Quote - An example for extreme skinnyness.

This is the kind of body detail I expect a modern Poser figure to be capeable of without the need for displacement maps.

I ran out of time to do an emaciated and obese look. I'm hoping that's something I can do in the near future for both of them.

Attached Link: http://vimeo.com/52668394

My point was not just having a discussion if a figure should be hiRes or lowRes. It is also about, that I just would like to have the choice. To have the tool to increase a figure's mesh resolution. C4D has that onboard tool allready for years. It's called Hypernurbs there. One click and that "meshcage" becomes a real mesh. My hope is, we can get to work something like that in Poser too.

Quote: "... (because people are never happy)... " Happyness use to make me lazy ;-)

Yeah that would be neat. I don't know of any plugins or scripts to do that outside of what Snarly wrote awhile back before we had subd in Poser. I don't think it retained any of the group or UV information though. I can't recall to be honest. HEY! You may want to look for that.

Actually....that may have been colorcurvature who wrote that script. It's bee so long.

I've totally hijacked this thread. I'm a heel. Sorry timberwolf.

Sorry my mistake. I wasn't referring to the Daz Genesis 2 when I mentioned the G2. But SimonG2 and SydneyG2 :)

I'm sure when the modeller looks at the model he thinks it looks good in the T pose. But then the rigger takes over and things starts to change. The usual problems are shoulders, elbows and buttocks when the figure starts to bend. The rigger only have the polys to work with that the modeller suply in the figure. So just blaime it on the rigger I think is wrong. The modeller should take this into account when making the model. Only solution now is to subD the mesh. V4 has just enough to not notice this.

Here you see a V4 in the background then SimonG2 and Alison 2 in front. The last two models are not very "rendering friendly" as the polys are too low in the shoulder area in this pose. And telling the artist that he just have to render the model in the T pose I think we all can agree is not the way to go :)

You know Timberwolf - unrelated to anything said in this thread - I absolutely loved that character from the comics. He was always one of my faves. Just had to say that.

Joequick, the point is not whether displacent maps can create more detail or not. They can.

The point is: Why ?

Why should we start using subdivided low res figures that have to rely on elaborate displacement maps that only very few professionals can create instead of using high polygon figures that are much more accessible to the average hobbyist user ?

What is the big advantage of low res figures ?

Render speed ?

Bending ?

What is it a low res figure can do better ?  In Poser ?

Teyon, thanks again for all the info you have posted here.

If we you DO want to add the fine detail (rather than the general definition in the base mesh) to Poser figures and render them in Poser, what is your recommended way of doing that?

Paul

Quote - Joequick, the point is not whether displacent maps can create more detail or not. They can.

The point is: Why ?

Why should we start using subdivided low res figures that have to rely on elaborate displacement maps that only very few professionals can create instead of using high polygon figures that are much more accessible to the average hobbyist user ?

What is the big advantage of low res figures ?

Render speed ?

Bending ?

What is it a low res figure can do better ?  In Poser ?

 

Do you not think that even the unmapped Freak 5 has good muscle definition? This gets back to what Teyon was saying, the right topology, with the right loops, can establish detail just as well as your 3rd gen daz meshes.

You're evading, joequick.

What are the advantages of SubD and displacement maps over polygons IN POSER ? For the end user ?

And no, I don't think the Freak 5 has "good" muscle definition.

None of the Genesis characters looks remotely as good as a high res figure without a displacement map.

The mesh tries to do the best with what few polygons it has, but to me it is very underwhelming.

I'm pretty sure with 70.000 to 100.000 polygons Genesis could be photorealistic.

But now it looks like a game avatar.

The very best 3D figure I have EVER seen had 6.300 polygons.

Now try to find that link again... Grrr....

Quote - I feel like when the polycount lack of definition conversation comes up, someone always mentions Freak 5, but then I don't know that anyone ever posts an image of MallenLane's big lug.  So here he is, mapless on the left, and then with normal, bump and displacement on the right.  To my eye he's got pretty good muscle definition.

