Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated:
1
00:00:00,256 --> 00:00:02,304
Hello there and welcome back
2
00:00:02,816 --> 00:00:07,936
What are the time has finally come I've alluded to it for the past 3 weeks but we're finally going to talk
3
00:00:08,192 --> 00:00:09,472
About colour space
4
00:00:09,984 --> 00:00:16,128
And my goal is were talking about this is to understand the practical application of
5
00:00:16,384 --> 00:00:17,408
Colour space because
6
00:00:17,664 --> 00:00:20,224
I can get very confusing and very technical
7
00:00:20,480 --> 00:00:20,992
Quickly
8
00:00:21,504 --> 00:00:25,344
And the goal is to have a practical understanding of it
9
00:00:25,600 --> 00:00:27,392
With a minimal scientific
10
00:00:28,672 --> 00:00:29,440
So let's gets
11
00:00:30,208 --> 00:00:33,792
Now before we talk about how colour space is implemented in nuke
12
00:00:34,048 --> 00:00:37,888
We first need to understand what exactly colour space is for
13
00:00:38,656 --> 00:00:42,752
Well our goal is to have our work myriad
14
00:00:43,008 --> 00:00:45,056
In order to help achieve that
15
00:00:45,312 --> 00:00:49,920
We should be working in an environment that mirrors reality as closely as possible
16
00:00:51,456 --> 00:00:52,992
Linear colour space
17
00:00:53,760 --> 00:00:59,904
What exactly does that mean because nuke works in a linear colour space under the hood that's what everything
18
00:01:00,123 --> 00:01:02,363
Subtitled by
-♪ online-courses.club ♪-
We compress knowledge for you!
https://t.me/joinchat/ailxpXoW3JVjYzQ1
19
00:01:02,364 --> 00:01:00,672
Is built on
20
00:01:01,696 --> 00:01:07,840
Well A linear colour space displays the relative brightness between objects as they exist in the room
21
00:01:09,888 --> 00:01:11,168
If we had two light
22
00:01:11,424 --> 00:01:12,448
Remote Bob
23
00:01:12,704 --> 00:01:15,008
Is twice as bright as like Barbie
24
00:01:15,264 --> 00:01:17,312
We will see the following luminance values
25
00:01:17,568 --> 00:01:19,360
In a linear colour space
26
00:01:19,872 --> 00:01:20,640
They would be
27
00:01:21,152 --> 00:01:22,688
And you would be .4
28
00:01:23,456 --> 00:01:29,600
The footage with a logarithmic colour space and will go into detail about exactly what this means
29
00:01:30,112 --> 00:01:31,392
We would see this Val
30
00:01:31,904 --> 00:01:33,952
Light bulb is that point
31
00:01:34,208 --> 00:01:36,256
Light bulb is that 0.7
32
00:01:37,280 --> 00:01:39,328
It's not like one Bob bryar
33
00:01:39,840 --> 00:01:45,472
Just being displayed in a colour space that isn't one-to-one with the actual intensity of the light
34
00:01:46,240 --> 00:01:47,264
So you may be asking
35
00:01:47,520 --> 00:01:48,544
Why does the smell
36
00:01:49,568 --> 00:01:52,896
Imagine a note was given to you to double the light intensity
37
00:01:53,152 --> 00:01:53,920
A bobby
38
00:01:54,176 --> 00:01:55,968
So that it matches like Bobby
39
00:01:56,480 --> 00:01:58,016
In a linear colour space
40
00:01:58,272 --> 00:02:00,832
That is easy as multiplying my Bobby
41
00:02:02,624 --> 00:02:05,184
Then suddenly in a logarithmic colour space
42
00:02:05,696 --> 00:02:07,744
Our calculation is not so straightforward
43
00:02:08,256 --> 00:02:14,400
In order to match the brightness of Bob we would need to multiply Ball by 1.1
44
00:02:15,424 --> 00:02:21,056
The strange number is required because our colour-space does not accurately match the real world
45
00:02:21,824 --> 00:02:24,896
We doubled our brightness with a seemingly random multiply
46
00:02:25,408 --> 00:02:28,480
And you can imagine how difficult managing exposures would be
47
00:02:28,736 --> 00:02:30,528
If we didn't have light up a door
48
00:02:30,784 --> 00:02:36,416
In this image you can see how a linear colour space properly matches are cameras exposure
49
00:02:37,440 --> 00:02:38,976
As we go from left to right
50
00:02:39,232 --> 00:02:43,584
Are camera is having the amount of light that it's letting into our lens
51
00:02:44,096 --> 00:02:50,240
The black number represents A linear colour space and the green number represents of logarithmic colour space
52
00:02:50,496 --> 00:02:53,056
And you can see that as we move from left to right
53
00:02:53,312 --> 00:02:58,432
The highlights in the darts in Ireland your colour space or perfectly have
54
00:02:59,200 --> 00:03:02,272
In the highlights in a logarithmic colour space
55
00:03:02,784 --> 00:03:04,064
It's been reduced by
56
00:03:04,320 --> 00:03:05,600
057
57
00:03:05,856 --> 00:03:12,000
In our Shadows in a logarithmic colour space the number of areas heavily in the first panel it's going
58
00:03:12,256 --> 00:03:15,328
.06 down in the next 1 x .07
59
00:03:15,584 --> 00:03:18,656
0.12
60
00:03:19,168 --> 00:03:25,312
And so as you can imagine those to make it very difficult to just double the brightness in a giver
61
00:03:26,592 --> 00:03:31,456
And imagine that we had multiple different cameras that were combined in footage from
62
00:03:31,968 --> 00:03:34,784
They are all going to have different colour space curves
63
00:03:35,296 --> 00:03:41,440
And you can imagine how difficult it would be to get the rolling off of the Shadows to look correct but
64
00:03:41,696 --> 00:03:44,000
Different cameras because they're Kershaw also did
65
00:03:44,512 --> 00:03:49,888
