[vtkusers] vtkFixedPointVolumeRayCastMapper features

[Mario Medved]

Hi Lisa,

Thank you for your prompt reply.

When I combined the source code and screenshots for your example I "noticed" a little bug. The problem is that we used vtkShiftScale for MIP raycaster which worked fine, fixed point MIP did not use vtkShiftScale, but there was some classification problem because of that. So forget about MIP problems:)

I must say that vtkFixedPointVolumeRayCastMapper is really good and it solves a lot of problems for us.

>switch the vtkFixedPointVolumeRayCastMapper over to using this
Is there any catch or I just have to copy the code from RayCastMapper?

Warm Regards,
Mario

  ----- Original Message ----- 
  From: Lisa Avila 
  To: Mario Medved ; vtkusers at public.kitware.com 
  Sent: Thursday, September 29, 2005 6:04 AM
  Subject: Re: [vtkusers] vtkFixedPointVolumeRayCastMapper features


  Hi Mario,

  For your first problem: I am unaware of this. Can you cause it on one of the VTK data sets? If so, can you send me the code that causes it so I can try it?

  On the second issue - yes, the vtkFixedPointVolumeRayCastMapper does use its own code for generating gradients since the gradient estimator is not able to process multicomponent data. Originally, the vtkFixedPointVolumeRayCastMapper relied on a specific gradient encoding since it used to interpolate normals (Phong-style shading) but that turned out to be too unstable so it interpolates shaded value (just like the standard ray cast mapper). So - it should be possible to expand the gradient estimator to handle multi component data, switch the vtkFixedPointVolumeRayCastMapper over to using this, and then everything will run faster on a multiprocessor system (although keep in mind that the code in the fixed point mapper tries harder to get a gradient in homogeneous regions - looking out at larger distances until it finds one with a sufficient magnitude - now that we are not interpolating gradients perhaps this is not necessary....). 

  Lisa


  At 08:53 AM 9/28/2005, Mario Medved wrote:

    Hi,

    I was experimenting with vtkFixedPointVolumeRayCastMapper which is great, 
    but I noticed 2 problems/features:

    1. MIP blend mode causes black spots over some images. I think this is due 
    space leaping, since vtkVolumeRayCastMapper does not have these problems.

    2. When the shading is turned on, the gradient estimator works slower then 
    in vtkVolumeRayCastMapper. On larger volumes this can be very lengthy 
    operation. I checked the source code and there is just block of code for 
    calculating gradient values directly without any multiprocessor support. 
    vtkVolumeRayCastMapper uses gradient estimator object which does all 
    calculations with miltiproc support, but I am not sure why this code is not 
    used in vtkFixedPointVolumeRayCastMapper.

    If anyone solved these problems, please let mo know, so I do not rediscover 
    hot water.

    Thanks,
    Mario 
    _______________________________________________
    This is the private VTK discussion list. 
    Please keep messages on-topic. Check the FAQ at: http://www.vtk.org/Wiki/VTK_FAQ
    Follow this link to subscribe/unsubscribe:
    http://www.vtk.org/mailman/listinfo/vtkusers
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.vtk.org/pipermail/vtkusers/attachments/20050929/c8ef85c2/attachment.htm>