Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated:
1
00:00:00,256 --> 00:00:02,048
Hello and welcome back
2
00:00:02,304 --> 00:00:05,888
In this video we take a look at our final composite in fundamental
3
00:00:06,144 --> 00:00:06,912
Perspective
4
00:00:07,168 --> 00:00:13,312
It is typically the hardest compositing fundamental to follow because it can be very tricky to identify
5
00:00:13,568 --> 00:00:15,872
That's wrong especially if you need a composite name
6
00:00:16,128 --> 00:00:22,272
Where can I inform you here on exactly what you should be looking for when your matching perspective to sit you on the
7
00:00:23,040 --> 00:00:24,064
So let's get star
8
00:00:25,088 --> 00:00:30,208
Now prospective is all about how an object or element spatially sits within us
9
00:00:30,720 --> 00:00:34,048
He needs optically make sense within its environment
10
00:00:34,560 --> 00:00:40,704
Lot of the time and were talking about perspective it's affecting more complex shops where we are replacing a back
11
00:00:40,960 --> 00:00:41,728
Entirely
12
00:00:41,984 --> 00:00:43,264
Integrating CG
13
00:00:43,520 --> 00:00:44,800
Shots of that nature
14
00:00:45,568 --> 00:00:51,712
Another prospective of your footage is determined by three main factors and that is the focal length
15
00:00:52,736 --> 00:00:58,880
The cameras sensor size and the positioning rotation of your camera if you have this
16
00:00:59,136 --> 00:01:02,701
Information for all the elements in a short it will make matching your perspective significantly easier because you'll be
17
00:01:02,703 --> 00:01:04,567
Subtitled by
-♪ online-courses.club ♪-
We compress knowledge for you!
https://t.me/joinchat/ailxpXoW3JVjYzQ1
18
00:01:05,536 --> 00:01:08,864
Mathematically determine how to position yourself
19
00:01:09,120 --> 00:01:11,424
Look at that later in this video
20
00:01:12,960 --> 00:01:16,032
Do these three factors all have to do with perspective
21
00:01:17,824 --> 00:01:23,968
Let's take a look at the two main characteristics that we need to match in order to have or correct perspective
22
00:01:24,480 --> 00:01:29,088
The focal length and sensor size will give us or lens compression
23
00:01:29,600 --> 00:01:34,208
And then the camera positioning and rotation will give us our vanishing
24
00:01:34,976 --> 00:01:37,792
If we have these two characteristics match
25
00:01:38,304 --> 00:01:40,096
We will have the correct prospect
26
00:01:40,352 --> 00:01:41,120
Within our shop
27
00:01:41,632 --> 00:01:46,240
So let's take a look at these two more in-depth and figure out exactly how we can find them
28
00:01:47,008 --> 00:01:51,616
Lens compression is about how compressed elements appear within your shop
29
00:01:52,128 --> 00:01:54,432
You can see in the skiff how focal length
30
00:01:54,688 --> 00:01:56,224
Effects lens compression
31
00:01:56,736 --> 00:02:01,600
Why do your focal length is the list compressed objects near you will appear
32
00:02:02,112 --> 00:02:04,160
And then the longer that you're focal like this
33
00:02:04,416 --> 00:02:06,976
Aka your lens be more telephoto
34
00:02:07,232 --> 00:02:09,024
The more compressed your objects
35
00:02:10,048 --> 00:02:11,584
Well why is this
36
00:02:11,840 --> 00:02:13,888
What's take a look at an example in 3D
37
00:02:14,656 --> 00:02:19,008
Here on the left we have two cubes been mugged at by a 20 mm camera
38
00:02:19,520 --> 00:02:23,872
And on the right we have the same 2 cubes being looked at by a 200 mm
39