Oh so in What's Included & Features

where it said Freak 5 Displacement Off/On

Tells you theres a zBrush Displacement Map.

Thanks

Quote - This one is not bad either at 8.300 polygons

her breast set awful low to be that small :unsure:

Quote - Joequick, the point is not whether displacent maps can create more detail or not. They can.

The point is: Why ?

Why should we start using subdivided low res figures that have to rely on elaborate displacement maps that only very few professionals can create instead of using high polygon figures that are much more accessible to the average hobbyist user ?

What is the big advantage of low res figures ?

Render speed ?

Bending ?

What is it a low res figure can do better ?  In Poser ?

 

Every thing.

Quote - Teyon, thanks again for all the info you have posted here.

If we you DO want to add the fine detail (rather than the general definition in the base mesh) to Poser figures and render them in Poser, what is your recommended way of doing that?

Paul

Not sure which you're asking. If you're asking do I prefer normal/bump maps or displacement maps this would be my answer:

I prefer both. Displacment to help capture the secondary shapes and a single bump or normal map to capture the tertiary detail.

And that is for low or high poly.  The model brings out the primary forms. The displacement can be used to achieve the secondary forms if the model doesn't allow for them and then the normal or bump map can be used to achieve tertiary detail like pores that would be too small to be worth doing on a displacement map.

If you're asking for material setups. I am no material room guru. So I wouldn't feel comfortable suggesting a setup here. 

Quote - It was Eva

Yup. She's quite nice.

Quote - You're evading, joequick.

Not intentionally.  In my mind, your argument goes "the third generation daz figures have the optimal topology among what's available to us".  You post images to support that argument. To my eye, those images don't prove your point.  So I respond. 

Generally, in terms of the sharpness of detail achieved, if it's that sharpness that matters to you, I see no real difference between 3rd gen figures and even a mapless Freak 5.  Mallenlane's might be a bit ambiguous in the serratus anterior compared to what I assume was Creek's, but I feel that that definition could be gained with small tweaks to the topology, and not by driving the polycount towards 100,000.  You show me an M2 muscleman, I counter that Mallen is sculpting more believable muscle anatomy with fewer polygons than what we saw dominating Millennium family figures generations 1-3.  You disagree.

You ask "why not multiply the polycount by three, our computers can handle it". I agree that more detail could be sculpted into those meshes. There's a lot more subtle detail sculpted into those faces than what I could do with the 4th and 5th gen figures.  But  the way that that detail distorted when combined with other morphs or when posed.  What seemed superior in a t-pose quickly deteriorated when other variables were added. I would think that it would only be worse with your 100,000 poly figure.  And with that hypothetical 100,000 poly figure I wonder if we would lose all sense of topology dictating form, which in my mind is useful, as it insures that my breast morph will jive with your breast morph, and no one will wind up with quadroboob or eighteen and half abs simply because one morph artist sculpted those details in a location that contradicts rather than compliments the work of another.  In my experience, the lower res figure with form defined by the topology, and detail defined by the mapping, just behaves better when posed and when combined with other morphs. Beyond that, I defer to the experts.  If this is what Mallenlane and Teyon are both advocating, if this is what we're seeing on cghub et all, then I'm fine with accepting that these professionals are wiser than me and adapting my workflow towards their idea of best practice.  An answer that I can't imagine is very satisfactory for you, but I feel no need to prove that I know more than the people who do this for a living. 

Sorry, but if you can't actually "see" that the 3rd gen morphs are sharper and more well defined than the Genesis morphs there is not much to discuss, is there ?

It's simple logic that a mesh with twice as many vertices allows twice as many different morphs than a fgure with only half as many vertices.

With a high res mesh, I can follow edgeflow, but I also can morph against it if I want.

That shapes can be much more well defined if I can create sharp creases because I can manipulate directly neighbouring edgeloops and not just any other.

