- Fri Mar 29, 2013 3:20 pm
#366571
Camera Height
Perspective is defined by many parameters.
Human perception can understand the scale of things by these parameters, although it is not a concious procedure.
In order to determine the missing data (in the case of La Ciudad de Las Ciencias by Calatrava) we can use the following pictures:
Here we can have an idea of the tile sizes.
![Image](http://t1.gstatic.com/images?q=tbn:ANd9GcSveObR1YnSAJplKYiqj73HE0M4qSDFd0t-tMxIDog_pBAb8NkO)
![Image](http://statics.erasmusu.com/originals/596ec99bbcfbec45cff16c15ea751841.jpg)
![Image](http://cdn1.gbot.me/photos/LQ/ed/1284662928/Esa%C3%B1a__Val%C3%A8ncia___Ciuda-Ciudad_de_las_Artes_y_las-3000000036470-500x375.jpg)
Here we have in the same frame, some walking humans, and the floor pattern, which could be clearly compared.
The white stick measures the height of the guy, and the gray horizontal stripe is the same measurement, to be compared to the floor tiles.
![Image](http://img195.imageshack.us/img195/8830/heightw.jpg)
In this enlargemente we can compare the height of the people with the floor pattern.
It seems that 16 floor tiles are similar to the height of the people in the picture.
Then we could assume that each floor tile is 10cm therefore the human height is about 160cm.
This is the key to determine the camera height in the backplate.
![Image](http://www.mesutcapkin.com/temp/spanish_comp_1826_01.jpg)
In this image we can see that the height of the camera is similar to 6 tiles therefore 60 cm.
This is the key to human perception to understand that the ball is 120 cm diameter.
If we do not take this in considration at the moment to create the scene we might fake scale of objects in the scene.
Perhaps for the people who had never been in the Ciudad de las Ciencias, it will be irrelevant, but people who knows it, will find the composition unreal.
FOV
The FOV, or Field Of View, is measured in degrees and is the angle of view.
Some use a vertical FOV others use a horizontal FOV and in the photographic market it is used a diaonal FOV.
It doesn´t matter which we use, but we have to be clear in order to avoid comparing diferent FOVs.
The image of the chrome ball could be usefull to determine the FOV.
We will use floor tiles again.
I could count 15 floor tiles at the base of the image. This would be 150 cm wide.
We knew already that the camera height was at 60 cm.
And that is all we need to calculate the FOV.
Our frame is 12 floor tiles by 15 floor tiles.
If we measure the width of the frame just where the ball is standing, we could count: 40 floor tiles
And the distance between the popint where the ball is standing to the lower frame border is: 24 Floor tiles.
In the following sketch, the full construction is represented, all measurements are in "floor tiles"
Each floor tile is estimated in 10 cm x 10 cm
![Image](http://img145.imageshack.us/img145/9541/28423800.png)
By trigonometry we can know that the horizontal FOV is 55 degrees.
This could correspond to a 35 mm lens (35mm negative standard) but since the proportion of the frame is 1:33 we are only using a 40mm lens.
Then if we use a longer lens in our Maxwell camera, the 3d object will look flattened, and on the other side if we use a shorter lens, for instance a 12 mm lens, our object will look stretched in its depth.
It all depend on the degree of realism that is being intended.
The concepts realistic or unrealistic, has to do with these parameters, and deppending on the observer cultural background, they will judge our work, realistic or not. By cultural backround I am not meaning a professional degree or anything like that, I only mean the previous knowledge that this person has on the objects in the image. As deeper the knowledge, the hability to detect an inconsistency in the image.
The only way to do a professional compositing job, is taking into accunt these problems. Otherwise we will be working by eye, and the realism of the final product will be judged by a diferent eye.
Ernesto
Perspective is defined by many parameters.
Human perception can understand the scale of things by these parameters, although it is not a concious procedure.
In order to determine the missing data (in the case of La Ciudad de Las Ciencias by Calatrava) we can use the following pictures:
Here we can have an idea of the tile sizes.
![Image](http://statics.erasmusu.com/originals/596ec99bbcfbec45cff16c15ea751841.jpg)
![Image](http://cdn1.gbot.me/photos/LQ/ed/1284662928/Esa%C3%B1a__Val%C3%A8ncia___Ciuda-Ciudad_de_las_Artes_y_las-3000000036470-500x375.jpg)
Here we have in the same frame, some walking humans, and the floor pattern, which could be clearly compared.
The white stick measures the height of the guy, and the gray horizontal stripe is the same measurement, to be compared to the floor tiles.
![Image](http://img195.imageshack.us/img195/8830/heightw.jpg)
In this enlargemente we can compare the height of the people with the floor pattern.
It seems that 16 floor tiles are similar to the height of the people in the picture.
Then we could assume that each floor tile is 10cm therefore the human height is about 160cm.
This is the key to determine the camera height in the backplate.
![Image](http://www.mesutcapkin.com/temp/spanish_comp_1826_01.jpg)
In this image we can see that the height of the camera is similar to 6 tiles therefore 60 cm.
This is the key to human perception to understand that the ball is 120 cm diameter.
If we do not take this in considration at the moment to create the scene we might fake scale of objects in the scene.
Perhaps for the people who had never been in the Ciudad de las Ciencias, it will be irrelevant, but people who knows it, will find the composition unreal.
FOV
The FOV, or Field Of View, is measured in degrees and is the angle of view.
Some use a vertical FOV others use a horizontal FOV and in the photographic market it is used a diaonal FOV.
It doesn´t matter which we use, but we have to be clear in order to avoid comparing diferent FOVs.
The image of the chrome ball could be usefull to determine the FOV.
We will use floor tiles again.
I could count 15 floor tiles at the base of the image. This would be 150 cm wide.
We knew already that the camera height was at 60 cm.
And that is all we need to calculate the FOV.
Our frame is 12 floor tiles by 15 floor tiles.
If we measure the width of the frame just where the ball is standing, we could count: 40 floor tiles
And the distance between the popint where the ball is standing to the lower frame border is: 24 Floor tiles.
In the following sketch, the full construction is represented, all measurements are in "floor tiles"
Each floor tile is estimated in 10 cm x 10 cm
![Image](http://img145.imageshack.us/img145/9541/28423800.png)
By trigonometry we can know that the horizontal FOV is 55 degrees.
This could correspond to a 35 mm lens (35mm negative standard) but since the proportion of the frame is 1:33 we are only using a 40mm lens.
Then if we use a longer lens in our Maxwell camera, the 3d object will look flattened, and on the other side if we use a shorter lens, for instance a 12 mm lens, our object will look stretched in its depth.
It all depend on the degree of realism that is being intended.
The concepts realistic or unrealistic, has to do with these parameters, and deppending on the observer cultural background, they will judge our work, realistic or not. By cultural backround I am not meaning a professional degree or anything like that, I only mean the previous knowledge that this person has on the objects in the image. As deeper the knowledge, the hability to detect an inconsistency in the image.
The only way to do a professional compositing job, is taking into accunt these problems. Otherwise we will be working by eye, and the realism of the final product will be judged by a diferent eye.
Ernesto
Last edited by Ernesto on Fri Mar 29, 2013 11:16 pm, edited 1 time in total.