In my experiments so far I've found that I get the color bleed from the Envsphere when IDL strength is between .35 and .4. This is a shame because I don't want that much color bleed. It would especially look bad for outdoor scenes using a sky image. Everything gets a blue tinge to it where is in reality our eyes white balance most of the blue out. I'm currently rendering at IDL intensity of .25 to .3. Anyone else's thoughts?

Change the colours/images on the EnvSphere?

I've been considering taking the image for the Envsphere and doing a greyscale on it or desaturaing it and if its and outdoor scene that shows some sky then I'd use a backdrop. Most of my images are indoor glamour shots but for another project I'll be doing outdoor images against apocalyptic broken buildings.

There's a saturation control in my EnvSphere shader - it's the HSV node. You don't need to modify the image outside Poser.

I GOT A SHADOW SPOT PROBLEM.

Here is an image of a necklace and bikini string around the neck area. There are numerous shadow spots.

• there is no AO on the figure or on the lights
• I've tried increasing the IDL samples
• turning off all shadows on all lights,
• turning off the light emitters on the figures/clothing,
• turning off cast shadows on figures/clothing.
• I've tried increasing the environment sphere size,
• i've tried stripping the bbenvsphere of any nodes on the  material and setting diffuse to 1.
• I've tried turning off SSS.
• I've tried stripping the figure down to the most basic nodes(texture only).

I've tried everything to possibly fix this but they won't go away. 

Increase the Irradiance Cache and experiment with your shading rate.

the scene may render better with GC checked.  then you can optimise various samples and shading rate.

Quote - the scene may render better with GC checked.  then you can optimise various samples and shading rate.

I'd rather fix the problem at the source than use a bandaid. 

So far by turning off visible in raytracing for bikini top & necklace has solved the issue but I'd rather not resort to that.

Quote - Increase the Irradiance Cache and experiment with your shading rate.

Well the IC is max, so I'll try lowering the shading rate.

I meant the Irradiance Cache under the IDL setting.

Quote - I meant the Irradiance Cache under the IDL setting.

Oh I see.  Man I'm trying that right now and the render time has skyrocketed. Maybe I can find a sweet spot.

Problem is still there :(

Damn man, seven days of hard work and the last issue seems to be impossible to fix unless I just turn off the raytracing on the conformed figures.

Just out of interest Zanzo, what does it look like with higher Pixel Samples = 7, Shading Rate=0.4, IDL Bounces=12, IDL Irradiance Cache=20?

Those are my standard settings... oh yeah, with IDL samples=200. I don't tend to find I need to increase IDL IC past 32.

I get pretty fast renders at those settings, in a simpler scene at least, and not too much in the way of blotchies.

My render times tend to go through the roof because of scene complexity I think. Hence I do try to stick to low IDL IC.

But, my lighting will play a part in that too... I suspect that I tend to override the IDL occlusion with ray traced shadows from the low intensity Point lights I use alongside any emitters...

In my experiments so far I've found that I get the color bleed from the Envsphere when IDL strength is between .35 and .4. This is a shame because I don't want that much color bleed. It would especially look bad for outdoor scenes using a sky image. Everything gets a blue tinge to it where is in reality our eyes white balance most of the blue out. I'm currently rendering at IDL intensity of .25 to .3. Anyone else's thoughts?

I'd have to see an example but "color bleed" is an art that Poser has just adopted with IDL. Impressionists brought it to light when they began to carry all their color pallette throughout the entire canvas. I would say we need more of that. I struggle with hdr quality, though and results are never predictable at first.. As bb said it's easy to desat them or just lower intensity.

The big issue for me here is where to clip the reflection intensity on EZskin.  I'm doing it one at a time. I was hoping snarly would have jumped in by now. May have to wait untill after the holiday... IDL intensity,, I love the logic. I'm using it but have not done a proper set of tests yet. My range has been 0.15 to 0.3. This is neither here nor there. This is getting us into a better realm for light,,, no doubt. Great stuff.

