[vtk-developers] strange behaviour in vtkVolumeTextureMapper2D

[Sebastien BARRE]

At 09:17 13/07/00 -0400, Lisa Sobierajski Avila a écrit:

>>Nope, I've tested that already, replacing the new code with the old 
>>(rectangle). No differences, bug remains.
>
>So if you simply alter the vertices (by using size instead of size-1) 
>everything works fine?

That's right.

But if you look at the portion of code :

     t[i*8 + 0] = (float)((size[0]*(xTile  ))  )/(float)textureSize[0] + 
offset[0];
     t[i*8 + 1] = (float)((size[2]*(yTile  ))  )/(float)textureSize[1] + 
offset[1];
     t[i*8 + 2] = (float)((size[0]*(xTile  ))  )/(float)textureSize[0] + 
offset[0];
     t[i*8 + 3] = (float)((size[2]*(yTile+1))-1)/(float)textureSize[1] - 
offset[1];
     t[i*8 + 4] = (float)((size[0]*(xTile+1))-1)/(float)textureSize[0] - 
offset[0];
     t[i*8 + 5] = (float)((size[2]*(yTile+1))-1)/(float)textureSize[1] - 
offset[1];
     t[i*8 + 6] = (float)((size[0]*(xTile+1))-1)/(float)textureSize[0] - 
offset[0];
     t[i*8 + 7] = (float)((size[2]*(yTile  ))  )/(float)textureSize[1] + 
offset[1];

     v[i*12 + 0] = origin[0];
     v[i*12 + 2] = origin[2];

     v[i*12 + 3] = origin[0];
     v[i*12 + 5] = spacing[2] * (float)(size[2]) + origin[2];

     v[i*12 + 6] = spacing[0] * (float)(size[0]) + origin[0];
     v[i*12 + 8] = spacing[2] * (float)(size[2]) + origin[2];

     v[i*12 +  9] = spacing[0] * (float)(size[0]) + origin[0];
     v[i*12 + 11] = origin[2];

I've just removed the -1 in the low part of the code (float(size[x]), not 
in the (*Tile+1))-1 above (the change you made initally concerned both -1). 
I've just tested right now that if you remove the -1 in that place also, it 
works fine too :( Which convinces me that I should better stop hacking this 
without understanding it perfectly, but I'm lacking time this week :)

>That change can be made (I'd prefer to wait until after finding the 
>current bug!)

Right :)

>>BTW2 : I've sent a mail to the VolumePro guys. Jesus it's expensive !  $ 
>>5500 ! (25% discount for university, but wow). Is it really worth it ? We 
>>are expecting a GeForce here, and I'm quite impatient to see how it deals 
>>with the volumeTextureMapper2D. I'm gonna test it
>>with Voodoo 3 this week also.
>
>Is that $5500 US? I thought it was about $3000 US if you got it without 
>any special add-ons.

The board is $3000 US, but the SDK is $2500, which is bad joke :( What can 
I do without the SDK ? I need the SDK to compile the VolumePro support in 
VTK, am I right ?

>Whether or not it is worth it depends on your data size, your application 
>area and your interactivity requirements. We do get about 10-30 frames per 
>second, and the quality is generally better than texture mapping (although 
>not as good as ray casting). When we are working with small data sets 
>(64^3 or 128^3) it is not that important since the texture map version is 
>usually rendering about 5-10 frames or so per second with either the full 
>volume or a half res version, and it fills in pretty quickly once you stop 
>interacting.

That's right, that's actually what I'm doing to visualize 256x256x100 : 
resampling it to 0.3333 in each direction or so, and use that resampled 
stack during interaction. BTW, I guess there might be some twicking to do, 
because when the full-rez is displayed, it is not exactly at the same place 
as the low-res, it looks shifted/rotated. I guess it has something to do 
with resampling, I will have to shift it manually by 0.5 or 0.333 voxel, 
andhave a look at how the ImageResampler works.

>We have two VolumePro boards here, and I feel that I can get things done 
>quicker with it (for example, simply loading some data in, setting the 
>properties, and generating an image).

And what about these SGI hardware 3D board (Oxygen ? Osiris ?). Well, I 
know it's expensive :)
I might send an email to the public list to check if anyone has been using 
another hardware  volume rendering solution. have you heard about anything 
else ?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/vtk-developers/attachments/20000713/988fd9e3/attachment.html>