Kari I've been experimenting with your suggestions on the two-pass radiosity
method, and I think I'm starting to get some decent results. I've made a
handy .inc file to switch between passes. Also, I applied the film curve in
photoshop and I think thats what really made the image come into its own.

George Pantazopoulos

Post a reply to this message

Attachments:
Download 'Athene_test1.jpg' (79 KB)

Preview of image 'Athene_test1.jpg'

> Kari I've been experimenting with your suggestions on the two-pass
radiosity
> method, and I think I'm starting to get some decent results. I've made a
> handy .inc file to switch between passes. Also, I applied the film curve
in
> photoshop and I think thats what really made the image come into its own.

Here is a version after some levels adjustment in photoshop I think it looks
better.

Post a reply to this message

Attachments:
Download 'Athene_test1_contrastcorrect.jpg' (26 KB)

Preview of image 'Athene_test1_contrastcorrect.jpg'

> Here is a version after some levels adjustment in photoshop I think it
looks
> better.

I prefer this one too, it looks at least just as good as brazil, though
still not quite realistic (?).. But why should it be necessary to use a
third party program, when POV can adjust levels on it's own, with
assumed_gamma plus the usual texture / light setup.

Regards,
Hugo

Post a reply to this message

>But why should it be necessary to use a
> third party program, when POV can adjust levels on it's own, with
> assumed_gamma plus the usual texture / light setup.

1. I didn't create the image.

2. I'm not as familiar with pov, giving that I use Moray a lot, as I should
be.

Post a reply to this message

George Pantazopoulos wrote:
> 
> Kari I've been experimenting with your suggestions on the two-pass radiosity
> method, and I think I'm starting to get some decent results.

Looks nice and smooth. Too much smoothing will erase detail though.


_____________
Kari Kivisalo

Post a reply to this message

Hugo wrote:
>
> But why should it be necessary to use a
> third party program, when POV can adjust levels on it's own

If a hammer is the only tool available, everything begins to look like a nail.


_____________
Kari Kivisalo

Post a reply to this message

Thomas, yes that looks better, it appears that you upped the contrast?

George

"Thomas Lake" <tla### [at] REMOVE-THISshawca> wrote in message
news:3ccf8c36@news.povray.org...
> > Kari I've been experimenting with your suggestions on the two-pass
> radiosity
> > method, and I think I'm starting to get some decent results. I've made a
> > handy .inc file to switch between passes. Also, I applied the film curve
> in
> > photoshop and I think thats what really made the image come into its
own.
>
> Here is a version after some levels adjustment in photoshop I think it
looks
> better.
>
>
>

Post a reply to this message

Kari,
   Yes youre correct. I've discovered ( in Pass 2) to get less smoothing,
you need to increase count, or else there can be some strange artifacts
(such as black or white lines at borders). With error_bound of 0.7 I've gone
as low as count 10 without artifacts (depends on the image) For this image I
used error_bound 0.400 and count 150. A Count of less than 100 tended to
produce artifacts around the feet. Raising count in Pass 2 does not seem to
improve image quality but it does prevent artifacts from appearing, so its a
matter of finding the magic number it seems. Its a balancing act because a
larger count slows down Pass 2
All the above info is for the second, quick smoothing pass. For Pass 1 I
used count 200, error_bound 0.180, recursion_limit 12 (both passes),
nearest_count 1 (both passes), and brightness 1.4 (both passes).

Thanks,
George Pantazopoulos


Kari Kivisalo" <pro### [at] luxlabcom> wrote in message
news:3CCFAE99.83E902B5@luxlab.com...
> George Pantazopoulos wrote:
> >
> > Kari I've been experimenting with your suggestions on the two-pass
radiosity
> > method, and I think I'm starting to get some decent results.
>
> Looks nice and smooth. Too much smoothing will erase detail though.
>
>
> _____________
> Kari Kivisalo

Post a reply to this message