Quote - Well, this is becoming more complex by the second. In order to do eyes, I'm going to have to rethink things... perhaps I need to create a Template Lips and a Template Face... since that PMC:Color_Tint node is only going to allow one output. Or am I barking up the wrong tree?

No, you don't want a separate template. Remember that we need to treat the lips and face as one zone so we can apply the lip mask gradient across that boundary.

The way to handle the Tint is to understand how to chain Blenders together.

The Blender that is making the lip color change tint has two input colors. Value_2 is the color of the lip tint. It only applies to the area defined by the lip mask. Everywhere else, the tint value is coming from Value_1. In other words, Value_1 is the tint for the rest of the face.

And there's the answer. The Blender:Value_1 is effectively the same thing as the Tint used to be: it defines the Tint color any part of the face except the lips. Value_2 is controlling the lip Tint.

It then is quite simple - design another Blender chain for the eye shadow Tint, and a mask that determines where eye shadow exists.

That new eye shadow blender should plug into your existing lip blender, in Value_1, since that controls the rest of the face.

Sheesh, Bill, you don't sleep much, do you?

Quote - The Gain<--Bias<--Mask chain is correct.

Oh. Well, I wasn't happy with the results, though... what it seemed to do was push my mask around - up and down, so that it didn't align properly with the underlying image. I must have been doing something else wrong, then... trying to go over the steps...

Quote - The result of the Gain is your new improved mask. You should think of it as a replacement of your mask even though it is using the mask as an input. So you want to connect that Gain node to all the places the Lip_Mask used to go.
That means into the Blender (which you did) and also into the other math node that is doing .25 + Lip_mask. That should be .25 + Gain, because you're trying to replace all occurences of Lip_Mask with Gain.So connect Math_Functions_23:Value_2 to the Gain node.

Which of course I wasn't doing! I'll do that then and see how I go...

Quote - Also, at this point you should be doing test renders to tweak those bias and gain values. I only offered .8 and .2 or .8 and .8 as suggestions.

Oh, I have been. I haven't figured out how to set up preview - the advice that IsaoShi gave didn't work for me (BIG surprise there, since I never stuff up!) so my previews look shocking, like I'm using the specular map or something, so I've been rendering every time I change things. Even saved a few, just for laughs. Good job I read about how to get decent renders cheaply in your VSS Discussion thingie...
 

Quote - But anyway, it worked. What I did was, create a new material zone called Template Lips, moved all the lippie stuff in there. Then in Template face, I created a similar set of nodes, but this time using a different mask.

Move it back and do the double Blender thing I just talked about.

If you make them separate, you run the risk of getting a mismatch between the zones, and you'll see the boundary between them as an abrupt change of some kind. In fact, I think I see that in your posted render.

If you make a template for every zone like this, you might as well throw VSS away, since you'll be back to editing N shader for N zones. The whole point of VSS is to have 1 template for N zones.

In face, if we were properly putting the lip mask file name on the target zones to begin with, instead of in the template, we could be back to using a single material for all skin zones.

When VSS copies a named Image_Map from a template to a target zone and finds that the target zone does not have a correspondingly named Image_Map, it replaces it with the Background color of the source Image_Map. So by setting up the Lip_Mask with a black background, you can run the same shader in all skin zones, even legs and arms. In those cases, there will be no lip mask, and the effective value of the Lip_Mask will be black. Black means "no lip here". See? So there's no need to have multiple skin templates, as long as you use the system correctly.

Quote - what it seemed to do was push my mask around - up and down, so that it didn't align properly with the underlying image.

I suspect the reason is that the original lip mask gradient is uneven. Since the bias and gain are adjusting the rate of transition in the area that is a gradient, any irregularity will be magnified.

Also, try Bias=.2, which will make the new sharp edge closer to the outer edge of the original softer gradient.

Or try Bias=.5, which has no effect. This will make the new sharp gradient right in the middle of the old softer gradient.

