Copy Morphs with Python?

[Lissa_xyz]

I don't have Poser installed anymore, and the latest version I did have was PP2014.
This is where google leads me, and the latest they have is 2012: Poser & Poser Pro User Manuals and clicking for 2012's Python reference just takes me to their support home page.

If you can get me a copy of it, I can at least take a look at the functions for you. Maybe it's possible to rip bits and pieces of existing functions apart and get something going that way.

 

[Lissa_xyz]

Come to think of it, I did this when I ported DVL's Doe for Dawn from DS to Poser, but it's PMD based.

Here's the top of my INJ to get Doe into Dawn's skinny jeans:

injectPMDFileMorphs :Runtime:libraries:pose:Doe for Dawn:DoeSkinnyJeans.pmd
createFullBodyMorph Doe
createFullBodyMorph DoeBodyOnly
createFullBodyMorph DoeGenitalCrease
attachFBMdial DoeGenitalCrease DoeGenitalCreaseFix

Here's what the skinny jeans PMD looks like:

The whole package is still on my sharecg if you want to crack it open yourself:
Doe for Dawn - Poser Edition - Poser - ShareCG

 

[Ken1171]

I don't have Poser installed anymore, and the latest version I did have was PP2014.
This is where google leads me, and the latest they have is 2012: Poser & Poser Pro User Manuals and clicking for 2012's Python reference just takes me to their support home page.
If you can get me a copy of it, I can at least take a look at the functions for you. Maybe it's possible to rip bits and pieces of existing functions apart and get something going that way.

The thing is, Poser Python doesn't let us even select a particular morph, and even less do something with it. What this would need is the ability to project a morph from figure to clothing, but those things were not exposed to Poser Python. This might be the reason why 3rd party tools that handle morph projections are all external tools that run outside Poser. An example of that is D3D's Morphing Clothes, that do exactly what I want, but it's an external tool.

 

[Ken1171]

Come to think of it, I did this when I ported DVL's Doe for Dawn from DS to Poser, but it's PMD based.

Yes, but this is copying morphs from Dawn to Dawn. What I need is from Dawn to Clothing, which means it's a morph projection, and not a copy.

 

[Hornet3d]

Oh I see! Well, there is a reason there is a dwindling number of Poser vendors. Personally, I blame SMS for their inability to position Poser in the market. I believe SMS has difficulties identifying who the Poser users and content creators are, and that became more evident with their pricing on the Pro version at $600, which by the time was close to the entry levels of the likes of Cinema4D. And then we had SMS advertising Poser as a "complete" professional character animation tool, when it can't even get IK to work properly, resulting into feet sliding everywhere. They advertised about Rooster Teeth adopting Poser in a commercial animated series, which [I think] was a bad example, because they quickly moved away from Poser after season 1, adopting Maya instead for the subsequent seasons, having to remodel all characters to be able to move on. SMS thinks that is good advertisement, while I believe that plays against Poser because Rooster Teeth has chosen a truly professional tool right after completing season 1, which to many speaks for itself.

But perhaps the biggest blow came from the release of Poser Pro 10 with unusable rigging tools. It was literally broken, and was never fixed. If you need to make a JCM using the in-house tools, good luck with that. If you try to delete a morph, you have no control over how many will actually be deleted - not only on the clothing, but on the base figure as well. I could keep going, but you get the point. That's when some content creators simply gave up on Poser. Lady Little Fox has written a note about this in one of her tutorials, claiming that was the reason she moved on from Poser. SMS has only decided to fix things a few months after Poser Pro 11 was released, because I was pressing them hard, showing them videos of my products getting crippled by Poser broken tools. Poser only got usable again in P11 SR-4 or 5, but how many people have given up since Poser 9?