And of course displacement maps interfere with joints because what looks good on a bend joint won't look good on a straight one because the displacement map is oblivious to the changes in shape of the underlying mesh.

No problem for a high res mesh that uses JCMs to give bend and straight joints completely different shapes.

Poser figures also have to work in scenes made of dozends of high res props, using clothes that multiply their original poygon count, so any "memory saving" advantage is theoretical only.

Here's what I think this is all about:

LowRes + SubD + Displacement maps are more "efficient" in a professional production pipeline where time is money.

For professional CGI artists who can easily paint a new displacement map as fast as the hobbyist Poser user spins a few dials, this is naturally "the way to go."

DAZ (And to a certain degree SM, too), both have this "dream" of entering the "Pro" market, that's why they all of a sudden push "Pro" techniques.

DAZ also tries for some time (Starting with V4 and her embedded magnets) to make cloth creation as efficient as possible.

Easier cloth creation = More clothes made = More $$$ earned.

The problem of course is:

1. Poser and Studio will never be a tool used by professionals (Except for some cheap pre-viz)

2. Figure and cloth realism suffers because there are less details possible and less elaborate rigging

3. Figure modification becomes less accessible to the amateur as now you need zBrush and a 3D painting app to create morphs and textures.

But to hell with amateurs, they are supposed to buy the stuff the "Pro" vendors create, not make their own, right ?

And to hell with realism and details, as long as the figures are "good enough to sell", right ?

Sorry, but as a Poser artist I'm opposed to every attempt to "dumb down" figures just to earn more $$$.

For me, it's quality over quantity.

I want figures that are photorealistic out of the box. Or at least as photorealistic as current tech allows.

I also rather pay double for clothes that took twice as long to make but bend correctly and look correctly.

Sorry, the 3rd Gen figures were made in a time when "money wasn't everything". When it was just about making the best figures possible.

I never said they were perfect. In fact, out of the box most are quite horrible.

But they are easy to work with and you can pretty much do with them whatever you want.

So unless someone comes up with a better mesh, I'll continue to use them.

The low res figures are an aberration. For Poser.

Even Daz has already noticed that.

(After finding out that Michael 5's shape is just a 10 minute dial spin of David 3, I've gotten quite cynical about DAZ' marketing. But not that cynical to think that Genesis 2 is only about the money, and not about improving realism, too)

So Genesis 2 has more polygons than Genesis 1.

And Dawn will have twice as much from the get go.

Given that it's now so easy to transfer rigging and morphs between different meshes, what is the problem having a dedicated High Res mesh and a dedicated Low Res mesh of the same figure ?

We had it in the past and absolutely noone complained.

Are profit margins so small that it's impossible now to let the end user decide ?

If you think that low res meshes are the bees knees, please and by all means, use them. Years ago I even posted a tutorial to show how to make the various LOD SubD meshes of V4 useable in Poser.

But do not try to push them as the only valid standard, nor claim they are superior without being able to give hard evidence.

I've seen pictures of nice figures and nice morphs. You all are very talented and please have fun with whatever you do.

But I've seen nothing that couldn't have been done better with high res figures.

And again, I'm ONLY speaking about photorealistic humans. Not fantasy, anime, stylized reality, whatever.

I'm not interrested in those typical ZBrush "Look how much detail I can squeeze into that impractical armor" demo reel sculpts.

My only point here is:

What is the best technical way to portray a realistic looking human in Poser.

Not in a game, not in a movie, not even in a Poser animation.

In a typical Poser still render.

With the change from high res figures to SubD low res figures, we made the change from:

"Best possible product" to "Good enough to still make a profit"

Correction:

"It's simple logic that a mesh with twice as many vertices allows twice as many different morphs than a fgure with only half as many vertices."

Never was good at statistics, but if I remember correctly I think it should be

2² = 2 x 2 = 4 times as many different morphs.