And that's another one of the reasons we want to convert Oliver elements into a linear colour space
66
00:03:50,144 --> 00:03:53,728
Bring them all in line with one another so any changes we may
67
00:03:53,984 --> 00:03:56,800
Are uniformly applied across your foot
68
00:03:57,312 --> 00:04:03,200
Singapore say we wanted to double our brightness we would just multiply all other elements by 2
69
00:04:03,968 --> 00:04:10,112
If we didn't convert our elements and I stayed in a logarithmic colour space we would need to figure out that way
70
00:04:10,368 --> 00:04:14,976
Multiplier for each other elements in order to technically double or brightness
71
00:04:15,232 --> 00:04:18,303
This is also directly related to C&G can pass
72
00:04:18,559 --> 00:04:24,447
Because 99.9999 9% of the time when you receive a CG exr
73
00:04:24,703 --> 00:04:27,007
It will be done in all in your colour space
74
00:04:27,263 --> 00:04:33,407
And unifying that with your footage is very important in it's going to be so much harder to get a CV
75
00:04:33,663 --> 00:04:34,175
You're real
76
00:04:34,431 --> 00:04:37,503
If a car is not properly set up in when you're colour
77
00:04:38,271 --> 00:04:42,367
The next I'd like to go over the three main types of colour spaces
78
00:04:42,879 --> 00:04:44,159
There is linear
79
00:04:44,415 --> 00:04:45,439
Logarithmic
80
00:04:45,951 --> 00:04:47,743
And Dewar aka video
81
00:04:47,999 --> 00:04:49,023
Colour spaces
82
00:04:49,791 --> 00:04:51,327
As we discussed before
83
00:04:51,583 --> 00:04:53,119
A linear colour space
84
00:04:53,375 --> 00:04:55,423
Imitates real world brightness
85
00:04:56,191 --> 00:05:01,311
In that it displays the relative brightness between the objects as they exist in Riyadh
86
00:05:02,335 --> 00:05:05,151
It should be the working space for every compositor using
87
00:05:05,407 --> 00:05:07,967
If you want to look at an image in its linear form
88
00:05:08,223 --> 00:05:10,527
It's actually a pretty ugly thing to look
89
00:05:11,295 --> 00:05:12,063
Don't worry
90
00:05:12,319 --> 00:05:13,599
Will just using this to work
91
00:05:14,111 --> 00:05:16,927
And you're not using it for the final viewing of April
92
00:05:17,695 --> 00:05:18,975
In a graph at the bottom
93
00:05:19,231 --> 00:05:22,815
You can see that on our x-axis we have the real world
94
00:05:23,071 --> 00:05:25,631
As it increasing in Dublin
95
00:05:25,887 --> 00:05:30,239
Footage RGB value is also uniformly increasing
96
00:05:30,751 --> 00:05:32,031
Giving us a link
97
00:05:32,287 --> 00:05:33,055
Colour space
98
00:05:33,567 --> 00:05:37,407
And one of the key components of a linear colour space is that it's floating point
99
00:05:37,663 --> 00:05:39,967
We can have values that are brighter than
100
00:05:40,223 --> 00:05:42,783
Yellow colour spaces generally Kent do
101
00:05:43,551 --> 00:05:45,343
So now let's talk about the
102
00:05:46,623 --> 00:05:49,183
So I wonder if the colour space AKA
103
00:05:49,439 --> 00:05:55,583
A large pillar space is used by high-end cameras to retain the most amount of data pass
104
00:05:56,351 --> 00:05:59,679
They do this by compressing the luminance values very tightly
105
00:06:00,191 --> 00:06:02,495
Typically between the values of .1
106
00:06:02,751 --> 00:06:04,543
Appoint 7 if you were the same
107
00:06:05,311 --> 00:06:11,455
This makes the relative luminance data not accurate to the real world when they recording with these cameras are trying to get them
108
00:06:11,711 --> 00:06:13,759
The man of dynamic range in their images pass
109
00:06:14,015 --> 00:06:16,319
So the flattening all these values
110
00:06:16,575 --> 00:06:18,367
And because of these flat values
111
00:06:18,623 --> 00:06:21,695
A login page will appear very washed out and
112
00:06:21,951 --> 00:06:22,975
As we can see here
113
00:06:23,743 --> 00:06:29,887
There are many different types of log colour spaces there is the Alexa V3 like sea there's a slug
114
00:06:30,143 --> 00:06:31,935
Dragon Wars Canon lore
115
00:06:32,191 --> 00:06:37,055
There is a slight variations between these colour spaces but they all follow the same basic print
116
00:06:38,079 --> 00:06:42,175
You can press the luminance that are used to retain as much detail as possible
117
00:06:42,943 --> 00:06:46,271
Auto you can retain large amounts of data while recording to log
118
00:06:46,527 --> 00:06:48,831
Thanks to High bit depth and cameras
119
00:06:49,087 --> 00:06:52,159
You are still limited to range between 0 and
120
00:06:52,927 --> 00:06:54,975
Anything beyond these values will
121
00:06:55,999 --> 00:06:56,767
And that's why
122
00:06:57,023 --> 00:07:00,607
The Kerb is so flat at the top to retain the brightest Bright Sparks
123
00:07:01,887 --> 00:07:05,983
Now the third type of colour space is when I called the Dior video colour
124
00:07:06,239 --> 00:07:11,871
This is the classification of colour spaces I've kind of come up with because it's a very Broad
125
00:07:12,639 --> 00:07:13,919
Ultimately
126
00:07:14,175 --> 00:07:16,223
If your colour is space isn't in
127
00:07:16,479 --> 00:07:19,039
When your colour space or a log colour space
128
00:07:19,295 --> 00:07:21,343
It's most likely in a viewer colour
129
00:07:22,367 --> 00:07:26,463
This colour space represents footage that has been artistically colour