To summarize, SMS has failed to place Poser in the market, failed to recognize who their customers were, and failed to keep Poser in working conditions for years. They have also failed to recognize DAZ as competitors, and as a result, Poser lost its position in its own market. But ultimately, they have failed to listen to their customers, and continued repeating the same mistakes from the past. As a result, when we look at the official Poser forums, we find as many feature suggestions as we find people infuriated by Poser's inability to reach nowadays expectations. Take, for example, how Reallusion took iClone from mediocrity to state-of-the-art after mere 2 version releases. Where was SMS when all this happened? I rest my case.

Thank you for the insight, as a basic user it is not something I was particularly aware of, mind you it only adds to my view that they should spend time more important aspects of the software than developing another figure that will have limited use.

Anyway I am causing the thread to meander so I will move on, I hope you find a solution to the problem you have.

 

[Lissa_xyz]

Yes, but this is copying morphs from Dawn to Dawn. What I need is from Dawn to Clothing, which means it's a morph projection, and not a copy.

Mine is to clothing. It transfers the Doe morph into the clothes and adds autofollow. Unless you're trying to add morphs from Dawn into like V4 clothing.

 

[Ken1171]

Mine is to clothing. It transfers the Doe morph into the clothes and adds autofollow. Unless you're trying to add morphs from Dawn into like V4 clothing.

Ohh, sorry - my mistake! I didn't know we could make a injection for this. Let me see if I got this right. The morphs in the PMD are from Dawn or from the Skinny Jeans? If it only works with Skinny Jeans, then it's not what I am looking for, because I cannot tell what clothing people will be using my morphs with, and I cannot make injections for all possible clothing that exists.

Would be cool if we could make an injection that does the actual projection, so it would work with ANY clothing. Is that even possible?

 

[Lissa_xyz]

Yanno, I'm not actually sure why I did it that way. Transferring Doe was my first real behind the scenes project for Poser. I might've used either the copy morphs function, or a Netherworks script, to initially transfer the morphs. I don't think it even occurred to me to try the injections on different clothing.

Doe herself is a PMD as well, so maybe it was just the easiest thing for me to get the hang of? No idea. lol

 

[Ken1171]

Then I think this brings us back to being able to make a Python call to Copy Morphs From, which is not possible as things are now. Somebody told me everything in the Poser menus should be able to be done with Python, but I suppose that was not quite true. There is a LOT from these menus we cannot access with Python.

 

[Lissa_xyz]

Unless you're up for trying the clothing injections in my download on other clothing. If it works, it's a viable option for what you're looking to do. I'd test it for you, but like I said, I don't even have Poser installed anymore.

 

[Ken1171]

PMDs are binary morphs containing deltas. They are vertex-dependent and will have no effect on other models. In theory, they should only affect the models they were created from. Like you said, it would be like trying to inject Dawn morphs on V4. It won't work on different geometry.

 

[PerpetualRevision]

I know this reply is coming in a few months late, but I just read this thread b/c it's relevant to something I was just wondering about.

Some V4 outfits from the DAZ Store come with separate pose files you're meant to apply to the clothing items in order to "inject" V4's body morphs into the clothing. There are separate FBM INJ pose files for each of the clothing items AND for each major FBM package (Morphs++, A4, G4, S4 and Elite).

A few years ago, I tried using the FBM INJ for one outfit's blouse to get A4 morphs into a different shirt, b/c I thought maybe that would be a nifty way to get certain FBMs into V4 clothing items that lacked them, but it didn't work. Not long after that, I upgraded to a version of Poser that had the "Copy Morphs From" menu command, so I forgot about the concept of using an FBM INJ pose for that purpose.

But just yesterday I returned to the concept b/c I discovered a problem with "Copy Morphs From" that does not occur when I use the FBM INJ file that came with a particular clothing item. And the reason the problem doesn't occur is that the FBM data in the INJ file has been customized to make the morphs work well with the clothing item, whereas any FBM data you transfer using "Copy Morphs From" can only replicate what happens in the original figure's body mesh. And when it comes to the way body meshes deform in order to create different shapes of breasts, you do NOT want the clothing item to deform in exactly the same way as the body. That's how you end up with tops that seem to be tucked up under V4's breasts!