Or shouldn't it be even : Number of polygons x number of polygons = number of morphs ?

Then a high res morph would allow A GAZILLION more morphs.

:-)

Does someone know the correct formula for the theoretical number of morphs between let's say a 20.000 poly mesh and a 40.000 poly mesh ?

One single 4096x4096 displacement map gives you 16.777.216 micro-polygons to work with for each and every material zone.

Combine a texture, displacement, bump, and a normal map, and study the possibilities.

And this is NOT the end. Poser supports 8192x8192 texture maps.

Imagine a 8192x8192 texture, displacement, bump, AND normal map on each and every material zone!!!!!!!!!!!!!!!!!

That is 67.108.864 x 4 micropolygons for each material zone.....

A total of 268.435.456 of pixels/micro-polygons for each and every material zone.

AND !!!!!!!!!!! You do not loose symmetry so fast.

Quote - It was Eva

they should hire this guy to make the next stock Poser figures. The only downside: The next PoserPro copy might cost 1500Dollars, haha ;-)

I don't use App's that don't have SubD or GoZ like C4D.
I don't use High Polycount meshes.
If a App wants to survive it has to keep up.

Tony, the moment I can:

1. Manipulate displacement maps directly in Poser as easily as I can mesh...

2. Create real undercuts with them...

3. Render them as quickly as I can render high res geometry...

4. See them in the OpenGL preview (VERY important for me)...

5. Easily animate them for joint deformation compensation...

...I happily scrap high res meshes and use low res meshes instead.

We can use displacement maps since Poser 5.

We have mesh smoothing to make low res meshes smooth since Poser 5.

Still in all these years both never caught on.

Ask yourself why ?

@Timberwolf: Eva is a nice figure, but I'm sure she wouldn't look and bend nearly half as good in Poser as she looks and bends in Max/Maya.

Nor would she be able to be morphed into that gazillion of different shapes Poser users expect a Poser figure to be capeable of.

Simply apples and oranges.

As for her looks, I don't think making a realistic sculpt takes "unaffordable" talent.

MIKI-I after all was a perfect photorealistic copy of Aneta Keyes (Minus the head), and still SM could afford her. (See pic)

Miki-I flat butt and awful bending, the rest was really great. She is on my top wishlist for someone converting her to a weight-mapped figure with smooth bendings. (I don't mean that single click converting) Aneta and Miki : Like both X-D

Quote - You're evading, joequick.

What are the advantages of SubD and displacement maps over polygons IN POSER ? For the end user ?

Unfortunately i cannot answer the "IN POSER" part, because i do not have subdivision in poser, but the disadvantage of having many polygons (for me as an end user) is definitely render time. Not render time in firefly, but render time in the preview. I do not really care if firefly renders 4 or 5 minutes, it is long enough that i have to get up and take a break anyway. But most of my time with poser i do not spent with rendering but with manipulation (posing, loading objects, materials etc) of meshes. And at least on my (not so new) computer the difference between 100K and 20K polygons is quite noticable especially if there are a lot of morphs involved. I did not measure it exactly, but it feels like manipulating a 100K mesh takes like at least 10ms more, giving me only 20 frames per second instead of 30, when bending an arm of a figure for example. Not the end of the world, but noticable. Doing the same with 10 figures instead of one gives me (felt) 5 frames per second which is unacceptable; as a consequence i never use 1Mio polygons in Poser but would rather spend some time, effort and if possible money to reduce that number. As long as Poser cannot deliver at least 50 fps with 1mio polygons, i see polygon reduction as a valid strategy to make it more usable. If i had a fast enough computer to do that (one day i will), i would like to render e.g. a vue scene in the background, while working with poser rather than to fill up the cpu with more polygons. And only if this works fluidly (Poser giving me, say 100fps, while running a render in the background) i would agree to having more polygons being a good thing without reservations. Until that happens the number of polygons will always be a compromise between quality and usability (my personal opinion, of course, other people probably have other priorities).

