It's top of my wish list for the material room.
I'm guessing that physically accurate translucency is somewhat difficult to implement, even more so now with advanced lighting features like IDL.

But I need it!!

Buy Carrara. It has a translucency channel ready for implementation.

I added that to a skin texture recently and it really helped bringing out more realism and some nice color variations.

-- Jeff

Poser 7 has two nodes, but I think they don't work.

Translucency root channel is really just an ambient channel. Nodes like skin and fastscatter are useless with Raytracing and have been for some time.

One of these days I am going to have to add another app to the 3d arsenal, but I have few funds to spare at present .

whilst there may be some minor differences in how non-zero ambient and translucent channels
display on a posersurface, IMVHO translucence is a volumetric effect that cannot be properly
displayed by a zero-thickness posersurface.  they've already enabled volumetric fx for a zone
bounded by a background plane, hence it should be easy for them to enable volumetric fx for
a zone bounded by both foreground and background planes in the case of transmitted
translucence (frosted glass, edge-viewed skin) or a zone of finite thickness introduced as a
temporary calculational aid in the case of reflected translucence (skin viewed between its normal
and its edge).

actually, you kind of have to be careful with translucency and skin.  translucency is effectively back scattering, and most of what skin shows is front scattering.  two different phenomena.  Poser does need both, but if you're using a tool with both, like D|S 3 Advanced, you need to be careful about volume.  even if the application can work translucence volumetrically, you need to learn to control its scale properly.  ears, hands, nose should show some translucence in very particular situations. and the rest of the skin shouldn't show any translucence unless you're doing a really fine close-up and the lighting can show you the thickness of the person's skin.  usually in the face, or in older people.

lots of renderers do not implement any type of volumetrics with their translucence, and it's really only designed for items with approximately no width.  like lampshades, curtains, leaves, etc.  i have no clue how Carrara's translucency works, but judging by the flatness of most skin i've seen out of it, even very realistic works, you might l want to experiment with the much more important SSS first if it has it.

Translucency in Poser is crazy. 

Quote - whilst there may be some minor differences in how non-zero ambient and translucent channels
display on a posersurface, IMVHO translucence is a volumetric effect that cannot be properly
displayed by a zero-thickness posersurface.

well, it depends on what you mean by the first.  there are differences in very specific situations.  in most of those situations, the display is completely wrong and terrible, with weird artifacts.  outside of those situations, no, there's absolutely no differences that anyone has produced in testing so far.  bagginsbill has a thread about it in the RDNA Node Cult.

actually, i'm not sure if i'd prefer actual volumetrics or a control map.  the reason a control map might be better is that it would allow you to account for changes in material, too.  just working volumetrically would assume a constant translucence, not variable.  one could make the ears more translucent than the fingers.  i'm sure they could work together, but my limited experience with volumetric translucence outside of Poser was that it was very resource intensive.  and i wasn't trying to use it with raytracing.  if control maps worked about as well as, but rendered 10x's as fast, i'd prefer them myself.  certainly, i'd prefer the choice.

I must confess although I have been aware of the distinction between front and back scattering, I tend to think about translucency simply as light passing through a surface diffusely - which is a gross simplification I know.
What often frustrates me with Poser skin is the old problem of the sharp transitions you often get at the terminator between light and shadow regions with bright directional lighting. I'm not sure whether control maps can even deal with this problem (perhaps I'm not understanding the problem). In some ways I kinda want a diffuse shader with a controllable terminator where you can push the diffuse effect past the normal cut-off point between light and shadow, if that makes any sense. I have a very limited understanding of the physics of these effects. Some are so subtle.

Quote - whilst there may be some minor differences in how non-zero ambient and translucent channels
display on a posersurface, IMVHO translucence is a volumetric effect that cannot be properly
displayed by a zero-thickness posersurface.  they've already enabled volumetric fx for a zone
bounded by a background plane, hence it should be easy for them to enable volumetric fx for
a zone bounded by both foreground and background planes in the case of transmitted
translucence (frosted glass, edge-viewed skin) or a zone of finite thickness introduced as a
temporary calculational aid in the case of reflected translucence (skin viewed between its normal
and its edge).

wellcome to CGI. ;)

how come other are doing it? 

yeah, it's two different phenomena.  and what you're talking about is front scattering.  so you have diffuse in Poser.  maybe the model in Poser isn't so great; certainly i see differences in the performance of its Clay/Oren-Nayar shading and Yafaray's.  but the actual spreading you're talking about is scattering.  light goes in, scatters back out.  considering the diffuse response to light as a gaussian curve peaking at a direct line to the light, the scattered response is the same gaussian curve lowered and widened.  like diffuse, light will eventually fall to zero, but due to scattering, more light will come back at a wider angle. now the material will determine how wide and how much scattering occurs.  but that will give you the softness you're looking for.

the thing is that only specular nodes have any control over spread in Poser.  which is why my current SSS fake combines both specular and diffuse techniques.  it's probably totally faulty, but i might be able to show you some examples, if you're interested.  but yeah, we just need true scattering, with a control on intensity and radius.

but you dont want only front scatter right?

when a strong light is behind someone's head then their ears are glowing red. this is also SSS.

i think poser is able to have a simple SSS. it doesnt have to be perfect and 100% physical correct.