Quote - The way to handle the Tint is to understand how to chain Blenders together.

The Blender that is making the lip color change tint has two input colors. Value_2 is the color of the lip tint. It only applies to the area defined by the lip mask. Everywhere else, the tint value is coming from Value_1. In other words, Value_1 is the tint for the rest of the face.

And there's the answer. The Blender:Value_1 is effectively the same thing as the Tint used to be: it defines the Tint color any part of the face except the lips. Value_2 is controlling the lip Tint.

It then is quite simple - design another Blender chain for the eye shadow Tint, and a mask that determines where eye shadow exists.

So I - what? - daisy chain another blender node off the Value_1 connector of that blender node that came off PMC:Color_Tint? Cool. I'll do that.

See what I mean by kludge? Helps to be talking to someone who knows! Thanks Bill... I'll have a go and letcha know... my turn to go to bed.

Quote - Sheesh, Bill, you don't sleep much, do you?

Actually I slept seven hours. It's morning here and I'm about to go to work. Nighty night. :)

Couldn't hardly wait to get up this morning to try all this new stuff. Of course, I did a fair bit of fumbling, plugging nodes together incorrectly at first (with absolutely shocking results), but I finally got it the way I think it's meant to be:

And no, Bill, this is not complete... I wasn't sure how I would apply the same gain/bias set-up for eyeshadow, or if I really wanted to. I'll leave it to you to comment on that (explaining the advantages)... in any event, I can see massive potential for creating multiple layers for eye-shadow, which is essential for a decent effect. Just keep daisy-chaining those blenders on out and yer laffin'. You can even introduce a bit more specular to the outer layers for that dramatic, sultry look... so this is only stage one, here:

Yep, I AM having way too much fun with this!! And I know just enough to be dangerous.

You did it perfectly! See it wasn't that complicated - besides the eyeshadow mask itself, it's just one more node - the Blender.

Now that you understand the technique, you can extend to as many different layers as you want. Rouge, lip liner, mascara - all can be done this way. You can also make moles, pimples, warts, and freckles this way.

I noticed you settled on a Bias of .5. It turns out Bias of .5 does exactly nothing. So if you like how its working, you can eliminate that node. Just use the Lip_Mask into Gain.

I did some math experimenting today and I came to the conclusion that for this particular purpose, the Gain node was going to be good enough on its own. You seem to have figured out the same thing, although your probably didn't know you did that.

As for bias or gain on the eyeshadow - that's not necessary at all. If you have the gradient just the way you like in your original drawing of the mask, then it's fine as it is. All that the Bias or Gain (or both) was being used for was to "fix" the mask so that you didn't have to go back into an image editor and sharpen it.

But for your own education, I'd play with inserting a Bias node between the eyeshadow mask and its Blender. Value_1 = 1 and connected to your eyeshadow mask. Value_2 = .5 for no effect, greater than .5 will increase the density of the eyeshadow, and less than .5 will decrease it.

In each picture there is a graph of a function on the left, and a mask on the right. The mask has lots of shapes with soft gradient edges. As I apply each function, you'll see how that affects the gradients in the mask.

This first image is just the pass-through do-nothing function f(x) = x

In other words, the values of the original mask are unmodified.

The red line shows the Bias function. The blue line is the original data. By observing the relationship between red and blue, you can understand how the function changes the data.

Notice how Bias works. The values 0 and 1 never change. You can see that the red line intersects the blue line at 0 and 1.

The 2nd argument to Bias is what value to produce when x is .5.

If you count the grid lines you'll see that midway across (when x is .5) the red line is at the 8th grid line, the one that corresponds to .8.

The red Gain line intersects the original blue line at 0, .5, and 1. It makes an S shape. Between those points, it curves around.

This is what we use to increase the sharpness of fuzzy edges. It waits to rise, then rises more rapidly than the original blue line does. Then slows down so as not to overshoot 1.