Sorry millighost, while I hear what you say (And are very picky about a good and fast OpenGL preview myself), I can't see this as a real problem anymore.

Using PP-2014 on an i5 Acer laptop with a NVIDIA GT-540m Geforce I'm pretty sure I qualify for "low end" user.

And there is no noticeable OpenGL speed reduction whether I use a 150.000 polygon figure or a 20.000 poly one.

Actually, using SubD in preview definitely slows Poser's OpenGL down. It's not drastic, but it is noticeable.

So, if I want to have a "nice" preview with smooth figures, the high res figure is actually faster than the subdivided low res figure.

Again, what you say is perfectly valid for older versions of Poser and (very) low end machines.

But definitely not what I experience in PP-2014 on a three year old cheap laptop.

Older versions of Poser had problems with high polycounts, but not anymore.

The bottleneck now are layered transparancy and high res textures, so fighting for a few polygons here or there really makes no sense anymore.

Besides, as I said above, there's still the option of having low res and igh res versions of the same figure.

Quote - Correction:

"It's simple logic that a mesh with twice as many vertices allows twice as many different morphs than a fgure with only half as many vertices."

Never was good at statistics, but if I remember correctly I think it should be

2² = 2 x 2 = 4 times as many different morphs.

Or shouldn't it be even : Number of polygons x number of polygons = number of morphs ?

Then a high res morph would allow A GAZILLION more morphs.

:-)

Does someone know the correct formula for the theoretical number of morphs between let's say a 20.000 poly mesh and a 40.000 poly mesh ?

You might be surprised to learn that theoretically there are exactly the same number of different morphs for 20K and 40K meshes. Basically because there are the same number of vectors of numbers as there are numbers. But this is really only theoretically, in a slightly more practical world the representation of the morphs (ie how many digits are in a pz2 file) is more relevant than everything else. Then the number of morphs would square, because you could combine every possible morph in the left 20k half of the 40k polys, with every morph in the right half, which is a lot more. But the 40k morphs would be twice as big as the 20k morphs, because on the average they cover twice as many vertices, so on any given computer with fixed amount of memory you could use only half the number of morphs, so the usable number is less for 40k than for 20k; for instance, if you have only 40KB of RAM you can decide if you fill it with exactly one 40K morph, or two different 20K morphs. Of course this only is theoretical practice. In practical practice you simply have the same number of morphs for 20K and 40K (but the 40K versions are still bigger).

Umm, what ?

A cube has eight vertices.

If I subdivide it, it has 28, but only the original 8 can be manipulated in Poser, as the 20 new ones are just "virtual".

If I "freeze" the subdivision I get a high res cube with 28 "real" vertices.

Morphing the SubD cube limits me to moving those 8 vertices around, as the others will be re-generated constantly.

With the high res cube I can move all 28 vertices completely independently from each other.

So of course I can have much more morphs with the high res cube.

Quote - Umm, what ?

A cube has eight vertices.

If I subdivide it, it has 28, but only the original 8 can be manipulated in Poser, as the 20 new ones are just "virtual".

If I "freeze" the subdivision I get a high res cube with 28 "real" vertices.

Morphing the SubD cube limits me to moving those 8 vertices around, as the others will be re-generated constantly.

The simplest object contains a single vertex. How many morphs could you make for this single vertex? There are only 3 coordinates you could modify, x, y, and z. A single morph can modify these x,y, and z in some arbitrary fashion. One of the possible morphs, for example can be x=1, y=2.78, and z=0. Every different combination of x, y, and z gives a different morph. How many different combinations are there? Poser uses a precision of approximately 10 digits. Because there are 3 values, there are in total 30 digits which makes a 10^(310) = 10^30 possibilities for a single morph.
In Poser each number uses approximately 4 bytes, so one morph for a single vertex uses 12 bytes. To store all of the single vertex morphs you would need 12 * 10^30 bytes. Even if you used new 500GB DIMMs (weight ca. 20g), you would still need (12 * 1030) / (500 * 230) = 210^19 of them or in other words 400 billion gigatons of RAM, which exceeds the specification of most computers, especially 3 year old laptops, and that is only for a single vertex! This number will not get smaller if you used more vertices, so the number of possible morphs is mainly determined by the size of a single morph, not by the number of vertices. So in practice: the smaller the morph, the more of them you can have. Of course, this does not take into account the usefulness of morphs, only their number.

