Stefan Viljoen <sviljoen@<removethis>polard.com> wrote:
> This image took about one year to render, first with a 1.8 system running
> 24/7 for about eight months, then a 3 gHz machine running for the remaining
> four, 24/7.
>
> Sorta an experiment (as regards isosurfaces, media, and media quality
> settings) that failed. I thought it would look nicer.
>
> Clearly, longer (or VERY long) trace times don't neccessarily lead to better
> images...
>
> Just thought I'd share the disappointment!
> --
> Stefan Viljoen
> Software Support Technician / Programmer
> Polar Design Solutions

Take heart Stefan - the road to media mastery is paved with patience, and I
am amazed at how patient you are.  I am currently 25% into the rendering of
a crystal object using photons, media, and radiosity.  With luck it will be
finished four more days, but I am already itching to abort and start
tweaking. I guess it's time for a multi-processor system (or several of
them!)

Post a reply to this message

Lonnie spake:

> Stefan Viljoen <sviljoen@<removethis>polard.com> wrote:
>> This image took about one year to render, first with a 1.8 system running
>> 24/7 for about eight months, then a 3 gHz machine running for the
>> remaining four, 24/7.
>>
>> Sorta an experiment (as regards isosurfaces, media, and media quality
>> settings) that failed. I thought it would look nicer.
>>
>> Clearly, longer (or VERY long) trace times don't neccessarily lead to
>> better images...
>>
>> Just thought I'd share the disappointment!
>> --
>> Stefan Viljoen
>> Software Support Technician / Programmer
>> Polar Design Solutions
> 
> Take heart Stefan - the road to media mastery is paved with patience, and
> I
> am amazed at how patient you are.  I am currently 25% into the rendering
> of
> a crystal object using photons, media, and radiosity.  With luck it will
> be finished four more days, but I am already itching to abort and start
> tweaking. I guess it's time for a multi-processor system (or several of
> them!)

Thanks - well, if I can scrape together the cash for another HT P4, I should
be alright. Then I can throw 4 processor at a trace, not just 2 as I have a
the moment... oh yeah!

I think I maybe have a bit too much patience thought - 4 days for a
media/photons/radiosity trace doesn't sound too bad.

-- 
Stefan Viljoen
Software Support Technician / Programmer
Polar Design Solutions

Post a reply to this message

> Looks kind of like you used a focal blur, did you?  If you take that out
> your render would probably go faster and your results a bit better, but
> that's my opinion.


Hee Hee.  And might even shave a WEEK of the render time.  Got to be worth a 
try!  Hee Hee

Post a reply to this message

What was the resolution of the render?
Why would not  render it in a thumb size(320x200, 160x... ) first, just to
check?

That is really maniacs idea to run that long without knowing the output.

> This image took about one year to render, first with a 1.8 system running
> 24/7 for about eight months, then a 3 gHz machine running for the
> remaining four, 24/7.
> 
> Sorta an experiment (as regards isosurfaces, media, and media quality
> settings) that failed. I thought it would look nicer.
> 
> Clearly, longer (or VERY long) trace times don't neccessarily lead to
> better images...
> 
> Just thought I'd share the disappointment!

Post a reply to this message

stingo spake:

> What was the resolution of the render?
> Why would not  render it in a thumb size(320x200, 160x... ) first, just to
> check?

Because even THAT would have taken about a week (or two...)

:0
 
> That is really maniacs idea to run that long without knowing the output.

I am a maniac!
 
-- 
Stefan Viljoen
Software Support Technician / Programmer
Polar Design Solutions
http://polard.com

Post a reply to this message