah.

1640
01:20:07,370 --> 01:20:13,152
So what that is doing is essentially
changing where the lens is focused,

1641
01:20:13,152 --> 01:20:14,360
and just moving it like this.

1642
01:20:14,360 --> 01:20:18,650
As the camera moves backwards
it changes that distance.

1643
01:20:18,650 --> 01:20:22,280
So when you look at the starting
position and the end position

1644
01:20:22,280 --> 01:20:27,380
you really can see
how soft it originally

1645
01:20:27,380 --> 01:20:29,720
was when we were very
close to the subject

1646
01:20:29,720 --> 01:20:32,630
compared to when we moved further back.

1647
01:20:32,630 --> 01:20:34,910
So this is all, again,
a wide angle lens.

1648
01:20:34,910 --> 01:20:38,180


1649
01:20:38,180 --> 01:20:39,760
So now we have a normal lens.

1650
01:20:39,760 --> 01:20:42,950


1651
01:20:42,950 --> 01:20:47,370
We start a little bit more out
of focus in the background.

1652
01:20:47,370 --> 01:20:51,485
The depth of field is
a little bit shallower.

1653
01:20:51,485 --> 01:21:17,490
AUDIENCE: [INAUDIBLE]

1654
01:21:17,490 --> 01:21:22,020
IAN: So if we do our
little comparison again

1655
01:21:22,020 --> 01:21:28,770
we can see that distance to subject is
really driving how much depth the field

1656
01:21:28,770 --> 01:21:31,352
we have between these two images.

1657
01:21:31,352 --> 01:21:34,330


1658
01:21:34,330 --> 01:21:37,570
And to do our due diligence
here is a telephoto lens.

1659
01:21:37,570 --> 01:21:42,230
Notice how out of focus
our friend the duck is.

1660
01:21:42,230 --> 01:21:46,210
And as we move through the scale we'll
see if we can get him to be in focus.

1661
01:21:46,210 --> 01:22:16,590


1662
01:22:16,590 --> 01:22:17,220
So not quite.

1663
01:22:17,220 --> 01:22:18,270
Not quite in focus.

1664
01:22:18,270 --> 01:22:24,270
The depth of field is still shallow
enough to keep the duck out of focus.

1665
01:22:24,270 --> 01:22:26,948
And this is again a telephoto lens.

1666
01:22:26,948 --> 01:22:28,740
And if we do the
comparison between the two

1667
01:22:28,740 --> 01:22:30,660
we can see that it's
incredibly out of focus.

1668
01:22:30,660 --> 01:22:34,020
Almost unrecognizable to
getting into some sort of shape

1669
01:22:34,020 --> 01:22:36,240
that we can understand a little bit.

1670
01:22:36,240 --> 01:22:48,880


1671
01:22:48,880 --> 01:22:50,919
What is happening?

1672
01:22:50,919 --> 01:23:02,803
AUDIENCE: [INAUDIBLE]

1673
01:23:02,803 --> 01:23:03,970
DAN: Don't steal my thunder.

1674
01:23:03,970 --> 01:23:05,580
IAN: OK.

1675
01:23:05,580 --> 01:23:07,170
We'll leave it for a second.

1676
01:23:07,170 --> 01:23:09,080
So this is what's called a dolly zoom.

1677
01:23:09,080 --> 01:23:12,560
This is where the camera is
moving back as the camera zooms in

1678
01:23:12,560 --> 01:23:13,950
at the same time.

1679
01:23:13,950 --> 01:23:18,440
So we maintain the exact same
frame over a camera movement.

1680
01:23:18,440 --> 01:23:24,798
And what you notice is we actually go
from a wide angle lens with this frame

1681
01:23:24,798 --> 01:23:26,590
to a telephoto lens
with this frame and you

1682
01:23:26,590 --> 01:23:33,830
can see the actual spatial
distortions that happen in real time

1683
01:23:33,830 --> 01:23:34,780
as you change--

1684
01:23:34,780 --> 01:23:36,410
OK, one more time.

1685
01:23:36,410 --> 01:23:37,128
OK.

1686
01:23:37,128 --> 01:23:39,920
DAN: We'll look at this next week
when we talk about video as well.

1687
01:23:39,920 --> 01:23:40,420
Yeah.

1688
01:23:40,420 --> 01:23:42,860


1689
01:23:42,860 --> 01:23:45,350
IAN: So this is--

1690
01:23:45,350 --> 01:23:48,440
I really wanted to show
it because it's awesome.

1691
01:23:48,440 --> 01:23:50,570
And one of the things that--

1692
01:23:50,570 --> 01:23:53,750


1693
01:23:53,750 --> 01:23:59,900
our final element of effecting depth
of field or controlling depth of field

1694
01:23:59,900 --> 01:24:02,960
is this idea that focal length matters.

1695
01:24:02,960 --> 01:24:08,540
That a longer full focal length
yields a shallower depth of field.

1696
01:24:08,540 --> 01:24:14,090
So a 150, a 250 millimeter lens on a
full frame camera, a telephoto lens,

1697
01:24:14,090 --> 01:24:19,088
has a shallower depth of field
than a wide angle lens, which

1698
01:24:19,088 --> 01:24:20,380
yields a deeper depth of field.

1699
01:24:20,380 --> 01:24:26,600
And I think we saw that when we looked
at how different each of the elements

1700
01:24:26,600 --> 01:24:32,840
were when they were in
wide, normal, and tele.

1701
01:24:32,840 --> 01:24:36,740
The wide angle lenses had more
in focus in the background

1702
01:24:36,740 --> 01:24:38,610
than the telephoto ones.

1703
01:24:38,610 --> 01:24:40,510
AUDIENCE: Regardless of the aperture?

1704
01:24:40,510 --> 01:24:44,900
IAN: Yeah, regardless of the aperture.

1705
01:24:44,900 --> 01:24:47,540
Well, yes.

1706
01:24:47,540 --> 01:24:51,230
So when we were changing aperture--

1707
01:24:51,230 --> 01:24:54,540
If we go-- let's-- whoops.

