Lots of photons here ....

Bruno

Post a reply to this message

Attachments:
Download 'experiment11.png' (413 KB)

Preview of image 'experiment11.png'

Am 11.02.2011 09:43, schrieb Bruno Cabasson:
> Lots of photons here ....

High dispersion count, too, I guess...

Looks awesome!

Post a reply to this message

clipka <ano### [at] anonymousorg> wrote:
> Am 11.02.2011 09:43, schrieb Bruno Cabasson:
> > Lots of photons here ....
>
> High dispersion count, too, I guess...
>
> Looks awesome!


ior = 1.5
dispersion = 1.1
dispersion_count = 768
gather_min = 128
gather_max = 512

127+ million photons (highly diluted far from lenses...)

I could not get rid of the interference fringes, probably due to spiral pattern
of the shooting. The solution could consist in making them sub-pixel, but this
means also lower resolution for a smooth result, or even more photons! Another
shooting method could avoid these fringes.

The actual diameter of the incoming beam (cylindrical light_source) is smaller
than the one rendered, and corresponds roughly to the center part of it. Due to
gathering?

Render time: about 8 hours on a dual Xeon 3Ghz. I had to mask the prism side
pointing to the camera, otherwise, rays get lost forever inside the prism.

Bruno.

Post a reply to this message

> clipka<ano### [at] anonymousorg>  wrote:
>> Am 11.02.2011 09:43, schrieb Bruno Cabasson:
>>> Lots of photons here ....
>>
>> High dispersion count, too, I guess...
>>
>> Looks awesome!
>
>
> ior = 1.5
> dispersion = 1.1
> dispersion_count = 768
> gather_min = 128
> gather_max = 512
>
> 127+ million photons (highly diluted far from lenses...)
>
> I could not get rid of the interference fringes, probably due to spiral pattern
> of the shooting. The solution could consist in making them sub-pixel, but this
> means also lower resolution for a smooth result, or even more photons! Another
> shooting method could avoid these fringes.
>
> The actual diameter of the incoming beam (cylindrical light_source) is smaller
> than the one rendered, and corresponds roughly to the center part of it. Due to
> gathering?
>
> Render time: about 8 hours on a dual Xeon 3Ghz. I had to mask the prism side
> pointing to the camera, otherwise, rays get lost forever inside the prism.
>
> Bruno.
>
>

Any jitter for the photons? It can greatly help getting rid of those 
"interference fringes".
Another thing that can help is to use more samples for your media. Here 
also, jittering can help somewhat.


Alain

Post a reply to this message

outstanding!

also, I feel your pain.

Post a reply to this message

Alain <aze### [at] qwertyorg> wrote:

> > clipka<ano### [at] anonymousorg>  wrote:
> >> Am 11.02.2011 09:43, schrieb Bruno Cabasson:
> >>> Lots of photons here ....
> >>
> >> High dispersion count, too, I guess...
> >>
> >> Looks awesome!
> >
> >
> > ior = 1.5
> > dispersion = 1.1
> > dispersion_count = 768
> > gather_min = 128
> > gather_max = 512
> >
> > 127+ million photons (highly diluted far from lenses...)
> >
> > I could not get rid of the interference fringes, probably due to spiral pattern
> > of the shooting. The solution could consist in making them sub-pixel, but this
> > means also lower resolution for a smooth result, or even more photons! Another
> > shooting method could avoid these fringes.
> >
> > The actual diameter of the incoming beam (cylindrical light_source) is smaller
> > than the one rendered, and corresponds roughly to the center part of it. Due to
> > gathering?
> >
> > Render time: about 8 hours on a dual Xeon 3Ghz. I had to mask the prism side
> > pointing to the camera, otherwise, rays get lost forever inside the prism.
> >
> > Bruno.
> >
> >
>
> Any jitter for the photons? It can greatly help getting rid of those
> "interference fringes".
> Another thing that can help is to use more samples for your media. Here
> also, jittering can help somewhat.
>
>
> Alain

I tried! But the result was always more or less splotchy. The only way I found
to get a smooth result is to use no jitter. It reduces the number of photons,
but the jitter pattern appears, mainly in highly diluted areas... It is just an
experiment.

Bruno

Post a reply to this message

"Bruno Cabasson" <bru### [at] alcatelaleniaspacefr> wrote:
> Alain <aze### [at] qwertyorg> wrote:

> > > clipka<ano### [at] anonymousorg>  wrote:
> > >> Am 11.02.2011 09:43, schrieb Bruno Cabasson:
> > >>> Lots of photons here ....
> > >>
> > >> High dispersion count, too, I guess...
> > >>
> > >> Looks awesome!
> > >
> > >
> > > ior = 1.5
> > > dispersion = 1.1
> > > dispersion_count = 768
> > > gather_min = 128
> > > gather_max = 512
> > >
> > > 127+ million photons (highly diluted far from lenses...)
> > >
> > > I could not get rid of the interference fringes, probably due to spiral pattern
> > > of the shooting. The solution could consist in making them sub-pixel, but this
> > > means also lower resolution for a smooth result, or even more photons! Another
> > > shooting method could avoid these fringes.
> > >
> > > The actual diameter of the incoming beam (cylindrical light_source) is smaller
> > > than the one rendered, and corresponds roughly to the center part of it. Due to
> > > gathering?
> > >
> > > Render time: about 8 hours on a dual Xeon 3Ghz. I had to mask the prism side
> > > pointing to the camera, otherwise, rays get lost forever inside the prism.
> > >
> > > Bruno.
> > >
> > >
> >
> > Any jitter for the photons? It can greatly help getting rid of those
> > "interference fringes".
> > Another thing that can help is to use more samples for your media. Here
> > also, jittering can help somewhat.
> >
> >
> > Alain
>
> I tried! But the result was always more or less splotchy. The only way I found
> to get a smooth result is to use no jitter. It reduces the number of photons,
> but the jitter pattern appears, mainly in highly diluted areas... It is just an
> experiment.
>
> Bruno

And I forgot: I used

global_settings {
  max_trace_level 15
  photons {count PHOTON_COUNT media PHOTON_MEDIA jitter 0 gather GATHER_MIN,
GATHER_MAX}
}

with PHOTON_MEDIA = 1e6

I think 1e6 is enough.

Bruno

Post a reply to this message