<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
<title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
But the image merging that takes place in
vtkRenderWindow::StereoRenderComplete takes two images and makes one
same-sized image out of them. It sounds like you want to have two
images and keep two images, right? Are they side by side, left eye view
taking up left half of render window?<br>
<br>
It sounds like your output image is different than, for example, the
merged output image of a VTK_STEREO_RED_BLUE render...<br>
<br>
Maybe a subclass of vtkRenderWindow that overrides the methods called
from vtkRenderWindow::DoStereoRender (StereoUpdate, StereoMidpoint,
StereoRenderComplete) would be easier for your sort of implementation?<br>
<br>
<br>
David<br>
<br>
Thomas Hansen wrote:
<blockquote cite="midb2c52c820506200918729095a5@mail.gmail.com"
type="cite">
<pre wrap="">Hi David,
Thanks for the swift response. I will take a closer look at how i can
use the setViewport method to maybe get the result i was looking for.
But I'm not sure it'll do exactly what I want.
When the Stereo modes are turned on (RedBlue, Dresden and Interlaced)
the result frame is created by merging two images (left and right) in
the vtkRenderWindow::StereoRenderComplete() method. The images are
rendered already at this point and are accessed through the
GetPixelData method.
This happens for each renderer so if I were to set the setViewport() I
would actually change the size of the renderer. I still need the
renderer to fill the whole window.
The only problem is that the two images I use to build one bigger
image really only need to be half as wide. Assuming a resolution of
800x600 on each screen i create a window 1600x600, but instead of
rendering two 800x600 pictures, I end up with two 1600x600 images of
which I have to cut 1/4 off each side and finally merge them.
Maybe i should change the whole approach, and use two separate
renderes keeping track of the camera positions with respect to each
other instead. It just seemed nice to facilitate the way that the
code already set up the two images. It makes using the stereo code
allot easier too.
I hope I'm being somewhat clear...
Tahnks,
Thomas
On 6/20/05, David Cole <a class="moz-txt-link-rfc2396E" href="mailto:david.cole@kitware.com"><david.cole@kitware.com></a> wrote:
</pre>
<blockquote type="cite">
<pre wrap="">vtkRenderer derives from vtkViewport. You can call
vtkViewport::SetViewport(xmin, ymin, xmax, ymax) to make the renderer
take up some fractional space of the render window... (Calling it with
0, 0, 0.5, 1.0 will make it take up the left half of the window, for
example. 0.5, 0, 1.0, 1.0 would be the right half, etc.) Then you have
to put another renderer into the other half of the window, or you'll end
up with an undefined area of your window.
HTH,
David
</pre>
</blockquote>
<pre wrap=""><!---->
</pre>
</blockquote>
</body>
</html>