A personal opinion posted to the artists who create content:

PLEASE stop making characters with exotic shaders that require exotic render settings. I'm getting really embarassed asking for refunds!

I've bought two now, and neither work worth a darn. The major issue with this sort of thing is, if you set the render engine for the character, then everything else in your render suffers, incl backgrounds and other characters that function better under more "normal" settings.

I understand that you guys are striving for more realism, but you're not getting there the way you are going. Please test your shaders and tweak them so that they will play nicely with the rest of a runtime.

My $.02 (worth less and less as time goes along)

DR

that sounds really, really weird.  i've never had render settings work for only a part of my scene.  can you describe your problem in more detail?

You misunderstand. The render settings affect the whole scene, so when you have to set them a certain way for a single character, the rest of the scene goes to crap. I'm suggesting that the designers make the characters work without requiring exotic shader settings.

no, i don't misunderstand.

i've never had render settings do anything negative to a part of my image.  i have had them improve parts.  that is, i have changed my settings to improve my results (and take more time), and sometimes portions have needed that improvement but others have not.  i've never, even seen a setting that would some how improve one portion of the image but not another, except maybe texture filtering.  and that's by material in 7 and up.

i've never seen anyone post render settings that gave problems that weren't computing power or a lack of quality.

so what settings are you finding make your scene "go to crap" but work for the figure?  could you show your results?  i'm very curious.

I, too, have no idea what you mean by "set the render engine for the character."

Can you give an example of an "exotic" shader that requires an "exotic" render setting?

 I hesitate to name a product for fear that Renderosity will step on me. (God forbit I should bad-mouth a specific product and hurt sales there-of!) The characters I'm refering to are very new and do all sorts of things on the advanced material panel like "math functions." In the readme for the characters, they plainly state that you must use a specific set of render settings entered on the manual panel. The problem is, these settings make the rest of the scene look like crap, and don't even think about putting a "regular" character in there with him!

BTW... I'm using Poser Pro 7, on a 64 bit XP platform with an AMD 6400+ and 8 gigs of ram. 

If a moderator will give me permission to name a character, I will.

If you don't want to name a character, show a picture with the background how you want it, and then with the effect of the recommended render settings, It may just be a matter of adjusting the rest of the materials, or the render settings. A pic showing the render settings would be helpful too.

can you in the name of god writte anything? is it a horse? is it a car? is it a cat with fur?

in the render settings you have dispalcement,pixel samples,raytracing,shadow only,.....

yeah, the point isn't the character looking good.  the point is that you say the scene looks bad.

i use advanced materials all the time, with frequent application of the "Math Functions" node, and have no problem whatsoever.  i have to change how i light to use gamma correction, but that has absolutely zilch to do with render settings.

LOL @ all the replies. I too am curious to find out about these "exotic" troublesome shaders.

SHOW US!!!! SHOW US!!!!  Hehe.

 Forget it. It's obviously me being an idiot after only using poser for 5 years. Sorry for the bother. Thread ended.

you dont want help or you can not say anything?

looks like you just wanted to complain and not find a solution to your problem.

I'm willing to go out on a limb and make the claim that nobody makes more exotic shaders than I do.  None of these require much in the way of specific render settings. Occasionally, you need to enable ray-tracing or displacement, but these render settings don't mess up anything. For example, this combination realistic skin and splattered paint requires both raytracing and displacement (to get the paint to be thick) and I'm pretty sure that you could render this figure with just about any light settings and render settings and still be happy with it. This was rendered with 1 bounce, min shading rate = 1, pixel samples = 3. These are not exotic settings by any means.

People want to help, not slam you. But so far, you've not given anybody anything concrete to react to.

wait, why?  because you feel like you're being jumped on, or because you fixed the problem?    if it's the second, then no problem.   but if it's the first, then please understand that we've just never seen this problem before. if you can demonstrate it, we might understand it better.  just saying "exotic" doesn't say what the problem is.  unless you've diagnosed the problem on your own, the only solution for a developer is to switch to P4 style materials.  that isn't very viable as a long-term solution.

so please, if you're still experiencing the problem, let us see specifically what settings are creating which problems.

