What exactly did you use a the test square ?
You can check the U and V gradients by either using a checker texture or using (seperately) the U and V coordinate nodes to drive the Diffuse colour or value.

i noticed you get very interesting results if you connect the Fractal Sum to the U and V hightlight size.

Looks good, but I still get that disruption of the texture. This is just the standard Poser Primitive one-sided square:

Might try it on the high-res square and see if my results vary.

Your results will not vary. I frequently encounter these abrupt discontinuities in the noise-based textures. There is some error in the implementation. I have usually had to offset the pattern or combine it with another copy to hide the artifact.

Please help figure this out as it drives me nuts often - I'm vacationing and can't do it at the moment.

Try different scales. Try using constants in the other two scales (an add node with a fixed number). Try using the V coordinate for Y_Scale.

Okay, different results. Different plane and it didn't happen. I'll have a go at your suggestions next, Bill and see how I go. BTW, you can see that artefact in my hair example above.

The underlying coordinates of the geomtrical object you use are what is driving the Fractal_Sum. So if you load a one-sided square, scale it or move it, export it as obj, and re-import it as a new prop, you will see a difference. But all that means is the model coordinates influence the presence and location of the artifacts. That doesn't tell you why they happen, where they happen, or how to get rid of them.

Different plane (the hi-res plane) and the texture didn't interrupt on the face. However, it looked completely different. Same settings, same shader, different plane (looks like poo):

Seems like the texture kinda repeats on the way down. So, might it be a mapping issue? Since the high-res square has more polygons and might be mapped differently? No idea.

I'll make my own square next, in Blender, UV-map an image to it, and see if I can attach the shader. But first, I'll have a go at your suggestions next, Bill and see how I go. BTW, you can see that artefact in my hair example above.

Made my own square, UV mapped it, gave it a material. Imported into Poser. Here's the shader setup:

s.Diffuse_Color = Blender(
  IColor(77,77,77),
  IColor(131,131,131),
  FractalSum(u_texture_coordinate() * .00004,4 * Add(0,0),4 * Add(0,0),8,.3,.5,.5)

Blender will stay the same... FractalSum will change.

s.Alternate_Specular = Anisotropic(
  FractalSum(u_texture_coordinate() * .00004,4 * Add(0,0),4 * Add(0,0),8,.3,.5,.5)

This is just recreating Ice-boy's shader:

Now, I'll take out the u_texture_coordinate node:
s.Alternate_Specular = Anisotropic(
  FractalSum( .00004,4 * Add(0,0),4 * Add(0,0),8,.3,.5,.5)

So, it has nothing to do with u_texture_coordinate - the break is still there.
Then, I did something weird (gee, what a surprise!). I went back into Blender3D, and extended the UV-map about a fourth to a fifth lower than where the texture was:

Saved as new .blend, exported as new object, loaded shaders, rendered:

The break is not a clean, straight line. There is a bit of a jagged change as well.

Now, where do we want math nodes inserting? between the u_texture_c oordinate node and the x scale channel of FractalSum? Here is where a solid understanding of what exactly u_texture_coordinate() does mathmatically (and why the plugging into y and z scale of Add(0,0) - was that just to bring the global XYZ to uv xyz? Per:
*But my favorite nodes, Fractal_Sum and Turbulence, do not have this checkbox. What to do!!?!?

Well I discovered another bit of magic, an undocumented behavior that at first was very peculiar and difficult to comprehend, but I now use it a lot. If you plug any kind of node into "x scale", "y scale", or "z scale", those parameters STOP USING Model space x,y, or z values. Instead, they use the node you plugged in!!!

So, for example, I can plug a P node into any of those and use World space X, Y, or Z as I see fit. You have to either use one P node and three Comp nodes (to extract each coordinate) or use three P nodes, each set up to only pull out one coordinate. For example, P(x=1, y=0, z=0) will extract only the x coordinate. Unfortunately, due to how vector to scalar math works in the nodes, this gets divided by 3. So you should use P(x=3, y=0, z=0) instead, which cancels the divide by three.*

I realise you are are going to be horror-struck by this, Bill, but at this point, since I have nil math basis or understanding what I'm doing, I'm just trying different nodes and watching for behaviour.

