As some of you know, I'm going to be selling creature characters on the side (through CP) and I'm curious on people's opinions on multiple UV tiles for a figure versus a single uv tile. With multiple uv tiles, you can get a lot more detail per a figure but you can also run into issues. Third party vendors without access to a 3D texture painting software will have a hard time matching up seams and getting things like displacement to line up well.  However, the visual yields can be significant without the need for large texture maps (though obviously larger ones allow for even more detail).  With single uv tiles, you have something that's easier to work on, line up and share but you're forced to use very large texture sizes to achieve details that hold up well close up.  The question I have is this: For creatures and monsters, where it's less likely that others will retexture it, is it still better to aim for ease of use or is visual fidelity the better way to go?

Thoughts? I am honestly torn.

That should read "With multiple uv tiles, you can get a lot more detail per a body part...". Sorry for the typo there.

"Thoughts? I am honestly torn."
Ah man, I knew it, I bloody knew it :-D

Nice one, Teyon, but on the subject of Multiple UV's, I personally despise them just as much as I despise multiple body part meshes, they're both a pain in the arse to work with even though they have their advantages.  I've always hoped that figures would have a single mesh and a single texture, so, imagine that, how nice and usable that would be compared to what we have now.  I nearly dropped dead in disbelief when I read what DAZ did with their Genensis, something about it not only having multiple UV's, but they all overlap a single space or something.

OMG, no, just, please no.

Couldn't you get Steve to make Poser able to hold larger textures because that would negate the need to split for better texture detail?

 The material room setup/ Only using the Diffuse texture and let Poser build all the rest inside the material room from this single Diffuse texture.

All of the above use the same texture with some material room work, and this is the "slightly adapted" default texture for the Poser Dev Lo Res Female.

@ Pumeco : Poser has NO texture size limit. Want a 8.192 x 8.192 pixel texture size? Make it and load it.

@ Teyon : The end user needs ONE high quality Diffuse texture.

YES , one 8.192 x 8.192 high quality diffuse texture would be preferrable to make material room work more end user friendly.

Poser is far better to build specular and displacement from the diffuse texture inside the material room then most content providers are.
Less textures to load, less memory consumption. **A High quality Diffuse texture is all what is required.
**
Ok, I agree : 8.192 x 8.192 textures are more difficult to build for the content creators (due to screen size and other applications limitations), but they are a lot more end user friendly then UV-tiles.

For dragons or other special 3D creatures, you might need a second or a third map.
For humans?

It would also help the "home alone" hobby end user that wants to experiment  if the texture UV was also symmetrical in lay-out.

Best regards, and keep up the good work my friend.

 @ teyon

For a dragon?
I would provide one Diffuse texture at 8.192 x 8.192 pixels
And one Displacement map at 4.096 x 4.096 pixels

For a human?
I would provide one Diffuse texture at 8.192 x 8.192 pixels
And one bump OR normal map at 4.096 x 4.096 pixels.

A displacement map can cause poke through the clothing at render time. => Bump and normal maps are both "fake 3D"  and do not cause poke through clothing.

Forget specular maps for every day skin; They are a thing of the past with the latest improvements in the material rooms possibilities.

For humans it makes complete sense to use a single map for the skin - that's why I've always suggested that at the office and part of why we moved to that for some of our more recent characters. It's just easier for the end user and for vendors who will be creating new textures for the character. That said, how many creatures get multiple texture sets from vendors who are NOT the creator? Not many, with the exception being the Millenium dragon (and last I checked, that used more than one map). Which is why I'm even asking the question.  I'm not working on a human - I legally can not do that outside of work. I'm working on a creature, so take your blinders off for a second, tie your knees down so they don't jerk during your reaction and consider the question. In film it's not uncommon to see multiple maps (at 8K!) being used for humans, creatures, doors, you name it.  The only reason we don't do that for Poser is to allow for ease of creation by third party vendors and the possible degradation in performance from multiple 8K maps running around all willy nilly. The average end user does not get into the nitty gritty of creating content - as much as it'd make understanding of the work involved and perhaps appreciation of value of the item(s) a bit more widespread.

I'm actually leaning toward providing both the multi-map approach and single map approach and then letting the voice of users decide. It won't make it for this first creature I'm making but yeah, I think I'm going to do that for the ones following.   Thanks for responding, you actually helped me decide a course of action.

