"gonzo" <rgo### [at] lansetcom> wrote:
> Jim Charter <jrc### [at] msncom> wrote:
>
> This is starting to look quite realistic.  As Steve said, it looks like a
> flash photo with backscatter from suspended particles in the water column.

I see you've stolen Gilles Tran's program which converts photographs to
scene files. :)

This would have to be the most realistic POV underwater image I've ever
seen.

Post a reply to this message

Jim Charter <jrc### [at] msncom> wrote:
> I have taken the approach of incrementally refining different aspects of
>   the scene rather than trying to perfect one thing, say the model,
> before proceeding to the next.  I am still tempted to use just a black
> bg for the scene though.

Bravo Jim, excellent image ! can't wait for the next step.
Maybe a grainy look should be interesting.
Why not try a black bg with a stronger and harsh light.

Rene
http://rene.bui.free.fr - online portfolio

Post a reply to this message

Jim Charter wrote:
> I have taken the approach of incrementally refining different aspects of 
>  the scene rather than trying to perfect one thing, say the model, 
> before proceeding to the next.  I am still tempted to use just a black 
> bg for the scene though.

well, i'm not much into "lifelike" critters in tracing, but this is 
nice. i like the chunks.

bob

Post a reply to this message

PM 2Ring wrote:
> "gonzo" <rgo### [at] lansetcom> wrote:
> 
>>Jim Charter <jrc### [at] msncom> wrote:
>>
>>This is starting to look quite realistic.  As Steve said, it looks like a
>>flash photo with backscatter from suspended particles in the water column.
> 
> 
> I see you've stolen Gilles Tran's program which converts photographs to
> scene files. :)
> 
> This would have to be the most realistic POV underwater image I've ever
> seen.
> 
That's quite a compliment!  Thankyou.

Post a reply to this message

Thomas de Groot wrote:
> "Jim Charter" <jrc### [at] msncom> schreef in bericht
> news:42dd4561@news.povray.org...
> 
>>I have taken the approach of incrementally refining different aspects of
>>  the scene rather than trying to perfect one thing, say the model,
>>before proceeding to the next.  I am still tempted to use just a black
>>bg for the scene though.
>>
> 
> Wow! Not only convincing, but terrifying! Very good job.
> 
Thankyou. As usual, what started out to be an easy exercise has quickly 
catapulted to the "edge" of my "envelop" of skills.

Post a reply to this message

Marneus Calgar wrote:

> 
>> I have taken the approach of incrementally refining different aspects 
>> of  the scene rather than trying to perfect one thing, say the model, 
>> before proceeding to the next.  I am still tempted to use just a black 
>> bg for the scene though.
> 
> 
> Looks very nice to me !
> Please don't put a black bg !!!!!
> 
Okay, but I gotta say, the tests done with a black bg are colourful and 
pretty in a way the media bg's are not.  So maybe I'll comprimise and 
promise not to post a black bg version without also doing a media version.

Post a reply to this message

Rene Bui wrote:
> Jim Charter <jrc### [at] msncom> wrote:
> 

> Why not try a black bg with a stronger and harsh light.

Many test runs are like that and it must be said they look very pretty. 
  The spot light with falloff alone gives a very nice sense of the fish 
emerging from the gloom and the color looks clear and crisp.

In this latest test I reduced the degree of large flopsum, added a fine 
grained pea and reduced the general scattering to almost black.  I also 
added real absorption to produce the color from the overhead light 
(formerly it was just a colored light overhead.)

Also, in response to gonzo I adjusted the whole scene so that the camera 
is now about 8 units from an 11 unit shark with a fade distance about 6 
units, fade power 7 or so and a lens and spotlight angle of around 60 
degrees.  The overhead light is about 100 units up with fade at 50 
units, power 4.  I also used the Rune/Shay contrast trick

There is also enough modelling changes in this one, mostly to the eye 
and vicinity and lower jaw, that I had to do a lot of work to fix the uv 
map.  But I am not entirely happy with the results so I attach a 
downscaled shot of the test here.