00:02:25,152 --> 00:02:27,456
Why is the compression different
40
00:02:28,224 --> 00:02:31,552
1 order to compensate for the increased focal
41
00:02:32,064 --> 00:02:37,696
We had to significantly back or camera up in order to achieve the same relative Framing
42
00:02:38,464 --> 00:02:44,608
What this means is that we have a significant disparity in the relative distances between or cubes
43
00:02:44,864 --> 00:02:45,376
Camera
44
00:02:45,888 --> 00:02:47,680
In a 20 mm example
45
00:02:47,936 --> 00:02:52,800
A background cube is about 2.5 times further away than the foreground cube
46
00:02:53,312 --> 00:02:55,360
Where in the 200 mm example
47
00:02:55,616 --> 00:03:00,992
Background Cubism a d 1.2 1.15 times further away
48
00:03:01,248 --> 00:03:02,016
NR4
49
00:03:03,040 --> 00:03:09,184
And this large discrepancy in our distance is actually the reason that the objects appear more can't
50
00:03:09,952 --> 00:03:14,304
The fog that self technically doesn't have anything to do with one compression
51
00:03:14,560 --> 00:03:20,704
Rather the focal lengths forces you to move a certain distance away in order to properly photograph
52
00:03:20,960 --> 00:03:24,288
Subject and the distance is what creates compression
53
00:03:24,800 --> 00:03:30,944
So I guess you could say fog leave does cause lens compression but it only does so indirectly by forcing
54
00:03:31,200 --> 00:03:32,736
Between you and your
55
00:03:33,504 --> 00:03:39,392
What's prove this really quickly in this Nick script I have a 20mm conduit in a 200mm
56
00:03:39,904 --> 00:03:45,280
These cameras are in the same position in are exactly the same except for the focal
57
00:03:46,304 --> 00:03:51,168
Now if I wanted to have our 20mm render match or 200mm render
58
00:03:51,424 --> 00:03:53,728
I need to scale up 10 times
59
00:03:54,496 --> 00:03:58,080
You know this woman Transformers up that is exactly the same
60
00:03:58,848 --> 00:04:04,480
That means that if we have footage that was shot on the exact same camera with different focal lengths
61
00:04:04,736 --> 00:04:09,856
We can automatically know how much to scale a background in order to match or lens compression
62
00:04:10,368 --> 00:04:13,184
The formula is to take your foreground focal line
63
00:04:13,440 --> 00:04:15,744
And / your background focal length
64
00:04:16,000 --> 00:04:19,327
And that will give you the amount you need to scale your background by
65
00:04:19,839 --> 00:04:25,983
For example if we had an 80 mm foreground in a 40 mm background we would do 80 /
66
00:04:26,751 --> 00:04:29,055
And that would give us a 2 time scale
67
00:04:29,567 --> 00:04:31,615
My plug lead into a transformer
68
00:04:31,871 --> 00:04:32,383
We would
69
00:04:32,639 --> 00:04:34,431
Perfectly match Orleans compression
70
00:04:35,199 --> 00:04:38,271
This is the ideal way to work and combine elements
71
00:04:38,783 --> 00:04:44,927
Unfortunately it's not always that simple if you're using multiple cameras because that means you're also using
72
00:04:45,183 --> 00:04:46,463
Portable sensor sizes
73
00:04:46,719 --> 00:04:49,023
And like I said earlier or lens compression
74
00:04:49,279 --> 00:04:51,327
Is caused by our focal length
75
00:04:51,839 --> 00:04:54,399
And or sensor size again
76
00:04:54,911 --> 00:05:01,055
Decompression is indirectly caused by these but it's caused by them none the less so what does the census
77
00:05:01,311 --> 00:05:02,335
Let's have to do with this
78
00:05:02,591 --> 00:05:05,407
What's a good look at this graph in front of us
