Began testing with all white pigment and default finish settings.  Made 
some  monochrome changes to the pigments, added different finishes and 
normals.  This is basically the OutdoorLQ settings from rad_def.inc with 
the count boosted 50% to 120.

The light-generating sphere is this:

sphere { 0, 2000 hollow
         pigment { average
                 pigment_map {
                        [ 1
                                 gradient y
                                 pigment_map {
                                         [ 0 rgb  20 ]
                                         [ 1 rgb -25 ]
                                 }
                                 scale 4000
                                 rotate x*180
                                 translate <0,-2000,0>
                        ]
                        [ .5
                                 spotted
                                 pigment_map {
                                         [ 0 rgb   25 ]
                                         [ 1 rgb  -15 ]
                                 }
                                 scale 900
                        ]
                 }
         }
}
The first image, "956", is under v 3.5  the second, v 3.6

Post a reply to this message

Attachments:
Download '956.jpg' (86 KB) Download '960.jpg' (87 KB)

Preview of image '956.jpg'

Preview of image '960.jpg'

Still nice shoes :-)

> The first image, "956", is under v 3.5  the second, v 3.6
>
I think the 1st one is the top one

I prefer it: It is cleaner on the edges, it is mostly noticeable on the
belt.

Marc

Post a reply to this message

Jim Charter wrote:

> The first image, "956", is under v 3.5  the second, v 3.6

There is very little difference in lighting quality but did you use same 
AA settings and same image output size? As Marc pointed out these is 
some jagged lines in the 3.6 image.

Severi

Post a reply to this message

Severi Salminen wrote:
> Jim Charter wrote:
> 
>> The first image, "956", is under v 3.5  the second, v 3.6
> 
> 
> There is very little difference in lighting quality but did you use same 
> AA settings and same image output size? As Marc pointed out these is 
> some jagged lines in the 3.6 image.
> 
> Severi
Yes exactly the same rad settings & run parameters.  I should explain 
that I ran the 3.6 example mostly as an afterthought.  To compare the 
two versions wasn't really the intention of the post. Mostly I was just 
posting because I'd spent so much time bringing the lighting as far as I 
had. But since radiosity was advertized as among that changes to the new 
release I thought I would try it in comparison.

I thought the main difference was the superior rendering of the 
highlighted buckle in the 3.6 example.

I am not very pleased with either outcome actually, And I am not quite 
sure where to turn next.

And in particular I have a difficult time understanding the dramatic 
difference in finish responsiveness between either of those tests and 
this one using HDR.  Obviously I have a lot to learn and running mind 
numbing numbers of tests trying to isolate the effects of contributing 
elements seems to be the only way. :)

Post a reply to this message

Attachments:
Download '999.jpg' (121 KB)

Preview of image '999.jpg'

"Jim Charter" <jrc### [at] msncom> wrote in message
news:40cd57ff@news.povray.org...

Excellent model.

I agree with you on the buckle, and the jagged lines in the 3.6 are I'm
sure easily corrected. Hard to tell if on the whole this in quite there
yet without the textures. There were areas in your "shoe study" post
(like the laces) which were truly eye catching and IMO perfectly
rendered. I can't see anything like that here yet, but, again, that may
just be the lack of textures.

The HDRI version looks bad, IMO. I much prefer the classic studio look
of the radiosity renders.

<tangent warning>
There is a technology which can take several frames of a crappy, VHS
security video and, by selecting only those pixels which demonstrate
integrity by remaining constant across several frames, produce a clear
picture of a license plate. The reverse effect can be seen when pausing
a recorded VHS tape of a semi-staticy television program. The moving
picture might have made sense, but the paused frame can be
unrecognizable. In this way, I think that studio lit (or even "flat")
pictures, while seeming less photo-real, in reality give a *more*
accurate picture of how our minds perceive a specific object. We can
more easily IMO perceive the fixed dimensions of length, width, and
height without the distraction of transient "dimensions" like time and
position.
</tangent>

 -Shay

Post a reply to this message