Hey guys,

I'm running into an issue with the new guide deform in 18.499 and I'm hoping someone can help me in the right direction. Maybe I'm using it wrong so please correct me if I do so.

In our pipeline we get a groom from Maya exported as alembic curves. These are all the render hairs. In Houdini I'm filtering a percentage of it that I can use for simulation or other things I need to do with them. When I'm done I can use these guide curves to deform the full groom. Obviously the Guide deform should be able to do this. The new guide deform in 18.499 is allready a lot better but I'm still getting artifacts.

So this is the full patch of hair as an example.


These are the guides I'm extracting:


The guides that are deformed:


So when I use the guide deform with barycentric weights it comes pretty close but there are artifact where the tips are not deformed correctly.


Even if I generate new hairs from the guides with the hairgen sop and tick on computes weights. And use the existing weights mode it still has the issue with the tips.



With the existing weights it is definitely following the guides better. I could even try to compute the weights myself on the original render groom. However I would still run into the tips issue.

In the past we just made sure we had enough guides and generate new render hairs from that. We sort of got away with that but it's not as accurate as deforming the original groom.

Hopefully someone can shed a light on it.

Thanks in advance!

Ralph

I'm on 18.5.351 still and I have the same issue.

Using Barycentric weights and no "Split Clumps" and no "Pick Single Best Guide" gives me the "best" results. But it still has issues like yours.
Every other method seems to be even worse. (eg. with "Split Clumps" turned on or the legacy Per Point Capture mode)


Example:
The deformed guides


the fully deformed groom


full groom deformed with guides (barycentric weights)


full groom deformed with guides (per point capture)


Best

J

Hello! I'm having the same exact issue, have you guys ever found a solution/workaround to this?

Nothing great at least.

I started setting everything up in a way where only the longest hairstrands are used as guides. (in my example from back than i'd only use the purple ones)
Therefore all the shorter ones follow the deformation okayish. Most of the time this works pretty well.

I also extend the guides by 20% before capturing the full groom. That way i make sure everything else is shorter and can be captured properly. However this is a very hacky way and doesn't work well when you need to have accurate collisions.

That's not sounding too hot, having the same issues now. Yuck.

same issue. any updates?