79
00:05:05,663 --> 00:05:08,991
The circle represents everything and lands could possibly see
80
00:05:09,503 --> 00:05:14,367
The boxes within the lines represents the sensor size and what we can actually record
81
00:05:14,879 --> 00:05:18,463
You're not the blue sensor size has a much larger field of view
82
00:05:18,719 --> 00:05:20,255
Then I read sensor size
83
00:05:20,511 --> 00:05:23,583
Despite having the exact same lens
84
00:05:24,607 --> 00:05:29,215
So the field of view is the total amount of area that a camera can see it's surroundings
85
00:05:29,471 --> 00:05:32,031
Based on the lens in sensor size
86
00:05:32,287 --> 00:05:38,431
End goal for Ireland's compassion is to have the same field of view for both of our objects
87
00:05:38,943 --> 00:05:44,063
It's a look at how that affects our formula when we have different sensor sizes on
88
00:05:45,855 --> 00:05:51,487
You'll see that it's not as simple as just dividing or focal length are foreground by the focal length of a background
89
00:05:51,743 --> 00:05:57,887
Now we need to find the percentage difference of a foreground sensor size compared to a background
90
00:05:58,143 --> 00:06:01,215
And multiply or background focal length by
91
00:06:01,727 --> 00:06:07,871
This may be a little confusing so let's take a look at an example say we have a 4 grand plate with 150mm
92
00:06:08,127 --> 00:06:10,943
25mm horizontal sensor size
93
00:06:11,199 --> 00:06:17,343
And then we have a 30mm background lens with a 20 mm horizontal sensor size
94
00:06:17,599 --> 00:06:23,231
Once we pipe fell into a formula the result is we need to scale or background play Up by 4
95
00:06:23,999 --> 00:06:30,143
I don't really expect people to have this memorise but I do write it down because even I occasionally forget
96
00:06:30,399 --> 00:06:31,423
What word goes in
97
00:06:31,679 --> 00:06:37,823
And if I have this info on the plates I'm using I definitely type it in because of the easiest start
98
00:06:38,079 --> 00:06:43,199
White to get your perspective looking correct it automatically make sure your compression is career
99
00:06:44,223 --> 00:06:49,599
The next characteristic we need to mash is our vanishing Point flash Horizon line
100
00:06:49,855 --> 00:06:54,975
We need to be able to see the exact point where all of the angles in Arsene or converging
101
00:06:55,231 --> 00:07:01,375
Because all of our elements will need to match these converging lines in point is determined by the
102
00:07:01,631 --> 00:07:04,703
Tilt of our camera 1 footage was shot
103
00:07:05,215 --> 00:07:09,567
Depending on your seen this may be difficult or very easy to determine
104
00:07:09,823 --> 00:07:15,199
What's take a look at an image of an officer in walkthrough how exactly we can find the vanish
105
00:07:15,711 --> 00:07:20,063
Image open in nuclear and that's where and how we can find our vanished
106
00:07:21,087 --> 00:07:22,879
Really quick
107
00:07:23,391 --> 00:07:27,743
I'm just going to put those two are GBX I want to actually draw some bits for wind
108
00:07:28,255 --> 00:07:30,815
And to do that we need to use the open spline
109
00:07:31,839 --> 00:07:37,727
And now that we can do is we're going to find the lines that we believe in real life are parallel with one another
110
00:07:37,983 --> 00:07:42,335
I'm going to start here and roll one very more online
111
00:07:42,847 --> 00:07:46,175
I need to make sure that it lines up all the way
112
00:07:47,455 --> 00:07:53,087