I replaced the weirdly UVmapped square with the normally UVmapped one.
I wanted to see something affect the z scale - which I take is the length of up-and-down (vertical?).
So yes, tried different functions between the v_texture_coordinate and the z scale channel of fractal sum.
This make a change:
s.Alternate_Specular = Anisotropic(
  FractalSum(
    .00004,
    4 * Add(0,0),
   ** 4 * Max(100 * v_texture_coordinate(),** v_texture_coordinate())**,
    8,
    .3,
    .5,
    .5)
  0,
  1,
  0)
**
Forgot to include that last anisotropic parameters bit last time.
And here's the change:

Well, that seemed like a band of another texture coming through the main texture. Took things down to simpler:
s.Alternate_Specular = Anisotropic(
  FractalSum(
    .00004 *** u_texture_coordinate()),
    4 * Add(256,0),
   ** 4 *** Add(0,0),
    8,
    .3,
    .5,
    .5)
  0,
  1,
  0)**

...and discovered this:

Brought up the render in The GIMP and cut off the bottom texture bar to see if it matched the top texture section:

Um, sorta. The pattern is identical, just a hair wider. Thinking memory?

No, hang on... I had the square tilted at 80 degrees - don't ask me why: I'm a nutcase. So, the texture is identical. So, I went back and tilted the square so it was um square to the camera... and got this!!   :

Um, now what?

Too many variables. So, square position: 0,90,0. Camera: Front. Render:

Okay, am I wasting my time? Is this whole experiment bogus? It would appear that shifting the camera angle shifted the texture disruption.

I'll start over.

Heheh. You're learning that there are many factors that alter your perception, including tilting the camera, that you didn't even think about - why would it matter?

This is why I know more about Poser materials than anybody - I don't draw conclusions after a few dozen tests. I do thousands. I change everything, one at a time, to learn how/if it matters.

Z is not up and down. Z is front to back. Y is up and down. X is left and right.

I'm going to turn into a monomaniac if I'm not careful.

Oops. Too late.
Zeroing in on the interaction between v_texture_coordinate and FractalSum. Left other settings at default:
FractalSum(
    .4 *** v_texture_coordinate()), 4,** 4**, 8, .5, .5, .5)**

...

Uploading all the next three hours worth of trials is more work than it's worth. However, I kinda think I might have gotten somewhere, after all:

Still a very long way to go. Might be a bogus solution. Knowing me, probably is.

FYI: v_texture_coordinate() is already a global - it's called simply V

Thanks, Bill... so I would reference it thus?
x_scale = .0004 * V

There appears to be some threshold where texture break up in a number of V- or U- related nodes. For instance, did:

s.alternate_diffuse = Sin(U * 500, .5) and it gave me this:

Don't blame me for the epileptic seizure, please.
Upping it to Sin(V * 10000, .5) and it gave me this:

I tried a lot of things to see if I could find that pattern. I found functions I'd never even looked at before. Keeping some of these for future reference... if I ever need astro-turf!!!

  Interestingly, pushing values didn't seem to affect horizontal patterns:
s.alternate_diffuse = Cos(Cellular(Scalex = 100, Scaley = 1, Scalez = 1) * 5, 1)  gave me:

s.alternate_diffuse = Cos(Cellular(Scalex = 10000, Scaley = 1, Scalez = 1) * 5, 1)  gave me:

The search continues....

Has nothing to do with any of those nodes. Here's s.Alt_Diffuse =  ImageMap("HairTxtr01.jpg", U_Scale = .5):

This:

**#=========================================================

- - GC Shader for Tequila Hair - version 04 - -

#      by Robynsveil (with ideas by IceBoy and BagginsBill)
#=========================================================

Creates a straight-hair and skull-cap material for Tequila Hair.