Quote - With the high res cube I can move all 28 vertices completely independently from each other.

So of course I can have much more morphs with the high res cube.

"Of course, this does not take into account the usefulness of morphs, only their number."

I don't doubt your theoretical numbers, but that's not how Poser works.

A cr2 only stores morphs at full strenght which are actually useful. The theoretical "trash-morphs" are all ignored.

A high res mesh can have more "useful" morphs than a low res mesh because it allows greater detail and you are not limited to predetermined edgeloops.

And once you go into custom morph creation with magnets or morphbrush, you have a lot more possible useful combinations with almost zero overhead, because only the finished morph is actually stored in Poser, not all possible combinations.

And of course, yes, a high res mesh uses more space for each morph.

But that's what injection morphs are there for.

:-)

Quote - "Of course, this does not take into account the usefulness of morphs, only their number."

I don't doubt your theoretical numbers, but that's not how Poser works.

Hm, i think that is exactly how poser works.

Quote - A cr2 only stores morphs at full strenght which are actually useful. The theoretical "trash-morphs" are all ignored.

But if a morph is useful or not, is dependent of the particular application. If i do a portrait, all the toe-wiggle morphs are not useful. How should poser know that the toes are not useful?

Quote - A high res mesh can have more "useful" morphs than a low res mesh because it allows greater detail and you are not limited to predetermined edgeloops.

Yes, but "not being limited" means more flexibility. And more flexibility always comes at a price. With software this price is often paid in memory, CPU power or usability.

Quote - And once you go into custom morph creation with magnets or morphbrush, you have a lot more possible useful combinations with almost zero overhead, because only the finished morph is actually stored in Poser, not all possible combinations.

I agree that single morph has almost zero overhead (like a single vertex more or less has zero overhead). The overhead comes into existence when you put many many of them into a figure. For example, on my machine i notice a slowdown when i am loading all of the V4++ morphs into V4 at once. Your laptop might be a bit better, so you might notice it a bit later, depending on the circumstances, but it surely would not go unnoticed. I even tried to subdivide V4 up to a million polygons once, because i thought it would be a good idea so i would not need displacement maps any more, but it was unusable (even without morphs).

Quote - And of course, yes, a high res mesh uses more space for each morph.

But that's what injection morphs are there for.

:-)

Joe your high polycount meshes are not going anywhere you can still use them.
For as long as you want.just don't force us to use them and we woun't force you to use are meshes.

DAZ started the high polycount mesh.
DAZ started the unimesh.

Now DAZ & Poser are going to 100% quad ,decent polycount SubD ,individual meshes.
why ?
Cause that's what HighEnd App's demand.

View all the high end gallerys and fourms.
We don't use tri'ed highpolycount unimeshes.

Poser can be a Hobbyist App or Poser can be a Pro App.
If they want to be a Pro App then they best have zBrush Vector Maps and Rigs as good as the High End App's.

It's all up to Poser ,Cause we don't need Poser.
Poser has to conform to us.

But to quote Ryan Kingslien "the sculptor with the most polys wins".

Been watching this thread with interest and there are good points from both perspectives.

I'm with JoePublic in this respect though - there is a place for high-res figures where secondary detail can be carried in the mesh. Maps are good for tertiary detail - veins, wrinkles, skin pores. 