I'm going to pitch in here, not to back up the OPs opinion that it's characters using 'exotic' shader setups that are the problem, but rather to suggest that it's usually background props not set up with mathmatically correct shaders that are the issue. I don't think render settings are really the problem.

I'm not a prolific buyer of content, but just about every prop I have bought and those I use from the Poser runtime have little more than a diffuse texture and perhaps a bump map attached. While I'm often surprised that these render as well as they do, they're not going to have the finesse in the final calculations that a skin setup like bb produces will have. Is that a fair comment?

A glance at my gallery here will show that I almost always render characters against a simple neutral prop (a photo studio type setup). That's because I'm primarily focused on the characters ATM and so have little interest in the background (kinda boring I know). This allows me to put all my efforts into the characters morph, pose skin and basic clothing/hair props (if used).
Now, I'm a self confessed bodger - I get there with the render by any means I know, and certainly not with any kind of  mathmatical prowess (unless borrowed from others in an adapted shader setup). This is fine for a single figure against a white backdrop, but as soon as I introduce any kind of complexity into the scene - well, that's where my approach usually crumbles.

OK, so recently I decided to take a look at bb's VSS skin shaders (PR3 - which I love btw). I've also been keeping up with some of the shader threads here which are just mind boggling from a non math mind perspective - I can just about keep up if I at least have an idea about what surface or effect someone is attempting to simulate. It's got me excited for the first time about the possibility of creating complete scenes in Poser where all the materials might be constructed on some sort of methodical principles about how light interacts with surfaces similar to how they do in the real world (limited of course by this being 3d simulation).
Problem for me is that since I can't honestly say I understand a lot of the math involved with material nodes or how they relate to simulation of RW surfaces (or understand the science of how light interracts with RW materials), making all my shaders for all the potential surfaces in, for example a typical living room, presents a few problems - especially if they are going to work well with each other given any particular lighting setup .

I guess what I'm getting at in my own rather clumsy way is that I kind of sympathise with what I think was the OP's frustration of multiple objects in a scene not playing well together in the final render. I don't agree with the suggested solution (to dumb down marketplace shader setups). Rather, if there is a systematic approach to material construction where materials can more closely simulate their RW counterparts, I'd suggest that ALL marketplace items are made that way so that when placed in a scene and rendered they all work well together.
This is problematic I think in that I suspect there arn't too many people with anywhere near the level of understanding of this stuff as bb has.

Some might say that this would be doing all the hard work for people, but some well conceived, mathmatically correct (as near as possible) shader presets for common materials would certainly speed things up and free up the creative juices for the really fun stuff like image conception and composition etc.

My apologies if I've got the wrong idea here.

Sorry its OT... just wanted to say...

BILL THAT GIRL LOOKS DISGUSTING!
...
...
i like it. girl playing with ketchup. CEWL :) mud shader?

what you say is entirely true, but also entirely independent of render settings (other than using raytracing and Firefly).  if basicwiz had said lighting, i would totally have understood.  as is, i'm really curious and confused.

I agree, it doesn't really make sense that the render settings would be an issue.

Attached Link: http://www.renderosity.com/mod/forumpro/showthread.php?message_id=3382187

 The OP vought a few things from me, but those were all three years ago, and none of them (mostly morph packs) uses "exotic" shaders. So I' can't help to clarify..

But I'm also rather curious. Most of my props these days are using procedural "exotic" shaders because it's much easier (to me) than to UV map and paint a suitable texture - so if THAT is exotic I may be to blame. but I'd just like to know the definition of "exotic"

Alas it seems like the OP too his pail and bucket and left the sandbox. Too bad. 

 One last time I will try and explain.