My concern with this color bleed has been in regards to creating a character/skin texture set potentially for the marketplace. I need to show the color of the skin texture as I created it. The one I'm working with now is pinkish. But I'm getting some orange in it at .3 to .4 IDL. I don't have many HDR images in my panoramics folder so if you've got some links (yep, looked at BB's links), it would be much appreciated!

My concern with this color bleed has been in regards to creating a character/skin texture set potentially for the marketplace.

There's your answer. You need to keep it in a pristine studio eviroment. A white room with direct Poser light or emitters. No envsphere unless its a desaturated fill light.

Like a photographer. I think kobaltkween is working on such an enviroment. Oh and sIBLs might give you more predictable results .

I am trying a .5 desaturation on the HDR I'm using and doing a test render to see what I get. I'd be interested in that project of kobaltkween's.

In my observations the environment contributes a lot of colour to surfaces via bleed & ambient & bounce light. IMO it's a mistake to try to eradicate it.

Take skin, for example. It takes on a different appearance on a cloudy day than in bright sunlight with a blue sky, and whether a person is stood in the shade of a tree or next to a red-brick building. So much of these variances in appearance are due to reflected light of one sort or another.

It is difficult to be scientific about the degree to which sources such as HDRi's and the systems we're using to simulate light in 3d are delivering this light (and colour) though. We sometimes look at a render and think it looks all wrong, but we're making assumptions based on the relationships and balances between surfaces that might not be at all well simulated. Sometimes it is because the systems are indeed out of whack, as bb has shown us with the relationship between diffuse and IDL in Poser.

I've been trying to re-examine the appearances of surfaces with regard to diffuse, reflection/specular. I think sometimes it's confusing that we mentally seperate light and surface properties as such. In Poser we think of reflection in terms of mirrors, metal and water. But in a way what we're simulating is all reflected light (that light that eventually reaches our eyes anyway) - that's how I look at it. The difference is the way in which and how much of that light is reflected by a surface. This gets incredibly complex in the case of skin.

My understanding isn't particularly scientific but it is based on obsevation. I'm very aware of the relationships between colours and contrasts in tones. That's why I know it's very easy to be decieved by what you're looking at. Take a colour picker over the surface of a photo in an image editor and you'll see some interesting colours in shadows and directly lit areas. You have to wonder what's influencing some of that colour - local surface colour or environmental light.

In the case of making a diffuse skin texture, I think we need to start deciding what system of simulating surfaces these textures are designed for - simple or complex. We have to be willing to accept that a diffuse map might not work outside the realm of use with SSS and fresnel reflection with environmental lighting. I have some diffuse maps that are virtually pale-grey with a yellowish tint - look like zombie flesh. But used with an SSS colour map and scatter they do a great job.

...like zombie flesh

That made me smile for some reason,,, probably because I had just read you post...;)

Quote - ...like zombie flesh

 

That made me smile for some reason,,, probably because I had just read you post...;)

i see Zombie flesh most times I look in a mirror. Too much sitting in dark rooms watching renders computing.

i see Zombie flesh most times I look in a mirror. Too much sitting in dark rooms watching renders computing.

^lol   looking in the mirror... nevermind. I'm sure it was a "funhouse" mirror at some point in its life.

Daniel, Are you using the original 240 shader? I meant to ask that earlier.

No. I'm tinkering with different ways of blending 2 or more scatter nodes at present, driving them with various other nodes & maps.

For the Lee-Perry Smith render I posted earlier I went back to a setup that really has no place in any concept of a physically based skin (the one I've posted here). It's incomplete as none of the reflection or specular is being modified by spec maps.

The blinn and reflection are derived from the 240 shader (that's the one that bb posted at RDNA ages ago, right?), but I'm using 2x scatter nodes out of a blender driven by...a diffuse node. Basically it gives me a scatter variation based on the diffuse light/shadow - higher scatter scale on the lit side, lower on the shadow side (so I get more scatter where there's less diffuse light). It seems to work quite well in brighter setups with the EnvSphere. Not really tried it in other scenarios. It's a fudge.

