V-Ray Next Update 3 available for download

(*) VRayALSurfaceMtl: SSS is not computed for materials seen through glossy refraction;
(*) VRayStochasticFlakesMtl: Stochastic flakes are not visible through glossy refraction;

Is this a Bug Fix?
Or is it not calculated to reduce rendering time?

Those are bug fixes, they were missing when looked through glossy refractions and now they should work as expected.

Thank you for the awesome work you put into this update guys, really appreciate everything that goes into it.

Unfortunately I just tested out the RTX integration on a quick test scene and it actually added 2 minutes and 20 seconds, running on a 2080ti with the new drivers. Ill test it out on more scenes I have and hopefully this isnt replicated.

Cheers

Thanks for the update. Why no video on your Youtube channel showcasing the new update? Usually you post such videos. It would be great to see what’s new and improved at a glance. :slight_smile:

I know, I know :slight_smile: There are a lot new nice things but check the list then try to market them in a video and you’ll see why :slight_smile:

Thanks. Maybe I’ll just install it and try it out.

I have unfortunately experienced the same with an rtx 2080 testing progressive and bucket… but still really nice work on fixing everything else.

Keep in mind that the first time you launch the RTX engine (per new build/driver) you wait 45 seconds on average for kernel compilation per GPU.

Our @KristinIvanova was reminding me that our excellent docs team has already prepared a page to better showcase the new filtering mode.
Go here for all the juicy details.

The first run is always slow for scene compilation, it should be fine with more runs
in my test scenes there have been a massive speed boost with RTX on

probably the bump delta scale issues

Indeed, but I also remember bump creating faceting or a hard shadow edge, although I remember that was solved but some people still had similar issues under some circumstances.

The hard shadow edge should be resolved, I’m not aware of any more issues related to it.

The bump delta scale issue it’s being worked on for months and I guess will be included in one of the next releases. It’s not a trivial one so it needed a lot of research since the solutions in the other renderers are not ideal.

I’ve done some further testing and the RT engine is still consistently slower than the CUDA engine (not by a major amount but still), if its not meant to be quicker as such or even on par then its not a problem but my understanding was that it would be quicker ?

It’s heavily dependent on the scene but yes - it shouldn’t be slower. Please send me such scenes so we can check if we can improve it on our side.

I ran a test yesterday and my render times went up from 4m16s to 4m54s…

Progressive Rendering VRay-Next (not GPU).

Then this is completely bizarre, since you’re not rendering with your GPU.

RTX rendering is a subset of GPU rendering that works on RTX cards. It doesn’t work with CPU rendering.

I think I have just spotted a big with the LIGHTCACHE PRESET animation/still preset (gpu)

Whatever you select on the drop-down, it just toggles between to two options.

So if I have it set to animation, click the dropdown and select animation, it will toggle to still and not stay/select animation.

Also, I don’t know if this is a big or a feature, but if I select animation, then edit the settings, selecting the default goes back to what I had set for the the animaiton.

So if I select animation then set the subdivs to 5, when I select animation again, it shows my subdivs as 5, not as whatever your default is.

In fact, it’s giving bizarre results and getting them confused now. You might want to look at this.

Keeping set values between modes is on purpose. You can always right-click on the spinners to set them on default values (as almost every V-Ray related setting that has a default value).

We will fix the first one.

Hello, been testing another scene with RTX and it consistently crashes, works fine on CUDA. Is there a specific email we should send our scenes to for analyzing the problem?

Thank you