"Thoughts? I am honestly torn."
Ah man, I knew it, I bloody knew it :-D

Nice one, Teyon, but on the subject of Multiple UV's, I personally despise them just as much as I despise multiple body part meshes, they're both a pain in the arse to work with even though they have their advantages.  I've always hoped that figures would have a single mesh and a single texture, so, imagine that, how nice and usable that would be compared to what we have now.  I nearly dropped dead in disbelief when I read what DAZ did with their Genensis, something about it not only having multiple UV's, but they all overlap a single space or something.

OMG, no, just, please no.

Couldn't you get Steve to make Poser able to hold larger textures because that would negate the need to split for better texture detail?

• As an aside, while I think Natalie Imbruglia very pretty, she really doesn't look like Roxie to me. I know who Roxie was based on and while I had to change the face before release because the face wasn't generic enough, I still see her in there, not Natalie. :-) Torn's a great song though. 

Also, you can rig a figure that uses only a single group. I had a video once showing how this could be done. It would mean all the bones were ghost bones and would make manual selection of body parts tedious, as you'd have to use the drop down menu for selecting everything.  However, it's totally do-able (just make sure everything is an affected actor of everything else in the body). I would not suggest it though. The Poser method, in some ways, is easier. At least for the end user.

To be honest I'm a total noob when it comes to rigging, in fact it's the thought of having to deal with multiple parts and textures that fills me with dread any time I fancy playing around and having a go at something like that.  It's been the biggest pain in the backside for me ever since my first Poser (Poser 3).  From early on I wanted to be able to just paint a texture onto my figure, but at the time  it was practially impossible to do due to the seperate texture maps.  And then there's morphs, it's just way better to load up a complete mesh in ZBrush and be able to work on the full thing without having to worry about technicalities.

That's really "the ideal" for me, but Genesis, for example, in comparison just sounds like an absolute nightmare to me due to all the technicalities of the way it's put together, mapped and rigged or whatever.  I was going to suggest that weird but simple technique where you use a single mesh and just define the texture unfold, but I can't rmember what it's called.

Thanks for the head's up on Natalie/Roxie, but man, you totally spoiled the fantasy, argrrrrrrrrr!
Suppose I'd better tell Roxie then, she's been disturbingly quiet since I brought it up, wondering about it no doubt :-D

 @ Teyon,

Sir, you actually answered your own question : "Why not so many texture creators?"
The seams.
The seams along different maps are the most problematic to make, and make it difficult for the less talented "home alone" end user / occasional content makers.

Single map will Always be more content creator AND end user friendly.

Going to 8k maps gives you a lot of area to work with on the details.
There is less * "lost space"* on a single 8K map then on 4 x 4k maps, so 8K maps are in a sence more economical too, and they offer more fflexibility in the uv lay-out.

Going to 8K is a win-win-win situation.

Imagine loading a 8k map for an eye colour, another for a different facial option on your character....Not practical in my opinion. How big will your files be? Also on that 8k UV, how many pixels are dedicated to the eye area for example? Will it be as sharp and clear as a 4096x4096 UV eye template? I am not seeing the benefits of one UV. Regardless, you will still have seams to deal with if you paint by hand on a single UV map. If you use projection mapping (which is how they texture a model these days, directly painting on the model), amount of seams don't really matter. So ask yourself what is the goal here.

Not everybody can paint textures across seams manually (regardless how many), and one UV is not really going to change that fact. In most cases on a single UV, the details are warped and not straight so manually editing them is harder than a split UV where the details are straighter on that same area. Texture editing is a skill that takes crafting and time.

 Euh, Zevo??
Single map means, euh,  single map. ALL is on a SINGLE Diffuse texture map.

And model, and UV-Map, and paint economically.
Example : For a pupil, iris, eyeball, you only need one.

Yes, so say I want to make an iris that is brown aka texture editing, it will load those huge dimensions of the entire diffuse just for that change? That is the point I am trying to make. SO while pupil and the rest use one huge uv map, the iris will use another.

 Ach, it s the same in every aspect of this hobby.
Everybody adds, and adds, but not a soul does a clean-up.
*How many clean out the old Spheres or Capsules when they are converted to Weightmaps?

Poser, like most 3D apps, would benefit when every aspect of the hobby would become more and more end user friendly.
And, so sorry to say, but some of this starts with the content creators.

Forget old habbits and start thinking  "end user friendly."

Model, UV-Map, and paint end user friendly.
Rig end user friendly.

Here, I completely understand Pumeco.

Lots of things that could be very simple, are overengineered and add to the confusion of most.

 @ Zevo
Want to change an eye color?
Same texture as before but;  Do the change in the advanced Material room.

For any entry level end user is is dead simple to add a mask and play with the Hue (Just giving an example here).

