Indigo Renderer Forum

something's a bit wrong, I'm sure.
the cylinders have SSS material. I'm sure that SSS converges faster when using the default render settings from blendigo.
I haven't got it installed at the minute, so I can't check.

Time elapsed: 1 h, 52 m, 14 s
Done 105310000.00000 samples (219.39583 samples per pixel)
15638.04184 samples / second (63.94663 micro-seconds / sample)

I would not be that positive, depends a lot on your advanced settings as well. BTW I always used Hybrid MLT till now, so my expertise is irrelevant. Objs or internal export ? I'd like to check your case, have you gott a ma to check please ? Is that .9t9 ?

It's the advanced settings I was wondering about actually.
For my own reference, the blendigo settings are:
LM probability 0.1
max change: 0.02
MNCR: 100
max depth: 1000
Supersamples: 2
hybrid threshold: 500000

I'll compare these when I get home.

ok, i committed a version to SVN with these settings, and also a test scene for SSS.
https://svn.hamsterfight.co.uk/listing. ... h=%2F&sc=0

this the SSS test scene. much more like what I was expecting to see yesterday.

as that's from blendigo, you should use an edgesplit-modifier

Kram1032 wrote:as that's from blendigo, you should use an edgesplit-modifier

no, it's not from blendigo. It's Maya -> MtI -> Indigo, but wuth the same default render settings as blendigo uses.

I've yet to fiddle around to emulate edgesplit with Maya (see the other thread).

ah, yeah, I already read that^^

although, the problem might actually be that the ground plane and the bottoms of the cylinders are occupying the same space.

edit: yes, this is the problem.