I bought Claes for M4 (Rendo can just have a cat because I named the name). When I unpacked it, the readme gives the following:

"I suggest to not use Auto settings in the render settings, but use the Manual one. I also have screenshot pictures of the settings in the product page about the render settings i use  

 POSER 7:

• Min Shading Rate: 0.10 or 0.00 ( If you get bad eyelashes, you maybe have this option in a high value)

• Pixel Sample should be set to at least 7 (I use 10-14 in my final renders)

• Raytrace Bounces: 4 (If you use AO, minimum should be 2)

• Min displacement bounds: 0,000

• Irradiance Caching: (I always use 100%) 

• "Use Displacement maps" CHECKED, if you have hair or other stuff with displacement map's in it.

• "Cast shadows" checked, if you want shadows 

• Check "Raytracing" also.

• Max Bucket Size: 32"

Fine. I loaded the character and setup my Poser Pro that way. The renders were passable, but nothing nearly as good as the promo images. I fiddled with the lights and finally got a decent render. At least it was decent until I added props and a background. Now, to make tghe character right, the other stuff is completely over-lit and washed out.

I looked at the materials page and found the following:

Given the statements by the author of the piece, I (obviously stupidly) assumed that the problem was with the way the render engine was interpreting the shaders. I posted here to air the issue and was greeted by rounds of  "give an example."

Reasonable enough, and that would have been fine except for the fact that the last time I criticized a product here, Rendo removed my posts as being "violations." (English translation, don't hurt sales of any items here.) I was hesitant to go down that road again. At that point everyone decides I'm an idiot and here we go again.

I was humbly suggesting that people who create these objects might give some thought to the fact that their creations do not exist in a vacuum. They must either play well with other things out there or they will not be used.

Again, sorry to inconvenience you good folk. It seems that every time I try to post in this forum I get my rear jumped, so this is the last time I'll bother anyone.  I will simply continue to shell out my money for items that catch my eye and ask for refunds when items do not work. I just thought a bit of attention to compatibility would have been a better plan.

DR

oh!

that has nothing to do with the render settings.  it has to do with your lighting.

given your process was:

1. loaded figure.
2. worked lights until the figure looked good.
3. props look washed out.

the lights are the problem.

what you want is a skin shader that works with other elements.  the general problem is that most character makers spend some time getting a good skin shader, while most prop makers don't bother to use anything but a texture and maybe a bump and displacement map.   developers can't hold up their character quality because prop quality is low.  generally speaking, what you need is for prop makers to improve their work,  or learn to translate material techniques to your props. 

that said, there's a lot of tools that can help.

Poser Pro's gamma correction should help overall.  are you using it?  also, this specific skin shader looks fairly inaccurate.  this is the kind of problem where it would help to know more about your lights (type, placement, intensity), your props (materials, placement), and what you're trying to achieve. 

bagginsbill is the man when it comes to lighting and materials.  and ice-boy follows his work closely.  you probably want to focus on what they have to say.

Ah, a lighting issue.

I too have run into this.  A set up (usually on skin) that requires more light or less light then something with a simple diffuse/specular set up.

I would like to see vendors (and some do) make two texture versions of their products. 

For one, if I am mixing characters from two different vendors, the skin set ups can be very different and I am forced to go through and manually change a lot of parameters.  I would be just as happy to start from scratch.

Two, if I am going to be taking this over into another application (such as Carrara) I really don't need all the "exotic" material shaders.

Still, I wouldn't want anyone to stop with really great shaders, just an option for with or without would be a nice touch.

 Those manual settings are interesting. Specifically:

-Does min shading rate affect texture maps? In my experience, no, only procedurals. A shading rate of 0.10?! Egads, this content creator has oodles of time on their hands. 

• Use displacement maps checked / Min displacement bounds 0.00 if you use displacements. What a great way to introduce holes in geometry!

Max bucket size 32 / Min Shading rate 0.10 = 10,240 micropolygons per bucket. Even without  raytracing and AO, I can only imagine how long such a render must take.

