Displacement and map channel issue

[romullus]

Build: Corona 5 2019 Oct 03

Old displacement method incorrectly renders maps with map channel 2 or above, if it's assigned to object which has those channels empty. New 2.5D displacement renders the same setup correctly, i.e. it does not displace geometry if channel is empty. BTW, if 1st channel is also empty, then both displacement methods renders correctly (does not displace).

In attached scene, all objects has the same material assigned. Switching between two displacement methods, will give different results.

P.S. Interestingly, old displacement is giving different results with CoronaMix, Mix and Composite maps. Might be part of the same issue. All 3 maps are present in the scene's slate editor, you can try to switch between them and see how it renders differently from each other.

[rowmanns]

Hi Romullus,

Thank you for creating this detailed report, I've confirmed the issue which you are describing and I will discuss it with our development team.

BTW, in my tests it seems the new displacement gives different results depending on whether CoronaMix, Mix or Composite maps are used.. Is this the same for you?

Cheers,

Rowan

(Internal ID=407335495)

[Juraj]

whether CoronaMix, Mix or Composite maps are used..
(Internal ID=407335495)

I would test this for bump map as well, I avoid using all these because they almost always made something go wrong in bump channel. Now we can (mostly?) fix that by plugging the final network into CoronaBumpConverter or how it's called.
But there isn't such for displacement.

[romullus]

BTW, in my tests it seems the new displacement gives different results depending on whether CoronaMix, Mix or Composite maps are used.. Is this the same for you?

I didn't notice that with new displacement, but i didn't made thorough comparison between those nodes either, just a quick run in interactive mode, so it's possible that i overlook it. On the other hand i'm not 100% sure that CoronaMix, Mix and Composite are set to be 100% identical in my scene, since modes naming is different in all of them and only Corona lets expicitly choose between RGB and sRGB. I better leave it to you guys, to decide wether there is issue or not :]

[rowmanns]

whether CoronaMix, Mix or Composite maps are used..
(Internal ID=407335495)

I would test this for bump map as well, I avoid using all these because they almost always made something go wrong in bump channel. Now we can (mostly?) fix that by plugging the final network into CoronaBumpConverter or how it's called.
But there isn't such for displacement.

Hi,

In my brief tests this seems to work correctly for the bump map with both the new and old displacement. Do you have a specific situation where these maps cause issues?

Thanks,

Rowan

[rowmanns]

I am closing this as resolved. The new 2.5d displacement manages the map channels correctly.

[romullus]

I'm sorry, but why? The bug is still present in RC1 Does that mean that classic displacement is now obsolete?

[rowmanns]

Hey,

So the bug will stay logged in our internal tracker.

But the idea with the 2.5d displacement is to replace the classic displacement, and at some point in the future the classic displacement will be removed.

Cheers,

Rowan

[romullus]

Understood. Thanks for the reply!