Secondary rays bias gretaer than 0 - Any drawback ?

[Pixelab]

I was wondering if there was a drawback to always put 0,001 for the value of the Secondary rays bias (instead of 0) I understand that the GI solution is not "perfectly" accurate with this parameter, bet when it's not barely visible, what's the problem ?

Any drawback in term of quality of your solution, render times, or other issues?

[percydaman]

sometimes it messes with your direct light shadows.

[vlado]

The problem would be if the offset is large enough to skip some surface in the rendering and cause incorrect shadows. For example, if you have a room enclosed with 1-surface walls only (no thickness), the ray bias might cause the rays to jump outside the room near corners producing bright edges.

Best regards,
Vlado

[Pixelab]

The problem would be if the offset is large enough to skip some surface in the rendering and cause incorrect shadows. For example, if you have a room enclosed with 1-surface walls only (no thickness), the ray bias might cause the rays to jump outside the room near corners producing bright edges.

Best regards,
Vlado

Ok, I understand. Thanks for the help

Hopefully, all my projects are modelled "volumetrically" but it may happen when importing models from crappy soft

[victor.nsy]

occasionally this happens

Dear all,

I occasionally get rendering with rays peeking thru my model even with
0.0 secondary ray bias ... is it due to the fact that my models may be intersecting each other? Turning on the check sample visiblity don't seem
to help ... if I am correct then it would be a modeling problem ... how can
I achieve a correct modeling tech. which is spoken here? I have always
modelled using minimum surface possible.

best regards,
Victor NSY