130
00:07:27,231 --> 00:07:29,791
It is not based on real-world value
131
00:07:30,047 --> 00:07:32,095
Rather what looks good to the eye
132
00:07:32,863 --> 00:07:35,167
When a colour has been working on a log burner
133
00:07:35,423 --> 00:07:40,031
Here she will commonly first apply in srgb or rec 709
134
00:07:40,287 --> 00:07:44,127
To have a baseline colour space conversion to make it visually appealing
135
00:07:45,151 --> 00:07:51,039
What's we have added to a shot that up the weeks follow the exact same principles by making off but it's more
136
00:07:52,319 --> 00:07:58,463
You can see visually this image looks the best of all the places we used so far but it's in 10
137
00:07:58,719 --> 00:07:59,231
To be viewed
138
00:07:59,487 --> 00:08:00,511
Not to be worked
139
00:08:01,023 --> 00:08:03,583
Because this also does not accurately represent
140
00:08:03,839 --> 00:08:05,375
The increasing and brightness
141
00:08:07,423 --> 00:08:12,031
Typically viewer colour spaces will not have that much data in there not floating point
142
00:08:12,287 --> 00:08:18,431
And they are typically the hardest colour space to convert to when you're because you don't know exactly what
143
00:08:18,687 --> 00:08:19,711
The colourist did
144
00:08:20,223 --> 00:08:26,367
In Newport working in a linear colour space or what were the viewing is a view or light on top of that
145
00:08:26,623 --> 00:08:28,671
So we can see what the final result will look like
146
00:08:28,927 --> 00:08:29,695
When it's cut
147
00:08:29,951 --> 00:08:31,999
Are the times of beer colour space
148
00:08:32,255 --> 00:08:38,399
Still running to claim being issues where you lose values because the colour of skull is it to retain data it's to make the image look
149
00:08:38,655 --> 00:08:39,423
Visually appealing
150
00:08:40,959 --> 00:08:44,543
So now let's start talking about how this applies to New
151
00:08:44,799 --> 00:08:50,943
I'm working in nuke are typically multiple color space conversions occurring and it's important to know exactly what
152
00:08:51,199 --> 00:08:51,711
Going on
153
00:08:52,223 --> 00:08:53,759
At the most basic level
154
00:08:54,015 --> 00:08:56,319
3 colour space conversions happy
155
00:08:56,831 --> 00:08:59,391
We have our read note input conversion
156
00:08:59,647 --> 00:09:02,719
The goal of this is to linearise our footage
157
00:09:03,487 --> 00:09:05,023
We have our right now
158
00:09:05,279 --> 00:09:06,559
Output conversion
159
00:09:06,815 --> 00:09:10,143
Which is going to return or footage to its original colour space
160
00:09:10,655 --> 00:09:12,447
And then we have our Buick invoice
161
00:09:12,959 --> 00:09:14,751
And this is the light on top
162
00:09:15,007 --> 00:09:19,103
Farm in your colour space so we can view it in an aesthetically pleasing
163
00:09:19,615 --> 00:09:22,431
So what's happened to nuke and see this in action
164
00:09:23,455 --> 00:09:25,759
So I brought week to spillage into
165
00:09:26,015 --> 00:09:27,295
To use as an example
166
00:09:27,807 --> 00:09:32,671
So what I want to do first is turn off all the colour space conversions that noob is currently
167
00:09:33,183 --> 00:09:34,207
Because there's two
168
00:09:34,719 --> 00:09:36,255
We have an input conversion
169
00:09:36,767 --> 00:09:38,559
Which has been very net from Alexa
170
00:09:39,071 --> 00:09:40,863
And we have a viewer converter
171
00:09:41,375 --> 00:09:43,679
Which is up here interviewer as srgb
172
00:09:44,447 --> 00:09:46,239
Sunday Turner input conversion
173
00:09:46,495 --> 00:09:50,591
We can set the Sterling here which means I want no colour space
174
00:09:50,847 --> 00:09:51,359
Used to happen
175
00:09:52,127 --> 00:09:52,895
And viewer
176
00:09:53,151 --> 00:09:55,199
We can set a colour space to none
177
00:09:56,223 --> 00:09:58,271
Now what we have is all washed out
178
00:09:58,783 --> 00:09:59,551
Log in
179
00:10:00,345 --> 00:10:02,248
Subtitled by
-♪ online-courses.club ♪-
We compress knowledge for you!
https://t.me/joinchat/ailxpXoW3JVjYzQ1
180
00:10:02,250 --> 00:10:01,343
Because that's what this was shot
181
00:10:02,111 --> 00:10:06,207
If you were to open this in any other program this is what the colour wood
182
00:10:06,719 --> 00:10:08,255
Nice and flight and washed out
183
00:10:09,023 --> 00:10:10,559
And so let's explicitly
184
00:10:10,815 --> 00:10:12,607
Input colour converter
185
00:10:13,119 --> 00:10:13,631
Not
186
00:10:13,887 --> 00:10:15,679
Using the colour safe
187
00:10:16,447 --> 00:10:18,495
I'm going to make a colour space node
188
00:10:21,055 --> 00:10:21,823
And you'll see
189
00:10:22,079 --> 00:10:25,407
We have our familiar list of colour space options here
190
00:10:25,919 --> 00:10:32,063
And ignore this d65 and srgb really you don't need any of these mobs except for the
191
00:10:32,319 --> 00:10:33,087
In and out tab
192
00:10:33,855 --> 00:10:35,391
And so if you look at the strap
193
00:10:35,647 --> 00:10:38,719
Either all of the default colour space conversions
194
00:10:38,975 --> 00:10:43,583
This list is the exact same as the list we have in our reader mode
195
00:10:43,839 --> 00:10:45,887
And your nearest the list is the same
196
00:10:46,911 --> 00:10:47,679
In r
197
00:10:47,935 --> 00:10:48,447
And help
198
00:10:49,215 --> 00:10:55,103
