Scatter slow-downs?

[aaouviz]

I've been pretty ambitious with my scatterings lately and now I've got myself into a bind. My scenes are getting very very very laggy and it's really affecting my workflow...

I'm just inquiring which of the following might be causing massive slow downs (mostly in scene parsing before rendering, and during IR).

- Scatter with texture maps applied for the density (specifically a black/white noise map)?
- Scatters with includes/exclude splines. Does this slow things down a lot?
   - Would calculation theoretically be slower if the curve has more vertexes?
   - Would calculation theoretically be faster if I use only 1 curve object with multiple splines rather than multiple curve objects?
- Scatters with edge trimming? Does this cause a noticeable slow down?
   - Seeing as the edge trimming requires proxy objects preview to be set to 'full mesh' mode, does this also slow things down? (separate to the scatter preview mode, of course).

Or is it just a result of me using all these on 10-20 scatter items in my scene? Too ambitious?

Any insight much appreciated!

[NazarVitkovskyi]

Hello,

In general, using various scattering techniques shouldn't have a noticeable effect on performance and shouldn't cause slowdowns. But please remember that there can be exceptions. Performance may occasionally be affected, especially when using advanced scattering techniques in complex, with a large number of scattered objects or with a lot of scatter items, there may be a noticeable impact on performance.

I did some simple, brief tests with the different settings you mentioned above, and the difference in performance was between 2-4 seconds. But again, the more complex a scene is - the more time it requires.

By the way, you can always share a problematic scene with us for further analysis. Please use this link as a file uploader: https://upload.chaos.com/d/qa_corona@corona-renderer.com

Also, would be great if you could share your PC/laptop specifications.

[Ink Visual]

From our experience, using texture maps for density is slowing Chaos Scatter down significantly on the more complex steups.

[aaouviz]

Thanks for the replies.

I've also noticed in my attempts at fixing the slow downs that there is a noticeable difference in calculation times when a scatter is hidden (or on a hidden layer) and when that same scatter is not enabled. Ie; it's a lot slower when scatters are simply hidden.

Is this known?

I guess this might be more of a feature request; but is it possible that this is not the case?

I've built a script for myself to quickly turn off all scatters, but I sometimes need some on/some off etc and the best workflow means having them all enabled (but some hidden).