 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Am 06.08.2013 09:54, schrieb Thomas de Groot:
> Not as easy as I thought (naively) ;-)
>
> This is the best I can do presently using Poser's Simon, by commenting out:
>
> [0.2 P_Iris_Shadowed]
>
> in the material, otherwise the iris becomes totally black. Also, the
> texture is not right even then.
>
> This probably has to do with /how/ the iris is uv_mapped in the first
> place. I shall have to go back to the figure and work upwards to see
> where I need to tweak things.
The current DAZ people use a UV mapping scheme for the eye textures that
has the irises in the bottom left and bottom right quadrant (each one
almost filling the quadrant entirely), while the sclerae separately
occupy the top left and top right quadrant.
The Poser people (at least G2) instead use a scheme where the irises sit
amid the sclerae in the left and right half of the texture; thus the
irises are at a totally different place, and also a lot smaller.
This might help to get you started:
-----------------------------------------
#declare M_Iris = material {
texture {
pigment {
uv_mapping
spherical
pigment_map {
[0.2 P_Iris_Shadowed]
[0.3 P_Iris_A ]
[0.4 P_Iris_A ]
[0.5 P_Iris_B ]
[0.6 P_Iris_B ]
[0.65 P_Iris_Shadowed]
}
scale 0.5 // <-- each copy of the iris is smaller
translate <1,2,0> // <-- iris centers are at different y
warp { repeat 2*x }
scale 0.25
}
finish { ambient 0 diffuse albedo 0.6 specular albedo 0 emission 0 }
}
}
-----------------------------------------
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thanks Christoph, I think I got it now with the following changes:
//--------------------------------------
#declare M_Iris =
material {
texture {
pigment {
uv_mapping
spherical
pigment_map {
[0.1 P_Iris_Shadowed]
[0.2 P_Iris_A ]
[0.4 P_Iris_A ]
[0.5 P_Iris_B ]
[0.6 P_Iris_B ]
[0.65 P_Iris_Shadowed]
}
scale 0.37
translate <0.99, 2.075, 0>
warp { repeat 2*x }
scale 0.25
}
finish { ambient 0 diffuse albedo 0.6 specular albedo 0 emission 0 }
}
}
//--------------------------------------
While this works for SimonG2, tweaking may be probably be necessary for
the other figures.
Note that a brownish line around the iris may appear. This comes from
the sclera map and shows a little part of the original iris as the iris
mesh does not cover exactly the image (or vice versa). This can
relatively well be solved by scaling down and translating a tiny bit the
pigment in the sclera material. This is done in the present image.
Thomas
Post a reply to this message
Attachments:
Download 'poserfiguretest.png' (409 KB)
Preview of image 'poserfiguretest.png'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Am 05.08.2013 23:45, schrieb Alain:
>> It's probably one of the best iris I've ever seen, just next to some
>> actual iris photos...
>
> Thanks. I was actually quite surprised how easy it was to achieve.
>
>> Care to share it?
>
> Here it is; note that it is a UV-mapped texture designed to fit the DAZ
> Genesis eyes; for other models, the main material (M_Iris) may need some
> tweaking. Also note that the DAZ Genesis model has separate geometry for
> the iris and the cornea; if this is not the case for your model, you
> will also need to tweak the finish.
>
> (Ah, and last not least, the slight "cat's pupil" effect in the image I
> posted is in the geometry, not the material.)
>
The iris been distinct from the cornea surely play an important role in
the beleivability.
Alain
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Two more versions. I'm not sure which lighting setup I like better - but
I do know that I'd really love to have MCPov's "render until I'm ok with
the noise" mode of operation right now... and/or faster SSLT, for that
matter.
Post a reply to this message
Attachments:
Download 'young_half_elf_woman_pov_scene 2013-08-07 0925.png' (481 KB)
Download 'young_half_elf_woman_pov_scene 2013-08-08 1300.png' (950 KB)
Preview of image 'young_half_elf_woman_pov_scene 2013-08-07 0925.png'
Preview of image 'young_half_elf_woman_pov_scene 2013-08-08 1300.png'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 8-8-2013 13:11, clipka wrote:
> Two more versions. I'm not sure which lighting setup I like better - but
> I do know that I'd really love to have MCPov's "render until I'm ok with
> the noise" mode of operation right now... and/or faster SSLT, for that
> matter.
>
I would prefer the first (0925) maybe with a tiny bit more fill-in on
the right. The other one is a bit too flat to me. Also, I am not too
sure about the dof in that case. Too extreme for my taste.
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
clipka <ano### [at] anonymous org> wrote:
> Subsurface scattering would probably not be necessary for a human
> character, but being a half-elf I might want her skin to have some
> surrealistic translucency, so I'll give that a try.
Are you kidding?
To me, skin is the only thing that really *needs* subsurface scattering,
especially on the face, which is one of the targets the human eye has been most
used and adjusted to scrutinize on earth under various lighting conditions.
Very nice procedural iris, considering it is an even more demanding challenge in
that respect. Do you mind if I use this Iris texture for the sample texture on
the blender to POV exporter WIKI? or if I have to adapt it as a basis for that.
Faster SSS would be more than awsome, actually a grail for attracting animation
focused users along with motion blur.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
>> Subsurface scattering would probably not be necessary for a human
>> character, but being a half-elf I might want her skin to have some
>> surrealistic translucency, so I'll give that a try.
>
> Are you kidding?
> To me, skin is the only thing that really *needs* subsurface scattering,
I remember reading a good article in GPU gems:
http://http.developer.nvidia.com/GPUGems3/gpugems3_ch14.html
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Am 08.08.2013 13:38, schrieb Thomas de Groot:
> On 8-8-2013 13:11, clipka wrote:
>> Two more versions. I'm not sure which lighting setup I like better - but
>> I do know that I'd really love to have MCPov's "render until I'm ok with
>> the noise" mode of operation right now... and/or faster SSLT, for that
>> matter.
>>
> I would prefer the first (0925) maybe with a tiny bit more fill-in on
> the right. The other one is a bit too flat to me.
Surprisingly enough, the first one uses a HDR light probe of an overcast
sky (albeit in a park with some trees around, so there is /some/
uniformity to the illumination), while the latter one uses a light probe
of a clear sky. (I suspect the creator of the probe didn't include a
shot with a short enough exposure time to capture the true brightness of
the sun; I'll try again with an added light.)
> Also, I am not too
> sure about the dof in that case. Too extreme for my taste.
I guess you're right. Then again, the acceptable dof might also depend
on how detailed the background is.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
clipka <ano### [at] anonymousorg> wrote:
> Two more versions. I'm not sure which lighting setup I like better ...
The darker one (2013-08-07 0925), definitely.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 8-8-2013 16:15, clipka wrote:
> Am 08.08.2013 13:38, schrieb Thomas de Groot:
>> Also, I am not too
>> sure about the dof in that case. Too extreme for my taste.
>
> I guess you're right. Then again, the acceptable dof might also depend
> on how detailed the background is.
>
It is mostly the out-of-focus part of the figure itself that I question.
With a more detailed background, I would prefer the figure to be sharp
front to back, or maybe only a little out of focus in the farthest part.
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