And now what I'm about to tell you is the most important concept to grasp about colour space in
199
00:10:55,871 --> 00:10:57,919
Each colour space auction
200
00:10:58,431 --> 00:10:59,711
Has two distinct
201
00:10:59,967 --> 00:11:00,735
Stuart
202
00:11:01,503 --> 00:11:02,783
It has an input Curve
203
00:11:03,295 --> 00:11:04,063
In an hour
204
00:11:05,343 --> 00:11:08,159
Sophie set Alexa to our in
205
00:11:08,671 --> 00:11:10,975
Were using input Alexa
206
00:11:12,255 --> 00:11:13,535
Change that back to one year
207
00:11:14,047 --> 00:11:15,071
If you're selling
208
00:11:15,327 --> 00:11:16,607
When using r output
209
00:11:17,631 --> 00:11:21,215
Very different things happened to our footage
210
00:11:21,471 --> 00:11:24,031
That's looking a quick recap of what this looks like on a car
211
00:11:24,799 --> 00:11:27,871
As you can see on the input conversion on her left
212
00:11:28,383 --> 00:11:31,711
It's the perfect opposite of the output conversion on her way
213
00:11:32,479 --> 00:11:35,551
And so if you were to apply both of these to your image
214
00:11:35,807 --> 00:11:36,575
The result
215
00:11:36,831 --> 00:11:38,623
Will it be no colour
216
00:11:40,159 --> 00:11:41,951
And so where we want to work in you
217
00:11:42,463 --> 00:11:43,743
Is right in the middle
218
00:11:44,255 --> 00:11:46,815
This is our linear colour space
219
00:11:48,351 --> 00:11:49,631
Let's go back into the
220
00:11:50,143 --> 00:11:53,727
I have my colour space known here I'm going to change the statue on here
221
00:11:54,239 --> 00:11:56,543
And I'm going to sat or input
222
00:11:56,799 --> 00:11:57,567
To Alexa
223
00:11:58,591 --> 00:11:59,871
Copy and paste
224
00:12:00,383 --> 00:12:01,663
And I have a look at here
225
00:12:01,919 --> 00:12:04,735
And I will change this colour space to Alexa as well
226
00:12:05,247 --> 00:12:06,783
Because the read node
227
00:12:07,039 --> 00:12:12,159
Colour space tab is looking for the input Curve of the selected colour space
228
00:12:12,415 --> 00:12:14,463
Selfie a b between these two
229
00:12:15,743 --> 00:12:18,815
There is absolutely no difference because the same exam
230
00:12:19,071 --> 00:12:20,095
Colour space Curve
231
00:12:20,351 --> 00:12:21,375
Is being up
232
00:12:23,423 --> 00:12:29,567
Converting your colour space like this is pretty unorthodox but it's still technically correct you could read all your footage
233
00:12:29,823 --> 00:12:33,407
And use colour space notes to do explicit transformation
234
00:12:34,687 --> 00:12:36,479
So what exactly is happening
235
00:12:36,735 --> 00:12:39,039
Whenever using input colour space
236
00:12:39,807 --> 00:12:41,343
The simplest explanation
237
00:12:41,599 --> 00:12:44,927
Is that were expecting a colour space that has a known Curve
238
00:12:45,183 --> 00:12:48,255
Emma telling Luke to x r footage
239
00:12:48,511 --> 00:12:51,583
With that inverted Curve give us a linear result
240
00:12:52,351 --> 00:12:54,911
That's why selecting the right colour space is
241
00:12:55,423 --> 00:12:58,495
Because they don't give you the most accurate linearisation pass
242
00:12:59,007 --> 00:13:01,567
Now let's take a look at the output Theory
243
00:13:02,079 --> 00:13:02,847
Of our
244
00:13:03,615 --> 00:13:04,383
Colour space
245
00:13:05,151 --> 00:13:06,687
So if we were to make
246
00:13:07,455 --> 00:13:11,295
Another colour space known for the sake of keeping this nice and organised
247
00:13:12,063 --> 00:13:13,343
What does coldest
248
00:13:13,599 --> 00:13:14,623
Input conversion
249
00:13:19,999 --> 00:13:24,863
Now if we wanted to die when your eyes are footage we would use the output flavour
250
00:13:25,119 --> 00:13:26,399
Of the Alexa
251
00:13:27,167 --> 00:13:28,959
And you're the one we turn it off
252
00:13:29,215 --> 00:13:32,031
There is 0 shift that happens in our food
253
00:13:32,799 --> 00:13:34,335
Label this output
254
00:13:34,591 --> 00:13:36,127
Conversion very quickly
255
00:13:36,895 --> 00:13:38,943
And so this our conversion
256
00:13:39,199 --> 00:13:40,735
Is what happens when you use
257
00:13:43,039 --> 00:13:46,367
When you decide to write this out and you still like a colour space
258
00:13:46,879 --> 00:13:48,671
It's using the output flavour
259
00:13:48,927 --> 00:13:50,207
Of your colour space
260
00:13:51,231 --> 00:13:53,279
It's Tricky when you're looking at this in New York
261
00:13:53,535 --> 00:13:57,375
Because you can actually see the colour space conversion that happens
262
00:13:57,631 --> 00:13:59,423
It's a current on x
263
00:14:00,703 --> 00:14:04,031
But this is the best way to visualise it with a colour space
264
00:14:05,055 --> 00:14:05,823
If you like
265
00:14:06,079 --> 00:14:06,847
You could
266
00:14:07,103 --> 00:14:08,639
Copy this output conversion
267
00:14:09,407 --> 00:14:10,175
Throw it here
268
00:14:10,943 --> 00:14:11,455
Candice
269
00:14:11,967 --> 00:14:13,247
Linearizing
270
00:14:14,271 --> 00:14:15,295
Make a right note
271
00:14:15,551 --> 00:14:17,343
And write it out in linear
272
00:14:18,111 --> 00:14:21,951
Again Lenny or means there's going to be no color space transformation happy
273
00:14:22,463 --> 00:14:23,487
So when you write this out
274
00:14:23,743 --> 00:14:25,023
What Racine in now
275
00:14:25,279 --> 00:14:27,071
Is exactly what we will get
276