And this is a little tricky to do is just one of you were and so we can actually make bots for the risen
113
00:07:53,343 --> 00:07:57,183
And we can do that by hitting the strap down kind of windows and typing in
114
00:07:57,439 --> 00:07:58,719
New comp viewer
115
00:07:59,231 --> 00:08:03,071
Then we can go ahead and drag it to the right over here and split or view
116
00:08:03,583 --> 00:08:06,143
So this is a little bit easier to adjust
117
00:08:06,911 --> 00:08:10,751
You can see we have our second even out here and you can see what it is connected to
118
00:08:11,007 --> 00:08:12,543
I'm going to zoom in here
119
00:08:13,823 --> 00:08:18,687
Just turn off this right of shape so that we don't see it cover image
120
00:08:18,943 --> 00:08:21,759
Change the width of our stroke to 2
121
00:08:22,783 --> 00:08:26,623
And on this for you I'm just going to drag or end point here
122
00:08:29,183 --> 00:08:31,743
Until we are all the way around
123
00:08:33,279 --> 00:08:34,047
Down
124
00:08:37,887 --> 00:08:38,655
You want to get there
125
00:08:39,935 --> 00:08:41,983
Right on the money here
126
00:08:43,775 --> 00:08:46,079
I'll wait till that is our first mine
127
00:08:46,335 --> 00:08:48,639
I know what I want to do is I'm just going to duplicate this
128
00:08:48,895 --> 00:08:51,967
What's go ahead and find another line
129
00:08:52,223 --> 00:08:54,015
This task here
130
00:08:55,039 --> 00:08:58,879
Now I'm going to you again break this wine until it's
131
00:08:59,391 --> 00:09:01,951
In line with third desk entirely
132
00:09:04,511 --> 00:09:07,071
Resume and we can make this nice and precise
133
00:09:09,887 --> 00:09:11,679
Something like that alright
134
00:09:11,935 --> 00:09:16,031
Starting to get a rough approximation of where are vanishing
135
00:09:17,055 --> 00:09:18,847
Is going to do three or four of these
136
00:09:19,103 --> 00:09:21,151
Duplicated again and
137
00:09:21,407 --> 00:09:26,015
To see if this works and put it right at our current intersection point and move this
138
00:09:26,527 --> 00:09:27,039
Here
139
00:09:30,623 --> 00:09:31,391
Closely
140
00:09:32,415 --> 00:09:34,975
Let's see how well this Lines
141
00:09:38,047 --> 00:09:43,423
Tiny little adjustments italics actually the same that's alright because there is one distortion
142
00:09:43,679 --> 00:09:44,959
Is a few facts
143
00:09:45,983 --> 00:09:47,519
Go ahead and
144
00:09:47,775 --> 00:09:48,543
Grab a
145
00:09:49,055 --> 00:09:49,567
Point
146
00:09:53,663 --> 00:09:54,431
This again
147
00:10:00,345 --> 00:10:03,912
Subtitled by
-♪ online-courses.club ♪-
We compress knowledge for you!
https://t.me/joinchat/ailxpXoW3JVjYzQ1
148
00:10:03,913 --> 00:10:05,183
Is that in line
149
00:10:06,463 --> 00:10:09,279
Megane what does going to Linus tapas bar
150
00:10:12,351 --> 00:10:13,631
A little bit
151
00:10:17,471 --> 00:10:20,287
Great so you can see we have some consistency
152
00:10:20,543 --> 00:10:24,127
And we're all of our lines are going approximately going
153
00:10:24,895 --> 00:10:26,175
Yeah I can see that
154
00:10:26,431 --> 00:10:28,223
Prudential e a l let's see it
155
00:10:30,783 --> 00:10:33,855
It does mean or vanishing point is approximately
156
00:10:35,135 --> 00:10:35,903
What does that mean
157
00:10:36,671 --> 00:10:41,535
Well we can draw line where these are in the setting I'm going to go right here it seems to be in the middle
158
00:10:42,815 --> 00:10:47,935
And what we can do now is we can click and drag this line and any other