Seems to me that these settings are overkill unless doing an extreme close up.

Yes, I'm using gamma correction, and render at high quality settings. I'm not expecting photographs out of draft.

"what you want is a skin shader that works with other elements.  the general problem is that most character makers spend some time getting a good skin shader, while most prop makers don't bother to use anything but a texture and maybe a bump and displacement map."

I disagree. "A good skin shader" is one that looks good under standard lighting and rendering and then goes on to look exceptional under whatever advances the maker has seen fit to incorporate. If it looks like crap when placed with other things (that all look fine together) there is a problem. At least, this is my view as the customer. 

Go ahead and push the envelope, but make sure the backward compatability is there for those of us who have neither the interest nor the technical background to make all the exotic settings. Trekkiegrrrl mentioned that she uses shaders. Fine. Her stuff looks good! She's obviously doing it right. This is what I'm talking about.

This is all a moot point now, however, as I've gotten my money back on this product, and I'm going to move on.

DR

Quote -  Those manual settings are interesting. Specifically:

-Does min shading rate affect texture maps? In my experience, no, only procedurals. A shading rate of 0.10?! Egads, this content creator has oodles of time on their hands. 
.

yes Min shading does affect texture maps. lower the shading rate, sharper the map renders. I found that out from BB.

and no. render times increase but not massively.

You were right to return that, but not because it has a shader. Return it because it has a bad shader, and one that doesn't work right with GC!

I can't see all the connections to know for sure, but I do know:

1) It uses FastScatter - which usually becomes the equivalent of ambient
2) It has little to no specular - the #1 feature of skin is the specular reflections

I'm sure you had to make a screwy lighting setup to make it look right, which would surely make other things look wrong. I agree with you.

And the suggested render settings are very over the top. Almost everything stated would:

1) Do things that are irrelevant to this shader, i.e. have no effect

1. Do too much work for no benefit

For example, why suggest high ray-trace bounces when the shader has no ray-traced effects? For almost everything, 1 bounce is enough. Only if you're looking through multiple layers of glass or have multiple mutual reflections would you need more than one bounce. Furthermore, it says

"Raytrace Bounces: 4 (If you use AO, minimum should be 2)"

AO has NOTHING WHATSOEVER to do with ray-trace bounces. AO works with 0 bounces. You simply have to enable ray-tracing and it works.

The min shading rate .1 or 0 will do nothing because all the default values on the body parts have .2. The higher value is used.

Anyway, these render settings will not make things look bad. They will look great for the most part, as much as render settings influence appearance, although most of these settings are unnecessarily aggressive.

The real problem is the shader was constructed badly, and requires that you put too much light in to make it look right. Other things don't work then.

I don't know why anyone should feel reluctant to name a product that does not work well.  If you buy something and it has a real technical problem, then you shouldn't feel like you will get punched in the groin if you let someone else know about it - although it is good manners to try to work this out with the vendor first of course!

Some comments about those render settings:
Min shading rate to 0.1 or lower is never ever going to hurt quality of the render.  It can only help.  It might cause higher memory usage problems on Poser 6 and older, but it won't ever make anything look the worse.

Increasing Pixel Samples mainly improves antialiasing.  It also can't ever hurt render quality to make it larger.  Raising it above 3 is good advice, and can help smooth out jagged pixelation edges and "stair stepping" effects.

Raytrace Bounces mainly affects reflection and refraction, although I suppose if you had an AO effect BEHIND a refracting surface, or visible in a refletive surface, it might matter - e.g. if Raytrace Bounces is set to 1, and you have a piece of "glass" with refraction enabled for that material, and behind it you're expecting to see some material-based Ambient Occlusion, you won't.  I don't know that it affects AO quality though, it's either on or off in this context.