1708
01:24:54,540 --> 01:24:55,460
Let's go back here.

1709
01:24:55,460 --> 01:24:58,890


1710
01:24:58,890 --> 01:25:03,780
So we were changing aperture here.

1711
01:25:03,780 --> 01:25:06,270
And you can see the effect
that it has on depth of field.

1712
01:25:06,270 --> 01:25:09,020
We didn't change focal length and
we didn't change focus distance.

1713
01:25:09,020 --> 01:25:12,600


1714
01:25:12,600 --> 01:25:15,690
When we go to here and we change--

1715
01:25:15,690 --> 01:25:18,690
let's do this one.

1716
01:25:18,690 --> 01:25:20,820
We don't change aperture.

1717
01:25:20,820 --> 01:25:23,220
We don't change focal length.

1718
01:25:23,220 --> 01:25:24,750
This is a wider angle of view.

1719
01:25:24,750 --> 01:25:26,672
This is the camera backing up.

1720
01:25:26,672 --> 01:25:28,380
This is just the
distance to the subject.

1721
01:25:28,380 --> 01:25:33,390
And you can see the difference
between the elements.

1722
01:25:33,390 --> 01:25:38,880
But that is to say that this
image compared to this image

1723
01:25:38,880 --> 01:25:42,210
are two different-- this is
us changing focal length,

1724
01:25:42,210 --> 01:25:44,940
while all the other
elements stay the same.

1725
01:25:44,940 --> 01:25:47,490
We still shifted over distance,
but this is its own unit.

1726
01:25:47,490 --> 01:25:50,293


1727
01:25:50,293 --> 01:25:52,710
So your rule of thumb for this
is halving the focal length

1728
01:25:52,710 --> 01:25:54,002
quadruples the depth the field.

1729
01:25:54,002 --> 01:25:56,270
Doubling the focal length
cuts it by a quarter.

1730
01:25:56,270 --> 01:25:58,020
That might be helpful
for you in the field

1731
01:25:58,020 --> 01:26:01,980
if you want to-- like I need
much more depth of field.

1732
01:26:01,980 --> 01:26:05,290
A little rule of thumb
to help you get there.

1733
01:26:05,290 --> 01:26:10,080
The interesting thing though
is that in this comparison

1734
01:26:10,080 --> 01:26:20,080
when we start at the beginning and the
end is the depth of field different?

1735
01:26:20,080 --> 01:26:23,832
And I think maybe we'll
do this side by side here.

1736
01:26:23,832 --> 01:26:26,820


1737
01:26:26,820 --> 01:26:29,160
So this is the dolly zoom that we did.

1738
01:26:29,160 --> 01:26:35,160
We started with a wide angle
lens very close to the subject.

1739
01:26:35,160 --> 01:26:37,740
And we didn't change the aperture.

1740
01:26:37,740 --> 01:26:40,410
We did change the focal
distance, but then we

1741
01:26:40,410 --> 01:26:46,860
moved the camera to a telephoto
lens further from the subject.

1742
01:26:46,860 --> 01:26:50,320
So we increased the focal
length, but at the same time

1743
01:26:50,320 --> 01:26:54,540
we also increased the
distance to the subject.

1744
01:26:54,540 --> 01:26:59,550
And if you look at this
it's sort of apparent,

1745
01:26:59,550 --> 01:27:01,740
but it's difficult because
we maybe don't see this,

1746
01:27:01,740 --> 01:27:04,320
but the parts that are
out of focus are sort of

1747
01:27:04,320 --> 01:27:09,190
similarly out of focus in
both of these positions.

1748
01:27:09,190 --> 01:27:13,795
So what that means is that as the
focal length pushed the depth of field

1749
01:27:13,795 --> 01:27:16,920
smaller and smaller the distance to
the subject pushed it larger and larger

1750
01:27:16,920 --> 01:27:19,590
and they sort of offset each other.

1751
01:27:19,590 --> 01:27:23,820
Remember that our rule of thumb was if
you change it by half it doubles it.

1752
01:27:23,820 --> 01:27:25,500
If you change it by half--

1753
01:27:25,500 --> 01:27:27,390
or if you change it by
half it quadruples it.

1754
01:27:27,390 --> 01:27:29,700
If you change it by
half it quadruples it.

1755
01:27:29,700 --> 01:27:32,910
And so those two values
actually offset each other.

1756
01:27:32,910 --> 01:27:36,720
So for the same frame
focal length doesn't

1757
01:27:36,720 --> 01:27:40,920
do a lot to adjust your
depth of field because you

1758
01:27:40,920 --> 01:27:43,200
have to change the focus distance.

1759
01:27:43,200 --> 01:27:44,160
Does that make sense?

1760
01:27:44,160 --> 01:27:47,610
And they actually offset each other.

1761
01:27:47,610 --> 01:27:52,932
So if you-- let's draw it.

1762
01:27:52,932 --> 01:27:54,140
Maybe that's the easiest way.

1763
01:27:54,140 --> 01:27:58,840


1764
01:27:58,840 --> 01:28:06,010
So if we have a frame here
and we have a frame here.

1765
01:28:06,010 --> 01:28:11,890
And they're the same frame, but for
this one the camera is very close,

1766
01:28:11,890 --> 01:28:14,830
and for this one the camera is very far.

1767
01:28:14,830 --> 01:28:18,340
And for this one we're at a wide,
and for this one we're at a tele.

1768
01:28:18,340 --> 01:28:19,540
We've zoomed the lens out--

1769
01:28:19,540 --> 01:28:21,082
this is exactly what a dolly zoom is.

1770
01:28:21,082 --> 01:28:22,270
It starts close.

1771
01:28:22,270 --> 01:28:26,710
It moves this way and as it
moves we zoom the camera in.

1772
01:28:26,710 --> 01:28:31,300
So what we've done is
this focal length is

1773
01:28:31,300 --> 01:28:35,860
getting larger, which we know
produces a shallower depth of field.

1774
01:28:35,860 --> 01:28:38,590


1775
01:28:38,590 --> 01:28:40,500
We're going from a wide to a tele.