00:14:28,095 --> 00:14:29,887
And so I'm going to delete this really quick
277
00:14:30,399 --> 00:14:36,543
And now maybe it makes a little more sense by the general rule of thumb is to have your read node input
278
00:14:36,799 --> 00:14:38,335
Space match your
279
00:14:38,591 --> 00:14:39,615
Write note
280
00:14:39,871 --> 00:14:42,175
Because they're going to cancel each other out
281
00:14:42,431 --> 00:14:43,967
And that colour shift your phone
282
00:14:44,735 --> 00:14:48,063
Now that we understand how input and output conversions will happen
283
00:14:48,319 --> 00:14:51,135
What's take a look at the final conversion that you
284
00:14:51,391 --> 00:14:52,159
Typically what we doing
285
00:14:52,415 --> 00:14:53,951
And that's are viewer
286
00:14:54,207 --> 00:14:54,719
Colour space
287
00:14:55,999 --> 00:14:57,535
So we bring in
288
00:14:58,047 --> 00:15:03,167
Forage and we'll linearize it like this it doesn't look good but this is technically correct
289
00:15:03,679 --> 00:15:06,495
If we like this one is reaching at 70
290
00:15:07,007 --> 00:15:08,799
Which is exactly what we want
291
00:15:09,311 --> 00:15:12,127
And you can see that nice little fall off of the light
292
00:15:12,639 --> 00:15:15,711
Because it's saying this left side is 16 times brighter
293
00:15:16,223 --> 00:15:17,503
Actually even more than that
294
00:15:18,015 --> 00:15:22,879
Tennis near right side and that kind of Luminous disparity is exactly what your
295
00:15:24,159 --> 00:15:26,975
If we didn't do a colour conversion on this
296
00:15:27,743 --> 00:15:31,071
Watch this is reading around 7
297
00:15:31,327 --> 00:15:33,631
And the right side is it .2 for
298
00:15:34,399 --> 00:15:40,031
And that is not even remotely close to the actual brightness difference on the left and right side of this part
299
00:15:41,311 --> 00:15:43,103
So we have an image in the right
300
00:15:43,359 --> 00:15:43,871
Space
301
00:15:44,383 --> 00:15:45,407
And by default
302
00:15:45,663 --> 00:15:47,711
Nuke applies of euro
303
00:15:48,735 --> 00:15:50,015
Srgb
304
00:15:51,039 --> 00:15:54,111
We talked a little bit about having a standard srgb
305
00:15:54,367 --> 00:15:56,159
Conversion isn't as good as ASUS
306
00:15:56,415 --> 00:15:57,951
For the sake of the 16th
307
00:15:58,207 --> 00:16:01,535
It's fine because this image doesn't look too bad - be
308
00:16:01,791 --> 00:16:04,863
In so what does colour spaces do in
309
00:16:05,119 --> 00:16:08,447
Is it taking the output Curve of the selected
310
00:16:08,703 --> 00:16:11,519
Colour space and applying it on of your own
311
00:16:12,031 --> 00:16:12,799
And you can see
312
00:16:13,055 --> 00:16:15,103
It's not actually changing the voice
313
00:16:15,359 --> 00:16:17,663
What's the sample on this ball
314
00:16:17,919 --> 00:16:20,479
We can see how bright this is we change
315
00:16:20,991 --> 00:16:26,879
Colour space for viewer that are RGB values are not changing
316
00:16:27,135 --> 00:16:28,159
And that's because
317
00:16:28,415 --> 00:16:33,791
This Villa is a post process it's not actually affecting or pixels under the
318
00:16:34,303 --> 00:16:37,631
It's just giving us something a little more visually pleasing to look
319
00:16:38,655 --> 00:16:44,799
And ideally you want this for your process to match your final colour environment as closely as possible
320
00:16:45,311 --> 00:16:49,151
And so we can Minack what is going on in vivo process
321
00:16:49,663 --> 00:16:51,455
What's at the Stena Line and turn it off
322
00:16:52,223 --> 00:16:54,271
Make another colour space node
323
00:16:58,367 --> 00:16:59,135
Oh my god
324
00:16:59,647 --> 00:17:01,183
And if we set are out
325
00:17:01,951 --> 00:17:02,719
2 srgb
326
00:17:03,487 --> 00:17:05,279
This is the exact same thing
327
00:17:05,791 --> 00:17:07,839
Is setting a reviewer to srt
328
00:17:08,607 --> 00:17:10,655
It's using the output Curve
329
00:17:10,911 --> 00:17:12,191
When you have a colour space
330
00:17:13,983 --> 00:17:17,311
What's take another look at this but will use aces this time
331
00:17:18,847 --> 00:17:20,895
We can go ahead and change your setting
332
00:17:21,151 --> 00:17:21,919
2 aces
333
00:17:23,711 --> 00:17:24,991
We're all Emmerdale
334
00:17:25,247 --> 00:17:26,015
You can set
335
00:17:26,527 --> 00:17:28,319
Colour space to raw
336
00:17:28,575 --> 00:17:32,159
What is easiest way of saying don't do any colour space conversions
337
00:17:32,927 --> 00:17:33,951
And we was set
338
00:17:34,207 --> 00:17:35,487
Or image here
339
00:17:35,999 --> 00:17:37,535
2 aces CG
340
00:17:37,791 --> 00:17:40,351
Which is the same thing as Mania
341
00:17:41,119 --> 00:17:44,703
And we'll see this looks exactly the same as when we had it in Luke's default
342
00:17:44,959 --> 00:17:45,471
Space
343
00:17:45,983 --> 00:17:50,591
And now we need to use aces waiver of color space transformation notes
344
00:17:50,847 --> 00:17:52,383
Witcher ocio colour
345
00:17:55,711 --> 00:17:56,991
And so again
346
00:17:57,247 --> 00:17:59,807
To make this in Alexa colour space
347
00:18:00,063 --> 00:18:01,087
Washington at 2
348
00:18:01,343 --> 00:18:02,879
Harry
349
00:18:03,903 --> 00:18:05,183
And now this is love
350
00:18:05,439 --> 00:18:07,231
And now we can go to work
351
00:18:07,487 --> 00:18:09,023
Srgb
352