This is Gain(Bias(x, .2), .8). First Bias runs, delaying the rise of X. Then Gain runs on that, making a steeper slope out of an S curve.

But notice now the intersection of red and blue is not in the middle at x = .5 anymore. It is postponed because of the Bias. Looking at the mask, we see the shapes have contracted and been sharpened.

First Bias accelerates the rise, because the value is .7. Then the Gain sharpens it by making the steeper S curve. In this case it is very steep because the Gain value is very close to 1. The intersection of red and blue is earlier now.

So now we have fatter shapes with very sharp edges.

i'm not a whiz at this like bagginsbill, face_off, Olivier, or semidieu.  heck, i'm not mec4d, with her technically wrong but amazingly realistic solutions.  but i'm still pretty decent at this material room thing, despite forgetting (more like repressing) just about all my math from college.  and that was quite a lot of math. 

and frankly, i never knew enough logic and algebra to be even a passable programmer.   i know enough programming to be dangerous, which is useful in my line of work, and may one day help me in the material room.  and, oddly or frighteningly enough, i have come across worse coders than i.  but if you're very comfortable with programming, you'll be leaps and bounds ahead of me in the material room in no time.  because at base, 3d rendering is about mathematics and physics, and the better you understand both, the more you can do.

and yet... i'm getting pretty good results.    more realistic than many.  i'm sure my adjustments to bagginsbill's shaders are "naive" (bagginsbill's polite way of refering to inaccurate material room efforts).  but i'm very comfortable trying things.

basically, i'm not the best, but i'm far from the worst.  i've gotten better realism than most who render with Poser, and certainly better than most product promos.

so i think you're doing a great job, and you should just keep trying and experimenting and having fun.

but looking at your results and gallery, i think it's time for you to start looking at both the art and realism of lighting.  you have a lot of images that look as if they use great shaders but really flat and somehow unrealistic lighting.  your gallery shows several images with the sun behind a figure who's lit entirely from the front and has hair that's got burned in highlights. 

i think you should definitely keep playing with materials, but when it comes time to promote your work, you'll be a lot more effective if you improve the drama of your compositions and lighting, the realism of your lighting, and clearly communicate concept or mood with your work.

i've attached an image just to give you something to comment on of mine (none of the work in my gallery is recent, it's months off).  i don't think i'm at all perfect, or that i embody my own ideals.  i'm actually commenting because i spent so long on technical aspects, i began to have images that were exactly the generic girl standing there type i'd never wanted to do.  i have no problem with plain portraits, but not.... same old boring ones.  i don't know how to explain the difference, but i'm looking harder at different photographers' works.  i'm just beginning to get back to feeling like i'm actually doing some interesting work.

Clamp(7 * (Bias(x, .7) - Bias(x, .6)))

It's not the best edge detector but it's cool how you can use two bias nodes for something interesting eh?

So the Math:Clamp function does that. It clamps the input value - anything above 1 is 1. Anything below 0 is 0.

Here is Clamp(2 * x)
 

Here is Clamp(2 * (x - .25))

Something like this does not make a lot of difference when you're using the function to make a mask - you can hardly see the difference. But if we were using this function to make a shape via displacement, this would make a very big difference.

This is Gain(Clamp(2 * (x - .1)), .99)

This will make a negative of the mask, while also sharpening the edges.

This is Gain(Clamp(2 * (.6 - x)), .99)

ooooo.  i think you should post this somewhere on your Google site.  i bet there's lots of people who aren't following this thread that would find this both helpful and inspiring.

I could write a book on this stuff. It is endlessly fascinating and so useful. People wonder why i can make shaders so fast with complex stuff. It's because I know these functions inside and out.

Here's a really really good edge detector.