1776
01:28:40,500 --> 01:28:43,300


1777
01:28:43,300 --> 01:28:48,900
And this focus distance, the
distance from here to here,

1778
01:28:48,900 --> 01:28:52,043
is getting larger, which we know
increases the depth of field

1779
01:28:52,043 --> 01:28:52,960
so it makes it deeper.

1780
01:28:52,960 --> 01:28:56,200


1781
01:28:56,200 --> 01:29:03,650
And in doing so they essentially
cancel each other out roughly.

1782
01:29:03,650 --> 01:29:07,500
AUDIENCE: And that's what gives
that whole effect [INAUDIBLE]

1783
01:29:07,500 --> 01:29:09,340
IAN: Yeah, so really
what we're only seeing

1784
01:29:09,340 --> 01:29:13,540
is the spatial relationships of
the foreground and background doing

1785
01:29:13,540 --> 01:29:16,900
that sort of expansion,
compression trick,

1786
01:29:16,900 --> 01:29:20,380
rather than large portions of the
image coming in and out of focus

1787
01:29:20,380 --> 01:29:24,740
as the depth of field shifts
throughout that entire element.

1788
01:29:24,740 --> 01:29:25,990
So that's just a little aside.

1789
01:29:25,990 --> 01:29:27,970
That's like a hiccup where if you--

1790
01:29:27,970 --> 01:29:31,390
focal length really matters for depth
of field if you don't with the camera.

1791
01:29:31,390 --> 01:29:33,390
Because obviously, you
start to move the camera,

1792
01:29:33,390 --> 01:29:38,100
it begins to do less because you change
your focal distance along with it.

1793
01:29:38,100 --> 01:29:43,641
Are there questions on depth of field
before we move on a little further?

1794
01:29:43,641 --> 01:29:50,120


1795
01:29:50,120 --> 01:29:54,950
So one of the other tools in your
arsenal besides a light meter

1796
01:29:54,950 --> 01:29:59,390
is this histogram, which
is basically a plotting.

1797
01:29:59,390 --> 01:30:03,970
It shows the distribution of
brightness values of any given image.

1798
01:30:03,970 --> 01:30:06,640
It can also display the
distribution of color values.

1799
01:30:06,640 --> 01:30:08,390
But for our purposes
right now we're going

1800
01:30:08,390 --> 01:30:12,170
to look at it as a luminosity scale.

1801
01:30:12,170 --> 01:30:16,340
And all this says is that they're--

1802
01:30:16,340 --> 01:30:17,720
it reads left to right.

1803
01:30:17,720 --> 01:30:22,280
So this is black, and this is
white, and everywhere in between

1804
01:30:22,280 --> 01:30:25,600
is some midtone of brightness.

1805
01:30:25,600 --> 01:30:31,100
And it just shows you how much of your
image is falling in certain areas.

1806
01:30:31,100 --> 01:30:35,090
And so is if this is
full black that means

1807
01:30:35,090 --> 01:30:36,860
this area is probably our shadows.

1808
01:30:36,860 --> 01:30:39,330
If this is full white this area
is probably our highlights,

1809
01:30:39,330 --> 01:30:41,540
and somewhere in here is our midtones.

1810
01:30:41,540 --> 01:30:44,690
So in looking at this
histogram we can see

1811
01:30:44,690 --> 01:30:49,550
that there's a lot of data
in the dark side of this

1812
01:30:49,550 --> 01:30:52,760
with a big spike of
white light at the top.

1813
01:30:52,760 --> 01:30:57,440
And there's sort of nothing at the
high delicate highlights or mid tones.

1814
01:30:57,440 --> 01:31:01,610


1815
01:31:01,610 --> 01:31:08,042
So this is a histogram What do
you think this is a histogram of?

1816
01:31:08,042 --> 01:31:09,860
AUDIENCE: [INAUDIBLE]

1817
01:31:09,860 --> 01:31:10,580
IAN: Yeah.

1818
01:31:10,580 --> 01:31:15,380
So this is a histogram of this.

1819
01:31:15,380 --> 01:31:18,170
Just plain white.

1820
01:31:18,170 --> 01:31:21,050
Everything is just
jammed up on that end.

1821
01:31:21,050 --> 01:31:24,650
So this should warn you when you
start to see your images sort of jam

1822
01:31:24,650 --> 01:31:26,120
up towards one end--

1823
01:31:26,120 --> 01:31:28,310
towards the right hand
side you're going to end up

1824
01:31:28,310 --> 01:31:31,550
with something that's very bright.

1825
01:31:31,550 --> 01:31:34,625
What about this one?

1826
01:31:34,625 --> 01:31:36,038
AUDIENCE: [INAUDIBLE]

1827
01:31:36,038 --> 01:31:37,116
IAN: Yeah, it's this.

1828
01:31:37,116 --> 01:31:37,930
It's just this.

1829
01:31:37,930 --> 01:31:38,510
Just black.

1830
01:31:38,510 --> 01:31:40,868
Inky blackness.

1831
01:31:40,868 --> 01:31:41,660
How about that one?

1832
01:31:41,660 --> 01:31:44,642


1833
01:31:44,642 --> 01:31:46,630
AUDIENCE: Is it 18%?

1834
01:31:46,630 --> 01:31:48,280
IAN: Yeah, this is middle gray.

1835
01:31:48,280 --> 01:31:50,230
It's this image right here.

1836
01:31:50,230 --> 01:31:52,000
Here's one for you.

1837
01:31:52,000 --> 01:31:52,607
What's this?

1838
01:31:52,607 --> 01:32:02,160


1839
01:32:02,160 --> 01:32:03,638
It's a lot of everything.

1840
01:32:03,638 --> 01:32:07,130
AUDIENCE: It's going to be, probably,
[INAUDIBLE] highlight [INAUDIBLE]..

1841
01:32:07,130 --> 01:32:08,810
IAN: Yeah.

1842
01:32:08,810 --> 01:32:09,537
Maybe.