159
00:10:48,447 --> 00:10:53,567
Horizontal lines that go to the standing shin point or going to line up we can see it's in line
160
00:10:55,359 --> 00:10:57,151
With this table
161
00:10:57,663 --> 00:11:01,503
With the boys on a swallow with the TV as we move this around
162
00:11:01,759 --> 00:11:04,063
If this is a lining up with your services
163
00:11:04,319 --> 00:11:06,879
That means you have a pretty accurate vanishing
164
00:11:07,391 --> 00:11:09,695
And this is exactly what we were
165
00:11:13,279 --> 00:11:13,791
Great
166
00:11:14,303 --> 00:11:17,887
Now there are two directions that this is currently receding
167
00:11:18,143 --> 00:11:22,751
You're going to want to follow this process again except for the right side this time
168
00:11:23,263 --> 00:11:24,031
Go ahead
169
00:11:25,055 --> 00:11:27,103
Make another landline
170
00:11:28,383 --> 00:11:30,431
Human hair so I can see what's going on
171
00:11:34,527 --> 00:11:35,295
Who
172
00:11:35,807 --> 00:11:37,599
Now we can grab this portion
173
00:11:38,623 --> 00:11:40,159
Scan
174
00:11:43,487 --> 00:11:44,255
Over
175
00:11:47,327 --> 00:11:51,679
It's a little bit tedious but this process is 100% worth it
176
00:11:51,935 --> 00:11:54,239
Can I help you align your seen so much
177
00:11:55,519 --> 00:11:57,823
It's worth the five to 10 minutes
178
00:11:58,079 --> 00:11:59,615
Text to set it up properly
179
00:11:59,871 --> 00:12:03,711
Call so we have that approximate intersection what's good for the ceiling now
180
00:12:04,223 --> 00:12:04,991
Peter Sagan
181
00:12:06,783 --> 00:12:07,295
Here
182
00:12:10,879 --> 00:12:13,439
And let's see how we can get this July
183
00:12:13,695 --> 00:12:14,207
Up
184
00:12:15,743 --> 00:12:16,767
Lower
185
00:12:20,095 --> 00:12:21,887
Even a little lower
186
00:12:22,143 --> 00:12:23,935
Literally finicky here now
187
00:12:24,703 --> 00:12:30,847
This is actually an excellent vanishing Point Deezer all lining up songs go ahead and just test that
188
00:12:31,871 --> 00:12:33,407
We will duplicate this
189
00:12:38,015 --> 00:12:41,599
How much has Tesla ever seen that lines up that lines up
190
00:12:42,367 --> 00:12:44,415
Checking period that lines up
191
00:12:44,927 --> 00:12:46,207
So excellent
192
00:12:46,463 --> 00:12:47,487
Very good very
193
00:12:47,999 --> 00:12:49,279
What's that extra 1
194
00:12:49,535 --> 00:12:51,327
In turn this Rolo shape on
195
00:12:51,839 --> 00:12:54,399
And now we can see all the vanishing
196
00:12:54,911 --> 00:12:56,191
Wine that we've made
197
00:12:56,447 --> 00:13:02,591
Let's go ahead and check or background play because this isn't the original we had in there and see her purpose
198
00:13:02,847 --> 00:13:04,127
Is stacking up
199
00:13:04,895 --> 00:13:08,223
We can go ahead and duplicate one of these again
200
00:13:10,015 --> 00:13:11,295
And we can chat
201
00:13:11,551 --> 00:13:13,087
Best perspective
202
00:13:15,391 --> 00:13:16,415
Lock screen last year
203
00:13:16,671 --> 00:13:17,951
And you can see
204
00:13:18,463 --> 00:13:20,767
The lines are not
205
00:13:21,279 --> 00:13:22,047
Lining up
206
00:13:22,303 --> 00:13:23,839
A man is not what we want
207
00:13:24,095 --> 00:13:26,399
Because assuming that this is built like a grid
208
00:13:26,655 --> 00:13:30,495
This should be square and flush with our building your
209
00:13:30,751 --> 00:13:32,543
So that is not correct
210
00:13:33,055 --> 00:13:35,871
We can also check out our other lines right here
211
00:13:36,383 --> 00:13:37,151