I've never touched Min Displacement Bounds, and from the material shown there, the displacement amount is so small that I don't see why adjusting this setting should be necessary.  Min Displacment Bounds is intended to avoid "patch crack" artifacts when dealing with a pretty un-subtle displacement effect, where the micropolygons created by displacement can fall outside of the bucket being rendered (bucket size of 32 means 32x32 grid of pixels, a pretty large area).  Read your manual for more info here.

The skin material you're showing has displacement, so if you want it to be visible, the render setting "Use Displacement Maps" would have to be on regardless of what else is in the scene ;)

Bucket Size does not have an effect on render quality unless you set it to something stupid like 1 (1 pixel by 1 pixel).  It affects memory usage and speed.

Quote - AO has NOTHING WHATSOEVER to do with ray-trace bounces. AO works with 0 bounces. You simply have to enable ray-tracing and it works.

I'm sure you're correct, I was guessing a bit when thinking about whether material based AO would show up behind a refracting surface.

By the way, in any situation where your results really differ greatly from promotional images, that's a red flag and you shouldn't feel uncomfortable complaining about that.  Any promotional images I do lately are not postworked - the Bargash promos are, because Renderosity made it a sticking point and gave me the impression the item would not be approved otherwise.  You aren't buying the vendor's skill at postwork.

Quote -
I disagree. "A good skin shader" is one that looks good under standard lighting and rendering and then goes on to look exceptional under whatever advances the maker has seen fit to incorporate. If it looks like crap when placed with other things (that all look fine together) there is a problem. At least, this is my view as the customer.

unfortunately, most props have materials that look bad (imho) under standard lighting, and something with an accurate shader looks good.  that's the problem.  having GC is a good start, but if the prop creator hasn't done anything with the specular or diffuse, it's still going to have problems because the diffuse is at 1.0 (default) and the specular is at 1.0 (iirc, default). that's simply not physically accurate.  then there are the props that just have the texture plugged into the specular channel.   expecting people to cripple their product to match someone else's mistakes doesn't make sense.  you just can't guess at how someone else is going to make mistakes or where they're going to cut corners.

pjz99 - i have to disagree about your images not matching the promos being a bad sign.  because i don't think every user is on par with every merchant when it comes to lighting and render settings.  and sometimes, you can't package the lights you use.  personally, i use bagginsbill's inverse square falloff material on any indoor light.  most of my materials have some bagginsbill tricks in them.  so i couldn't distribute them, let alone sell them.  and, imho, when i use spots without inverse square falloff, it looks like butt.  lots of people start with sets they bought and adjust them significantly.  again, not something they could distribute. 

just because tons of people can't light Stonemason's products well doesn't mean there's anything wrong with his products.  some customers have more money than skill, and i think it's great that even a newbie has access to high quality items.  but i've seen lots of low quality renders of high quality stuff.

not that i think there isn't a problem with this product, or with saying something when you do have a problem with commercial products.

the vendor of OP's product may have specified some settings due to small details like the eyelashes, for which OP didn't supply a screenshot to allow checking the imagemap node settings.  the eyelashes would probly look spotty and clumpy with default settings.  the shading rate is a value that allows an estimate of the relative number of calculations per screen pixel.  if all the other settings are such that FFRender is doing n calcs per screen pixel at shade rate = 1, then FFRender is doing 5n calcs per screen pixel at default posersurface shade rate = 0.2.

Attached Link: http://www.k-3d.org/wiki/RenderMan_Controls

Quote - Strangely, the best place to read about Firefly parameters is not in the Poser manual.

what do they mean with :

''Tip: *A good setting for Shading Rate is 1 if you do high-res work'' ? 

that i should use 1 when i render a big picture?

Yes, what they're getting at there is when you render at e.g. 10,000 x 10,000 there may not be much value in setting shading rate lower than 1.

i actually found that with depth mapped shadows, too low a shading rate and a high resolution means that the weird way the shadows are calculated kind of breaks.  it becomes thin lines that look like a topographic map rather than a solid shadow.  it's as if they need a certain amount of inaccuracy to be one blurred blob.

