LuxRender v0.7RC3 released

[tomb]

I apologize for my English.

when will implement GPU rendering LuxRender?

For version 8?

Greetings from Colombia

That is the plan yes - see the GPU subforum here viewforum.php?f=34 for the current experiments

[heretic]

I just realized that some Magician integrated group based particle support for RC3.

Great job! Just a pitty it happened so silent - but cool!

[SATtva]

I just realized that some Magician integrated group based particle support for RC3.

His name is Atom.

Just a pitty it happened so silent - but cool!

It's not if you watch what happens in the repository from time to time. Sorry, I ceased my usual progress reports (in weeklies) during the later phase of RCs as releases are already quite frequent, and currently I don't have much spare time.

[binarycortex]

Does anyone else get a craptonne of "[2010-06-10 07:46:08 Warning: 43] Inconsistent shading normals" messages using 0.7 RC-Whatever?

[jeanphi]

Does anyone else get a craptonne of "[2010-06-10 07:46:08 Warning: 43] Inconsistent shading normals" messages using 0.7 RC-Whatever?

That means some faces of your mesh have vertices with the normal on one side of the face while the remaining vertices have their normal on the other side of the mesh. If such a face is visible it will most probably produce shading artifacts, and if such a face uses volumetrics it will most probably ruin your render.

Jeanphi

[heretic]

...and currently I don't have much spare time.

Its just that this change greatly supports my "style" and reduces quite a significant chunk of additional work for me.

PS: are those instanced? If so, Id try somethign very crazy.

[heretic]

...and currently I don't have much spare time.

Its just that this change greatly supports my "style" and reduces quite a significant chunk of additional work for me.

PS: are those instanced? If so, Id try somethign very crazy.

They are perfectly instanced - looking at the size of the geometry on my tests am impressed...

PS: some figures - 20Mill vertices on 400K (200 children on 2000 parent objects) instanced objects are handled by LuxBlend/LuxRender with ease; the whole geometry-file is only 140 MB, so on my system I could probably go as far as 200 Mill verts on 4Mill instnaced objects. Thats sufficient for now

[binarycortex]

Does anyone else get a craptonne of "[2010-06-10 07:46:08 Warning: 43] Inconsistent shading normals" messages using 0.7 RC-Whatever?

That means some faces of your mesh have vertices with the normal on one side of the face while the remaining vertices have their normal on the other side of the mesh. If such a face is visible it will most probably produce shading artifacts, and if such a face uses volumetrics it will most probably ruin your render.

Jeanphi

Thanks Jeanphi I'll look into that. But that also means that the luxball4.blend has incorrect normals as well. Perhaps if I can correct it I'll release luxball5 into the wild.

[hurleyman77]

LuxRender 0.7RC3 is refusing to run under any conditions. I'm using Win7 x64, Blender 2.49b, and the LuxBlend for 0.7RC3 (I think, the download page was kinda vague). I've tried exporting a scene that ran just fine with 0.6 just 20 minutes ago, but 0.7RC3 won't even run by itself (just trying to open the exe without Blender). The 32-bit of 0.7RC3 works fine. Beautiful GUI, btw. Oh, and thanks so much for fixing that bug in the "Update Interval" input field. It was really obnoxious to try and enter 3 digits between refreshes.

Anyhow, there's a bug for ya, but it seems minor to me... not knowing anything about programming.

Also, I've suddenly begun running into a problem with other versions as well. I've been distributing all day, working on a scene. I decided to add an HDRI skymap via LuxRender. Now LuxRender runs fine, but when I try to distribute to eithe rof my other computers LuxRender apparently requests that Runtime terminate in an unusual way and immediately crashes. The other computers sit there with a new session ID, but never do anything else. With the older versions (I've tried 0.6, 0.7RC2, and now 0.7RC3) I could say "Ignore" and Lux would continue happily chugging away on my scene, and I could even try to add the other computer, again "ignore" and again Lux keeps puttering away, but neither of the other computers will render anymore. One of them is running Linux x64, the other is running Vista x86, I'm running Win7 x64. I've confirmed that they were are running the same versions with each attempt.

Thanks for the hard work. Y'all are doing a great job. The functionality and speed are fantastic, it's just these little details that keep getting in the way.

Hurleyman77

[Montee]

I compared the results of the bumpmapping scene from the testsuite ( core_regression_tests\tests\materials\bumpmapping\ ) between 0.6.1 and 0.7rc3. The bump effect of 0.7rc3 is much weaker. Is that intended ?