Swapping eyes for frustum in stereoscopy ?

[Cédric Le Cam]

Hello everyone,


I've come across a method to calculate the frustums for stereoscopy I don't understand. In this method, the image sent to one of the eyes is rendered from the opposite camera, for example the left camera renders the image sent to the right eye.
Is this method relevant? When asking for details, I have been told that depending on the devices, it may be necessary to swap the eyes.

What do you think about this?

[Sébastien 'Cb' Kuntz]

Hi C�dric,
it happens quite often that a VR system needs that the eyes are swapped,
so it's always useful to be able to swap the eyes.
The method should allow for both though, so I'm surprised your method is
always swapping the eyes ?
What's the method ?
cb

Le 07/08/2013 11:25, C�dric Le Cam a �crit :

> --
> You received this message because you are subscribed to the Google
> Groups "VR Geeks" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to vr-geeks+u...@googlegroups.com.
> Visit this group at http://groups.google.com/group/vr-geeks.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

[Cédric Le Cam]

> What's the method ?

Let's take the right eye for example:

First, the frustum associated with it is calculated from the left, right, top, bottom values, all based on the position of this eye in the scene.
Then the resulting projection matrix is set to the left camera.
That's all there is to say about it.

>The method should allow for both though, so I'm surprised your method is always swapping the eyes ?

It can be modified through the code, but it is definitely not highlighted.

Cedric

[Naem BARON]

Hi Cédric

Usually, the swapeyes parameters is present because some hardwares do not know how to address the correct buffer to the targeted eyes... some others just always invert the eyes.

Especially with quadbuffer with back buffer left and right in the OpenGL API but depending on the implementation of the graphic card those buffers can be inverted.

The same goes for split-screen stereo, it's depends on the display implementation.

In those cases, you just use the swapeyes parameter to set everything in the right position. As Seb stated, it's strange that the eyes are inverted by default.

C U

Naëm

return 0;

[Cédric Le Cam]

Hi Naëm,

Is there a way to determine the behaviour of a system, other than trial and error?


Cedric

[Sébastien 'Cb' Kuntz]

The only automated way I can think of is to have two webcams looking
through the glasses, displaying a different color for each eye, and
check that the output of the cameras is as expected :)
Other than that you have to check with your own eyes.
The problem can come either from the projector, from the sync-emitter or
from the glasses themselves,
so anyway solving the problem will involve a human.
cb

Le 07/08/2013 13:58, C�dric Le Cam a �crit :
> Hi Na�m,

>
> Is there a way to determine the behaviour of a system, other than
> trial and error?
>
>
> Cedric

[Cédric Le Cam]

>solving the problem will involve a human

Good thing I have some at hand  :D
Thanks for the answers.

Cedric