00:18:09,279 --> 00:18:13,375
And there is a colour space you can flip through these to see which of
353
00:18:13,631 --> 00:18:14,655
Looks nice
354
00:18:17,727 --> 00:18:18,239
You typically
355
00:18:19,007 --> 00:18:21,055
And then what changes to rot again
356
00:18:21,567 --> 00:18:23,615
Make another ocio now
357
00:18:27,711 --> 00:18:30,271
Now if you want to mimic or viewer colour space
358
00:18:30,527 --> 00:18:32,831
We can grab any of these output curves
359
00:18:34,111 --> 00:18:34,879
And applied
360
00:18:35,647 --> 00:18:37,695
And then this is exactly what we were seeing
361
00:18:37,951 --> 00:18:38,719
Interview
362
00:18:39,487 --> 00:18:41,791
And there's something I would like to make very clear
363
00:18:42,303 --> 00:18:46,143
Even though this category of colour spaces is called output
364
00:18:46,911 --> 00:18:51,263
Each of these colour spaces has a distinct input and output
365
00:18:52,031 --> 00:18:55,615
The naming convention I think is a little more confusing than it needs to be
366
00:18:55,871 --> 00:18:59,711
Just keep that in mind that is really two curves for these output colour space
367
00:19:00,479 --> 00:19:02,527
Go back to talking about your process
368
00:19:03,295 --> 00:19:08,159
Disadvantage of grading of your process this way is that as recycling
369
00:19:08,415 --> 00:19:11,231
This actually does change the relative value
370
00:19:11,743 --> 00:19:13,023
Go ahead and Singapore
371
00:19:13,279 --> 00:19:16,351
In this area again it's Rena pt99
372
00:19:16,863 --> 00:19:17,631
Turn this off
373
00:19:18,655 --> 00:19:20,191
It's back up to 70
374
00:19:21,727 --> 00:19:24,031
Best to rec709 interviewer
375
00:19:24,799 --> 00:19:25,567
It doesn't
376
00:19:26,335 --> 00:19:28,639
And so that's why you don't generally work
377
00:19:28,895 --> 00:19:29,663
Like this
378
00:19:30,431 --> 00:19:34,527
And remember we made that for your process in weeks 1 and 2
379
00:19:35,039 --> 00:19:37,599
Your head and do that really quick
380
00:19:38,623 --> 00:19:40,159
With a colour space node
381
00:19:40,415 --> 00:19:42,463
We'll go to our view our settings
382
00:19:43,487 --> 00:19:46,047
If you can't retrieve your settings just make sure you clear
383
00:19:46,303 --> 00:19:49,119
All of your properties that ass in the viewer
384
00:19:49,887 --> 00:19:51,167
We will copy the name
385
00:19:52,191 --> 00:19:52,703
Paste
386
00:19:52,959 --> 00:19:53,471
Here
387
00:19:56,287 --> 00:19:56,799
It's very well
388
00:19:57,311 --> 00:19:58,079
Okapi
389
00:19:59,103 --> 00:19:59,871
Is the name
390
00:20:00,639 --> 00:20:01,919
Now we have argument
391
00:20:02,687 --> 00:20:03,967
We can change this
392
00:20:04,735 --> 00:20:05,247
2
393
00:20:05,759 --> 00:20:06,527
Rec 709
394
00:20:08,063 --> 00:20:09,343
Turn this one off
395
00:20:10,111 --> 00:20:11,903
And you'll see this also does not
396
00:20:12,159 --> 00:20:14,207
Change the value because it's a poo
397
00:20:14,463 --> 00:20:15,743
That is happy
398
00:20:16,255 --> 00:20:20,351
So you can make all the work colour-space adjustments you want but your underlying values
399
00:20:20,607 --> 00:20:21,887
I'm using these two
400
00:20:22,143 --> 00:20:22,911
Will not
401
00:20:24,959 --> 00:20:26,495
I'm going to delete this
402
00:20:27,519 --> 00:20:28,031
And
403
00:20:28,287 --> 00:20:28,799
Keep the cell
404
00:20:29,055 --> 00:20:30,079
Srgb
405
00:20:31,615 --> 00:20:33,407
Actually let's change this to rar
406
00:20:33,663 --> 00:20:36,479
I want to talk about something that upsets
407
00:20:37,503 --> 00:20:38,527
And that is
408
00:20:39,039 --> 00:20:39,807
Effective
409
00:20:41,087 --> 00:20:41,855
Mouse learning
410
00:20:42,367 --> 00:20:44,415
How to apply colour space
411
00:20:45,183 --> 00:20:47,743
This node really threw me through a loop
412
00:20:48,255 --> 00:20:49,279
Because you would think
413
00:20:49,535 --> 00:20:50,559
Colour space in
414
00:20:50,815 --> 00:20:52,863
Would it be applying input
415
00:20:54,143 --> 00:20:54,655
Would be
416
00:20:54,911 --> 00:20:55,679
Sign and helper
417
00:20:56,447 --> 00:20:57,471
Actually
418
00:20:57,727 --> 00:20:58,239
Exact
419
00:20:58,495 --> 00:20:59,007
Opposite
420
00:21:00,031 --> 00:21:02,079
I'm going to go back to standard
421
00:21:02,335 --> 00:21:03,359
Colour space for this
422
00:21:08,479 --> 00:21:08,991
Go back
423
00:21:14,623 --> 00:21:15,647
Wallpaper
424
00:21:15,903 --> 00:21:16,928
Filled in here
425
00:21:18,464 --> 00:21:20,000
Or grab or lip from weed
426
00:21:24,096 --> 00:21:26,144
And this looks correct this is well lyrics
427
00:21:27,424 --> 00:21:29,728
So let's start doing some
428
00:21:30,240 --> 00:21:31,520
Collarspace adjustment
429
00:21:31,776 --> 00:21:33,568
Like we would be doing if you were composite
430
00:21:35,872 --> 00:21:37,664
So the input Curve is Alexa
431
00:21:38,176 --> 00:21:40,224
And now we having your eyes
432
00:21:41,760 --> 00:21:43,040
Now in a vector file
433
00:21:43,552 --> 00:21:44,832
This one that be applied
434
00:21:45,088 --> 00:21:46,880
Correctly because it's expecting
435
00:21:47,136 --> 00:21:48,416
Lenient
436
00:21:49,952 --> 00:21:53,024
So you can actually use this colour space in
437
00:21:53,280 --> 00:21:54,048
To apply
438
00:21:54,304 --> 00:21:54,816
The out
439
00:21:55,072 --> 00:21:57,120
