Pixel Plow and GI Caches

[digitalguru]

Hi, me again  with some GI cache questions

I see Pixel Plow can generate GI caches as part of the render (and does it automatically, it seems) Do most of you guys using PP let this run automatically or supply your own caches? It would all add to the total job time.

I bit the bullet and upgraded my Internet connection, so the size of the upload isn't an issue anymore.

PP defaults to a cache per frame, but would a sparse cache be better? Most of the shots have a fairly fast moving camera and I'm thinking of creating a cache every 5 frames with Number of files to blend set to 3 as a starting point for tests.

Does supersampling have a positive or negative effect when using GI caches?

It's worth mentioning that there are no clouds in the scene (sky is turned off in fact) so caching Easy clouds is not an issue.

[Oshyan]

I usually use PP for GI cache, the prepass is usually fairly quick compared to final render so cost/time addition is often not much. They do support sparse caches and I recommend people use that in almost all cases. PP has docs on doing sparse cache but basically you can type something like "1:100 by 5" and it'll do every 5th frame. Pretty cool.

I always use Supersampling. Always. Matt has feelings about this, but has not demonstrated compelling reasons not to.

- Oshyan

[WAS]

I always use Supersampling. Always. Matt has feelings about this, but has not demonstrated compelling reasons not to.

- Oshyan

Have you played with fine details? When I used supersample on grainy sand like displacement, the final result doesn't look as rough or detailed with shadows and such.

[digitalguru]

Thanks again guys, all good info.

Have you played with fine details?

I will test supersampling this week, most of the shots are quite motion blurred so that may not be an issue.

Another quick question, do you have detail blending on by default - I did have it turned off for some of my tests and enabling it adds a fair bit to the render times. I did a quick test last night (on a smaller sample scene) with it turned off and didn't see any adverse effect - what should I be looking out for?