This GC shader requires the following file exists: TH_StBlack.jpg

It can be made in the GIMP with noise that is then motion-blurred

and unsharp-masked a few times.

 
 
def makeSurface(region):
  # Colour and other Maps
 
  if region == "skullcap":
    transMap = ImageMap("TequilaCap_Tr.jpg")
    colourMap = ImageMap("TequilaCap_Spec.jpg")
  else:
    transMap = ImageMap("TequilaHair_Tr.jpg")
    colourMap = ImageMap("TH_StBlack.jpg")
 
  # convert colourMap to linear colour space
  gcVal = Add(2.2)
  linearColourMap = Color_Pow(colourMap, gcVal)
  
  # Plugging Add(0,0) to y_scale and z_scale is an Ice-Boy-ism - might have
  # something to do with changing model xyz toglobal XYZ or something.
  # Not sure why this is being done...
  hairTxtr = FractalSum(.0004 * U, 4 * Add(0, 0), 4 * Add(0, 0), 8, .3, .6, .2)
  setSpecular = Anisotropic(hairTxtr, 1, .5, .15, 0, 1, 0)
  setTexture = Blender(setSpecular, linearColourMap, linearColourMap * .5)
  # convert baxk to sRGB
  diffuseOut = (setTexture ** (1/2.2)
  
  
  # Make the surface
  s = Surface(1,0,0,0)
  s.Diffuse_Color = colourMap
  s.Transparency = transMap
  s.Bump = colourMap * .05
  s.Alternate_Diffuse = diffuseOut
 
  return s
  
hairBit = [
  "Straight",
  "Skullcap",
 ]
 
 
for mat in hairBit:
  outputs += [
    "-" + mat, makeSurface(mat),
]**

...gives me this...

The artefact is quite noticeable and seems camera and light-angle dependent:

BTW, if the code seems a bit wordy, I was preparing for hair texture sets where different regions use different maps (I've encountered one hair set with 5 maps, not including the skullcap). Might be just sending all that as parameters... still coming to grips with that one.

i plugged the fractalsum into the alt-diffuse and got the same artifacts as in your renders RobynsVeil, i did this after several other tests to find out what was going wrong.im still playing with other nodes to get a similar effect... so im off to play,lol

Can someone answer me this question: does grey of any type have to be gamma-corrected? I know, this is a stupid question, but I'm just not finding the answer anywhere...

Not sure where the problem is, Shed... it's not just the maths node: the image node is affected as well... see my last render before I posted my shader.
The hair seems okay at a distance - it's just close-up it looks a bit average. And even at a distance, the hair seems a bit blurred. Overall, I'm not even slightly satisfied this so-called shader I've cobbled together represents what I want to achieve.

Quote - Can someone answer me this question: does grey of any type have to be gamma-corrected? I know, this is a stupid question, but I'm just not finding the answer anywhere...

You know the answer already. The only values that are the same in linear color space and sRGB color space are 0 and 1. Everything else is different.

So I have to GC shades of grey. Dang!
Thanks, BagginsBill. Hope you're enjoying Paris. I lived there for 7 years, back in the 70s... yep, I'm that freakin' old. I understand it's a lot different now... still miss the smell of coffee roasting of a morning waiting for a train at Gare du Nord...

You lived here? How great that must have been. I live near Boston (and lived in it for quite a while). Boston is wonderful because of its age - there is so much history and traditions that are fun, as well as the eclectic community created by the incredible variety of immigrants. I went to school in Boston (far from home) and decided to spend the rest of my life there because of the culture.

But Paris is better than Boston in every way I can measure. I'm in love with it. I've been toying with the idea of buying something here, and trying to convince my daughter to go to school here. She will start college this fall in Maine and will probably spend one semester in Dijon. She wants to be a doctor, and my wife and I are trying to convince her to apply to the medical school at the Sorbonne. She's been in French immersion since kindergarten.

P.S: OK I found one way that Boston is better than Paris. The Indian restaurants in Boston will serve you food that is hot as hell, which I love. The Parisians seem to have an aversion to really spicy hot food, and the Indian restaurant we went to didn't go really hot. However, they were incredibly pleasant and the food was delicious, nonetheless.

