The future of Poser is the Fitting Room



  • @erogenesis You point, then, appears to be lost by the title to your post and the fact that it's TL;DR.

    A character artist does not use the fitting room.

    And, I am not being ugly, just giving your some facts. What I see in your renders, as a vendor, is ill fitting clothing with lots of pokethrough, textures stretched beyond recognition, and textures that have been in use on cheap clothing for too many years. Those are not commercial quality. There's no clothing vendor, worth their value who would submit those conversions to a store. First, they wouldn't call them commercial quality, and second they would never pass QA in a reputable brokerage.



  • @Deecey said in The future of Poser is the Fitting Room:

    @erogenesis

    The fitting room does not create commercial quality content. It's OK for personal use, but for commercial quality stuff I wouldn't even think about it.

    The first thing that needs to be fixed is when you save a figure that is rigged inside the Setup or Fitting room, it saves an OBJ file that is not split at the seams, does not contain extra vertices, and has the same vertex count and vertex order as the original OBJ.

    Which is why I say: weaponize the fitting room. Make it really smart. I use it all the time and sure its not amazing, but its that bad either. Yes there are instances where you need to fix things, which is why vendors might still have to do the rigging themselves, but it need to be easier.

    Ideally I envision something like this:

    • Go to library
    • drag vendor prop onto figure
    • Poser asks: do you want to rig this to PE?
    • yes
    • done

    In the meanwhile, Poser picked up on the morph to fit to PE (because the vendor provided it), Poser kept all the morphs the vendor provided, like open shirt or something, then its gone through the cloth room behind the scenes, for example, and the cloth room has returned FBMs, JCMs and perhaps helped refine the WM to work with the figure. There are ways of doing this pretty cleverly. Color Curvature made scripts that did the pre transform thing years back and he did it using iterations. Itrations can slim the WM of a piece of clothing to a certain distance from the host figure and perhaps with certain pre-set angles refine these bends. Etc etc. Its just ideas, but I;m positive something can be done. Vendors might have to step in here and there, but it can be automated here and there.



  • @erogenesis It already IS that easy.

    I have the base rig in a piece of clothing within 10 minutes with Copy Joints From.

    That's not where the talent comes in. The talent comes from a vendor who takes the time to check and refine every joint, every morph, every bend, twist and side-side rotation. And then, when that is done, you repeat the process for every morph.

    That takes eyes and time. And you can't automate it. Because it depends on the model, the modeling, the density of the model, and on and on and on.



  • @Glitterati3D said in The future of Poser is the Fitting Room:

    The talent comes from a vendor who takes the time to check and refine every joint, every morph, every bend, twist and side-side rotation. And then, when that is done, you repeat the process for every morph.

    Yup.



  • Ok, do yo guys firstly agree that it would be awesome if this while compatibility thing could be reduced? My guess is yes, because I would find it strange if you wouldn't want this to happen.

    So, now how could we do this? My idea is stated above. I think out of the box, and fair enough, according to you it will not work at all. I think there's potential to speed things up, great potential. Cloth room could play a big role in this. Yes I remember you picked on that bad example I posted (which btw is the same as some vendors to it), thank you for reminding me.

    But ok that's my idea...

    How would you do it then? What would you recommend? Try be positive.



  • @erogenesis I would do it the same way I do it every day. By taking the time to rig my products to the best of my ability.

    I think the part you are missing is automation doesn't work because every model is different. Every modeler is different. You can't automate that for commercial quality conversions. Don't believe me? See WW and Crossdresser conversions.

    Those conversions may work for you, for your purposes. They will not ever work for a vendor and QA at a quality brokerage.



  • @erogenesis said in The future of Poser is the Fitting Room:

    Ok, do yo guys firstly agree that it would be awesome if this while compatibility thing could be reduced?

    Yes and no.

    As much as I would like to see more automated ways to convert content, I can't think of a single "automated" method that addresses every single issue that needs to be addressed in releasing commercial content. Things you have to think about ...

    .... If the figures have different grouping, they will have to be regrouped in the fitting room, and automatic grouping usually sucks.
    .... Unless the shapes of the two figures are similar, UV mapping in the clothing will become distorted.
    .... Morph transfer will get you part of the way there, but I have yet to see clean morph transfers in any automated method. The best results are usually with Dimension3D's "Morphing Clothes."
    .... As Glitterati mentioned, the process is not over after you convert the clothing. If you want to release it as a commercial product, every joint rotation for every morph has to be checked and tweaked. There is no automatic way to do this.