Gain(Clamp(Clamp(10 * (x - .1)) - Clamp(10 * (x - .2))), .99)

I use this to make lip liner from the lip mask.

mmmmm.  tasty!  as much as i've expurged the math from my mind, i still have my tendency to want to start with the basics and derive everything possible.  it's much more elegant and pleasing to have a lip liner equation than need two different maps that have to synch up. 

I'll have to admit: the only maths I was ever any good at was geometry, much to the despair of my poor father, who could do incredible mental maths and just assumed his offspring would inherit those traits.
Alas, that's not how it works. I'm a great speller, though. I guess that's more customary for one of my gender, anyway. Hate to make generalizations though: my FoxPro programmer friend was a girl and a whiz at maths!

The diagrams and formulas you've presented here make your math functions a lot clearer - for me, especially - in a practical, Poser-material-room sense. I'm sure there are infinitely more applications for these formulas for non-Poser artists - I'm just glad we've got you on-side to bring the material room up to potential.
To paraphrase CobaltDream, the idea of creating precise art using judiciously-applied formulas is infinitely appealing. I've been mulling over the directions this all could go, and find it incredibly exciting.

I must admit, though, that whilst I am beginning to see how one could create colour curves and swirls that would end up as eye-shadow or body-art, I think I'll still go the route of layered masks for the time being.

Also, I'd love to develop - or see developed - a more sophisticated approach to hair, which is still Poser's weak link. No one has come close to emulating Kozaburo's technique, which is why hair all looks like it's made of wire.

@CobaltDream: I agree completely with your assessment of my ability to light effectively. I have none. Again, not for lack of trying, but because I really didn't pay much attention to what the experts were doing. So, you see evidence of a bright light where you really don't expect to see any, and it creates a sense of falseness, that something's not quite right, here.

Lights are one of my main weaknesses.

I could say: "well, all I really want to do is develop make-up with shaders and masks, so lights don't really play a role." That's until I look at the work "shaders". Implies lights. So, it is time I studied it.

Any of you have any views on a book I recently bought: Practical Poser 7 by Denise Tyler? It was primarily for the tutorials/workshops on lights that I invested in it. Well, regardless, gonna go curl up with that tonight and have me a read on that subject.

Anyway, thank you for your constructive critique. I have a very long way to go - and I'm enjoying every minute of it! - before I finally achieve what I want to achieve: a strong sense of emotion evoked by viewing one of my images... one where technical accuracy blends well with powerful visual beauty.

 koz har is absolutely great, even among commercial products.  but, imho, not the only good stuff out there.  his modeling technique is pretty straight forward, and how most people make hair.  you take a bunch of strips and then bend them like you want.  that way your map is always flat.  more strips means more realism, but kills resources if you need it to cast shadows (especially raytraced ones) .  all of my hair shaders so far use the same diffuse map as a base, because everything but skull caps and some of koz's maps (which tend to be singular, rather than having a separate map for non-vertical maps) can take a plain flat texture.  i mean, i seriously respect kozaburo's work and have everything he's done.  he is an incredibly talented and generous man.  but there are quite a few talented modelers.

what i'd say koz locked actually was cute, simple hair styles.  he didn't need them to be edgy and razored or wispy.  so he actually worked with fairly few strips (lower resources), but achieved easier realism.  at the same time, his styles aren't at all unlikely, though a few i've never seen in real life. and his morphs are better than most.  i can't count how many products i haven't  been able to use because it won't fit my pose and simulate gravity because instead of forward being forward, it's curving to the right and forward.  or left, or up, or whatever.  that said, i've had the same problem with Koz hair, and to an extent, it's simply a problem with using someone else's polygon hair with my own poses.

personally, i've gotten the best results out Wild Hair by Digicalermo, but that was before i started trying to use raytraced shadows (bagginsbill's advice) and render at 2700 x 3600. 