Quote - > Quote -  Those manual settings are interesting. Specifically:

-Does min shading rate affect texture maps? In my experience, no, only procedurals. A shading rate of 0.10?! Egads, this content creator has oodles of time on their hands. 
.

yes Min shading does affect texture maps. lower the shading rate, sharper the map renders. I found that out from BB.

and no. render times increase but not massively.

Especially when you have a UV mapped figure that's been textured right up to the edges of the UVs without any overlapping. The lower the minimum shading rate, the less likely you'll see a visible seam, which would look like a gap in this case..

Quote - Especially when you have a UV mapped figure that's been textured right up to the edges of the UVs without any overlapping. The lower the minimum shading rate, the less likely you'll see a visible seam, which would look like a gap in this case..

Right, the space between the textures are affected, not the texture itself. I fail to see how (bitmap) texture resolution improves with decreasing the shading rate. Said another way, if a single pixel of a bitmap in shader space is broken into "sub-pixels", how could one introduce greater resolution? They're all still the same color.

No, the texture's clarity is pretty obviously affected by min shading rate, here's a good example from a while back:
http://www.renderosity.com/mod/forumpro/showthread.php?message_id=3189633&ebot_calc_page#message_3189633

I used to think the same, because according to a lot of documentation it really should not behave this way, but it does.

I'm no expert on REYES, so I can only parrot what I read.

This is what I've read.

All geometry is sliced up into micropolygons. The shader (and any texture map lookup involved) happens at the vertices of these micropolygons. As the micropolygon is rasterized, the shader is not consulted for interior points - it interpolates between the values sampled at the vertices.

Now since the overall size of a micropolygon is determined by the shading rate, it follows that shading rate has some influence on the number of times that a shader is evaluated in a given space number of pixels. If you set the shading rate such that a micropolygon is larger than a pixel, then you get smearing of the texture (or procedural shader) across those pixels. Even if it is the same size as a pixel, it may not exactly fit in a pixel (probably never does) so the data for a single pixel will come from two or more micropolygons that collectively cover that pixel. So you get more smearing.

If a true polygon in your geometry is already smaller than a pixel, I suppose that the the actual polygon is used as the micropolygon, in which case the sampling will happen at each real polygon vertex.

OS X uses sub-pixel screen rendering when displaying anything (graphics or text).  I dunno how it works, but perhaps it affects how one perceives a poser render.  does vista also do sub-pixel screen rendering?

p.s. wouldya lookit the tongue on pj's avatar.

Isn't anti-aliasing a form of sub-pixel screen rendering? The pixels are softened at the edges to give a smooth line which runs between pixels. Blurred, I guess. You only have so many pixels and they can't be subdivided, but the edges can be blurred together or softened to make it look that way.

Quote - Isn't anti-aliasing a form of sub-pixel screen rendering? The pixels are softened at the edges to give a smooth line which runs between pixels. Blurred, I guess. You only have so many pixels and they can't be subdivided, but the edges can be blurred together or softened to make it look that way.

Depends on how it's done. I don't know how Firefly does it, but I know of at least one ray-tracer that handles anti-aliasing by shooting more rays in the general direction of each pixel, each slightly off center, then averaging the resulting pixel colors. The effect is controlled adaptively by comparing the color of the pixel after the inital ray to the colors of finished neighbors; if the color is off by a certain threshold amount, the extra rays are shot.

Not to be nitpicking or complaining, but the problem of understanding this lies in the definition of "pixel".  People often using this as a rather unsharp definition.

Technically a PIXEL is a dimensionless indivisible thing. Only together with a length dimension it's making sense. Especially as a DPI measurement.. That means "Dot per Inch", there dots are used same as pixel. So 300 DPI give you 300 pixels on one inch or 2400 DPI gives you 2400 pixels on that same inch, so much more dense.