Come to think of it. It can all be done inside the advanced Material room, even without a mask.

Not all can be done via material room, and I am well aware of just editing colour values. My point is what if somebody wants to add in different actual texture details that cannot just rely on colour manipulation? Say for example they want to add in a different facial blemishes, can't just change colour there, you have to swap out the texture or add an overlay, which, will be on a huge UV because well, all uv assets are all on one and it will have to match exactly where the source area is being changed. So instead of just stacking a face uv, you are stacking the entire single UV map dimensions. My point is, one huge UV map will cause more problems than it's current structure in this market. Single Uv map is fine for one option, but for those who might want to include multiple variations of details etc, it can become problematic and could potentially use way more rig resources than the current format.

*Image added after writing the post.
This one is using a second texture, using the Blender node and a mask to "dress" V4.
I"ll search for the one that does not use a mask, but where Poser creates the mask for me. 

@ Zevo
There are soo many options and variables to play with.

Example :
Start with one high quality 8K Diffuse texture.

Now you want to some changes.

1. You "can" use a second map at 8K

2. You "can" Blend a second map with the first map.
   But thet second map (with your changes) does NOT have to be 8K at all.
   As long as the proportions are correct you can "BLEND" a 2K, a 4K, or an 8K or even a 16K map with the Original 8K map.

3. You can use  the Blender node (as above), but you can also use the "Add" node, or the "Ramp" node, or the "Color_Math" node, all at different map sizes.

4. You "can" use UV offset if required on you secondary maps.
   In the Diffuse texture node ; Set the U_Scale, the V_Scale, the U_Offset and the V_Offset to get on the aera required.

5. You "can" create masks inside the advanced material room to cover about any change you can think about.

=> Did you see how I put a corset on V4, just using the material room? => I"ll search for the example in the archives and post it back here.

There are so many ways to blend or drive secondary maps at different sizes, or to get on different area's on a map.

One of the huge advantages of the advanced material room is its flexibility.

I prefer a single map. Poser can handle big maps, it makes stuff easier not to handle with a bunch of them.

 Found, from a post I made in 2011.

(Click to enlarge)

Here I do the oposite, I Blend a  HUGE secondary texture over the Default V4 texture.. 

The Math_Functions nodes build the mask automatically for the secondary diffuse texture in the lower right hand corner.

This mask drives the Blender node to determine where and how the blending between the MAIN and SECONDARY have tp be.

The Seconday texture is 8k and the Main texture is 4k.

**PS; And now for the NICE STUFF !!! LOL

DID you figure out how I changed the Color of the corset Yet????

In this example it was easy. Just jut a color In Input_2 of the Blender node. Tada!!!! Colored corsets.

Poser is such fun, fun, fun.**

I like the DAZ gen3 type UV maps where you can tell if you're painting stockings on the legs. With the Vicki4 UV maps it's hard to know where you're painting.

But, I prefer people and creatures with many material zones so users can add glove textures to forearms with just a simple click of a Poser shader preset.

If you want to make everyone happy, you could include multiple versions. 

I would rather have more zones. But at the same time you can add what you need in Poser.

Both ways of doing it are valid. And in the end it is up to the creator to pick based on the tools they have, and what they think the end user will want out of the character.

Victoria is probably the most popular figure series for Poser, and she has a lot of material zones from 3 up.

Many of the Poser characters have single zones, and we all know how much support they got in the past.

Support has very little to do with with things like texturing, as the V series of characters has already shown.

Does that mean that a figure should be multiple zones? Well, that depends on a few things.

How are you going to license the figure? Will it be allowed to used outside of Poser, such as in Unity as a game character?

If so, then the past may be completely irrelevant. Unity does not support tons of materials, nor does it support wire frames above 65K without breaking them down.

Will you offer multiple LOD wireframes? That is another thing to consider if it will be usable outside of Poser. Good low poly topology is far better than decimated.

The future of successful Poser figures may have more to do with what you can do with them outside of Poser than in it,

Yes it will have to be fully functional in Poser, but making it so it is functional outside of Poser expands its usability and its user base at the same time.

Things like rigging for other apps could be sold as addons for the character if it will be licensed outside of Poser.

Flexibility in a figure is what will sell it, and that flexibility has little to do with the way it bends....

As far as the number of texture maps for non-human characters, the skin (or equivalent) should be all one map. Other body parts such as horns, fangs, eyes etc that more than likely wouldn't have textures that need to be lined up to match the skin could have different maps. This is coming from someone who might edit textures for her own personal use but does not make content that involves texture maps.