Flavour of our Alexa in here
440
00:21:57,376 --> 00:21:58,400
So that it looks
441
00:21:59,168 --> 00:22:01,472
So despite be named the same thing
442
00:22:01,728 --> 00:22:03,776
The colour space in Tab in your
443
00:22:04,032 --> 00:22:08,128
Sheffield does the opposite of the intab in the colour space
444
00:22:08,640 --> 00:22:10,176
The colour space in knob
445
00:22:10,432 --> 00:22:11,456
In vector field
446
00:22:11,712 --> 00:22:12,992
What apply in
447
00:22:13,760 --> 00:22:16,320
Really close space out tab in the vector field
448
00:22:16,576 --> 00:22:17,344
What apply
449
00:22:17,600 --> 00:22:18,624
An input Curve
450
00:22:19,392 --> 00:22:21,952
Sofia to change this to srgb
451
00:22:23,232 --> 00:22:25,792
That we haven't srgb input applied
452
00:22:26,560 --> 00:22:27,840
And remember or viewer
453
00:22:28,096 --> 00:22:29,120
Help
454
00:22:29,632 --> 00:22:31,424
We change this to srgb
455
00:22:31,680 --> 00:22:34,752
They cancel each other out and give us our desired
456
00:22:35,264 --> 00:22:37,568
This looks exactly as we would expect
457
00:22:38,336 --> 00:22:39,104
With Charlotte
458
00:22:40,384 --> 00:22:42,944
This is very confusing and it just
459
00:22:43,456 --> 00:22:49,088
Really bothers me that it's not more explicit it makes sense if you're looking at just the vector field now
460
00:22:49,344 --> 00:22:51,648
But when you look at it in the context of nuke
461
00:22:51,904 --> 00:22:54,208
Name in at the exact same thing as another name
462
00:22:54,464 --> 00:22:55,744
Having to do the opposite
463
00:22:56,256 --> 00:22:57,280
Really doesn't make any
464
00:22:58,304 --> 00:23:01,376
And that will be the end of my vector Fields colour split screen
465
00:23:02,400 --> 00:23:05,472
And so now that we know how exactly colour space works in you
466
00:23:05,728 --> 00:23:08,032
Let's talk about how we select Cars
467
00:23:08,288 --> 00:23:12,640
Spacers your ability to pick an accurate colour space is entirely dependant
468
00:23:12,896 --> 00:23:14,688
On what you know about your foot
469
00:23:15,456 --> 00:23:19,040
We're going to go over the 4 questions I ask when I'm trying to do
470
00:23:19,296 --> 00:23:20,576
What colour space to use
471
00:23:21,600 --> 00:23:25,952
Is there any medi at that explicitly says what colour space or gamemaker
472
00:23:26,208 --> 00:23:27,488
Was used on the foot
473
00:23:29,024 --> 00:23:29,536
2
474
00:23:30,304 --> 00:23:35,424
Do we have any set reports are written documentation on what colour space was used for your cam
475
00:23:36,960 --> 00:23:37,472
3
476
00:23:37,984 --> 00:23:41,056
Do we know at least what brand of camera this was shot
477
00:23:41,568 --> 00:23:43,872
Limit our search to colour space Curve
478
00:23:44,128 --> 00:23:45,408
Specific to that camera
479
00:23:46,688 --> 00:23:48,736
And then if you don't know any of the above
480
00:23:48,992 --> 00:23:52,064
You need to identify the current category
481
00:23:52,320 --> 00:23:56,160
Colour space that is in and do your best to approx
482
00:23:56,928 --> 00:24:03,072
If you don't know any of this information it can be very difficult in potentially impossible to discover the correct colours
483
00:24:03,584 --> 00:24:08,960
A lot of furniture that is my colour corrected it really is impossible to properly clean your eyes
484
00:24:09,472 --> 00:24:13,312
And that's one of the reasons that's stuck but it can be very difficult to let me rise
485
00:24:13,568 --> 00:24:14,080
Because
486
00:24:14,336 --> 00:24:17,664
There's not really any way of knowing how the colour has been
487
00:24:19,200 --> 00:24:24,320
Let's see this in action in talk about the colour space that I shot for this class and how I feel
488
00:24:25,344 --> 00:24:28,160
Cottage for the first 4 weeks of this class was shot
489
00:24:28,416 --> 00:24:29,440
Black magic
490
00:24:30,720 --> 00:24:35,328
This camera has its own flavour of a log Curve that uses black magic
491
00:24:35,840 --> 00:24:37,376
An Asus really only has
492
00:24:37,632 --> 00:24:39,680
Hi and cinema cameras to choose
493
00:24:39,936 --> 00:24:40,448
Input
494
00:24:41,472 --> 00:24:42,496
So what do you do
495
00:24:43,264 --> 00:24:47,616
What you need to find a curve that closely approximates the one that you're using
496
00:24:48,384 --> 00:24:49,152
In this instance
497
00:24:49,408 --> 00:24:54,784
The blackmagic log Curve is very similar to the area Alexa log CV3
498
00:24:56,320 --> 00:24:59,904
And so I ended up using that to get the most accurate linearization
499
00:25:00,160 --> 00:25:01,952
Possible with an Asus
500
00:25:02,464 --> 00:25:04,768
It's definitely not a perfect one to one match
501
00:25:05,280 --> 00:25:07,840
But I know that it is very close
502
00:25:08,096 --> 00:25:08,864
To an Alexa
503
00:25:09,120 --> 00:25:11,680
And it's definitely within the same category of War
504
00:25:12,192 --> 00:25:14,496
And so it was good and ok too
505
00:25:15,264 --> 00:25:18,336
What's talk about another type of camera that's commonly used
506
00:25:18,592 --> 00:25:21,152
That it may be a little tricky to find the colour space
507
00:25:21,408 --> 00:25:22,432
I'm talking about
508
00:25:22,688 --> 00:25:23,712
Otters
509
00:25:24,224 --> 00:25:30,368
Dslrs modern ones do you have some log curves but a lot of them don't they just
510