1843
01:32:09,537 --> 01:32:10,730
AUDIENCE: [INAUDIBLE]

1844
01:32:10,730 --> 01:32:11,630
IAN: Yeah.

1845
01:32:11,630 --> 01:32:14,810
It's actually just this gradient.

1846
01:32:14,810 --> 01:32:18,200
It's equal parts of
every sort of element

1847
01:32:18,200 --> 01:32:21,380
and it renders a very flat histogram.

1848
01:32:21,380 --> 01:32:23,942
So what's interesting about
that is this flat histogram

1849
01:32:23,942 --> 01:32:26,150
tells you that there's a
completely even distribution

1850
01:32:26,150 --> 01:32:30,410
of tones, which sort of suggests
that that's exactly what this is.

1851
01:32:30,410 --> 01:32:32,240
It's about as even a
distribution of tones

1852
01:32:32,240 --> 01:32:35,180
as you can get in any kind of image.

1853
01:32:35,180 --> 01:32:39,590
DAN: Well, [INAUDIBLE],, just to hammer
around the point of how to read this,

1854
01:32:39,590 --> 01:32:43,250
this gradient is on the left
black, on the right white,

1855
01:32:43,250 --> 01:32:46,830
which is how a histogram
represents its luminosity as well.

1856
01:32:46,830 --> 01:32:49,992
But if you were to reverse
this image the histogram

1857
01:32:49,992 --> 01:32:51,950
does not read left to
right like an image does.

1858
01:32:51,950 --> 01:32:53,242
It would be the same histogram.

1859
01:32:53,242 --> 01:32:54,050
IAN: Yes.

1860
01:32:54,050 --> 01:32:55,310
Yeah.

1861
01:32:55,310 --> 01:32:58,247
Actually that-- yeah, exactly.

1862
01:32:58,247 --> 01:33:00,080
'Cause it's just basically
saying that there

1863
01:33:00,080 --> 01:33:03,865
is a certain number of
dark pixels in this image

1864
01:33:03,865 --> 01:33:05,990
and a certain number of
light pixels in this image.

1865
01:33:05,990 --> 01:33:07,670
It doesn't care where they are.

1866
01:33:07,670 --> 01:33:11,220
Just there's this much value.

1867
01:33:11,220 --> 01:33:13,600
So let's go back to the
histogram that we saw before.

1868
01:33:13,600 --> 01:33:16,610
What do you think this
is a histogram for?

1869
01:33:16,610 --> 01:33:20,078


1870
01:33:20,078 --> 01:33:21,120
Keeps coming back around.

1871
01:33:21,120 --> 01:33:22,340
It's like a boomerang.

1872
01:33:22,340 --> 01:33:24,840
AUDIENCE: [INAUDIBLE]

1873
01:33:24,840 --> 01:33:26,590
DAN: Alec says an outside shot.

1874
01:33:26,590 --> 01:33:27,690
IAN: An outside shot.

1875
01:33:27,690 --> 01:33:28,815
That's sort of interesting.

1876
01:33:28,815 --> 01:33:30,585
Why might we say that?

1877
01:33:30,585 --> 01:33:32,960
DAN: Alec also feel free to
unmute and just shout it out.

1878
01:33:32,960 --> 01:33:34,085
IAN: Yeah, we can hear you.

1879
01:33:34,085 --> 01:33:38,270


1880
01:33:38,270 --> 01:33:43,550
AUDIENCE: My guess looking at that would
be like your outside shot that you had

1881
01:33:43,550 --> 01:33:45,860
'cause there was so much
shadow detail from--

1882
01:33:45,860 --> 01:33:47,320
like in the trees and everything.

1883
01:33:47,320 --> 01:33:48,112
IAN: Yeah, totally.

1884
01:33:48,112 --> 01:33:48,940
It's the same shot.

1885
01:33:48,940 --> 01:33:50,320
It just keeps coming back.

1886
01:33:50,320 --> 01:33:51,778
AUDIENCE: You've gotta make points.

1887
01:33:51,778 --> 01:33:53,470
IAN: Yeah, that's nice.

1888
01:33:53,470 --> 01:33:56,680
So there's not a lot of
highlight value in this image.

1889
01:33:56,680 --> 01:33:58,900
There's a little bit of
that snow, but most of it

1890
01:33:58,900 --> 01:34:03,580
is this dark tree value, which is
pushing the majority of our tones

1891
01:34:03,580 --> 01:34:08,740
into this shadow area, but we see a
pretty decent distribution of tones,

1892
01:34:08,740 --> 01:34:13,900
and we have some highlight values,
and we have some shadow values.

1893
01:34:13,900 --> 01:34:17,540
DAN: And the very rightmost
pixel here, on the right side,

1894
01:34:17,540 --> 01:34:19,040
we also have this indicator up here.

1895
01:34:19,040 --> 01:34:20,380
And it looks different
in different software,

1896
01:34:20,380 --> 01:34:22,130
but this is the clipping
indicator meaning

1897
01:34:22,130 --> 01:34:25,630
that something is fully overexposed,
meaning that it's true white.

1898
01:34:25,630 --> 01:34:28,437
IAN: Yeah, and it honestly is the--

1899
01:34:28,437 --> 01:34:29,770
it's the middle section in here.

1900
01:34:29,770 --> 01:34:32,320


1901
01:34:32,320 --> 01:34:35,952
Like there's some small amount of
that that's clipped out pretty hard.

1902
01:34:35,952 --> 01:34:39,160
AUDIENCE: So when you say clipping that
means like there's like [INAUDIBLE]??

1903
01:34:39,160 --> 01:34:40,040
IAN: Yeah.

1904
01:34:40,040 --> 01:34:46,060
So as you try to render
tones in an image

1905
01:34:46,060 --> 01:34:50,080
you can essentially push up
against the maximum value, which

1906
01:34:50,080 --> 01:34:55,750
is like it's so bright that it hits the
maximum recordable value for the camera

1907
01:34:55,750 --> 01:34:58,630
and then there's just no more data
to record so it just is at max.