So Ken, I'm wondering two things: (1) might it be possible for you to create FBM INJ pose files that users could use to inject your FBMs into clothing? Poser Pro 11 has the option to export a "morph injection," so maybe that's how to create such a thing? And (2) might a better solution be to offer "adjustment" morphs that help clothing fit in the most commonly morphed areas, given that clothing shouldn't be shaped exactly like the skin underneath it?

I'm attaching two preview renders I did to document the difference between using the Veranil top's FBM INJ pose file and using Poser's "Copy Morphs From" command to get Aiko into the top. The difference is dramatic!

 

[Ken1171]

So Ken, I'm wondering two things: (1) might it be possible for you to create FBM INJ pose files that users could use to inject your FBMs into clothing? Poser Pro 11 has the option to export a "morph injection," so maybe that's how to create such a thing?

It's a nice suggestion, but it would be impractical for me to create an injection for every possible clothing that exists and include them with my body morphs. It is not possible to make one injection that would work in all existing clothing, because morphs are vertex-dependent.

In addition, I have tried to use Python to automate "Copy Morphs From", but it was just to find out that this feature is not available from the Poser Python API. Maybe it's not available because SMS wanted to make this a Poser Pro-only feature, but your guess on this is as good as mine.

And (2) might a better solution be to offer "adjustment" morphs that help clothing fit in the most commonly morphed areas, given that clothing shouldn't be shaped exactly like the skin underneath it?

I'm attaching two preview renders I did to document the difference between using the Veranil top's FBM INJ pose file and using Poser's "Copy Morphs From" command to get Aiko into the top. The difference is dramatic!

It's again impractical for me to offer morph adjustments for every conforming outfit that has ever existed. I cannot make 1-fits-all morph injection because morphs are vertex dependent, so those made for one clothing will not work with another. What you are showing me in the renders is when CAs create dials to support specific body morphs from a figure. The starting point from them is the same as "Copy Morphs From", but the difference is that CAs clean-up them morphs afterwards. When I add Dawn morphs support to my clothing, that's what I do, and this can be quite time-consuming. The idea is that morph injections will only work with 1 set of clothing. We cannot make one that would work with all clothing for a given figure.

In general, "Copy Morphs From" is just a starting point that in most cases will require some manual work to clean it up. This is especially true with Poser, because DS applies a smoothing pass afterwards that removes most distortions, while Poser doesn't. In addition, DS applies a mesh-collision detection pass to attempt to correct poke-throughs, while Poser doesn't. This is the root of the distortions you see from using "Copy Morphs From" in Poser. DS was created 10 years after Poser, so they had the chance to improve it in some areas.

The bottom line is that morphs are vertex-dependent, so injections only work in specific geometries. If we add or remove even a single vertex, it will invalidate all morphs created for it. This means injections need to be created for every individual conforming outfit, and we cannot make one work with all of them. That would be convenient, but impractical as things are.

 

[PerpetualRevision]

In addition, I have tried to use Python to automate "Copy Morphs From", but it was just to find out that this feature is not available from the Poser Python API. Maybe it's not available because SMS wanted to make this a Poser Pro-only feature, but your guess on this is as good as mine.
...
It's again impractical for me to offer morph adjustments for every conforming outfit that has ever existed. I cannot make 1-fits-all morph injection because morphs are vertex dependent, so those made for one clothing will not work with another.

I think what I was trying to point out is that the problem you were describing isn't that you can't get Poser Python to access the "Copy Morphs From" features. Even if you could, and you made a script that allowed non-Pro users to utilize the "Copy Morphs From" functionality, the user would still end up with the same problem you mention elsewhere: the morphs copied over into the clothing will not necessarily deform the clothing in an appropriate way and may need additional cleanup w/ the morph brush.