00:25:30,624 --> 00:25:34,720
Standard rec709 those different hacks for dslrs to get curry
511
00:25:35,232 --> 00:25:37,280
But I can be very difficult to figure out
512
00:25:37,536 --> 00:25:39,328
What exactly you should be you
513
00:25:40,096 --> 00:25:41,120
If you're using the
514
00:25:41,376 --> 00:25:42,144
Default
515
00:25:43,424 --> 00:25:47,264
From your DSLR ads are it's an srgb or x 709
516
00:25:47,776 --> 00:25:49,056
And that's how you shouldn't
517
00:25:49,312 --> 00:25:50,336
Colour space
518
00:25:50,592 --> 00:25:55,968
The trick with the srgb my 1.8 all of these kind of viewer
519
00:25:56,224 --> 00:25:57,248
Public spaces
520
00:25:57,504 --> 00:26:02,880
Is a typically don't let your eyes very well and I don't give you a large range of luminance value
521
00:26:03,136 --> 00:26:03,648
To play with
522
00:26:03,904 --> 00:26:06,720
And that's just a casualty of working with a lower
523
00:26:08,000 --> 00:26:09,280
And unfortunately
524
00:26:09,536 --> 00:26:15,680
This goes for a lot of sacrilege as well anything that you get that visually appealing that it's been colour corrected
525
00:26:16,192 --> 00:26:16,704
Good
526
00:26:16,960 --> 00:26:20,288
You're probably have to use one of these if you were colour spaces
527
00:26:20,544 --> 00:26:21,056
The transfer
528
00:26:21,824 --> 00:26:26,176
And that's why the general brought is if you don't know the colour space and its been coloured
529
00:26:26,688 --> 00:26:27,712
Use a sword
530
00:26:28,736 --> 00:26:30,016
That's why in week 2
531
00:26:30,272 --> 00:26:31,552
Use srgb
532
00:26:31,808 --> 00:26:33,344
For that stock photographs
533
00:26:33,600 --> 00:26:35,136
And then week 3
534
00:26:35,392 --> 00:26:36,160
Freddie Starr
535
00:26:36,416 --> 00:26:42,560
Video we use that was also srgb because I really wasn't anyway to tell what camera it was
536
00:26:42,816 --> 00:26:44,608
Looks like it has been coloured in
537
00:26:45,120 --> 00:26:46,144
So we use dove
538
00:26:48,448 --> 00:26:51,008
When you using CGI elements as exr
539
00:26:51,264 --> 00:26:57,152
Like we said earlier they were almost always be linear or 99.9999 9% of the time
540
00:26:57,408 --> 00:27:03,552
They're going to be linear and that's a pretty good rule of thumb for x ours if you don't know what colour space
541
00:27:03,808 --> 00:27:04,320
They should be
542
00:27:04,576 --> 00:27:06,112
It's probably linear
543
00:27:07,392 --> 00:27:10,208
And now is where I do have to make a confession
544
00:27:10,976 --> 00:27:12,000
In week 1
545
00:27:12,768 --> 00:27:14,048
We should not have used
546
00:27:14,304 --> 00:27:15,840
The srgb colour space
547
00:27:16,352 --> 00:27:17,376
Technically
548
00:27:17,632 --> 00:27:19,168
Snot linearizing
549
00:27:20,960 --> 00:27:24,800
I did this mainly because we were just getting started in
550
00:27:25,312 --> 00:27:29,408
And I didn't want to even begin to dip my toe into colour space
551
00:27:29,664 --> 00:27:33,248
Because of its complexity and how obstructed is I wanted us to
552
00:27:33,504 --> 00:27:39,648
Compositing and if we would have had it in Alexa the three like see we would have had a play with our colour because srg
553
00:27:39,904 --> 00:27:42,464
You were colour space doesn't show the highlights very well
554
00:27:42,976 --> 00:27:45,280
There are a lot of reasons that I didn't want to use
555
00:27:45,536 --> 00:27:46,816
Since we were just getting
556
00:27:47,584 --> 00:27:49,888
And since we were really effect
557
00:27:50,144 --> 00:27:52,704
Colour and brightness to match in the image
558
00:27:52,960 --> 00:27:55,264
I felt that it was OK to cheat
559
00:27:55,520 --> 00:27:58,592
Well if we were starting that shot now with the knowledge we have
560
00:27:58,848 --> 00:28:01,152
We would have read it in as Alexa
561
00:28:01,664 --> 00:28:07,552
Most likely in aces and then beauty with aces srgb or cases rec709
562
00:28:08,320 --> 00:28:09,600
And now that I have that
563
00:28:09,856 --> 00:28:11,648
Half of my chest I just want to recap
564
00:28:11,904 --> 00:28:14,720
Why we use colour space and that was a lot of information
565
00:28:15,232 --> 00:28:17,536
I think this is the hardest
566
00:28:18,048 --> 00:28:18,560
Concept
567
00:28:18,816 --> 00:28:19,328
Under
568
00:28:20,864 --> 00:28:21,888
And it is
569
00:28:22,912 --> 00:28:26,240
Because it will bring or object into the real world in terms
570
00:28:26,752 --> 00:28:31,360
And it's going to unify Oliver elements and how they react
571
00:28:31,616 --> 00:28:32,128
The changes
572
00:28:34,176 --> 00:28:36,224
You know you've messed up your colour space
573
00:28:36,480 --> 00:28:37,760
When you increase your game
574
00:28:38,016 --> 00:28:41,088
Appoint one and then suddenly your son
575
00:28:41,344 --> 00:28:42,624
Is 20 x 2
576
00:28:43,648 --> 00:28:44,160
What is
577
00:28:44,416 --> 00:28:47,488
That may seem like an over exaggeration but I've seen it
578
00:28:48,256 --> 00:28:50,304
And it's why colour space is so
579
00:28:51,328 --> 00:28:56,448
So we're here we're happy and thank you very much for sticking with me through this very long very video
580
00:28:56,960 --> 00:28:58,240
And I'll catch you guys in and
43348
Can't find what you're looking for?
Get subtitles in any language from opensubtitles.com, and translate them here.