To describe a image map it's useless to do it like it's often done  (look into all marketplaces)  "with lots of  4000  *  4000  pixel maps".  This does say nothing about the map quality cause  you don't know  how big the map  is really.  I assume the creator is meaning  a image with 4000  DPI  * 4000 DPI, in x and y direction. Than it's a hires map.
But 4000 * 4000 pixels could also be on a map of real 10 * 10 inch size and than it's only 400 DPI in each direction.

So with that in mind the shading rate have to be seen according to the DPI of the given shader, i.e. a texture map.

Looking to the document linked by Bagginsbill there is following definition:

Quote - The Shading Rate tells the renderer the frequency at which the shaders should be run on the geometry in your scene. Confusingly, it is expressed as an area in pixels, not as a frequency per pixel, as one might be trapped to guess from its name.

Quote - A Shading Rate of 1 means that no micropolygon's size will exceed the area of one pixel. Imagine how even the simplest objects will generate thousands of micropolygons and render beautifully and without any nasty polygonal silhouette edges. Tip: A good setting for Shading Rate is 1 if you do high-res work.

*   meaning hires INPUT !!!
*> Quote - Always keep in mind that Shading Rate determines area respective size of a micropolygon, not its edge length! At a Shading Rate of 9, micropolygons should end up with a maximum edge length of around 3 pixels on screen, namely the square root of 9.

If I interpret that right a shading rate of 1will stay with one pixel. And cause a pixel can't be divided we have to come to a 1:1 conclusion with the DPI rate of the texture.

And so in fact of a shading rate of 0.1 we end with 1/10 of a pixel but the dimensionless pixel can not be divided and so the only way is to increase the DPI rate of the rendering. So instead of 100 DPI output with shading rate 1, we will end at 1000 DPI output with a shading rate of 0.1.

And so indeed a lesser shading rate will highly sharpen the look of the output ... if it could be given by the input texture, for sure.

And the same with sample pixels rate for the antialising. Thinking of 2D imaging antialising will NOT increase the DPI count of the image. It is smoothing the hard edges by averaging the colors and/or greyscale of neighbors. But this is done with a loss of sharpness cause pixel can't be divided. So more antialising means for example to smooth between pixel 1 and pixel 5 and to loose the information of pixels 2,3,4 cause these are now needed as intermediat steps between 1 and 5.
Also with this if the overall DPI of a image is higher, antialising will work better and it seems that you loose less sharpness. But you loose the same 3 pixel content in above example. But they represent less real detail each by each as with less DPI.

You can reproduce this simply with a scanner. Scan a simple straight line with 100 DPI and with 2400 DPI. You may find that first the line will be covered by 10 pixel and second with 240 pixel. So if you antialise the edge of the line by the given 5 pixel, you will loose 2 pixel of the line (and 2 of the background) for that. In total (left and right) you loose 4 of 10 pixel detail of the line.
With the higher DPI you loose 4 of 240 pixel detail.

So with this it's pretty clear that a smaller shading rate will give a sharper image and than also a higher sample pixel will do less unsharpness with smoothing the seams.

How this simple 2D work is produced to corresponding micropolygons (mp) isn't my part of the theory. But I understand that in a manner that with shading rate below 1 the mp's will cover 1:1 to the texture pixels. And with shading rate bigger 1 depending of the quality of the texture map it can happen quickly that 1 pixel of the map is used to cover more than 1 mp. That means smearing and the output will look rather unsharp.

Just my 2 cent ...

Quote - assume the creator is meaning  a image with 4000  DPI  * 4000 DPI, in x and y direction. Than it's a hires map. But 4000 * 4000 pixels could also be on a map of real 10 * 10 inch size and than it's only 400 DPI in each direction.

.... That's - uh - complete nonsense.  DPI has zero, zip, nada, nunquam, nothing whatsoever to do with pixel "size".  Bitmap pixels do not have an intrinsic size, they're a unit of convenience only.  Size comes into play when you get to a printing device or display device.  4000x4000 pixels is exactly the same whether DPI is 1 or 1 million.