So if your goal is to give non-Pro users the ability to inject a figure's clothing with the FBMs used in the figure -- so that the user ends up with basically the same result as using "Copy Morphs From" in Pro -- then why wouldn't it work to give those users an FBM INJ pose file they could use that does the same thing: puts the figure's FBMs into the clothing item? In either scenario, the FBMs won't work perfectly in the clothing b/c they haven't been customized for the clothing's geometry. So whether a non-Pro user uses the INJ pose or a Pro user uses "Copy Morphs From" to get a figure's FBMs into a clothing item, wouldn't they end up with the same result -- FBMs in the clothing that may need further adjustment?

Or maybe I'm missing something about how "Copy Morphs From" works and/or how FBM INJ poses work?

I understand that the ideal situation is when the clothing creator has already included the FBMs in the clothing item AND adjusted them as needed based on the clothing's geometry. But I thought you were looking for a solution for situations where a non-Pro user wants to use a clothing item that doesn't already have FBMs in it, so that the non-Pro user wouldn't feel the need to upgrade to Poser Pro just to get "Copy Morphs From" functionality. If I'm wrong about that, my apologies! I'm still trying to understand all this stuff myself.

 

[Ken1171]

I understand that the ideal situation is when the clothing creator has already included the FBMs in the clothing item AND adjusted them as needed based on the clothing's geometry. But I thought you were looking for a solution for situations where a non-Pro user wants to use a clothing item that doesn't already have FBMs in it, so that the non-Pro user wouldn't feel the need to upgrade to Poser Pro just to get "Copy Morphs From" functionality. If I'm wrong about that, my apologies! I'm still trying to understand all this stuff myself.

I think it's a bit of both. I wanted my Poser morph injection for the body types to automatically make the clothing follow, like it has always been with DS Auto-follow. Poser users have to copy the morphs from the figure, and I wanted to automate that part. Like you said, the results aren't perfect, but I just wanted things to be easier in Poser (if I could help).

However, this cannot work as an injection because the base figure morph needs to be "projected" to the clothing, and injections cannot do that. Injections can only load morphs made for the specific geometry, while morph projections dynamically create a new morph over any geometry. Those 2 are very different things. Injections are just not an option.

 

[kobaltkween]

What you two are sort dancing around is the strengths and weaknesses of automatic morphs vs. custom morphs. Yes, auto-morphing tends to be inferior to the point of often being unusable. Custom morphs are often much higher quality. But an automorphing tool means not having to worry about new figure morphs. About 75% of the reason I don't use conforming clothes is my tendency to use every morph I've managed to find to dial characters. That tends to mean some free morphs, some custom morphs from characters, and some random utility morphs. I'd get the very best conforming clothes out there for women (the top of the line for men tends to be _much_ better), and find that even with as many morphs as their quite capable and generous creators included, I still couldn't fit my characters. Before I started making my own clothes, I'd end up rendering nudes after trying and failing to get some of the best clothes out there to fit my characters.

Ken1171 wants his morphs to be usable for the majority of the Poser community who a) use conforming clothing and b) have clothes for Dawn that don't support his morphs either by creator choice or by circumstance. The complication of getting conforming clothes to work with morphs, especially popular morphs in the future that _cannot_ be supported by current clothes, is a major issue that holds back the Poser content community. DAZ solved it, if not as cleanly as custom morphs, precisely because it's crucial to building a figure's content ecosystem. It's as foundational as being able to add new morphs by injecting them.

As I said before, it makes sense that it's not in the API, because that would mean any Python programmer could offer the feature to the community cheaply or for free. If you buy into the idea that this feature should only be in the pro version, exposing it in the API would reverse that. The problem is that it's not a "pro" feature, it's a "I just want to make a render without a hassle" feature. The problem is the reason it "has always been [like that] with DS Auto-follow."

 

[Ken1171]

I never understood SMS' position to limit access to something as essential as "Copy Morphs From", keeping it as a "Pro" feature. This just makes no sense. It's something all of us need in every scene, unless you only render nudes. It should be there, and it should be for everyone because it affects Poser's usability.

