I’m a bit confused about scene parsing. I noticed that my CPU is just using around 12%-13% of its power during scene parsing. Is this normal? What is your experience?
In my view this is too slow, and I think there must be an issue, maybe the setup of my scene or c4d/Corona.
So I’m interested if you experienced something similar and how you deal with it.
I did some tests as well. RAM usage less than 50%.
First test: Multithreaded parsing off.
Geometry parsing: 0:09
Scene parsing: 8:55
Second test: Multithreaded parsing on.
Geometry parsing: 0:08
Scene parsing: 8:55
I would expect a speed difference here. Is that how it works or maybe an issue?
Win 10
Latest daily
R20
Another issue by the way: Put a proxy under a clone object and change it to multi-instance. It renders fine in IR, but it doesn’t show up in PV.
@btw issue: What is your visualization method? For wire/solid bounding, full mesh, I can see the objects in the viewport, for point cloud I cannot - I reported that internally (Internal ID=631403480).
I’ll let others talk about the performance thing.
Jan
What I mean: When I render them in IR they are visible as expected. But they disappear in final render.
See the right side is completely black, no proxies are rendered, they disappeared.
About the performance thing: Parsing time is now much shorter than yesterday. (less than 1min, compared to almost 9min yesterday) I’m not sure what the main reason for that is. Maybe computer restart, maybe some optimizations of the scene.
I also noticed that parsing time for IR is longer than for final render. Hope I can do some testing and give you some feedback.
Hi, In regards to your proxy scene, can you share it with us and let us know the file name? I have tried to recreate this but it’s working as expected. I’m also using the latest V7 daily and Win 10 Pro & S22.
That does seem to be the issue! Thanks for telling us, I can reproduce the issue on R20 but not on R21+ and <R20 do not have multi-instance, so a very deserted island, this one.
I’ll report it internally (Internal ID=632122022)
Hey, I looked into it and there is some foul play afoot - I reported it to the developers.
Try hooking the Fusion into the UVW randomizer rather than the UVW randomizer to the Fusion, maybe you’ll get better results that way.