1908
01:34:58,630 --> 01:34:59,505
AUDIENCE: [INAUDIBLE]

1909
01:34:59,505 --> 01:35:02,020
IAN: It's at full, which is usually--

1910
01:35:02,020 --> 01:35:04,280
like clipping over exposure.

1911
01:35:04,280 --> 01:35:06,820
It will show up as white
because it's maximum red,

1912
01:35:06,820 --> 01:35:08,200
maximum blue, maximum green.

1913
01:35:08,200 --> 01:35:10,540
It's just pinned at the top.

1914
01:35:10,540 --> 01:35:12,070
AUDIENCE: [INAUDIBLE]

1915
01:35:12,070 --> 01:35:13,220
IAN: Yeah, exactly.

1916
01:35:13,220 --> 01:35:15,902
And the opposite is where it is
just crushed down to that black.

1917
01:35:15,902 --> 01:35:18,610
And that's what those two extreme
examples of the white histogram

1918
01:35:18,610 --> 01:35:20,312
and the black scram were.

1919
01:35:20,312 --> 01:35:22,270
Where there was-- like
all the values were just

1920
01:35:22,270 --> 01:35:25,210
pinned at either the left or the right.

1921
01:35:25,210 --> 01:35:32,640
Either the maximum highlight value
or the minimum darkness value.

1922
01:35:32,640 --> 01:35:35,380
And so once you clip though you
can't get that information back.

1923
01:35:35,380 --> 01:35:35,900
It's lost.

1924
01:35:35,900 --> 01:35:37,317
There's no way to bring that back.

1925
01:35:37,317 --> 01:35:41,020
If you try to do any correction
on that and bring the tone down

1926
01:35:41,020 --> 01:35:47,570
it's just going to shift in
grayness 'cause there's no detail.

1927
01:35:47,570 --> 01:35:48,850
It's just bright white.

1928
01:35:48,850 --> 01:35:51,550
It's a flat white field,
and if you bring it down

1929
01:35:51,550 --> 01:35:56,530
it just will be a flat gray field or
you go all the way down to a flat black.

1930
01:35:56,530 --> 01:35:59,290


1931
01:35:59,290 --> 01:36:03,310
So if we look at this
here histogram, which

1932
01:36:03,310 --> 01:36:08,775
is the idea of a low key histogram
it's pushed to which side?

1933
01:36:08,775 --> 01:36:09,689
AUDIENCE: The left.

1934
01:36:09,689 --> 01:36:10,570
IAN: Which is?

1935
01:36:10,570 --> 01:36:11,320
AUDIENCE: Shadows.

1936
01:36:11,320 --> 01:36:14,050
IAN: Shadows.

1937
01:36:14,050 --> 01:36:19,480
And we've got some midtones.

1938
01:36:19,480 --> 01:36:21,293
Not really a lot.

1939
01:36:21,293 --> 01:36:22,210
AUDIENCE: Very little.

1940
01:36:22,210 --> 01:36:23,102
IAN: Yeah.

1941
01:36:23,102 --> 01:36:24,310
And we have some highlights--

1942
01:36:24,310 --> 01:36:26,800
AUDIENCE: And there is a
peak of light seek of white.

1943
01:36:26,800 --> 01:36:28,390
IAN: What's that?

1944
01:36:28,390 --> 01:36:31,300
AUDIENCE: There is a peak of
white somewhere in there as well.

1945
01:36:31,300 --> 01:36:34,360
IAN: Yeah, there is a
little bit of brightness.

1946
01:36:34,360 --> 01:36:37,330
You can see it ticks up just at the end.

1947
01:36:37,330 --> 01:36:39,370
Just a little bit there.

1948
01:36:39,370 --> 01:36:41,635
So what do we imagine
this image looks like?

1949
01:36:41,635 --> 01:36:44,960


1950
01:36:44,960 --> 01:36:47,340
DAN: [INAUDIBLE] says, underexposed.

1951
01:36:47,340 --> 01:36:49,743
AUDIENCE: A white dot on a dark wall?

1952
01:36:49,743 --> 01:36:51,160
IAN: A white tall and a dark wall?

1953
01:36:51,160 --> 01:36:55,180
Yeah, something like that.

1954
01:36:55,180 --> 01:36:56,200
Yeah.

1955
01:36:56,200 --> 01:37:00,973
So there's some small amount of
brightness in a large dark field.

1956
01:37:00,973 --> 01:37:02,140
That was a pretty good read.

1957
01:37:02,140 --> 01:37:04,930


1958
01:37:04,930 --> 01:37:07,588
This is probably our white values.

1959
01:37:07,588 --> 01:37:10,630
There's the tiny bit of mid tones that
we were getting and the rest of it

1960
01:37:10,630 --> 01:37:15,040
is falling off into these very
distinctive shadow details.

1961
01:37:15,040 --> 01:37:18,580


1962
01:37:18,580 --> 01:37:22,510
So you might have looked at-- if you
look at this histogram, like all tools,

1963
01:37:22,510 --> 01:37:24,640
it can be fooled.

1964
01:37:24,640 --> 01:37:30,730
This was an intentionally exposed
image to have all of the values--

1965
01:37:30,730 --> 01:37:32,260
to have this be dark.

1966
01:37:32,260 --> 01:37:35,800
To have everything pushed
to the left because that's

1967
01:37:35,800 --> 01:37:40,410
the sort of the composition that
this photographer was going for,

1968
01:37:40,410 --> 01:37:42,410
but if you were sort of
just looking a histogram

1969
01:37:42,410 --> 01:37:43,600
you'd be like no that's not right.

1970
01:37:43,600 --> 01:37:45,550
I'm looking for an even
distribution of midtones.

1971
01:37:45,550 --> 01:37:48,760
And I think you'll find a lot of people
suggest that that's the correct way,

1972
01:37:48,760 --> 01:37:51,160
but it's not always.

1973
01:37:51,160 --> 01:37:54,190
It does matter about your intention.

1974
01:37:54,190 --> 01:37:57,670
So the flip side of this
is a high key histogram.

1975
01:37:57,670 --> 01:38:00,438
It's pushed to which side?

