I am having trouble getting mograph setups cached so that I can render it via teamrender. I tried mocache tag on the cloner object (and on each and every matrix/clones in the scene). Also I tried baking as alembic which wouldnt work since the projection on the visible clones is not UV.
What am I missing? Any help appreciated.
The basic scene file is attached.
Cheers
Rogurt
Matrix_Cache_trouble.zip (535.3 KB)
Are you caching internally or externally? I would try internally first (if you havent already).
Mo-cache has always worked fine on TR for me in the past, although I have heard some complaints over the past couple of years of this sort of problem happenning with team render and cached mograph.
TBH, I would probably move on from TR if you can and go with another render manager.
We’ve moved to deadline and very happy with that decision.
Cool! We were just talking about AWS stuff. Deadline looks pretty robust. You guys use Shotgun too, don’t you?
(not to derail…but…) yes deadline is working great. We have AWS set up too and thats amazing—it just took us about 4 months to get AWS working properly for us. Can be complex. We’ve looked at shotgun and have tried a few other P.M. apps, but not using it currently.
@JoelDubin Tried internal and external cache (path accessible via network). Neither did work. Tried different mograph setups also. First setup had a matrix deformed by standard deformers. Second setup only via mograph effectors. Both did not render ok via TR. Not sure about swithing to another render pipeline…
Thanks anyway…
Another vote for deadline.
Stop wasting your time with Team Render it has never worked reliably and likely never will. Net render worked so much better.
Yikes, really? Fortunately our coder is built for that :). He’s been pushing for AWS integration for a while. That’s his problem now :).
Hi Rogurt. Sorry you are still having troubles. I havent looked at your file yet tbh, but will take another look asap.
One thing I noticed. Although it shouldnt matter if you add a mocahce tag and cache the sequence, your order of operations in the object manager should ideally be organized so the thing being referenced is at the top of the manager list, and the cloner is below.
So I would put your cloner i this case at the bottom of the list.
ALSO–I see you have the “kugel” object set as a chld of the “2 Matrix Kugel” object, and you have that matrix object targeting the child . I would actually reverse the heirarchy so the matrix is a child of the kugel (or at least above the matrix in the OM). I would think this would create a conflict as the matrix is trying to clone onto its own child.
I would also try making your clones render instances if posible.
You should only really need to mo-cache the cloner (and do it internally if possible)
@JoelDubin Thanks for your advice. In the meantime I tried a different route that seems to work for my case. I took the cloner object and converted it to poly object which I then broke apart by polyFX effector. This construct can be tweaked by all the effectors I already set up and then can successfully be baked as alembic (moCache wont work).
As the viewport performance is way higher with the polyFX instead the Cloner I only bake just before TeamRendering.
I will have a look at your thoughts anyway. Hope I´ll understand how things work with mograph and cache.
Cheers
Addendum: A method by ilirbeqiri over at c4dcafe I stumbled upon to get mograph animations baked into PLA that are otherwise problematic
Did not try myself though…
Important notice for german dudes. DO NOT use “Umlaute” (ä.ö,ü etc) in Alembic Filenames. Gave me a hard time figuring why it just would not render via teamrender while rendering on the workstation itself works fine.