@Vilters
I'm not gonna sit here wondering what colour her corset is because I already noticed that two are 'sheer' - and that's fun enough for any man :-P

@RedPhantom
I think that is key, that's pretty much the way I see it, I'm sort of between what Vilters and Bantha prefer and what you just mentoned.  My own decision would depend on how transmapped resources work (and I'm unsure about that).  For example, if putting the textures for parts that require transmapping all onto one texture, would be a resource hog, then I'd put all continuously connected stuff onto one texture, and all the discontinuous transmapped stuff onto a smaller one.

The only thing I would never do, not in a million years, is put different parts of what is supposed to be a continuous surface, onto different maps - it just seems ridiculous and like Vilters said, totally unnecessary - it's been the biggest most pointless pain in the arse ever since I started using Poser.  Just my opinion, but I think all skin should be on one UV and texture map.  I personally have no problem with eyes and fingernalis being on a different map because they're not supposed to be part of a continuous surface like the skin is anyway, therefore painting those would not cause any problems with edges, them being discontinuous from the skin in real life anyway.

It just seems the logical, sensible way to do it.

 I love the idea of "end user" friendly ;-) with that in mind one big~ish map is great for beginners

for the more experienced so long as its unwrapped well and has sensible mat zones,all we need is the .obj and a uv mapper(classic works blender is better lol ) we can export at any size we like ! we can export by material[other] at any size we like

as vilters  shows there's a lot of "tricks" in the mat room to extract info from the diffuse and apply it to other channels 

Problem with one, single, big 8000 pixel texture for a human: when all I'm rendering is the face and hands, I still have to load a massive texture that contains data for parts that are invisible. It's a huge waste of computer resources if you occasionally, I dunno, clothe your figures. Or do close up portrait head shots. Not a problem if you're doing bikini pinups of one gal. But for a big battle scene of 50 dudes (and dudettes) in various kinds of armor, I NEED to be able to load just the head for many of them. It would be better, for my purposes, to have even more maps than we currently do in general.

So: If your monster wears clothes, then multiple maps may not be a bad idea. If he's pretty much always nude, then I guess an enormous single map will only inconvenience people who want to do a render that isn't full-body.

Some reactions here/

1. For populated scenes with lots of background figures, the end user can reduce texture size at will. No issue.
2. Poser uses texture caching to read textures. Again, no issue.
   3; The number of material zones ? - Default as few as possible.
   Why?
   The end user can add material zones at will inside Poser.
   => But once inside Poser;  it is IMPOSSIBLE to get rid of the extra material zones once they are in the figure.

Removing or merging, or reducing extra material zones can only be done externally on the obj file and also requires cr2 work. => Pretty work intensive

Conclusion:
Adding material zones is dead easy inside Poser.
Reducing the number of material zones is impossible inside Poser.

That is why => Default should be : as few material zones as practical.

"It's a huge waste of computer resources if you occasionally, I dunno, clothe your figures."

But how many actually clothe them?

I might be wrong but I honeslty think only a small percentage of the figures out there get fully clothed. There's a lot of clothed figures in the galleries here, but likely a million more times unclothed figures doing the rounds on the net.  One other thing to consider is that clothing your figure doesn't always negate the need to worry about UV seams, where the head map meets the body map being a good example of this.

So, even clothed, splitting what is supposed to be a continuous surface into a multiple surface, can still be a problem for the end user even when clothed.  I remember years ago, I did render where the figure was actually clothed (yes, I actually did one with clothes :-P), but I had to give her a neck band just to hide the stupid seam between the head and body maps.  I didn't want her to have a neck band, but the top she had on was close to being a V-top so it was either that, a scarf, or have her look like she'd done a head swap :-D

I think the point a lot of people are missing is that what was a problem when Poser 4 was a "thing" is not a problem today. Anyone serious about texturing should have access to tools that can do the job.  Correct me if I'm wrong here, as I stopped paying attention to the program years ago, but I'm pretty sure even Blender offers painting directly on a model. So really, other than just pure hatred of new things or an honest dislike of an interface, there's no excuse for not being able to remove visible seams from your textures in the present. The past is dead, let it be so.   So while I can understand how back in the day, you saw a seam and you screamed bloody murder, in today's world, that just shouldn't even be a problem anymore. 

Now here's a big question for you large map types - I have a large, 8K map of a head and body (all on one single map). The head's even scaled up a bit without overlapping on the body, so as to give it a little more texture space.  Then I have an 8K map of a head and body (on separate maps), both scaled up to take the maximum space available on their respective maps. Which set is going to give me better detail?

