Talk:Volume rendering

Promotion
This page should not promote any particular commercial solution. I have removed a link (and its replacement) from two different companies--neither one is suitable. If there is a good non-commercial gallery available showing various techniques, that could be used as an external link. DGG (talk) 23:21, 2 June 2008 (UTC)

Where would a good place for a volume rendering image gallery be? The gallery will have some large images. It should be editable by all concerned - and operate with some consensus. If we use something like http://commons.wikimedia.org, vendors would need to be comfortable with the level of copyright. Bodysurfinyon 2008.06.02 —Preceding comment was added at 00:35, 3 June 2008 (UTC)

the descriptions of the various rendering processes seem very vague to me. i would like to see some more detail. maybe it is expecting a bit much to be able to write a volume renderer from reading the wikipedia article, though maybe a little bit more detail could be included for orientation too. Inhahe (talk) 21:03, 9 April 2009 (UTC)

Page too heavy to load
The total size of this page is 6.80MB!! This is because it contains an image of mouse skull (http://en.wikipedia.org/wiki/File:VolRenderShearWarp.gif) which is of size 6.51MB. It either needs to be replaced or optimized somehow so that the page can load quickly.

Abhijeet Pathak (talk) 11:02, 5 December 2009 (UTC)

deleted images
This images of a skull and of crocodile has no volume imaging; just a regular surface rendering. Therefore I deleted it. If someone restores it again, please thoroughly explain how it demonstrates volume rendering. That the engine used is called "volume rendering" is irrelevant to the article topic. The images must show the volume data throughout the volume. Otherwise one sees no difference from surface-based rendering. —Preceding unsigned comment added by 71.146.69.71 (talk) 03:25, 4 May 2010 (UTC)

[UnsignedIP] This images of a skull and of crocodile has no volume imaging

[Stefanbanev] This statement alone is so telling about level of your competence. It is 2010 and VR these days is way better then it was 10 years ago. The major advancement was how accurately VR may visualize the most minute scalar field fluctuations and it became possible thanks to very deep level of supper sampling by volumetric ray casting. Texture Mapping with super-sampling 16+ samples per cell is insanely expensive and therefore not practical only adaptive ray casting allows to do-so interactively. Please stop project you ignorance upon the world if you have any sense of mercy... —Preceding unsigned comment added by Stefanbanev (talk • contribs) 17:08, 5 May 2010 (UTC)

[UnsignedIP] If someone restores it again, please thoroughly explain how it demonstrates volume rendering

[Stefanbanev]Please read below more elaborated explanation:

[Stefanbanev] The major advantage of volumetric ray casting comparatively with Texture Mapping is the accuracy to visualize the most minute scalar field fluctuations and it became possible thanks to very deep level of supper-sampling which is possible with volumetric ray casting and insanely expensive with Texture Mapping (TM), TM with super-sampling 16+ samples per cell is exceedingly expensive and therefore not practical, only adaptive ray casting allows to do-so interactively. The Skull Image from gallery is a good example to elaborate this point, this example illustrates well the different output for different ray profile along thin film upon bone tissue. It is possible only with very high sampling rate to compute integral accurately for such thin film of tissue upon bone; to visualize the difference in rays profile along this film - very aggressive Color/Opacity gradient of TF should be set and it is exactly recipe to have artifacts for Texture Mapping (TM). Therefore TM adepts do not use to such naturalistic crisp VR images and perceive them as a surface rendering Tricks. The Bottom Line is that adaptive volume ray casting renders accurately with low/high opacity and/or aggressive color ramp Transfer-Function settings because it may adaptively change the precision while TM is good only for Low/Mid Opacity due to even-sampling limitations of TM what makes it exceedingly expensive for cases when high sampling is essential. I hope this explanation is sufficiently thorough. Thanks, Stefan. —Preceding unsigned comment added by Stefanbanev (talk • contribs) 19:31, 5 May 2010 (UTC)

[Lev.kozlodoev] It's also important to mention that if you're volume rendering an opaque surface of a model via an isosurface of the scalar field representing this model, it is completely different from rasterization-based rendering since there are no triangles, vertices, and vertex attributes (normals, texture coordinates) involved. The reason why a volume rendering approach to render opaque objects is often preferred over rasterization/triangulation is because a lot of data (like from MRI/CT scans, even some geological data) will come in the form of scalar fields volumes so rasterizing triangles is simply out of the question here. What I'm trying to say is that, UnsignedIP is a fucking idiot and has absolutely no moral right to be contributing to this page as he clearly has absolutely no grasp on what volume rendering is.