Just going to
212
00:13:37,407 --> 00:13:39,711
Grab the sand move around
213
00:13:41,247 --> 00:13:42,527
These ones actually
214
00:13:43,039 --> 00:13:46,111
Too bad this is kind of lucky that
215
00:13:46,367 --> 00:13:48,415
This is winding up as well as it is
216
00:13:48,671 --> 00:13:53,279
But the issue remains that not over vanishing lines are in line with one another
217
00:13:53,791 --> 00:13:55,583
So this doesn't quite feel
218
00:13:55,839 --> 00:13:56,351
Correct
219
00:13:57,375 --> 00:14:01,215
And so I've actually drawn finishing lines already
220
00:14:01,727 --> 00:14:03,007
For this background
221
00:14:03,775 --> 00:14:08,127
You can see it lines up we just had the exact same process that we did
222
00:14:09,151 --> 00:14:09,663
And
223
00:14:09,919 --> 00:14:12,735
Are foreground image all of these are lined up
224
00:14:13,247 --> 00:14:16,063
We have our two distinct vanishing points
225
00:14:16,319 --> 00:14:20,927
And so now what we need to do is line or pairs of vanishing points up with one hand
226
00:14:22,463 --> 00:14:24,511
Arbroath shape here
227
00:14:25,023 --> 00:14:28,607
And what I'm going to do is just roughly drawabox
228
00:14:28,863 --> 00:14:31,679
Where are the centre of our vanishing lines are
229
00:14:31,935 --> 00:14:34,751
And I'm going to select or background please
230
00:14:37,567 --> 00:14:39,359
And move our anchor
231
00:14:40,127 --> 00:14:41,919
Into the vanishing point
232
00:14:42,943 --> 00:14:43,711
And let's
233
00:14:43,967 --> 00:14:46,015
Go ahead and move this right
234
00:14:47,807 --> 00:14:49,087
On top of
235
00:14:52,415 --> 00:14:56,255
Let's see what this has done for us with another perspective check
236
00:14:57,535 --> 00:14:59,583
You can check some lines and here now
237
00:15:01,887 --> 00:15:03,423
And now we can see
238
00:15:03,935 --> 00:15:07,519
This is nice and mine that anywhere that we check it
239
00:15:08,799 --> 00:15:11,871
Which is excellent this is exactly what we want
240
00:15:14,431 --> 00:15:18,015
And we can go ahead and see if this is working for our other line
241
00:15:18,527 --> 00:15:20,063
McGregor vs here
242
00:15:20,575 --> 00:15:22,367
Start investigating
243
00:15:22,623 --> 00:15:23,647
A little bit
244
00:15:25,439 --> 00:15:26,975
How to tell exactly
245
00:15:27,487 --> 00:15:28,511
What is
246
00:15:29,535 --> 00:15:33,375
Here but you can see there's a little bit of a discrepancy here
247
00:15:33,631 --> 00:15:36,447
This is why scaling by eyes so difficult
248
00:15:36,959 --> 00:15:38,751
When your trying to align your backgrounds
249
00:15:39,007 --> 00:15:41,311
It doesn't quite a line with her other finishing
250
00:15:41,823 --> 00:15:43,103
Weekend again
251
00:15:43,359 --> 00:15:45,407
Follow the exact same process
252
00:15:46,431 --> 00:15:46,943
Weekend
253
00:15:47,199 --> 00:15:49,503
Look at the vanishing lights here in a cycle
254
00:15:50,015 --> 00:15:52,831
We need to scale this app to measure of location
255
00:15:53,343 --> 00:15:56,927
What's go ahead and start scalene and we can rotate
256
00:15:57,695 --> 00:15:59,743
Are vanishing point a little bit
257
00:16:00,255 --> 00:16:02,815
And now they're right on top of you
258
00:16:04,863 --> 00:16:07,423
So let's take a look
259
00:16:07,679 --> 00:16:08,447
The Snail
260
00:16:09,727 --> 00:16:12,799
And let's check all of our lines again
261
00:16:13,567 --> 00:16:15,871
So now that's very accurate
262
00:16:16,639 --> 00:16:19,199
Horizontal building is accurate