That is not to say that Auto-Follow is the way, because it either works, or it doesn't. There are pros and cons about trying to make it fully automated. I like the option to do it manually in Poser, for even if it doesn't provide the best results, the Morphing Tool is there to fix anything. We don't have that option in DS. Instead, we have global fixing solutions like the Smooth or the Push modifiers. If you only have a pokethru on the chest, those modifiers will affect the entire outfit, which is never a good thing. Here again, either these modifiers will help, or maybe not. If not, you are out of luck. In addition, those modifiers are dynamic, meaning they can slow your scene down.

Personally, I prefer "Copy Morphs From" + Morphing Tool. Even though Auto-follow in DS produces much smoother results than Poser, it lacks control when things go wrong. That is, I prefer control over automation. But the idea that "Copy Morphs From" is only present in Poser Pro makes no sense to me. Where did they get the idea that only "professionals" use morphs and conforming clothing? Whoever made that decision sound like they never used Poser.

 

[kobaltkween]

I never understood SMS' position to limit access to something as essential as "Copy Morphs From", keeping it as a "Pro" feature. This just makes no sense. It's something all of us need in every scene, unless you only render nudes. It should be there, and it should be for everyone because it affects Poser's usability.

That is not to say that Auto-Follow is the way, because it either works, or it doesn't. There are pros and cons about trying to make it fully automated. I like the option to do it manually in Poser, for even if it doesn't provide the best results, the Morphing Tool is there to fix anything. We don't have that option in DS. Instead, we have global fixing solutions like the Smooth or the Push modifiers. If you only have a pokethru on the chest, those modifiers will affect the entire outfit, which is never a good thing. Here again, either these modifiers will help, or maybe not. If not, you are out of luck. In addition, those modifiers are dynamic, meaning they can slow your scene down.

Personally, I prefer "Copy Morphs From" + Morphing Tool. Even though Auto-follow in DS produces much smoother results than Poser, it lacks control when things go wrong. That is, I prefer control over automation. But the idea that "Copy Morphs From" is only present in Poser Pro makes no sense to me. Where did they get the idea that only "professionals" use morphs and conforming clothing? Whoever made that decision sound like they never used Poser.

Also, the quality is an independent issue. That DS Auto-Follow produces better results in general just means the Poser version has room to improve, not that it has to work without user control.

IMHO, the general problem with the Poser team's decisions is that starting back at P5 the team decided that Poser wasn't a tool for helping illustrators work more quickly (its original purpose), it was a 3D animator's tool. And ever since then, every feature has been based on the idea of users who create their own content. Unfortunately for them, it's easier to animate in just about any 3D studio app (they just have more resources and better ties to the industry), and even if it wasn't, it's easier to stay in any 3D studio app than transfer all your content to something non-standard and stand-alone. Maya to Unity or Unreal- that's a different story, because almost anyone making a game needs content that works on those platforms. But after making a huge deal of the RWBY crew and literally two decades of ignoring content community artists, the RWBY crew dropped Poser from their pipeline (they've been pure Maya for a while) and Poser sales are dying with the content community.

Funny thing is, I _don't_ need the "Copy Morphs From" tool for my own work. Only for products. If I used Poser in the way its developers obviously intend- as a 3D professional making all my own figures and other content- I wouldn't need such a tool at all. For that matter, I wouldn't need Poser at all. I only use Poser still so I can make things for the average users their design excludes- which I've been told explicitly is on purpose. DS dynamics get used because their controls are designed for people making stills as the primary use. I've been told explicitly by Poser customer support that Poser dynamic cloth is only for animators, and that people unfamiliar with 3d animation and industry standards shouldn't use it. Meanwhile, TrekkieGrrrl, who has _made_ dynamic clothes since P5, had no clue "Steps per frame" was sim quality, let alone that 2 was _the_ lowest quality setting it could possibly be (I routinely use 8 to 16).

Since Curious Labs days, to the eFrontier and SM days, the Poser team has focused on the customers they wanted instead of the customers they had. And ignored that the customers they want, professional 3D animators, have zero use for even a magically perfect version of Poser. Meanwhile, illustrators have moved from Poser to DS and stock photos/photomanipulation.