Quote -
Anyway, thank you for your constructive critique. I have a very long way to go - and I'm enjoying every minute of it! - before I finally achieve what I want to achieve: a strong sense of emotion evoked by viewing one of my images... one where technical accuracy blends well with powerful visual beauty.

oh, i know what you mean, except i've spent a lot of time on my lights.  as i said, i'm actually more to working on artistry, but i have miles to go.  i would say this: if you want strong emotion, you can't do what everyone else does.  not that i know how to do that (i don't), but in a world glutted with images, to move someone you have to stand out somehow.  and if you have good lights and good compositions in your promos, more people will buy because this is an impulse based market.  most people buy because they want to acquire the coolness of your (promo) renders.  so the better you can make the actually ancilliary stuff, the better the important stuff looks.  sort of like how people think drinks in glass bottles have better ingredients than cheap plastic bottles. 

Quote - For anyone else who downloaded the VSS_Template_Skin_PR3_WIP.mt5 file and applied it to their VSS Template Skin shader.... After I synchronised with VSS, there were no skin textures displaying in Preview, although it rendered okay. To fix this, I just added the link shown in the attached piccie in the Template Skin shader, and saved the prop. Leave the diffuse value at zero. Also, the eyes are showing just white. Although I have not seen this before in VSS, I can't imagine it's anything to do with the new material. Looking into it...

Um, when I wired the diffuse_color into color_pow (as your illustration suggests) my previews were quite scary - looked like bump-map and specular map were struggling for dominance over the skin, and neither was winning. However, when I wired diffuse_color into color_map (as bagginsbill suggests in the post following this one) it worked fine. I wasn't all that worried about a preview, but I was getting a bit over seeing a plastic mannequin staring back at me... besides, it's harder to pose a character when the skin texture colour isn't there to help reinforce the look.

I keep coming back to this thread.

And of course, it's completely above my head, although I purchased Poser Pro largely to take advantage of work like yours, Bill... but why oh why isn't your pro shader ready yet? I'd love to have the time to do what RobynsVeil is doing, but realistically... I'm investing far too much time in this hobby already. I need a more "out of the box" solution while I get up to speed. Maybe I can get a better understanding of lights in the meantime....

:

Holy cow! I just got the VSS running and I am totally blown away. I've only done ONE render with it, but the effect is dramatic.  

This is a huge leap forward. Thank you for sharing it with us.

just to say, because it's come up again elsewhere...

your render is better due to the default shader provided with VSS.  VSS is not the default shader.  it is a tool for managing multiple materials.  with VSS, you could have multiple figures in a scene and control such properties as SSS and specularity for all of them at once.  i recently spent ages dealing with the P.I.C.K. components.  because it uses bitmaps, they have been designed not to scale in order to work together.  so for each P.I.C.K. unit of space you need, you need another prop.  with even a small room, i had over 10 pieces.  they all had walls, some had glass (multiple zones), some had molding (multiple zones), and just about all had floors and ceilings (some were just that).  because i haven't taken the time i should to learn to use VSS (i'm sort of waiting on the final version and learning Blender 3d in the meantime), it took me ages just to manage the floor and walls correctly, let alone details. it was a big PITA.  if i had taken a bit of time to learn VSS, it probably would have taken me minutes. it would have been a moments work to, say, change the color of the room, rather than taking ages as i change the wall color on each prop and figure.

with that, bagginsbill, when you sell VSS, let me suggest you sell either bundled or separtely a set of defaults for the P.I.C.K. components.  you seem to have good RDNA connections, and i think the P.I.C.K. system has an even greater need for good materials than figures do, just because they're so varied (glass, stone, paint, wood, metal, and plastic).

Heheheh.

gomcse - Show us your render.

cobaltdream - P.I.C.K. is a perfect example of where VSS really shines. Semidieu has done some scripts to automate stuff for P.I.C.K. I think he did some bulk material utilities. The "P.I.C.K. Creator" product literature says "P*.I.C.K. Creator is a set of scripts that allows you to completely build a scene, arrange various parts and apply materials with ease."