Ultimately, as I said before, I came to the conclusion that I will offer both types going forward in my personal work. I think it just makes sense. I've been considering a lot of different ways of improving upon what I do - some of it means going back to ways of doing things I had stopped because of the "ease of use" idea. I think, for what I'm trying to do - what I HOPE to do - it will benefit everyone down the road. Poser's a great little app and the community is so very spirited and always seeking the best, I want to try and see if I can pull out better quality stuff given the time so as to honor that. Not saying the stuff I've done for work isn't good - it is, I just feel like I can do better with more time and this is an opportunity to do that. To explore, where I may otherwise not have the time.  The first character I release won't benefit from this but don't hold that against me. :) Anyway, good talk gang, glad to see the various points of view - they all have their strengths and weaknesses. 

 Detail equals the number of pixels per head.
The more pixels/head gives more details.

Exactly. You'll get more pixels per head when it isn't competing for space with the body.

 You could also try the U_Offset and the V_Offset trick in the material room to get four 4k maps on a single 8K map.

I know some end users that build those 8K maps from the default 4K maps for Alyson and for Roxie.

it makes material room work soooo much easyer.

And if you want to adapt the textures in Photoshop or Gimp, one never has to worry to get them all axactly right.

When combined into a single 8K map, they are similar = no seams danger.

Want an example?

 In the math room? Typing this from memory so if I misstype one please forgive.

From an 8K map :
U_Offset 0 and V-Offset 0 is the left upper corner 4k map of the 8k map
U_Offset 0 and V-Offset 1 is the right upper corner 4k map of the 8k map
U_Offset 1 and V-Offset 0 is the left upper corner 4k map of the 8k map
U_Offset 1 and V-Offset 1 is the right upper corner 4k map of the 8k map

So some end users combined the 4 main Diffuse 4k textures into a single 8K map to play this trick.

And now they can adjust the 8K map  in Photoshop or Gimp at will.

4K detail is maintained, but texture work and the material room work is simplified.

*From the above description..

This (reduced) example of a 80192x4096 Alyson texture shown the idea.

Here I only use the U_Offset in the material room to choose the left or right area of the 8.192x4.096 map.

But in Photoshop or Gimp, I can rework the texture without fear of having colors differences.

For human figures, this seems like a good compromise.
And, yes, there is room to spare to put the eyes and inner mouth also. :-)

Best regards, Tony*

 Arch, there is no way to add the Nudity flag any more.... Mods?

@Teyon
You're preaching to the choir with what you just said, you can't be everywhere but I've said a few times I'm a big fan of your work (you'll just have to believe me).  You have that knack of making things look human and that's one of the reasons I like Roxie.  I realise Roxie was made to a deadline and you'd have loved to spend more time on her, so I think it's great that you're doing your own thing on the side and I wish you the best of luck with it.

I'm totally clueless about the technicalities of maps v rigging though, but like I said, I'm only stating it from this persons perspective, I'd take a large map and ease of use over multiple maps any day, but that's only because it suits the way I prefer to do things - even though it's not necessarily the best way.

I tell you what though, Teyon, it's a shame Roxie was released the way she was (without morphs) now that you intend to do it independently.  The thing about Roxie is that she fills a gap in the Runtime that needed filling, and had you made her more refined and developed a set of default morphs for her, I think she would have sold very well as "An Independent Figure by Teyon".  I'd have bought her in a heartbeat and I hardly ever buy stuff for Poser.  Even the name "Roxie" is a good one - sounds sort of cheeky and mischievous - perfect for the sort of figure she is.

in contrast, I'm not sure how big the market is for creatures but I hope it gives you a good start at being an Indie character dev, enough to keep you at it.

"Arch, there is no way to add the Nudity flag any more.... Mods?"

No, Tony, there isn't, so I'm going to report you to Shane for posting dirty pictures without a naughty tag  :-D :-D :-D

 @ Pumeco

Using the material room trick I mentioned and demonstrated above, you can make your own 8K map for Roxie.(Or Alyson, or any other multiple map figure).

Then take the combined map into Gimp (I think you are a Blender and Gimp fan) and work the combined texture over.

Then use the U and V_offset trick in the advanced material room at will.

PS : Blender lets you paint over seams of older figures to correct them, and resave the textures too.

Cheers Tony, I've actually bookmarked this thread for the stuff you posted on the previous page in case I forget where I saw it when I need it :-)

I see no true value to one map for the entire figure unless you are catering to people who are new to poser and daz. At the very least I would think 1 for the body, 1 for the face, 1 for the eyes, 1 for misc (tongue, teeth, etc...)...