no, you don't want only front scatter.  but back scatter is only relevant in very, very particular places, and even then, only very particular situations. if you're using it everywhere, your skin won't be at all accurate. not even a little.

no, Poser isn't able to have even simple SSS.  that spread carodan's talking about is absolutely crucial.  as is the blur of micro and macro features.  check out the papers on BSSRDF, like:

http://graphics.stanford.edu/papers/bssrdf/
http://graphics.ucsd.edu/~henrik/papers/skin_bssrdf/

at a distance, and low resolution, for things with a relatively low scattering radius and smaller scattering component than diffuse component, just colored diffuse and specular work.  and in some cases, Edge Blend (i think, at least that's what some very soft-looking skin textures/materials use).  but do a closeup, or make something like soft wax or gelatin that's almost all scattering, and you get a really flat, really problematic results (especially around the nose).  and the more bump and detail you have, the worse it looks.  it looks like painted wood or plastic. the best, most realistic Poser close-ups without postwork have this problem.  and that's just skin. 

there's a reason bagginsbill's candle shader required a geometry dependent fake, and couldn't react to light in general.  or that it's hard to make a white porcelain which is easily distinguished from equally shiny white paint.  for Poser, we've all come a long, long way.  but compared to even free renderers, we're just beginning to address realistic materials.

I've played with various combinations of diffuse and specular nodes in Poser to try and find a balanced solution to this, but always the hard line of the terminator remains, particularly with side lighting when IDL  is used. I think the problem is we end up needing both front and rear scattering effects to deal with the terminator.
BB's previously suggested way to partially soften the line away from the shadow region using a blender controlled by a smoothstepped diffuse node also falls down when using global lighting - the harsh terminator line is tere again.
You can't force any light past the terminator in a diffuse manner, but you also don't seem to be able to diffuse the transition in the other direction, away from the shadow region.

The only material based way I've found to even slightly soften the terminator is to disturb the object surface using bump or displacement, but this offers only a very slight diffusion and doesn't always produce desirable effects.
Using light arrays also offers a partial solution but sharply increases render times.

I keep hoping I'm missing something, but we seem to need a new material node to even fake this.

In revision of my own post - we need more accurate front and rear scattering simulators that work with Raytracing.

That is exactly what Poser's translucence channel does.

What are these channels on the root node for?

Quote - > Quote - lots of renderers do not implement any type of volumetrics with their translucence, and it's really only designed for items with approximately no width.  like lampshades, curtains, leaves, etc. 

That is exactly what Poser's translucence channel does.

this could be used for plants right?

i think i will use this for plants.

carodan:

I have never seen those channels before. In what version of Poser are they found?

Have you clicked the question mark (?) to see what it says?

Thanks,

LMK

I noticed them in the material room screenshot stewer posted. I don't have any version with those channels.
Very curious though.

That's the diffuse node doing that. It has nothing to do with the Translucence channel.

Have a look at this - just using the built-in diffuse.

carodan:

I see those extra channels also on BB's screen shots. I suppose that these extra channels may be in the Poser Pro 2010 closed beta version (for the private beta testers).

Maybe Poser will finally have some programmable channels and nodes?

lmk

Quote - I noticed them in the material room screenshot stewer posted. I don't have any version with those channels.
Very curious though.

That's a new feature in Poser Pro 2010. PPro 2010 can render to layers, if you save the render as a Photoshop document. The custom channels default to various useful things but you can connect them to nodes to make them do any sort of layer you want.

The default layers possible are:

Normal
Toon ID
Z Depth
Position
Texture Coordinates
Custom 1 (defaults to diffuse only)
Custom 2 (defaults to specular only)
Custom 3 (defaults to shadows only)

The custom ones will do whatever you plug into them if you override them with nodes.

The thread where I fully demonstrate the behavior of the Translucence channel is here:

http://www.runtimedna.com/forum/showthread.php?t=26024&highlight=translucence&page=2

It is extremely esoteric and unless you're willing to read every word and build a pretty complete model in your head about what to expect versus what I demonstrate, you can just skip to my conclusion. (Stefan should ready it very thoroughly, though.)

Conclusion - the translucence either acts ordinary (like ambient) or does something that isn't very useful. In addition, there are situations where it doesn't even do its intended behavior for a completely meaningless reason. There should be no distinction between using two identical but separate Diffuse nodes versus using the same node for two channels.

I say stay away from it.

Quote - That's the diffuse node doing that. It has nothing to do with the Translucence channel.

  Have a look at this - just using the built-in diffuse.

 
Try ticking "normals forward" on your material and the translucency will disappear.
The translucency input will evaluate the nodes plugged into it with the normal reversed. A diffuse node plugged into it will only show light when it's lit from the backside and not from the frontside. In My scene, if you were to move the camera around to look at the plane from the other side, it would appear plain white and not with the beige tint that it shows on the backside.

It's not supposed to subsurface scattering or volumetric rendering, what it does is allow to set up a shader that determines how the surface reacts to light coming from behind. 

Thx for the extra channel info bb - sounds good.

stewer - Give us SSS...pleeeease!

Quote -
stewer - Give us SSS...pleeeease!

i would sell my soul for SSS in poser :) 

lol...you may need to - I sold mine for IDL.

Carodan the Soulless.

HA!

It's worse than you think....they sent it back claiming it was defective. The cheek!