    As ambitious as your idea is, nothing can replace taking the time to do it manually.



  • @Deecey And I would add that the worst conversions I see/attempt is when converting something rigged with spherical rigging converted to weight map rigging. Almost without fail, the conversions are unusable.



  • Anyway, moving on. I'm just trying to come up with something useful.

    Yes, fair enough, a nice piece of clothing does require more detailed work in making sure all of it bend nicely. Not all vendors pay that much attention to this, but its great if vendors do. In the meanwhile, much can still be automated IMO:

    • Parking the prop file plus morphs on a figure and making it inherit the rig in one go (as opposed of going through the setup room or fitting room)
    • refining can happen thereafter (groupings have to go anyway) (BTW, what if it could all be done in the pose room, like the morph editor? No fitting room but a fitting editor, that would be cool!)
    • original morphs are maintained all the time
    • a feature that runs the item through the cloth room to make it endure pre-defined angles to help refine the bends even more
    • the same feature helps make nice FBMs for the item, like cloth suspended between bigger boob morphs

    and that's just for clothing. UV maps and texture transfers can also work this way. The figure creator can use a tool to make a template, and then refine that template like blacksmith3D does. And yes, yes, yes, yes yes I know there's always a refining you have to do with regards to clothing. I regret not making a big point of that in the first post, but at least we can try to speed some of it up.



  • @Glitterati3D
    Another good point.
    When you're converting non-weight mapped clothing to a weight-mapped figure, that's another thing that has to be checked and tweaked.



  • Ok now lets think positive. What ideas do you have?



  • @erogenesis said in The future of Poser is the Fitting Room:

    Not all vendors pay that much attention to this, but its great if vendors do

    Sorry, but this is just not true. Perhaps it was true with V4 whose bends were so bad there was no making them look decent, but I don't know a single vendor who doesn't take great care to ensure their clothing bends, twists and rotates beautifully.


  • Poser Ambassadors

    That's why I wrote some time ago: If you need to write a manual on how your figure works?
    You are too far away from "standard operating procedures" for most vendors to become economically profitable overnight.
    It will also make any support item created for PE harder to use on other figures because either vendors have to "add, or remove" too many things.

    That and: The way some sites are pushing vendors in "their" direction.

    On the Poser side of things :

    • Never forget that you have at least a 10 year backlog against V4.
      If more then 3 clicks are needed to convert something, and it does not work the first few times for an average home alone Poser end user? (The one that never reads a readme because he/she already had 10.000+ of those to read) She's history.

    Another discussion here was on "old tech V4".
    V4 is not old tech at all.
    Only her rigging became obsolete over time, and that's been taken care of multiple times.

    The only "old tech" that's left over is her too high polycount that does not "fit" modern SubD and MicroHD morphs any more. Customers HAVE to use her "as is" because a) no way to reduce polygons easy, and b) too dense when SubD.

    While your demo renders look very good, with a lot of "flesh detail", I fear the PE "handbrake" to modern technology is gonna be the high polycount.

    What you certainly are gonna need right from the start, and what is often forgotten, are multiple high quality texture sets.
    => The quality of a rendered picture comes from the texture set.

    A mesh is only the coathanger to hang the textures on; And any mesh can be morphed by anybody with a mouse.
    The rig has to give realistic bending, but the final quality of the "rendered output"? => Comes from the textures.

    +/-70.000 polygons is a "time proven" concept.
    But it's the only thing that's 10 years old.
    All the rest moved forward.

    END USER FRIENDLY should be hammered in STONE.

    What's the difference between old and obsolete content and "modern" visually acceptable content?
    (Remember, I am in this "game" from Poser 1.)
    Uv_Unwrapping and texture quality.

    Even from "old and obsolete content" the mesh can always be recovered, properly grouped and welded, Unwrapped properly, textured properly and chased through the fitting room.

    The Fitting room basics.
    Only requires some basic knowledge to work properly.

    • A properly grouped and welded obj file
    • The "clothing to fit" HAS to have +/- the same polygon density as the donor figure to get an optimum result.
    • Fine tuning tools are provided inside the Fitting room.