I never will understand why only in Poser do people make "off size" image maps.
The whole professional 3D world works in powers of two for textures, and there's an actual physical reason for that, but nooooo... not for Poser. Poser has to be different, I suppose.

You mean 4096x4096, rather than 4000x4000?

DPI is Dots Per Inch and refers to the number of dots that a printer makes per inch. It has nothing to do with pixels.

PPI is Pixels Per Inch and has something to do with how big an image will be when you print it.

You can print a 300 PPI image on a 2400 DPI printer - all that means is that a pixel will be printed as 64 (8x8) dots.

Some scanners have a dialog box asking you for what DPI you want to scan at - this is incorrect terminology. The scanner should be asking you for PPI. DPI has nothing to do with PPI.

And rendering parameters have nothing to do with PPI either. You render to a certain number of pixels - period. How you print them (later) is not involved in the calculation.

And a shading rate of 1 does not mean that the texture map associated with the geometry will be rendered at 1 texture pixel = 1 screen pixel, nor does it mean that a micro-polygon is exactly covering one pixel in the render. With displacement, a micro-polygon can be sheared to cover many pixels. Furthermore, I can scale an entire 4K by 4K image map to fit in one pixel in the render. Simply make a one-sided square cover only one pixel, and attach an image. The entire square will be rendered into one pixel. Now, how many pixels of the attached image will be consulted? Do you think all 16 million are consulted to produce an average color for the whole image? Nope. Only 4 samples will be taken - one at each corner of the square, because the square is so small the area is already the size of a micropolygon and so will not be further sub-divided.

As for the 4000 number versus 4096, that's just what some sloppy people do. There are plenty of texture makers who produce 4K (4096) dimensions. And some who actually mean 4K say 4000 by mistake, but actually have produced a 4096 dimension image.

And ... as for whether using 4000 is stupid and bad practice, it just means you could have had 96 more pixels of detail, but you don't. The renderer still expands the image to a power of 2 before doing mip-mapping.

Quote - And ... as for whether using 4000 is stupid and bad practice, it just means you could have had 96 more pixels of detail, but you don't. The renderer still expands the image to a power of 2 before doing mip-mapping.

But the reason was an ancient (possibly Mac only) problem which limited the max texture size to 4000 (Poser 3 IIRC) and like a lot of other Poser related things (e.g. MAT poses) it's become engrained.

Quote -
And ... as for whether using 4000 is stupid and bad practice, it just means you could have had 96 more pixels of detail, but you don't. The renderer still expands the image to a power of 2 before doing mip-mapping.

Graphics cards render textures in real time only in powers of two, and if a texture is, say, 4K, the GPU has to internally resize it to 4096. I don't know if it does that every time the camera is moved, or only once as the texture is loaded, though. If the GPU is doing that ALL the time, there will probabaly be a considerable display performace hit, especially if you have a scene with a whole lot of off-size textures.
It probably doesn't make a big difference in Poser, but I've been told it's an absolute necessity for game textures and animations, so this is why people make textures in powers of two.
The former limit isn't the only reason people are doing it (probably most people don't even know about tthet) - I think they do it because they see everyone else doing it that way, started by DAZ. Also, there are plenty of 3000x3000 and 3500x3500 and oher sized textures out there, too, so that's not explained away by any former 4K limit.
And when GPU's are being  used for actual rendering, along with the CPU, it will probably become more of a necessity to have "correct" image sizes. Well, not like Poser is ever going to get GPU rendering anyway....

At any rate, if nothing else, it would just lseem more professional to use industry standard texture sizing rules, no?

But aside from that, if the renderer is expanding an image size, isn't it getting stretched slightly? Why not just make it that size to begin with? I can't see any good reason why not.

I just re-read your post, bagginsbill. I really wasn't arguing with you, I seem to have gotten your reply mixed up with nruddick's reply.