263
00:16:19,455 --> 00:16:20,479
Everywhere that we check
264
00:16:20,991 --> 00:16:22,783
Perfectly aligns with a reindeer
265
00:16:23,807 --> 00:16:26,367
So now if I turn the stereo off
266
00:16:28,671 --> 00:16:29,951
And I just look at
267
00:16:30,207 --> 00:16:33,535
Final result only get rid of that second viewer
268
00:16:35,327 --> 00:16:39,423
All right so this is probably at there still one more issue
269
00:16:39,679 --> 00:16:42,239
Because we have are receding Lines
270
00:16:42,495 --> 00:16:43,775
All nice and mind
271
00:16:45,055 --> 00:16:48,383
We still have our vertical lines you can see here
272
00:16:48,639 --> 00:16:52,735
Tilting out words and this doesn't quite line-up
273
00:16:52,991 --> 00:16:55,295
With are edges here they should be perfectly
274
00:16:55,551 --> 00:16:58,111
Vertical along with our sins
275
00:16:58,367 --> 00:17:01,439
There will be a little bit of bowing of the lens that takes place
276
00:17:01,695 --> 00:17:03,231
So what do I do to fix that
277
00:17:03,743 --> 00:17:05,791
Is I make a corner pin
278
00:17:10,655 --> 00:17:12,447
And what I'll do is I'll turn it off
279
00:17:12,959 --> 00:17:13,983
And let's move
280
00:17:14,239 --> 00:17:17,311
These points into the corner of our window
281
00:17:19,103 --> 00:17:20,383
Just like this
282
00:17:20,895 --> 00:17:25,503
And then I will copy the two points just a snapper from there
283
00:17:25,759 --> 00:17:28,319
And now I turn this on is going to be no effect on her image
284
00:17:28,575 --> 00:17:29,855
Which is exactly what
285
00:17:30,623 --> 00:17:36,767
So now I just want to skip these to the left or to the right to a line or verticals so I'm going to grab
286
00:17:37,023 --> 00:17:38,047
Q4
287
00:17:39,071 --> 00:17:40,863
And just screwed it to the right
288
00:17:41,119 --> 00:17:42,143
A little bit
289
00:17:43,423 --> 00:17:44,191
I'm here
290
00:17:45,983 --> 00:17:51,103
Mrs make sure our verticals are nice and consistent this entire time
291
00:17:51,359 --> 00:17:53,407
Looking at the stash line
292
00:17:53,919 --> 00:17:55,967
Like a slave gone a little too far
293
00:17:58,271 --> 00:18:01,599
Just like that then on a right side here
294
00:18:01,855 --> 00:18:02,879
Chequer building
295
00:18:03,135 --> 00:18:06,719
Using a corner train because it's not consistent throughout the entire frame
296
00:18:06,975 --> 00:18:08,511
So grab r23
297
00:18:09,279 --> 00:18:11,327
And to the left a little bit
298
00:18:12,863 --> 00:18:15,167
That's nice and lined up
299
00:18:17,215 --> 00:18:20,543
And Derry girl let's go ahead and turn it off and on
300
00:18:21,311 --> 00:18:23,103
The Adjustment is saddle
301
00:18:23,359 --> 00:18:27,711
Stacks of subtle differences make a huge impact on the final
302
00:18:27,967 --> 00:18:30,783
That's got a head and full screen SSE
303
00:18:31,295 --> 00:18:32,575
This is very words
304
00:18:32,831 --> 00:18:34,623
Dropping the image in
305
00:18:34,879 --> 00:18:37,183
And this is with our perspective aligned
306
00:18:37,695 --> 00:18:43,071
Now off about this may not have not too bad but clearly when your ab with our align perspective
307
00:18:43,327 --> 00:18:45,119
This fits with an art scene
308
00:18:45,375 --> 00:18:47,423
Much much better
309
00:18:47,679 --> 00:18:48,703
Van Basten
310
00:18:49,215 --> 00:18:53,055
The perspective is all about the subtleties and getting them to align
311
00:18:53,311 --> 00:18:59,455
And we've not only allander vanishing points but we've actually also fixed or compression by her