1 8K map might seem like a good idea, but you had better be a uv master (game quality) to get the benefits.

The biggest gain is in "end user friendly". Something that seems to be forgotten all too often.
And fexibility.
And usability.
And content creator friendly too.
And making the most of the 8k surface. Reducing "lost" and unused space without loosing quality.

**On a QUALITY note ;
**Any end user can Always reduce texture size if required. (Thinking background figures here, to reduce memory load)

Increasing texture size is IMPOSSIBLE without loosing quality.

One thing that has to be considered when UV mapping is the percentage of the texture space that is actually on the figure. (Tony touched on this, and the Alyson texture is a good example)

If you have an 8k texture and only 50 to 60% (or less) of the texture space is used, it is a complete waste of usable texture space.

The advantage of multiple textures is getting more pixels on each polygon. You still have a lot of wasted space, but you can get more pixels onto each polygon this way. Alyson's tongue has a higher pixel per poly than her ears do. Great for dental renders, not so much for ear ring ads.

UV mapping in general, is a waste of texture space. Hence alternatives like Ptex, which resolves far more than UV issues..

Poser does not have Ptex, so careful UV planning is a must to use the highest percentage of the map without making the texture to hard to edit.

You can't always use a high texture space percentage due to the wire frame and where the UV seams are,  so using multiple textures is the easiest way around that so you get a high pixel to polygon count..

End user friendly is one thing, terrible pixel to poly count is another.

I would rather have a UV setup that focuses on high pixel to polygon count over all crammed on one map and making it pixelize on closeups. Depending on what you are doing, you may need more than one texture to do so. People are one of the things that need a high pixel to polygon count if you want detail close up..

Picking up on shvrdavid's last comment - The benefit of V4 4K face map is realised because people want to render portraits,  the area of the face fills most of the picture frame and the extra detail is used even when displayed on a computer screen. Compare that with a horse, most people will render that at a much further distance from the camera so making use of the extra LOD on a seperate head UV is more difficult to realise so of less use.

Care must be taken with multiple UV maps when shared via the Internet as they increase download size and what tends to happen is people compress the textures to support slow Internet connections and this has the overall effect of a lower LOD for the whole model.

Some reactions here/

1. For populated scenes with lots of background figures, the end user can reduce texture size at will. No issue.
2. Poser uses texture caching to read textures. Again, no issue.
   3; The number of material zones ? - Default as few as possible.
   Why?
   The end user can add material zones at will inside Poser.
   => But once inside Poser;  it is IMPOSSIBLE to get rid of the extra material zones once they are in the figure.

Removing or merging, or reducing extra material zones can only be done externally on the obj file and also requires cr2 work. => Pretty work intensive

Conclusion:
Adding material zones is dead easy inside Poser.
Reducing the number of material zones is impossible inside Poser.

That is why => Default should be : as few material zones as practical.

I wonder if you actually make a lot of complex scenes with multiple, fully clothed characters interacting -- because the only way to really know what the problems are is to discover them yourself. Sure, you can lower the texture resolution for background figs to practically nothing. Some will be fine with procedural skins, even. And having a single map will make reducing those textures very easy.

But what do I do about my foreground figures? Suppose there are six of them, and I need highly detailed facial skins (face, teeth, tongue, eyes, etc) but NOTHING else? I can't reduce the resolution. In fact, I'd probably wish I had more if I were rendering for print size. But I'm stuck loading at least 3 8k maps for each figure (and the 32 bit displacement maps that must be in tif format because Poser won't recognize 32 bit jpgs are huge) -- just for the faces.

Now, you're telling me that I need to create new material zones, cut up those 8k maps, and hope I can figure out how to get them to line up properly -- and THAT'S supposed to USER FRIENDLY?

If you were joking, however, then forgive me. :D :D

Oh, I am not joking at all.
But, forgive a question before going into detail.

What size do you render?
The question is Always the same, be it for polycount or for texture size.

• Do you NEED a 70.000 polycount figure for a 4" high rendered figure?
• Do you NEED 3 or 4 diffuse maps at 4K, PLUS  3 or 4 specular maps at 4K each, PLUS 3 or 4 difsplacement maps at 4K each, for a 4" high rendered figure?

Second question;
Do you prefer to load ONE High Quality 8K Diffuse map and let Poser build the rest for you in the material room, thus maintaining the 8K quality.
Or?
Do you prefer to load 3 to 4  less quality diffuse 4k maps, PLUS some 4k less quality specular maps, PLUS some 4K less quality Bump or Displacement maps?