Here's the way I'm thinking - every Poser figure is multi-mapped. Rex and Roxie have a head and body map, V4 has body, limbs and head. A good displacement map should be 16 bit (over 65,000 'steps' of gray as compared to 256 for an 8 bit grayscale map) - and the file size increases accordingly. So to carry crisp secondary detail on a figure like Roxie with 26,000 quads will take two 16 bit grayscale displacement maps at something like 2000 pixels square each. That carries a memory overhead.

Worth considering as well that, to paraphrase Scott Spencer, using a displacement map on a low-res mesh is like trying to project detail through a blanket. If you use render subD to get a finer mesh then it's like projecting through silk - but again, that carries a memory hit as each subD level quadruples the mesh plus it doubles again when the render engine triangulates it.

So when a high level of detail is required from a figure, it would seem a lot simpler to just start with a base resolution that can support that goal. One size fits all is an approach that just doesn't work in practice, and based on what I have seen and done in Poser and Zbrush, my opinion is that ultimately mesh detail beats map detail.

This is a great conversation, however everyone is talking about greating displacement and normal maps like it's part of their pipeline, and I suspect they have never actually done it.  As someone who has been actually rendering Poser figures with displacement and normal maps, and subd'ing, in my experience...

Creating displacement and normal maps for Poser figures in ZBrush is incredibly difficult - because the material zone of the figures skin are split.  You need to remap the figure to get anything of a high quality.  How many full body displacement mapped figures are on the market?  0?  Why is that?  Because even if you get around the multiple UV zones, the Poser joint deformation system will kill your displacement map.  The amazing graphics coming out of games is happening because they have one material map for the figure - not 5 or 6.  [Caveat....maybe the new Poser Pro 2014 skinning system kixed the joint deformation problems with displacement maps - I don't know].

I've been trying a new workflow in PP2014.....GoZ a posed Roxie figure, subd in ZB 3 times, add all the detail, generate a normal map, GoZ the level 0 subd back to the Poser, and apply the normal map to the materials.  But of course this utterly fails, because the eye UV's overlay the skin UV's, so ZB adds some normal mapping info for the eyes on the skin map.  The multiple material zones is killing realism renders in Poser :-(  So I have re-UV'ed Roxie, created one material zone for "skin", and the above works.  But it requires pythons skills and a lot of effort, and I doubt anyone else could be bothered.  And the end result?  You can still see polygon edges on the figures shoulder. 

12 months back there was a post here about what features the next gen figures should have, and everyone said - non-overlaying UV's and one material zone for skin.

One skin material.  Non-overlapping UVs.  More polys.  That's what we need.

http://www.renderosity.com/mod/gallery/index.php?image_id=2451124 - subd'd TWICE and I stll don't think it has enough polys for the detail I'm after.

Paul

Quote - This is a great conversation, however everyone is talking about greating displacement and normal maps like it's part of their pipeline, and I suspect they have never actually done it.  As someone who has been actually rendering Poser figures with displacement and normal maps, and subd'ing, in my experience...

Creating displacement and normal maps for Poser figures in ZBrush is incredibly difficult - because the material zone of the figures skin are split.  You need to remap the figure to get anything of a high quality.  How many full body displacement mapped figures are on the market?  0?  Why is that?  Because even if you get around the multiple UV zones, the Poser joint deformation system will kill your displacement map.  The amazing graphics coming out of games is happening because they have one material map for the figure - not 5 or 6.  [Caveat....maybe the new Poser Pro 2014 skinning system kixed the joint deformation problems with displacement maps - I don't know].

I've been trying a new workflow in PP2014.....GoZ a posed Roxie figure, subd in ZB 3 times, add all the detail, generate a normal map, GoZ the level 0 subd back to the Poser, and apply the normal map to the materials.  But of course this utterly fails, because the eye UV's overlay the skin UV's, so ZB adds some normal mapping info for the eyes on the skin map.  The multiple material zones is killing realism renders in Poser :-(  So I have re-UV'ed Roxie, created one material zone for "skin", and the above works.  But it requires pythons skills and a lot of effort, and I doubt anyone else could be bothered.  And the end result?  You can still see polygon edges on the figures shoulder. 