    And if all else fails? Do a "pre-fitting" in an outside 3D app like Blender or equivalent, and only use the Fitting room to transfer the rig. (And replace the obj file with the good one after the save to Library. Don't be left with unwelded obj files).

    That's all end users need to get a good fitting room result.



  • So yeah, those are my ideas.

    Thanks for your input Vilters.

    I think I'm gonna go back to making a Poser figure now :P or render something fun for my fans...



  • @erogenesis said in The future of Poser is the Fitting Room:

    Anyway, moving on. I'm just trying to come up with something useful.

    Yes, fair enough, a nice piece of clothing does require more detailed work in making sure all of it bend nicely. Not all vendors pay that much attention to this, but its great if vendors do. In the meanwhile, much can still be automated IMO:

    • Parking the prop file plus morphs on a figure and making it inherit the rig in one go (as opposed of going through the setup room or fitting room)
    • refining can happen thereafter (groupings have to go anyway) (BTW, what if it could all be done in the pose room, like the morph editor? No fitting room but a fitting editor, that would be cool!)
    • original morphs are maintained all the time
    • a feature that runs the item through the cloth room to make it endure pre-defined angles to help refine the bends even more
    • the same feature helps make nice FBMs for the item, like cloth suspended between bigger boob morphs

    and that's just for clothing. UV maps and texture transfers can also work this way. The figure creator can use a tool to make a template, and then refine that template like blacksmith3D does. And yes, yes, yes, yes yes I know there's always a refining you have to do with regards to clothing. I regret not making a big point of that in the first post, but at least we can try to speed some of it up.

    Here's an extreme example. I've been doing a lot of work on Pauline, and have her morphed into several different ages. The first one (on the left) is the default Pauline. The others are morphs that I've created.

    Two issues here. First, the more the figure deviates from the original size and shape, the less accurate the joint centers are when the "Match Centers to Morph" feature is used. So the joint centers have to be tweaked.

    Secondly, look at the clothing. These are untouched "Copy Morphs From" morphs. Again, the more the figure deviates from the original size and shape, the more they need to be fixed and smoothed out.

    These are things that will be difficult to achieve automatically. I'm not a programmer, I don't know how easy it would be. All I know is, these clothing morphs were created in Poser and are far from "prime time" quality.

    0_1502547993811_paulines.png



  • And another extreme example. This one to show you how much depends on the model itself. This "mannequin" works great for transferring rigging from Dawn. But it is intentionally low poly, and therefore does not accept morphs well at all. It was not designed to accept morphs the way I am using it here. And this is just ONE morph (bulk) of many I wanted to transfer.

    0_1502548500743_DawnResourceBulk.jpg



  • While I can agree that the fitting room, and perhaps any automated solution, will not produce commercial quality results, I have to agree with Eros main point that striving towards greater universality of assets could be the backbone of poser's future. It's always going to need some help and skill but to me this seems like an area that will pay off big time in investing dev resources.
    The point I take away from ero is that the smaller you make the mountain of a chore of transferring assets from one figure to another, the less the content creators have to waste their energy on it over making awesome stuff



  • @KittyBrown That may be true in some instances, but it will be so rarely.

    Frankly, if I want to convert something from say V4 to Dawn, it's much faster to take my ungrouped, original mesh into the modeler, load the Dawn mannequin and begin the conversion process, then group and rig to the figure.

    The automated conversion just give me more work to try to correct the poor conversion.



  • @Glitterati3D said in The future of Poser is the Fitting Room:

    @KittyBrown That may be true in some instances, but it will be so rarely.

    Frankly, if I want to convert something from say V4 to Dawn, it's much faster to take my ungrouped, original mesh into the modeler, load the Dawn mannequin and begin the conversion process, then group and rig to the figure.

    The automated conversion just give me more work to try to correct the poor conversion.

    That's been my experience as well. It's usually a lot less work to refit and regroup something manually than it is to fix an automatic conversion.



  • @erogenesis said in The future of Poser is the Fitting Room:

    Ok now lets think positive. What ideas do you have?

    In order to determine the solutions, you have to know what the problems are. What I've offered here is constructive criticism and the problems that I've seen while using the automated conversion tools.