hughes b wrote:

> Hi there Johann,
> 
> Same as 'Raf256', I couldn't see your attached image (maybe that "8-bit"
> encoding?). 

Well, sorry for that unreadable attachment. I switched 8-bit-encoding off - 
hope this will fix it.

> [...]
> I really wonder about that problem of the missing photons in media if it
> isn't enclosed, something either not right or I missed a reason for it. I
> won't say it's a bug unless others find likewise.

Maybe this problem is from the same type as those "no laser effect"-threads 
I just discovered.

> Now, about that dispersion. Reason that looked sharp-edged for you I guess
> is because the default of 7 colors was used, so adding a
> dispersion_samples 14 would double that.

Thanks for that hint :)

> About that media again. I was surprised to find how the box container
> addition slowed it down terribly. I couldn't seem to speed it back up to
> it's original render-time, maybe something to do with the previously
> missing media photons and now it shows all bouncing around from reflection
> and refraction. Still was much, much slower than I'd have expected. So
> slow in fact I quit trying to check on it.

I guess, I will try it anyway. Using a linux-cluster should speed it up a 
bit. 
Speaking of clusters: Currently, I am using shell scripts to distribute the 
load. Unfortunately, this does not speed up the photon map creation. Has 
someone an idea how to parallelize photon maps?

> Only half helpful perhaps.

It was all the help I hoped for ;)


Thank You,

Hans

Post a reply to this message