As note on that shader I just posted, what I like about it is that I can have two seperate maps with different colours and levels of detail that appear more prominently depending on whether there is more or less diffuse light - that's the theory anyway.

It's more an artistic interpretation of the layering of skin rather than a physical simulation - a half-way house perhaps. That render was with IDL intensity at 0.4, with EnvSphere intensity about 2x normal and fresnel reflection reduced to balance.

None of this would be possible without the D3D render script - essential rendering tool.

Very interesting interpretation, Dan.  This is similar to the way photographs capture all tones and colors that the human eye misinterprets because of selective optical observation.  The shock of Nature's absolute truth.  It would seem that there is much to learn on how to manipulate Poser into rendering an image with artistic qualities which looks correct to our interpretation rather than delivering a very realistic stark reproduction of true natural lighting.

Something I've been thinking about in terms of diffuse maps lately is that there's rarely enough variation in colour - another factor that might be affecting how we read resulting renders.

It only takes a slight indication of some blue/green veins to sing out nicely against the warmer flesh tones, perhaps some slight cooling in areas around the eye-sockets, temples and chin etc (subtle, like). Very often diffuse skin maps have just an all-over, averaged colouration with variations in tone - not enough for a convincing impression of skin layers and the underlying bone structure.

Ideally though It'd be nice to see some specialist maps for attempting some of that layering in a more physical way - local colour, SSS colour, veins etc, along with blow-out displacements.

I'm not sure the scatter node itself gives us enough control to play with some of these layering techniques properly; or the custom scatter. What we can do is pretty cool though, even if it does take some playing with.

Quote - Very interesting interpretation, Dan.  This is similar to the way photographs capture all tones and colors that the human eye misinterprets because of selective optical observation.  The shock of Nature's absolute truth.  It would seem that there is much to learn on how to manipulate Poser into rendering an image with artistic qualities which looks correct to our interpretation rather than delivering a very realistic stark reproduction of true natural lighting.

I think we need the scientific model of simulation based on RW physics to help us understand what we're doing artistically - I'm just not hard-wired that way to get through all that science & math. Honestly, my renders would be pretty awful without all the contributions of the math & science savvy people.

I've been using a black (noir) skin to see the reflection better. On indirect inside lighting, I've hacked it (fresnel) down 85% so far. This is of course an approximation. But the results seem waaay better than before.

I appreciate your explanations.

Quote - I've been using a black (noir) skin to see the reflection better. On indirect inside lighting, I've hacked it (fresnel) down 85% so far. This is of course an approximation. But the results seem waaay better than before.

I appreciate your explanations.

Good idea breaking it down like that, seeing the strengths in isolation. I'm using slightly higher reflection values but then the IDL intensity/diffuse lighting balance is more in a mid range also - that's the compromise (risk a little glow in places).

Bump.

This was one of the best threads we've ever had. So it's over a year later. What are you all doing with this?

I am lazy. I just routinely set IDL intensity to .65 and just keep doing what I was always doing with skin. I gave up on reconciling Poser with my photo-measured 1/8 diffuse value.

Yes, I miss discussions like this one! Right now I am so frustrated with Poser 2012 that I could scream. It's not been good for me since Serial Update 3. But I've ordered Poser 10 and it should come in the mail in a few days. Then I'll have something new to gripe about, maybe. Until then, I have been rendering with IDL set at .7 and using 4 or 5 bounces. I understand that in Poser 10/PP2014 that IDL strength is still too high when set at 1.0, so I will be trying those numbers when I get Poser 10.

Quote - Bump.

This was one of the best threads we've ever had. So it's over a year later. What are you all doing with this?

I am lazy. I just routinely set IDL intensity to .65 and just keep doing what I was always doing with skin. I gave up on reconciling Poser with my photo-measured 1/8 diffuse value.

I'm lazy too; too lazy even to push the 5, so my IDL intensity is at 0.6.

This was a good thread for me but left me frustrated with all the anomolies still in Poser light. The issue of 1/8th diffuse could use a thread of its own...