We’re having some issues with vellum in Houdini engine for Maya.

Running 5 sim substeps with the playback set to int frames results in a good simulation with resolved collisions.

Running 5 sim substeps with playback set to 0.2 step gives us very different results with collisions looking like they’re not working and pin to anima drifting off.

So it looks like maya isn’t feeding Houdini engine the correct subframe geo.

Any tips?
Thanks
Christian

Could you please attach the hda in question?

I’m afraid not. I quite like my job

But where as I’d expect to see better collision results over substeps, I’m seeing the opposite. I’m presuming that’s because there’s some funny time offset going on when we’re using subframes in mayas timeline.

I guess a good question to have answered is. If we have maya timeline set to integers and we use 5 substeps on the hda, Are we actually getting animation moving over those substeps, or does it freeze on the integer frame and solve 5 times…