Has anyone else noticed some strange things going on with motion blur on hair grooms in Redshift ?

I’ve been trying to get motion blur to work on the hair of an animated character without succes. But on some tests with a basic sphere it sometimes works, just to stop working on some occasions after reloading the scene.

Also; is it normal that hair simulations explode when loading a simulated hair scene when you’re working on another one ? I’ve noticed several times that the only way to load a scene with hair simulation without problems when you’re already working on another hair simulation, is to restart Houdini.

R.

I posted this issue at the Redshift forum and it seems there's an issue with the way Houdini calculates the point count in the hair object.

There's a different point count before- and after the camera shutter close. So Redshift can't calculate deformation blur. That means it needs a “v” attribute and the hair has to be cached.

Now it seems the “v” attribute is not included when I cache the hair simulation. So as a newbie I was wondering how to cache the hair with the “v” attribute included ?

Did anyone else use deformation blur with Redshift on Hair in Houdini before and figured this out ?

Thanks,

R.

Just in case anyone else runs into this issue. Seems the “v” attribute is written in the cached groom after all. But I didn't know one has to manually add it in the “Point Attributes” for the “Guide Attributes Transfer” and the “Output Attributes” as well.

So caching the hair groom, loading the cache and adding these attributes in both input fields, next to activating the “Mesh Deformation Blur From Velocity Attribute” in the Redshift render settings will work with deformation blur in Redshift.

cheers,

R.

toonafish
There's a different point count before- and after the camera shutter close. So Redshift can't calculate deformation blur. That means it needs a “v” attribute and the hair has to be cached.

You shouldn't be getting different point counts on the hair at different times. Try to debug that first. The hairgen, when given the right guide rest attributes, should always output the same number of points and primitives.

jsmack

toonafish
There's a different point count before- and after the camera shutter close. So Redshift can't calculate deformation blur. That means it needs a “v” attribute and the hair has to be cached.

You shouldn't be getting different point counts on the hair at different times. Try to debug that first. The hairgen, when given the right guide rest attributes, should always output the same number of points and primitives.

Hey thanks.

It was because I didn't do the official little Houdini fur dance before running the simulation. After adding the “guidedeform” and a “stack” node for the static object the point count was stable.

R.