Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated:
1
00:00:01,806 --> 00:00:04,306
In this clip we're going to review how to analyze a piece of
2
00:00:04,306 --> 00:00:07,806
greenscreen footage before we begin to pull keys.
3
00:00:07,806 --> 00:00:08,234
Now,
4
00:00:08,234 --> 00:00:12,139
it's really important for us to analyze this footage in
5
00:00:12,139 --> 00:00:14,806
order to strategize our workflow in advance.
6
00:00:14,806 --> 00:00:19,805
In this way we can figure out what the most appropriate tools are going to be
7
00:00:19,806 --> 00:00:23,805
and how to use those tools for the most successful outcome.
8
00:00:23,806 --> 00:00:26,377
So a couple things that we want to take a look at,
9
00:00:26,377 --> 00:00:28,805
of course, is the quality of the footage.
10
00:00:28,806 --> 00:00:31,806
So, was it shot cleanly?
11
00:00:31,806 --> 00:00:32,806
Is it crisp?
12
00:00:32,806 --> 00:00:37,806
Was something out of focus that we did not want it to be?
13
00:00:37,806 --> 00:00:40,806
How about the evenness of the greenscreen?
14
00:00:40,806 --> 00:00:42,805
So let's take a look at that.
15
00:00:42,806 --> 00:00:45,806
We can look at it in a couple of different channels and we can already
16
00:00:45,806 --> 00:00:48,805
see that the greenscreen is a little bit uneven.
17
00:00:48,806 --> 00:00:49,806
And this is pretty natural.
18
00:00:49,806 --> 00:00:54,472
This was shot in a green screen room and it's just the difference
19
00:00:54,472 --> 00:00:56,806
between the back wall and the ground plane here.
20
00:00:56,806 --> 00:01:03,169
So it is possible that we may need multiple keys in order to
21
00:01:03,169 --> 00:01:06,805
complete a final alpha channel for the shot.
22
00:01:06,806 --> 00:01:10,806
Now let's take a look at a couple of other issues that we might have.
23
00:01:10,806 --> 00:01:11,139
So,
24
00:01:11,139 --> 00:01:16,694
when we pull this key because of all of that uneven light
25
00:01:16,694 --> 00:01:20,805
in here we might have some artifacting.
26
00:01:20,806 --> 00:01:26,028
The definition of an artifact in visual effects is an undesired or
27
00:01:26,028 --> 00:01:31,431
unintended alteration in the data and is usually introduced by a digital
28
00:01:31,431 --> 00:01:34,806
process and in this case it was through keying.
29
00:01:34,806 --> 00:01:37,805
So we do want to be mindful of that.
30
00:01:37,806 --> 00:01:41,005
Artifacting can also be any detail in the edge work
31
00:01:41,005 --> 00:01:42,805
that we may have issues with.
32
00:01:42,806 --> 00:01:43,234
Now,
33
00:01:43,234 --> 00:01:47,306
this piece of footage was denoised and if you're not familiar with
34
00:01:47,306 --> 00:01:51,905
that process we will be going through it in a few more modules but
35
00:01:51,905 --> 00:01:54,806
for now we're just going to work with this denoised piece of footage
36
00:01:54,806 --> 00:01:57,806
and analyze the rest of the shot.
37
00:01:57,806 --> 00:02:01,806
Now the second thing that we want to take a look at is the reflection.
38
00:02:01,806 --> 00:02:06,606
So we take a look at the original piece of footage again and we see that
39
00:02:06,606 --> 00:02:11,948
there's a strong reflection in the glass over here and over here in the
40
00:02:11,948 --> 00:02:17,306
watch face and we take a look at that alpha we can see that we have almost a
41
00:02:17,306 --> 00:02:23,306
hole in our alpha and another hole over here and the reflection that has
42
00:02:23,306 --> 00:02:25,091
reflected on the table cloth so,
43
00:02:25,091 --> 00:02:28,805
definitely something to be mindful of when you're shooting.
