Car Paint Flake Density Above 20

[brickyard]

Hello,
I'm running Maya 2014 Linux with Vray 2.45.01, and using the car paint shader. It seems that flake density is disallowed from going above 20, is this a bug in my cut, or is there a reason for this? I'm new to the car paint shader, but watching a tutorial of it, such as one here:

https://www.youtube.com/watch?v=lQXYSlcxwGs

He is using a density of 1200 (in maya 2013). Right now with my density maxed out at 20, my image simply has a few large dots with a lot of distance between them. I am working at scale of 1 maya unit = 10 cm real world.

Also, I have already set the flake scale down to .001 in an effort to fix this, and I have tried with a flake map size of 4096 but that also did not help.

Please let me know if this limitation is on purpose, and if so why and what the workaround is, and if this is a bug, which newer cut will address this. Thanks.

[kristianturner85]

Hi,

What type of mapping are you using, Explicit will use your UV map for the flake placement and Triplaner will project in world space/scale. Here is a test I just did using Triplaner and turning off the base colour, lit with a simple HDR.

I hope this helps,

-K

[Svetlozar Draganov]

There were some changes on that regard and in the latest Vray versions that parameter is limited to 20.
Let us know if this is too low according to your opinion and we can suggest increasing its boundaries.
Would it be possible to send us example scene if that limitation is preventing you to get the effect you are looking after?

Thank you very much in advance.

[brickyard]

Thanks Kristian and Svetlozar,
It looks like switching to explicit mapping seems to be doing the trick. The docs say this gets computed from the object space normals, does that mean this is recomputed every frame? IE, if the surface moves through space or deforms, will the flecks "swim" around or stay locked using this method? As to the density limiting, of course I have no idea on the reason for the limit, though I would guess too high might cause memory overruns or such, but it would be preferable to not have a hard limit on this, or at least to make it significantly higher, possibly with a soft limit on the attribute slider along with popup help for the suggested range, but yes, it would be nice to be able to go above 20, especially if I end up having to use explicit mapping which I assume was the cause of the problem due to my scale. Thoughts?

[tashko.zashev]

Hi,

When you use the explicit mapping channel the flakes will be mapped through the object UVs. If you deform or animate the geometry, the flakes map will be changed according the UVs. There should be no flickering or jumps over the frames.

[brickyard]

Sorry Tashko,
I mistyped in my last post, what I meant to say was switching to triplanar mapping did the trick, the explicit mapping was coming out with the flecks much too far apart, I assume due to the truncated density setting which I can't set higher than 20. So, my questions were all in regard to triplanar, I'm assuming they will swim around, which is not a good solution for me. Therefore, if you guys allow the density to go much higher than 20, will this fix my problem and also remain locked? I will assume that in fact is the case and therefore request that be modified if possible. Let me know. Thanks!

[tashko.zashev]

Thank you for your fast reply.

One of the reasons that the flakes density values has been limited in order to prevent high memory consumption.
Yes, the triplanar mode will lead to map flickering in case that the object has animated topology changes. Therefore the explicit mapping should be better for those cases.

Would it be possible to share a scene file with the Carpaint shader applied to a simple primitive (or part of the original geometry might be better). So if there is now way to increase the flakes density for a specific setup I will forward the case to our developers.
Thanks.