Hey, regarding shades of gray, you're not talking to a bump map, right? The values in a bump/displacement  map are not luminance values, and so should not be GC'd.

If you love the Paris of today, there must still be a lot that is much the same as in the 70s. And you are there at the very best time of year, too: Springtime (au printemps).
Most other visitor's experience of the French notwithstanding, I've found that if you make the least amount of effort to speak the language, even Parisians are really quite decent. And there's nothing like getting up really early in the morning, going down to the local Cafe des Sports for a cafe creme (cafe au lait) and a tartine and listening to the oldies at the zinc discussing whatever.
'Bistro' was just not the same in Calif, no matter how hard they tried: it's a unique, quintessentailly French institution. Have you had croque-monsieur?

Jeez, just talking about all this takes me back.

With regards to shades of grey: no, this is for the alternate_diffuse channel. Since specularity also does depend on colour to some degree, I'm going to assume that needs GC-ing as well. Just those channels that use image or colour information as data - bump (incl gradient), displacement - you don't want to be doing gc-ing to. I suppose I should have done the experiment on grey before asking - I'd puzzled over this for quite some time, and arrived at an erroneous conclusion without confirming it with maths.
Thanks for taking the time to answer this, and please do enjoy Paris and the French: it is my favourite place in the world, although Queensland is quickly catching up.

Impressive, Ice-Boy! Love to see the node set. You're doing something different, aren't you?

almost the exact settings like here.V_highlight is 0,27. and different colors.

And you get nil artefact with your FractalSum(x scale) set so low? Which version of Poser are you using? I recreated this exact setup and used it on a plane and got the results I posted earlier.

I guess it's okay for non-close-up work, where you can't really see the artefacts.

poser pro.
but  there are some artifacts similar to yours.

oh and before i forget the skullcap is still textured with the original image files, this is because i have yet to attack those....although this will be soon...lol
oh and thanks to RobynsVeil,BagginsBill,stewer and Jentron for the ideas that have been pushing me forwards in something i didnt think was possible...

Did you get artefact? If not, how did you get around it, shedOfJoy?

please click to enlarge the images...
of course the only image map used on the hair is the Transmap,the hair texture is node driven...
But i have a question for BagginsBill....
most hair has a skullcap material zone, all the other material zones require lines from top to bottom to create the hair strands... the skullcap does not, it requires the strands to eminate out from the center point of the texture. is there a way using nodes to create these strands ? as currently the skullcap is using a texture map, although you cant see the skullcap that well its still there....

It would depend on the angle. If they are perfectly horizontal or vertical, it really doesn't matter, except perhaps to make a darker groove down (or across) the middle.

If they form a V shape, then we have to do a little math. Not hard, but it is "intermediate" work.

But you guys are having so much fun learning to fish instead of be fed a fish, I'm reluctant to give the answer.

Perhaps I give some clues instead.

Consider the mathematical expression Abs(U - .5), i.e. the absolute value of the distance from the current U value to the fixed value .5. What does that do?

Now consider procedural hair that forms horizontal hairs, using the V node to drive an fBm or Fractal_Sum. Suppose instead of using V to drive it, you used:

V + k * (Abs(U - .5))

where k is a small fraction such as .1 or .2. What does that do?

More hints.

A Math:Sub node computes subtraction. Value_1 - Value_2 is computed.
Consider plugging a U into Value_1, setting Value_1 = 1 and Value_2 = .5. Plug that into ?.

A Math:Abs node computes absolute value. Whatever is plugged into Value_1 is multiplied with Value_1 and then the absolute value is computed. So if you were to plug a Math:Sub as described above into a Math:Abs, you'd have

Abs(U - .5)

Of course, in matmatic you simply type:

k = .1
hairdriver = V + k * Abs(U - .5)

If the hair was going the other way (vertical stripes instead of horizontal stripes) just swap the U and V values

k = .1
hairdriver = U + k * Abs(V - .5)