1976
01:38:00,438 --> 01:38:01,480
AUDIENCE: The highlights.

1977
01:38:01,480 --> 01:38:03,880
IAN: The highlights.

1978
01:38:03,880 --> 01:38:06,610
DAN: And just to say
plainly we're looking

1979
01:38:06,610 --> 01:38:08,320
at an overlay of
several histograms here.

1980
01:38:08,320 --> 01:38:09,460
Different color channels.

1981
01:38:09,460 --> 01:38:11,690
And then the gray one
represents the luminosity.

1982
01:38:11,690 --> 01:38:12,190
IAN: Right.

1983
01:38:12,190 --> 01:38:14,050
DAN: But that just depends on
which software you're using

1984
01:38:14,050 --> 01:38:15,550
and what options you have turned on.

1985
01:38:15,550 --> 01:38:18,370
IAN: Yeah, and so you can get
a variety of different scopes.

1986
01:38:18,370 --> 01:38:19,760
And actually that's a good point.

1987
01:38:19,760 --> 01:38:23,050
So the earlier version
that we were looking at it

1988
01:38:23,050 --> 01:38:28,540
was the histogram from Photoshop
and this is luminosity and color

1989
01:38:28,540 --> 01:38:31,270
like a compound
histogram from Lightroom.

1990
01:38:31,270 --> 01:38:34,868
And so you have different options
that you can turn on and turn off.

1991
01:38:34,868 --> 01:38:36,910
And sometimes for certain
images having the color

1992
01:38:36,910 --> 01:38:41,510
on shows you a cast or a skew that might
be in there especially if you did not

1993
01:38:41,510 --> 01:38:43,233
set your color temperature.

1994
01:38:43,233 --> 01:38:44,650
So what does this image look like?

1995
01:38:44,650 --> 01:38:48,500


1996
01:38:48,500 --> 01:38:49,520
AUDIENCE: Bright.

1997
01:38:49,520 --> 01:38:50,240
IAN: Bright.

1998
01:38:50,240 --> 01:38:50,740
Right.

1999
01:38:50,740 --> 01:38:53,170
I think we can safely assume that.

2000
01:38:53,170 --> 01:38:56,980
There's not a lot in here that
is actually a deep dark color--

2001
01:38:56,980 --> 01:38:57,970
dark tone.

2002
01:38:57,970 --> 01:38:59,560
It's mostly whites.

2003
01:38:59,560 --> 01:39:01,180
Bright sky highlights.

2004
01:39:01,180 --> 01:39:03,850
There's some grays in there.

2005
01:39:03,850 --> 01:39:05,950
There's not even really a solid black.

2006
01:39:05,950 --> 01:39:08,890
Maybe a little bit of a
shadow detail in there.

2007
01:39:08,890 --> 01:39:13,150
And so when we look at the
histogram you really can see this.

2008
01:39:13,150 --> 01:39:17,740
And so again we're not
using an even distribution

2009
01:39:17,740 --> 01:39:21,220
but we understand that this
is correct for our subject.

2010
01:39:21,220 --> 01:39:24,030


2011
01:39:24,030 --> 01:39:26,110
Sweet.

2012
01:39:26,110 --> 01:39:27,235
So putting it all together.

2013
01:39:27,235 --> 01:39:30,250


2014
01:39:30,250 --> 01:39:34,360
Here is an image that is overexposed.

2015
01:39:34,360 --> 01:39:38,620
This image is mostly white.

2016
01:39:38,620 --> 01:39:43,840
The values up here are
completely clipping.

2017
01:39:43,840 --> 01:39:50,440
It's solid white, but it's OK
because it's the point of the image.

2018
01:39:50,440 --> 01:39:55,720
So this is an intentionally
overexposed image.

2019
01:39:55,720 --> 01:39:57,670
We could meter this and
the camera might try

2020
01:39:57,670 --> 01:40:01,780
to tell us to make this a middle gray
because it thinks that's what we want,

2021
01:40:01,780 --> 01:40:04,360
but we know we're smarter than it.

2022
01:40:04,360 --> 01:40:08,680
So we're going to increase the exposure
so that pushes up to white and in fact

2023
01:40:08,680 --> 01:40:09,870
clips.

2024
01:40:09,870 --> 01:40:14,230
And we get an interesting
shot of these sunglasses.

2025
01:40:14,230 --> 01:40:17,680
Solid marketing.

2026
01:40:17,680 --> 01:40:27,200
So then the flip side is this
is intentional under exposure

2027
01:40:27,200 --> 01:40:33,220
where we've decided to
not expose for this value,

2028
01:40:33,220 --> 01:40:39,090
but actually for this white value
and give ourselves a silhouette.

2029
01:40:39,090 --> 01:40:40,840
DAN: Do you have a
histogram for this one?

2030
01:40:40,840 --> 01:40:42,700
Like what does this histogram look like?

2031
01:40:42,700 --> 01:40:44,620
IAN: I don't.

2032
01:40:44,620 --> 01:40:47,560
I could get it in a second, but I don't.

2033
01:40:47,560 --> 01:40:49,123
I don't have it.

2034
01:40:49,123 --> 01:40:51,040
DAN: It's split though
right? 'Cause we have--

2035
01:40:51,040 --> 01:40:51,330
IAN: Yeah, it is split.

2036
01:40:51,330 --> 01:40:54,413
DAN: --very little in the middle 'cause
it's almost all largely the bottom

2037
01:40:54,413 --> 01:40:57,640
because the histogram represents
like 100% of the pixels in the image.

2038
01:40:57,640 --> 01:40:59,473
Most of them are in the
dark so that's going

2039
01:40:59,473 --> 01:41:02,380
to be where our biggest mountain
and the highest peaks are.

2040
01:41:02,380 --> 01:41:04,570
And then we'll have also a
big spike up on the right

2041
01:41:04,570 --> 01:41:08,740
because the white screen in that
image was almost fully overexposed.

2042
01:41:08,740 --> 01:41:10,930
So histograms I think--

2043
01:41:10,930 --> 01:41:13,510
most cameras have histograms
on them when you're shooting.