I"ll take the single 8K High Quality Diffuse map any time of the day, and use the material room to build the rest for me, thus maintaining the 8K high quality.

But? And this goes for ALL figures.
There is SOOOOOOOOOO much empty space on all of the existing Poser and DS figures maps, it is unbelievable.
If the UV layout had been better, you would not need all those maps at all.

Seeing what is done on some Game figures UV lay-out, I often wonder what Quality "could" be achieved if Poser or DS figures where UV mapped that way.

All I expect from a vendor:

• A high quality object file.
• A high quality rigging
• A high quality Diffuse map.
• If required, and I write again; If REQUIRED ! ! ! !  A high quality displacement map.

Please, do not bother with the rest.
Bump and Normal maps are fake 3D, so the Delete buttons is close.
Specular maps?
Hey Poser can build better ones inside the material room, so . . . . . . .That button that was close here above, just came a bit closer. LOL.

AFAIK there is no direct correlation between the size and quantity of texture maps you feed Poser and the subsequent render speed and memory usage. As Stewer posted in another thread, Poser coverts all textures into tiled mipmapped 32 bit OpenEXR's prior to rendering. In the one test I did, the difference between rendering with an additional 4k 8 bit normal map or not was: no change in render speed (averaged over 3 renders) and an increase in RAM use of 20-30Mb when the extra map was used. 

You cannot derive every map type required from a single diffuse map without compromising a lot on quality. Each map type should be considered separately for what it adds to the final render - specular, colour, diffuse, height (bump or displacement), normal, opacity - all have specific purposes, and each should be made with that purpose in mind. Converting an 8 bit JPG diffuse map (as that's what seems to be most commonly used in Poser content) into a specular or height map in the material room can be done, but the results would be better if those maps were made properly in the first place.

On UV's - the best thing from my POV would be if Poser had a means of swapping UV data without affecting anything else, eg. being able to associate UV data with a material collection. Sometimes a single map is best, other times multiple tiles is the way to go. Then you've got the tradeoff between efficient use of UV space vs seams vs texture distortion ....

There is no one right way in 3d, period. The only consideration is what works best for the intended end use - what works best in a realtime engine for games is different to what would be best in an offline render for print or film; having multiple seams to maximise UV space and reduce distortion is a bad choice if you're using normal and/or displacement maps etc - it's all about the context.

Thought I'd post this here. It's characters for a commercial. It was done by Bad Clay Studio. I had nothing to do with it but it felt relevant to the discussion at hand.

Caisson is right that there's no one right way to do it and the end use is often the deciding factor. The problem with Poser content creation is that there is no one end use for what we make. That's why I think offering both styles of UV layout is the best way to go for Poser content. At least until we get some kind of UV swapping method. Anyway, thought I'd post this to show that I'm not crazy and people do actually use this multimap method too. :D

Hello Teyon, from your example? I count 7 UV-lay-out-maps.
So?
7 diffuse maps to load at 4k each
7 displacement maps to load at 4K each  => (or 7 bump or normal maps, take your pick)
7 specular maps to load at 4K each

That are 21 maps to load at 4K each?

See where this is going?

And look at all the lost space on, and between each map?

Load a figure in Poser, load the 21 maps at 4K each, and now create an fbx and bake all that into a single 4K texture?

Even reducing the displacement and specular maps to 2K each, that still 21 maps to bake.

I'd like to see the result of that . . . . .. . . . . . . . . .

Now go for the extra detail and ADD bump and normal maps.

Then you have 35 maps to bake.... Oeps , out of memory. . . . . . . . .

There must be more economical ways with a far better and higher quality outcome.

Yup, have to say I still agree with Vilters on this one and I know that opinion isn't going to change, I would still take a single-map skin over a multiple UV one.  Vilters might be responsible for some of the scariest faces I've seen on a Roxie, but I do admire that strict obsession he has for being sensible and pushing for optimization.  The good thing about pushing for optimization is that if you can optimize well, then you're better placed to deal with anything that is thrown at you.  Speaking from a strictly technical point of view, I'm pretty sure Vilters could create an extremely efficient scratch-built figure.  If I was to actually commision one, I'd jump on Vilters to do it in a heartbeat.  He wouldn't be my first choice for the sculpting, we have creatures called Teyons for that, but this is interesting because we have the opinions of two people where their expertise is clearly in different areas of figure creation.

A figure designed and scratch-built by Vilters and sculpted by Tenon would be truly superb both in it's efficiency and it's looks.

