VRay 1.46.05 available

[vlado]

Bug fixes:

(*) QMC sampler behaves in a more similar way to the 1.45 versions without changes in settings. Note however, that the QMC is still different from the 1.45 version, so differences in render times/quality are possible.
(*) Lightmaps and photon maps will take more samples on average compared to 1.45 versions which may lead to slightly longer render times (although results may be cleaner). Reduce the subdivs to decrease the render time.
(*) VRayProxies did not provide material id's properly (did not work with multi/subobject materials) .
(*) Total internal reflections were not correct.
(*) Caustics photon map was used when present, even if caustics are off.
(*) The "Sub-pixel weight" channel was incorrect (improved compatibility with PSD manager).
(*) 2d displacement could cause infinite cycling on meshes with degenerate faces.
(*) Better handling of opacity maps - both with VRay materials and with Standard materials.
(*) Parts of VRayLights could disappear during rendering.
(*) Caustics did not appear on Standard materials.
(*) Photon mapping did not work properly with VRayMtlWrapper with the Matte option turned on.
(*) The exposure control peview in the Environment dialog crashed with VRay.
(*) Thumbnails in the ME on multi/sub materials were not rendered.
(*) In distributed rendering mode, VRay tries to delete any temporary .max files after rendering is complete.
(*) The VRaySpawner attempts to delete any files left over from previous sessions.
(*) 2d displacement did not work with motion blur.
(*) Added a section in the help with a list of 3dsmax features which do not work or work only partially with VRay. See the Unsupported 3dsmax features section.
(*) The mouse tracking mode of the VRay VFB did not work properly when enabled at the start of the rendering.

Thanks to all who submitted bug reports and helped us resolve these issues; if you find any other problems, please send them to vray@chaosgroup.com or post them in the Bug reports section.

Best regards,
Vlado

[percydaman]

thanks for the hard work vlado. question: Alot of people and myself included, had issues with lightmaps hanging during initial calculation. seemed it happened to me if I had ANY standard max materials in the scene. Has that been adressed with this version? didn't see any specific word of it above.

[Robert]

very happy to see this bugfix so soon...
continue with that and we get even more happy...

Thanks a lot.

regards,

robert

[Gijs]

Thanks Vlado for working through the weekend. I really appreciate that.

gonna test it right away...

[vlado]

thanks for the hard work vlado. question: Alot of people and myself included, had issues with lightmaps hanging during initial calculation. seemed it happened to me if I had ANY standard max materials in the scene. Has that been adressed with this version? didn't see any specific word of it above.

This was due mostly to standard materials with opacity less than 100. Should be a lot better now; at least for the scenes we got for testing, it seemed ok.

Best regards,
Vlado

[plastic_]

wow, that was FAST.

[acid]

Chaos Group : congratulation !!!

Chaos Group : congratulation !!!


today the new name for powerfull is vray !!!
nice work Chaos Group, and now i can say:"I 'm happy to Purchase vray advenced!!!" the best GI solution that i see at this time... (quality/rendertime)


thanks for this new build !!!

[Geddart]

I did a few tests concerning the QMC Sampler under different
circumstances with three different Vray Versions. 1.45.70, 1.46.04 and
1.46.05. (All pictures are uncompressed png's so you can save them and compare them with your image viewer) Here are the results:

Athena on VrayPlane with glossy reflections.



The 1.46.05 Version really behaves like the 1.45.70 version in this case.

Same Scene with Settings really cranked up: (only 1.45.70 and 1.46.05)


When the Settings are tuned for really high quality 1.45.70 traces more samples. (longer rendering time and finer image)
This means one could need higher samples to get final image quality!

Lightmap only Scene



The 05 Version shoots more samples in the Lightmap = slightly higher rendering time and slightly better image quality)

Same Scene with Irradiance Map only (only 1.45.70 and 1.46.05)


No significant difference here.

Scene with multiple glossy reflections and fuzzy effects: (the reason why vlado changed the qmc sampler)



Here Vlado's optimizations take place. The Image in the 1.46.05 Version took less time than the one in 1.45.70 and have better image quality. Good work here Vlado.

Conclusion: The 1.46.05 Version behaves like 1.45.70 in most cases. Only with Lightmapping and with really high glossy samples it behaves different.

Note: Displacement is NOT affected by the change in the QMC Sampler so rendertime should be exactly the same in all versions. The only thing that could affect your rendertime is AA cause it is depends on the QMC Sampler.

[percydaman]

thanks for the hard work vlado. question: Alot of people and myself included, had issues with lightmaps hanging during initial calculation. seemed it happened to me if I had ANY standard max materials in the scene. Has that been adressed with this version? didn't see any specific word of it above.

This was due mostly to standard materials with opacity less than 100. Should be a lot better now; at least for the scenes we got for testing, it seemed ok.

Best regards,
Vlado

hmmm... dont recall any of my materials standard materials having less than 100% opacity, but Ill recheck.

[dennis]

garrhh now i have to make all my 1.46.04 scenes from scratch again !!!

hehe just kidding of course..this is great. Thanks.

[Robert]

Interesting test Sascha..

Saves me lot of time.

I did some tests with scenes done in 1.09 with lot of blurry reflection too. Rendertime was 50% higher and some glossy reflection were incorrect (Omni's didn't produce highlights.)
Anyhow scenes that crushed with MotionBlur in 1.09 did render fine and fast.

I noticed even a crash when an Object with animated visibility hit an object with blurry reflection. Both objects where behind another, so strange but true. I'll going to make these test's on Monday in a new scene as the old one is, like most time, confidential.

regards,

robert

[travistlo]

FYI:

If you use a vraymtlwrapper with matte shadows turned on. And use lightmaps, max will crash.

Vlado is aware.

[vlado]

Yeah, the bug hunt continues...

Best regards,
Vlado

[algato]

good advices

Thank you Sascha for the interesting tests.

Could you please post some setting comparisons?
what have you changed to get similar rendertimes with vray 1.45.70 <-> 1.46.05 ?

Kind regards
Alexander

[Geddart]

Good my tests are of some interest for ya guys.

@algato: The Settings of all Scenes are absolutely the same. I did not change anything. The one who changed things was Vlado. He made the QMC Sampler internally more like the one from 1.45.70. So all your 1.45.xx Scenes should render as fast as they used to be in Vray 1.46.05. (not exactly, but almost )