Post a reply to this message

Attachments:
Download 'img.00111.jpg' (52 KB)

Preview of image 'img.00111.jpg'

bob wrote:
> Jim Charter wrote:
> 
>> I have taken the approach of incrementally refining different aspects 
>> of  the scene rather than trying to perfect one thing, say the model, 
>> before proceeding to the next.  I am still tempted to use just a black 
>> bg for the scene though.
> 
> 
> well, i'm not much into "lifelike" critters in tracing, but this is 
> nice. i like the chunks.
> 
> bob
Thanks.  As I am sure you realize the chunks are just an added media 
layer with a granite density pattern that's mostly black.  Looks okay as 
long as a "chunk" doesn't overlap the figure.  Then the chunk looks 
faded for reasons I am clueless about.

Post a reply to this message

Jim Charter <jrc### [at] msncom> wrote:
> map.  But I am not entirely happy with the results so I attach a
> downscaled shot of the test here.

Jim, it's better and better. The environment, the lighting and modelling are
great and remoras add a life-like touch. The head is very convincing too.
But I think there is something wrong with the skin texture. It seems like a
sort of metallized plastic. Honnestly, I don't know how to fix it, maybe
with the normal-map (?). I think Grey-nurse (or Sand-tiger) skin have many
more spottings than you did. And also, the fins could be a bit darker with
color variations...etc.
http://www.dinosoria.com/diapo_requins/images/requin_taureau_jpg.jpg
http://www.dinosoria.com/diapo_requins/images/requin_taureau_002_jpg.jpg

Rene
http://rene.bui.free.fr - online portfolio

Post a reply to this message

Rene Bui wrote:
> Jim Charter <jrc### [at] msncom> wrote:
> 
>>map.  But I am not entirely happy with the results so I attach a
>>downscaled shot of the test here.
> 
> 
> Jim, it's better and better. The environment, the lighting and modelling are
> great and remoras add a life-like touch. The head is very convincing too.
> But I think there is something wrong with the skin texture. It seems like a
> sort of metallized plastic. Honnestly, I don't know how to fix it, maybe
> with the normal-map (?). I think Grey-nurse (or Sand-tiger) skin have many
> more spottings than you did. And also, the fins could be a bit darker with
> color variations...etc.
> http://www.dinosoria.com/diapo_requins/images/requin_taureau_jpg.jpg
> http://www.dinosoria.com/diapo_requins/images/requin_taureau_002_jpg.jpg
> 
Hi Rene, I truly appreciate the support.  It's at this late stage in a 
work where the final 10% of quality seems to cost 90% of the effort that 
my determination usually breaks down.

One thing I may need to do is look up a post someone recently made about 
using jpeg's and their gamma adjustment, because in fact you wouldn't 
believe how dark and contrasted this color image_map actually is.  It's 
to the extent that while editing it I can just barely see the markings I 
am adding.  Also, I am having some better success if I also average the 
markings from the bump_map layer into the image_map layer so that some 
darkness corresponds to the depressions.  But it comes at a cost.  That 
merging process tends to soften the markings that were intended to show 
the color changes alone. But you photos seem to show that I am getting 
fairly close to the overall color.

Right now the bump_map is also being averaged with a procedural granite 
normal. This is meant to show the overall skin texture under the random 
gouges. I have spent several evenings experimenting with different brush 
definitions when drawing the bump_map.  It has resulted in improvements 
but it is not yet "right"  Though actually it shows up better at higher 
resolutions.

As far as the finish goes, part of the problem is that the metallic 
look, which is mostly do to the high specular I think, is close to how 
it actually looks to me in different photo references and on some TV 
documentaries I've been seeing.  I tried to use a texture_map with a 
slope pattern pointed at the camera to get the specular to fall off more 
dramatically but unfortunately that increases the render time 
exponentially.  I also tried a layered texture to get more control over 
the color markings you mentioned but that also makes the render time 
untenable.  Right now the little-sphere-over-the-camera trick is what I 
am resorting to to increase the apparent contrast.  Also, in the render 
I showed you, I may not have adjusted the fade-power enough to 
compensate for the lessening of scattering and extinction.

Anyway I am taking a break from the skin texture issues and writing a 
macro to generate some little fishes in various random postures.

Thanks again for taking the time.

-Jim

Post a reply to this message