Me, I've honestly had a very good think about it, trying my best to appreciate the benefits of multiple UV over single UV, and while I do appreciate it's benefits, I still don't think that for the average Poser user, that multiple UV would be more welcome than a well optimised single UV.  Again, I put myself in Teyon's situation and genuinely questioned what I would do if I was faced with a choice - and it is this:

Two Maps:

• Map 1 for all the skin (ensuring all continuous surfaces are on a single map and therefore problem free for any user to paint on, no matter what they use).
• Map 2 for the eyes, nails, teeth (ensuring easy customisation without having to consider the main map).

If I ever release a figure of my own, that's definitiely the way I'll be doing it.  As Teyon points out, he intends to do both, and choice is good so a thumbs-up for offering it, but I still think a two-map method with a continuous single-UV skin is the one I personally think is the best way to do it.  Done right, and optimised well, it offers problem free ease of use and still separates things where it would be wise to separate them.  But like I said, no way, not in a million years would I ever release a skin map on seperate UV's, I've always hated that so for me personally, it would feel wrong to incorporate such a design into one of my own figures, it's just something I would never do, don't care what benefits there are for multiple UV, because I still feel they are outweighed by the benefits of a single UV, and for me, there's nothing kills creativity quicker than technical issues getting in the way.

I have Blender and I have ZBrush, and neither of those things change my opinion, because the point is, I'm still having to deal with seams simply because the designer of the product decided that I'll have to.  In contrast to that, I'm the end user (a stubborn one who prefers things that "just work") - so this end user can't be bothered to deal with seams, I shouldn't have to when there was a way to do the same figure without having to concern myself with them.

@ Pumeco
@ Teyon

Sorry Pumeco, I have to disagree to agree. :-)

Teyon build the Original Roxie and Rex Object files. He is an expert at what he does and I have the highest respect for Teyon.

Their retopology has been done by Darrel (ex SM) and both object files are of exellent quality exept some minor sloppy retopology around the elbow tips.

Unfortunately the good news ends here.
All of Rex and Roxie errors are in the cr2, and somebody, somewhere at SM STANDS on the handbrake and BOLDLY refuses to repair the rigging and clean the cr2.
None are Teyons faults, he did an exellent job on both the object files.

But the native Poser figures are a completely different discussion not suited for this forum.

For object file, rigging, and UV-Mapping there is only ONE rule:

**KISS, KISS, KISS, taken to infinity and back.

Guard "end user friendly" with your life, and refuse cr2 pollution at ALL cost.

**

What ... WTF ... I never said anything was Teyon's fault :-D

I said a scratch-built figure by you, but sculpted by Teyon, would be superb.  The problems you point out are the problems you clearly know about, otherwise you wouldn't be able to talk tech about them in such detail and with such confidence, would you?

That's my point, you either genuinely know your stuff or you're talking crap just to look smart!
I'm assuming you know your stuff, so what I said was a compliment, you crazy person!

 Everything I write can be checked and cross-checked by every other individual.
My only advantage is that I have been repairing native Poser figures from Poser1 till now, so my old fingertips are pretty worn.. LOL..

Rex and Roxie are the very - very first realy GOOD object files. All issues are in the cr2's.

• Alyson's arms where too short, and she had 42 magnet errors.

• Alyson2 corrected the short arms, but at the cost of loosing ALL object file symmetry and adding +5.000 loose vertex. (all vertex at all welds are double.) And the 42 magnet errors are still there inside Alyson2.

• Miki4, the perfect figure, untill they added a bunch of completely out of symmetry magnets and dependencies renderring the figure useless to content creators.
  Well, not useless, but lots of hair was lost trying. LOL.

• Roxie, wonderfull object file. Then they added breast bones AND breast magnets. => Well, they are in a war with each other.
  And unfortunately the moving breasts are build on dependencies from those **** magnets.
  Again creating a PIA for content builders in that area.
  Had they used the breast bones to build the moving breasts, there would not have been an issue at all.

Do I need to go on? There's a lot more you know? Everybody who digs in, can see all of these and more.

This post was about textures.
Textures are a minor issue. All you need is an economical UV-map Lay-out. Teyon knows it, that's why he asks the question.
Rigging is a MAJOR issue for end user and content creator usability.
It is the rigging that needs the KISS .

Ha-ha-ha-, thanks for the complement, but I am no good at creating, but pretty confident in cleaning and repairing.
It's what I'v been doing the last 20 years or so.

it is only by being very vocal about all of the above that I can get someone to wake up and increase the usability of the "future" native Poser figures..

Best regards, Tony