*But there's no doubt VSS could be used on it to do the same things. The real question is where are you gonna get an awesome set of glass, stone, paint, wood, metal, and plastic materials. Hmmm.

Selling materials, what a concept. Heheh.

Derailing current topic to ask a probably redundant question, but...

Using Vss and the included light set, what exactly would I tweak to wring out slightly darker shadows than what default provides? 

thanks!

The included lights are way too hot. My first pass was accidentally done in isolation (I should know better).

Reduce the IBL intensity by half to start. You can (and probably should) decrease the intensity of the other lights as well.

i'd pay $30 easy for a parametrized set of templates based around the P.I.C.K. zones.  heck, i'd pay that much for indoor and outdoor each.   and i'm not thinking mega shaders, like your proceedural wood or loom for rugs.  i'm fine with using bitmaps for stuff like that.  i just would love to be able to control color, shine, reflection, etc. in a consistent and realistic way.

personally, i'd pay $20 for different types and thicknesses of glass shaders and $40 for different types of water shaders.   i managed to create glass i could see through that stopped light from getting through.  outside came out fine, but i had to make the glass pure transparent

my horribly, awfully embarrassing recent satin efforts have taught me that i obviously need a lot of help with Poser specularity and real life material properties.  i can't even guess at what i'd pay for a fabric templates.

but that's probably not as fun as building the tools.  if selling materials was something you wanted to do, you would probably already do that.  and you've given us so much already, it's probably rude of me to say anything.   but it would be sooooo nice, and just PR2 requires either throwing everyone else's shaders out or doing some heavy adjustments to them to make them consistent.

i was thinking maybe using the orb shader for the eyes? with vss skin it would look pretty cool

Sure - you might want to try the one that has a little bump in it.

I haven't spent any time on the eyes yet. The Orb shaders are the first ones I published that correctly implement the true Fresnel reflection equations. The results are remarkable I think.

the results are amazing.
i dont know if you remember but weeks ago or even months ago i asked you for gamma corrected ''lame'' shaders. GC diffuse plus reflection. and it looked good . but nto realistic enough.
here you went did something amazing.
its GC with reflection and refraction

Outstanding render ice-boy. He looks real.

Hi, I wanted to post the images but my ftp site wasn't working. I got it working now. The first image is WITHOUT the default VSS. The second is WITH the default VSS. I was using 4BlueEyes_Classic xSoft IBL lights here (because I love them in Poser 7).

And here is the second one:

 
Now I need to work on lights that will work with VSS, or I need to modify 4BlueEyes_Classic.

Dave

are you sure that's the latest release?  it looks a bit like the PR1 instead of PR2 shader.   that said, i never did use my BlueEyes lights.

Quote - Outstanding render ice-boy. He looks real.

yeah the poser background really doesnt work.it needs to fit with the lighting and everything. i was using here the VSS2

the lens flare is just for the cool factor. :)
i was using here the orb shader for the eyes.

Under The Lights

They both look good to me I do prefer the AO one but I have just started using AO and have it "on the Brain". I viewed it on both a Sceptre 19" LCD and and LG 22" wide screen TV that is RGB equiped and can be a monitor and both show it looking very good. The TV one looks slightly better with better skin tone. but most people do on it.

Anyway, I put together a few materials. Quite a tedious process. Test rendering a scene like this is excrutiatingly slow. There are so many material zones to set up, even with VSS free it's a pain in the butt. I need to finish VSS Pro so I can use a nice multi-select drag-and-drop GUI. The narrow pinhole-of-a-user-interface afforded by the Python buttons to add rules and copy nodes one by one just doesn't cut it.

I used an IBL, and infinite for the sun. I  added two point lights with inverse square falloff and a serious shadow blur to try to get the general sky glow to flow through the windows. Not entirely successful on that. I think I'd need to use something like 16 point lights to get that to look good, because we don't have GI or area lights in Poser. Sigh.

Click for full size.