12 months back there was a post here about what features the next gen figures should have, and everyone said - non-overlaying UV's and one material zone for skin.

One skin material.  Non-overlapping UVs.  More polys.  That's what we need.

http://www.renderosity.com/mod/gallery/index.php?image_id=2451124 - subd'd TWICE and I stll don't think it has enough polys for the detail I'm after.

Paul

If I'm putting displacement maps on my own meshes that I can make non-overlaying UV's for and one material zone for skin for.

Then what's your thoughts about having a displacement map in Poser Pro 14 ?

I personally would like to see a figure covered by a single map - given that Poser can handle up to 8k maps; given that for a lot of renders I do (full body shots at max. 3000 pixels) multiple maps are a waste of memory as that level of detail isn't seen; and given that I am in a minority who use Zbrush with Poser (and ZB is easier to use with a single map). But, like I said I know I'm in a minority!

@ Paul - have you tried using polygroups? In that tab, UV Groups will create polygroups based on the UV tile, so Roxie gets separate groups covering body (inc. nails), head (inc. tongue and inner mouth), eyes and teeth/gums. If you use Auto Groups with UV it'll create groups for each UV shell. So to make a skin map I'd use UV Groups first, then control-shift click on the head to hide it and show the eyes and teeth. Mask the eyes and teeth then Visibility - ShowPt, then invert the mask, then HidePt, then Del Hidden to get rid them. Then I'd use Auto Groups with UV to split the mesh further and hide and delete the inner mouth, tongue and nails. Active Point count for Roxie should now be 19,764. Hit UV Groups again and you have two polygroups, head and body. Subdivide and sculpt - based on 1 poly = 1 pixel, for a 2k map you need min. 4 million polys for each map (4k map = 16 million polys). I can go to 6 subD levels which gives 20 million polys with Roxie, but only 4 in the head and 16 in the body, so if I wanted to paint a detailed skin and make 4k maps I'd have to use HD geo. When you're done then control-click to hide one group, then generate the maps you need from the visible mesh, then do the reverse to create the other map.

If you already use that then ignore me, but it may be useful info for someone!

Paul - have you tried using polygroups? In that tab, UV Groups will create polygroups based on the UV tile, so Roxie gets separate groups covering body (inc. nails), head (inc. tongue and inner mouth), eyes and teeth/gums. If you use Auto Groups with UV it'll create groups for each UV shell. So to make a skin map I'd use UV Groups first, ....

Thanks Caisson - that's a good workflow.  I'll give it a try.

Paul

I get a lot of my info from books, ended up buying most of those for Zbrush. Scott Spencer's Character Creation is particularly good.

Great render you posted BTW. Was the hair Poser's dynamic hair or a Fibremesh import?

The bodyhair was fibremesh import.  I'm still working on getting Poser dynamic hair looking 100% in Octane, so using ZB Fibremesh for the moment.

BTW Caisson - have you worked out how to LOAD a normal map into ZB and apply it to a subd mesh (from Poser GoZ)?  So I have GoZ from Poser to ZB, SUB'd x 4, tweaked, generated norrmal map, GoZ subd0 back to Poser.  If I shutdown ZB and restart - how do I get back to the Subd x 4 mesh (which is a combination of the Poser base mesh and the normal map)?

Paul

Hmmm, don't know how to do that, not tried it. I think it should be possible to do it with a displacement map, not so sure about a normal. I'll try to have a look tomorrow, gotta sleep now :)

ZB does auto-save, so you might find the last few projects under Lightbox. The best backup though would be to save as a ZTool or better still as a Project after generating the normal.

You can load a displacement map in ZBrush and apply it to a mesh.  While you can load a Normal map into the program (it's just a texture afterall), you can not apply it so that you gain the detail from it onto your sculpt.