Hi,
This is my last render trying to improve indoor lighting and radiosity.

The main parameters used in this scene are:
 Resolution: 1800x1200 (resized to 800x533)
 Antialias: +a0.1
 Lights: 3 spot_light area_light 5x5, 1 point_light (fill light)

Radiosity:
 count 50
 error_bound 0.5
 rec.lim. 1
(I've used low values for radiosity parameters in order
 to improve the render time)

Render time: 2h30m
CPU: 3.4GHz, 1GB RAM

Comments? :)

JSR

Post a reply to this message

Attachments:
Download 'bath.jpg' (247 KB)

Preview of image 'bath.jpg'

It is a nice scene.

Sven

Post a reply to this message

You can come and clean *my* bathroom any time! ;-)

Seriously though... nice atmosphere.

Post a reply to this message

Nice - what did you use for the toilet & taps etc?
Depressingly good modelling and a successful lighting test.

There's still something a little cg about this image though and I'm not 
sure if it's the lighting or textures looking a little too "clean" or 
just the perfectly straight lines on the tiles that are ever so slightly 
aliased. (that said a photo of that much perfect tiling might look cg 
anyway :-)

Post a reply to this message

Verm <pov### [at] thirteeendynucom> wrote:
> Nice - what did you use for the toilet & taps etc?
> Depressingly good modelling and a successful lighting test.

Hi,
All objects are made in AutoCAD, imported to the application I use (owner
application which can import AutoCAD models, render in OpenGL and add
lights and radiosity parameters, then export to .pov, .inc)

> There's still something a little cg about this image though and I'm not
> sure if it's the lighting or textures looking a little too "clean" or
> just the perfectly straight lines on the tiles that are ever so slightly
> aliased. (that said a photo of that much perfect tiling might look cg
> anyway :-)

I'm trying to improve my lighting techniques with key/fill lights and/or
radiosity, also trying to learn the correct parameters for materials
(reflection, refraction, etc..).

The only way I get good antialias is rendering a big image (>3000 pix
weight) and after resize. I always use just +a0.1, because other options
like +a0.1 +am2 +r7 +j takes too long to render and no improvement is
obtained.

Thanks for all comments,

JSR.

Post a reply to this message

"JSR" <jsr### [at] gmailcom> wrote:
> Hi,
> This is my last render trying to improve indoor lighting and radiosity.
>
> The main parameters used in this scene are:
>  Resolution: 1800x1200 (resized to 800x533)
>  Antialias: +a0.1
>  Lights: 3 spot_light area_light 5x5, 1 point_light (fill light)
>
> Radiosity:
>  count 50
>  error_bound 0.5
>  rec.lim. 1
> (I've used low values for radiosity parameters in order
>  to improve the render time)
>
> Render time: 2h30m
> CPU: 3.4GHz, 1GB RAM
>
> Comments? :)
>
> JSR

Hi JSR,

   Wow, I'm commenting on two bathroom images in the same week :-) I think
the modeling is exceptional, and you're doing well. Nice use of the towel
and soap bottle. I'd love to see more of those little touches.

The lighting could use some warming up, in my opinion
(same thing I saw with the other bathroom image). I'd recommend checking
out Jaime Vives Piqueres' excellent, and free LightSys at
http://www.ignorancia.org/en/index.php?page=Lightsys

Speaking of towels, Jaime also has some VERY interesting cloth renders on
his homepage :-)

Rock on,
George

Kick your render speed into high gear with MegaPOV XRS:
http://www.gammaburst.net/xrs

Post a reply to this message

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

JSR wrote:
> The only way I get good antialias is rendering a big image (>3000 pix
> weight) and after resize. I always use just +a0.1, because other options
> like +a0.1 +am2 +r7 +j takes too long to render and no improvement is
> obtained.
> 
	And with +a0.1 +am2 +r1 ? This should be faster than just +a0.1
with no quality loss (note that +am2 +r7 may take up to 1849 times
more samples than if you specify neither +am nor +r !! no wonder you
find it slow)

		Jerome
- --
+------------------------- Jerome M. BERGER ---------------------+
|    mailto:jeb### [at] freefr      | ICQ:    238062172            |
|    http://jeberger.free.fr/     | Jabber: jeb### [at] jabberfr   |
+---------------------------------+------------------------------+
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (GNU/Linux)

iD8DBQFFbzMdd0kWM4JG3k8RAoI+AJ0aP9mL8z9J1kE3ROAkARZRXc8fsACfZThY
AbNsFTivilfLjauRVDZMngg=
=tnwo
-----END PGP SIGNATURE-----

Post a reply to this message

JSR nous apporta ses lumieres en ce 30/11/2006 05:04:
> Verm <pov### [at] thirteeendynucom> wrote:
>> Nice - what did you use for the toilet & taps etc?
>> Depressingly good modelling and a successful lighting test.

> Hi,
> All objects are made in AutoCAD, imported to the application I use (owner
> application which can import AutoCAD models, render in OpenGL and add
> lights and radiosity parameters, then export to .pov, .inc)

>> There's still something a little cg about this image though and I'm not
>> sure if it's the lighting or textures looking a little too "clean" or
>> just the perfectly straight lines on the tiles that are ever so slightly
>> aliased. (that said a photo of that much perfect tiling might look cg
>> anyway :-)

> I'm trying to improve my lighting techniques with key/fill lights and/or
> radiosity, also trying to learn the correct parameters for materials
> (reflection, refraction, etc..).

> The only way I get good antialias is rendering a big image (>3000 pix
> weight) and after resize. I always use just +a0.1, because other options
> like +a0.1 +am2 +r7 +j takes too long to render and no improvement is
> obtained.

> Thanks for all comments,

> JSR.



+r7? That's 7 level of recursion. Exept in some very special cases, it don't 
improve anything. The default recursion level of 3 is usualy suficient. If not 
suficient, you can go up to 4, or 5 if you are VERY picky and pacient.

-- 
Alain
-------------------------------------------------
If Barbie is so popular, why do you have to buy her friends?

Post a reply to this message

Hi,
I think that LightSys (from Jaime Vives Piqueres) is a very good tool, but
at the moment I prefer to use my own light setup in the scenes, in order to
understand the parameters and their behaviour.

The antialias parameter: +a0.1 +am2 +r1 is faster than just +a0.1 and the
result has the same quality, thanks guys!

I've discovered the importance of good textures for walls and floor instead
of just simple tiles. The image attached is an example.

JSR

Post a reply to this message

Attachments:
Download 'bath2.jpg' (103 KB)

Preview of image 'bath2.jpg'

"JSR" <jsr### [at] gmailcom> wrote:
> Hi,
> I think that LightSys (from Jaime Vives Piqueres) is a very good tool, but
> at the moment I prefer to use my own light setup in the scenes, in order to
> understand the parameters and their behaviour.
>
> The antialias parameter: +a0.1 +am2 +r1 is faster than just +a0.1 and the
> result has the same quality, thanks guys!
>
> I've discovered the importance of good textures for walls and floor instead
> of just simple tiles. The image attached is an example.
>
> JSR

Wow! What a massive improvement in visual impact! (Thomas De Groot: mind if
I borrow your expression? ;-)) I can understand wanting to learn on your
own and not just wanting to plug someone elses system in all the time.
Speaking of that, I think Jaime has a good grasp of textures (which you're
obviously learning fast). His room and cloth renders might help you.

Render on,
George

MegaPOV XRS: Renders faster than light.
http://www.gammaburst.net/xrs

Post a reply to this message