2044
01:41:13,510 --> 01:41:17,740
If you pull up your digital screen and
push the button to pull up the display

2045
01:41:17,740 --> 01:41:21,250
you can cycle through different
overlays on your screen.

2046
01:41:21,250 --> 01:41:23,620
And so histogram is an option,
but I find that they're

2047
01:41:23,620 --> 01:41:25,810
much more useful in post-production.

2048
01:41:25,810 --> 01:41:27,790
When you're actually
shooting I think the thing

2049
01:41:27,790 --> 01:41:30,480
you typically want to look
at is your light meter

2050
01:41:30,480 --> 01:41:34,160
and to know if you're getting
a good exposure or not.

2051
01:41:34,160 --> 01:41:36,580
And obviously you'll
either intentionally

2052
01:41:36,580 --> 01:41:41,367
add exposure over or under, but
at the end of the day histogram

2053
01:41:41,367 --> 01:41:43,450
is helpful when you want
to look at overall trends

2054
01:41:43,450 --> 01:41:44,783
once you get to post-production.

2055
01:41:44,783 --> 01:41:48,820
And really to check if you are clipping
any information at the highlights

2056
01:41:48,820 --> 01:41:49,570
or in the shadows.

2057
01:41:49,570 --> 01:41:52,330
I think that's really where
the histogram is best served.

2058
01:41:52,330 --> 01:41:55,930
IAN: Yeah, and I think
actually, I also tend

2059
01:41:55,930 --> 01:41:59,920
to check the histogram
early when I'm shooting,

2060
01:41:59,920 --> 01:42:03,970
but when I'm pushing exposure in
an image like this or an image

2061
01:42:03,970 --> 01:42:11,080
where it's really bright and I'm going
to maybe push up against overexposure

2062
01:42:11,080 --> 01:42:13,750
or clipping I will look at
the histogram at that moment.

2063
01:42:13,750 --> 01:42:16,990
When I know that I'm compensating
and I want to make an image brighter,

2064
01:42:16,990 --> 01:42:19,660
and I want to push it up
towards that bright white value

2065
01:42:19,660 --> 01:42:22,330
I want to make sure that
I don't clip because I

2066
01:42:22,330 --> 01:42:26,290
can't get that information back
so I want to get as close to it

2067
01:42:26,290 --> 01:42:28,870
while still maintaining
some detail in the image.

2068
01:42:28,870 --> 01:42:31,430


2069
01:42:31,430 --> 01:42:34,900
So it's a really good
tool for when you begin

2070
01:42:34,900 --> 01:42:37,870
to experiment with pushing
and pulling your exposure

2071
01:42:37,870 --> 01:42:41,020
away from what the light
meter is telling you.

2072
01:42:41,020 --> 01:42:44,650
At the end of the day 90% of the scenes
that you photograph the light meter

2073
01:42:44,650 --> 01:42:49,148
is going to do an amazing job at
calculating some calibration for you.

2074
01:42:49,148 --> 01:42:50,690
AUDIENCE: It'll get you really close.

2075
01:42:50,690 --> 01:42:53,900
IAN: Yeah, it'll get you really close,
but you then have to make a decision.

2076
01:42:53,900 --> 01:42:58,122
Do I accept this, or do I
push one way or the other?

2077
01:42:58,122 --> 01:43:01,330
DAN: And the other thing I'll just say,
since we're talking about histograms,

2078
01:43:01,330 --> 01:43:04,837
is the useful thing-- the
indicator that popped up.

2079
01:43:04,837 --> 01:43:07,170
In post, and we haven't covered
Lightroom in this class,

2080
01:43:07,170 --> 01:43:09,545
but if you're using something
like Lightroom or Photoshop

2081
01:43:09,545 --> 01:43:12,780
typically if you hover over the
indicator that you're overexposed

2082
01:43:12,780 --> 01:43:16,720
it'll show you an overlay on the screen
where which portion of your image

2083
01:43:16,720 --> 01:43:17,830
is clipping.

2084
01:43:17,830 --> 01:43:21,190
And it's just helpful to get a read
on exactly which part of the image

2085
01:43:21,190 --> 01:43:23,650
is over or underexposed.

2086
01:43:23,650 --> 01:43:29,980
So I think that's the tail end of
our conversation about exposure.

2087
01:43:29,980 --> 01:43:32,920
I want to stop for a
minute and see if there are

2088
01:43:32,920 --> 01:43:37,390
questions from anyone in the audience.

2089
01:43:37,390 --> 01:43:40,180
What questions do you have?

2090
01:43:40,180 --> 01:43:41,500
AUDIENCE: I have a question.

2091
01:43:41,500 --> 01:43:43,793
IAN: Yeah, go for it.

2092
01:43:43,793 --> 01:43:46,960
AUDIENCE: When you introduced histograms
programs Dan just briefly mentioned

2093
01:43:46,960 --> 01:43:52,770
this, that we can see a
histogram in our viewfinder

2094
01:43:52,770 --> 01:43:55,320
or on the screen in our cameras?

2095
01:43:55,320 --> 01:44:01,050
IAN: You can in a large
number of cameras.

2096
01:44:01,050 --> 01:44:04,500
I think this is now turned off.

2097
01:44:04,500 --> 01:44:07,012


2098
01:44:07,012 --> 01:44:08,470
So let me just fire up this camera.

2099
01:44:08,470 --> 01:44:09,540
We can take a look.

2100
01:44:09,540 --> 01:44:15,690
So this is the output of
the 5D and it's currently

2101
01:44:15,690 --> 01:44:18,630
shooting this bright white wall.

2102
01:44:18,630 --> 01:44:21,800
And you can see that
the histogram is pinned,

2103
01:44:21,800 --> 01:44:27,360
except when I walk in front of
it, right to the right hand side.

2104
01:44:27,360 --> 01:44:31,290
But if we do something like
maybe introduce another tone--

2105
01:44:31,290 --> 01:44:36,210
again, this is over exposed-- we can
see that the histogram starts to move--

2106
01:44:36,210 --> 01:44:37,800
this is a darker tone--

2107
01:44:37,800 --> 01:44:42,480
in real time for what
it is that it's seeing.