312
00:18:59,711 --> 00:19:02,527
Windows on top of each other
313
00:19:03,039 --> 00:19:06,367
So this is a great technique to use when you don't have
314
00:19:06,623 --> 00:19:09,439
That information to do the mathematical School
315
00:19:09,951 --> 00:19:13,023
What happens if you don't have such a friendly shark
316
00:19:13,535 --> 00:19:15,327
What if you are able to determine
317
00:19:15,583 --> 00:19:16,351
Any van
318
00:19:17,887 --> 00:19:21,215
Well the first thing you need to do is try to establish your camera
319
00:19:21,983 --> 00:19:27,871
Your camera height is where receding angles will become perfectly flat as you can see in this image here
320
00:19:28,127 --> 00:19:34,271
Any angles that are above or below the camera will have a more and more pronounced v-shape
321
00:19:34,527 --> 00:19:35,039
A receipt
322
00:19:35,807 --> 00:19:40,927
If you can identify and then align your camera height between your two pieces of footage
323
00:19:41,183 --> 00:19:42,719
You're off to a great start
324
00:19:43,487 --> 00:19:48,863
Then you need to look at available visual Cues that inform you about your lens compression
325
00:19:49,375 --> 00:19:54,239
Is images undergo more lens compression the receding angles become flatter
326
00:19:54,495 --> 00:19:55,775
Where in wiper images
327
00:19:56,031 --> 00:20:02,175
We can see the receding angles become significantly more pronounced and again the culprit behind this
328
00:20:02,431 --> 00:20:08,063
Difference in the intensity of a receding angles is all about the relative distance to the cam
329
00:20:08,575 --> 00:20:12,415
Veneer corner of a stack of boxes in the 50mm lens
330
00:20:12,671 --> 00:20:14,975
Is a decent about closer to the camera
331
00:20:15,231 --> 00:20:16,767
My other two corners
332
00:20:17,279 --> 00:20:22,911
Put on the 150 mm lens the relative distance between them is negligible
333
00:20:23,167 --> 00:20:24,959
And that's why they appear much
334
00:20:25,215 --> 00:20:27,263
Weather in closely packed together
335
00:20:28,287 --> 00:20:34,431
Do you need to take all the visual Cues that you can determine how compress the images and do your best to ski
336
00:20:34,687 --> 00:20:35,711
Your background recording
337
00:20:36,223 --> 00:20:42,367
It can be very difficult to scale correctly we do have an advantage if we don't have many visual Cues to a
338
00:20:43,903 --> 00:20:49,023
Also means there are many visual Cues for the viewers to tell it is off as well
339
00:20:49,279 --> 00:20:54,399
So it's a little more forgiving but still important to try to get it as accurate as we can
340
00:20:54,911 --> 00:20:58,239
And that wraps up our final compositing fundamental
341
00:20:58,495 --> 00:21:04,639
That is prospective it is definitely by far in my opinion the hardest to read
342
00:21:04,895 --> 00:21:05,663
Turn up an eye for
343
00:21:05,919 --> 00:21:11,807
It took me several years of doing incorrectly before I was able to see it immediately
344
00:21:12,063 --> 00:21:18,208
And so don't get discouraged if you're having a hard time with that it's one of those things that is little more
345
00:21:18,464 --> 00:21:20,000
And will come with experience
346
00:21:20,512 --> 00:21:23,328
Don't get discouraged and thank you very much for listening
347
00:21:23,584 --> 00:21:24,864
And I'll catch you in the next one
28530
Can't find what you're looking for?
Get subtitles in any language from opensubtitles.com, and translate them here.