44
00:02:28,806 --> 00:02:35,234
We will have to do some extra work to refine those areas and we can see that,
45
00:02:35,234 --> 00:02:38,806
over here, we've just analyzed it.
46
00:02:38,806 --> 00:02:43,806
So we can see, again, in the glass and in the watch face.
47
00:02:43,806 --> 00:02:45,706
Now the second thing that we want to do is we want to
48
00:02:45,706 --> 00:02:47,806
take a look at the color values.
49
00:02:47,806 --> 00:02:53,806
I've already preselected these colors and if we wanted to analyze these colors,
50
00:02:53,806 --> 00:02:55,806
I'm just going to show you over here,
51
00:02:55,806 --> 00:03:01,306
it is a jpeg versus the original was an exr but if you hold down
52
00:03:01,306 --> 00:03:04,806
Ctrl+Shift+drag you're going to get the numbers down here and that is
53
00:03:04,806 --> 00:03:07,806
the read out for the values on the greenscreen.
54
00:03:07,806 --> 00:03:11,606
If you just want to sample a single pixel hold down
55
00:03:11,606 --> 00:03:14,806
Ctrl+click and it will select a single pixel.
56
00:03:14,806 --> 00:03:18,806
Hold down Ctrl+Shift+click again to get rid of that box.
57
00:03:18,806 --> 00:03:24,806
So in evaluating this we can see that the red values are somewhat similar,
58
00:03:24,806 --> 00:03:29,806
the blue values are somewhat similar but the green are definitely different,
59
00:03:29,806 --> 00:03:30,806
in fact,
60
00:03:30,806 --> 00:03:36,583
in some cases it's 10% different so that shows us again not only visually that
61
00:03:36,583 --> 00:03:40,806
these greens are slightly different but with the numbers and numbers never lie
62
00:03:40,806 --> 00:03:44,806
so it's always great to check in with those numbers.
63
00:03:44,806 --> 00:03:50,006
Let's take a look at a couple of other last things and that is,
64
00:03:50,006 --> 00:03:53,805
of course, where we're going to sample the greenscreen.
65
00:03:53,806 --> 00:03:58,806
Some areas to keep in mind are fine hair detail.
66
00:03:58,806 --> 00:04:04,806
So this will create wispies and or transparencies are in a matte.
67
00:04:04,806 --> 00:04:09,472
These areas down here where maybe there isn't a lot of
68
00:04:09,472 --> 00:04:12,139
greenscreen that's showing and because of that,
69
00:04:12,139 --> 00:04:17,106
some of the edge work will spill over into the greenscreen and
70
00:04:17,106 --> 00:04:19,806
it will be definitely hard to key this area.
71
00:04:19,805 --> 00:04:25,805
We may get some transparency but not 100% transparency like we'd like.
72
00:04:25,806 --> 00:04:28,806
The other area of course is motion blur.
73
00:04:28,806 --> 00:04:29,575
So again,
74
00:04:29,575 --> 00:04:35,139
we want to keep this to be nice and transparent and keying it on a
75
00:04:35,139 --> 00:04:39,605
motion-blurred frame such as this it'll get a sample close to the
76
00:04:39,605 --> 00:04:42,806
element but not actually sampling it,
77
00:04:42,806 --> 00:04:46,806
otherwise we won't get the correct key and so we want to sample on a
78
00:04:46,806 --> 00:04:50,805
motion blur frame versus a non-motion-blurred frame.
79
00:04:50,806 --> 00:04:52,805
So again, in looking at this,
80
00:04:52,806 --> 00:04:57,806
we do have some details that we should pay extra attention to
81
00:04:57,806 --> 00:05:01,806
and we may need multiple keys in order to complete this
82
00:05:01,806 --> 00:05:06,806
particular shot and take it to final.
83
00:05:06,806 --> 00:05:10,806
Now that we've evaluated the greenscreen image or the image sequence,
84
00:05:10,806 --> 00:05:16,806
we can start to begin to determine the most appropriate workflow.
7618
Can't find what you're looking for?
Get subtitles in any language from opensubtitles.com, and translate them here.