2108
01:44:42,480 --> 01:44:44,430
So in this image I would
look at this exposure

2109
01:44:44,430 --> 01:44:48,960
and I would know that if what
I want is a bright white field

2110
01:44:48,960 --> 01:44:52,860
I have successfully exposed this image.

2111
01:44:52,860 --> 01:44:58,490
But if I don't want that--
if I want some kind of detail

2112
01:44:58,490 --> 01:45:03,273
I'm going to adjust some values.

2113
01:45:03,273 --> 01:45:05,190
Like maybe I'll close
the aperture down, which

2114
01:45:05,190 --> 01:45:07,410
you can see I'm doing on the bottom.

2115
01:45:07,410 --> 01:45:11,130
And I'll change the exposure so that
now I don't have a bright white field,

2116
01:45:11,130 --> 01:45:15,510
but I have a middle gray field.

2117
01:45:15,510 --> 01:45:17,460
And I know that it's
rendering as middle gray

2118
01:45:17,460 --> 01:45:20,902
because right in the middle of the
histogram there's this giant peak.

2119
01:45:20,902 --> 01:45:25,820


2120
01:45:25,820 --> 01:45:29,425
I think-- can you push the talk?

2121
01:45:29,425 --> 01:45:30,092
AUDIENCE: Sorry.

2122
01:45:30,092 --> 01:45:30,310
IAN: That's all right.

2123
01:45:30,310 --> 01:45:33,980
AUDIENCE: So I assume you have to be
on a manual mode in order for this--

2124
01:45:33,980 --> 01:45:37,520
to see a histogram and alter
all of these different--

2125
01:45:37,520 --> 01:45:40,760
IAN: It's actually just
one of the info features.

2126
01:45:40,760 --> 01:45:44,510
I can actually turn off
all of that clutter.

2127
01:45:44,510 --> 01:45:46,970
If I press it again I get
it without the histogram.

2128
01:45:46,970 --> 01:45:48,038
I get some extra data.

2129
01:45:48,038 --> 01:45:49,580
And then I can turn the histogram on.

2130
01:45:49,580 --> 01:45:53,810
So your camera may or may
not have this feature.

2131
01:45:53,810 --> 01:45:56,520
Most do at this point.

2132
01:45:56,520 --> 01:46:00,950
So there should be either a way to
turn it on in the menu or a button

2133
01:46:00,950 --> 01:46:04,490
that functions to allow you
to turn this on so you can get

2134
01:46:04,490 --> 01:46:06,350
a sense of what you're exposing for.

2135
01:46:06,350 --> 01:46:10,100
DAN: And Lorna, your camera may
not have it in automatic mode.

2136
01:46:10,100 --> 01:46:12,230
You might need to be in
a different camera mode.

2137
01:46:12,230 --> 01:46:14,910
And we did record a short
video on camera modes

2138
01:46:14,910 --> 01:46:17,720
so check that out after this
lecture if you have more questions.

2139
01:46:17,720 --> 01:46:21,110
And if I can just speak experientially
for a second with histograms.

2140
01:46:21,110 --> 01:46:23,480
Like I said I think when
you're going out to shoot--

2141
01:46:23,480 --> 01:46:25,687
I don't find them all that
useful when I'm shooting,

2142
01:46:25,687 --> 01:46:28,520
but the times that they are useful--
if you're shooting a bright sun

2143
01:46:28,520 --> 01:46:30,228
and can't quite get
a read on your screen

2144
01:46:30,228 --> 01:46:33,770
and you want to know if you're
overexposed on your highlights it's

2145
01:46:33,770 --> 01:46:38,090
really helpful in that moment, but for
the most part when I go out and shoot I

2146
01:46:38,090 --> 01:46:40,280
am using exclusively the camera meter.

2147
01:46:40,280 --> 01:46:40,867
IAN: Yeah.

2148
01:46:40,867 --> 01:46:42,200
And I think that's a good point.

2149
01:46:42,200 --> 01:46:46,400
Just like it's sometimes very hard
to assess your focus on a very

2150
01:46:46,400 --> 01:46:48,320
tiny monitor or through a viewfinder.

2151
01:46:48,320 --> 01:46:52,683
It can be difficult. It can also
be very difficult to see a LCD

2152
01:46:52,683 --> 01:46:55,600
screen in bright sunlight, which is
I think what you're talking about.

2153
01:46:55,600 --> 01:47:00,500
And that the histogram sort of proves
to you what is happening in the image

2154
01:47:00,500 --> 01:47:02,630
because it's not based
on some visual cue.

2155
01:47:02,630 --> 01:47:04,490
It's based on the actual
data in the field.

2156
01:47:04,490 --> 01:47:08,210


2157
01:47:08,210 --> 01:47:11,900
So again there's a video
on light metering more

2158
01:47:11,900 --> 01:47:14,210
generally and how to trick
and fool your light meter,

2159
01:47:14,210 --> 01:47:17,000
and also one on camera modes, which
I would encourage you to watch

2160
01:47:17,000 --> 01:47:19,520
for the next assignment because
there's a couple-- we're

2161
01:47:19,520 --> 01:47:22,880
going to ask you to experiment with
the different elements of exposure.

2162
01:47:22,880 --> 01:47:27,230
And you can use either aperture
priority mode or shutter priority mode

2163
01:47:27,230 --> 01:47:31,010
to help you play with that
as well as full manual

2164
01:47:31,010 --> 01:47:33,230
if you're feeling brave
and adventurous, which

2165
01:47:33,230 --> 01:47:37,010
I encourage you to feel at this point.

2166
01:47:37,010 --> 01:47:38,990
Any final questions or parting thoughts?

2167
01:47:38,990 --> 01:47:41,800


2168
01:47:41,800 --> 01:47:43,550
Well, we'll stick
around for a few minutes

2169
01:47:43,550 --> 01:47:48,370
and let us know, but thank you all
very much and we'll see you next week.

2170
01:47:48,370 --> 01:47:49,705