 

[Ken1171]

I never knew if such decisions came from the Poser dev team, or from the SMS men-in-suits. Either ways, I have been calling attention for years to how distant Poser was becoming from its community, with little to no communication, and poor PR. I didn't really understand how deep the gap was until I joined their beta testing and started working with the dev team directly. Having used Poser since the days of Fractal Designs and MetaCreations, I could see SMS was driving Poser away from its community. I have submitted large lists of suggestions since their very first Poser 7 re-release, but at the time, I didn't understand why I couldn't get through.

My first clues were the way SMS was positioning Poser in magazine ads as the ultimate character animation program, and also how they priced Poser Pro to $600+. Both were bad news to the community in my humble opinion, showing that SMS didn't seem to understand WHO their public was. The 2nd clue was when SMS publicly declared they didn't see DAZ3D as competition, and that they didn't care what they were doing. Oh my, big mistake. That has marked the beginning of the more visible decline.

It became clear to me that SMS had lost track of the market, making it look like they were trying to reinvent Poser and make a new market that never came true, because Poser wasn't up to the task. It was like SMS made up what they thought Poser was, and tried to place it in a place where it didn't belong. It's not like DAZ never did that, for they had no issues dropping support for the entire Poser side of the market, which would mean loosing sales, and generate a fair share of community resentment. They thought they could make it up by attracting the game developers industry, because they have deeper pockets than this hobbyist market. That wasn't working for DAZ - until Reallusion stepped in with CC3 and Transformer. This may be the game-changing factor for them.

In the meanwhile, Poser is still paying the price for having neglected its own surviving community for too long. It took quite a beating thanks to a series of poor decisions, and it is still unclear if they have learned anything from it. Looks like Poser development has finally started up again, even if slow and STILL with very little communication with the community. The fact that SMS' first talk about Poser development started with a read-only thread is a sign that things may not have changed after all. They are still not willing to talk to us.

 

[kobaltkween]

Well, to me it's pretty easy to spot where it's coming from. At least in the sense that it's the people who have been a constant since the problem started, not the people who came along mid-way. For one, eFrontier's Poser 7 had about 21 advances (at least according to ad copy), and I think all but one of them was only relevant to animation. So not SMS suits, because that was before SMS. For two, you can see the animation and professional focus in your face in the Cloth Room and Hair Room UI design and workflow. Those are from Curious Labs days, even before eFrontier. For three, if you've had any dealings with the developers, they wax on about animators and animation standards and gaming modeling and denigrate anything outside of those standards. Stefan openly posted at CG Talk that Renderman's only benefit over raytracers is animation workflow modularity. Firefly is a _terrible_ render engine for stills where you need real caustics and GI that doesn't blotch, but OK for animation that's fine with lots of fast faked funk. Firefly is a terrible choice for people making images, and that was chosen at P5, when there were better _free_ rendering engines out at the time.

SMS bought Poser as an animation program because that's what the team sold them. SMS suits are just business people. They sell the software for what the _team_ tells them it is. And their support, their developers, and even the ambassadors those developers pick from the community all look down on the content community and at _best_ humor us as flawed 3D artists who should learn to model and rig instead of understanding us as mostly 2D artists who find modeling and rigging just get in the way of the goal of making pictures.

That's not me, but I know as a 3D artist I'm in the vast minority. And that there's a whole lot bigger demand for premade content among people who can't make it themselves than people who can but want to save time.

And it's not the SMS suits making a particular thread read-only. No high-level manager pays that kind of granular attention, and any manager above the head of the Poser dev team is high level. That's the Poser dev team, that thinks it's great when their "ambassadors" beat on vendors and try to push Poser users to do everything themselves, openly saying that content isn't worth money.

They lost their way at P5 so thoroughly that for _years_ I assumed that there was a much larger but invisible to us group of 3D animators buying their product. Because it was dead obvious to me back when they were headquartered in Germany what community they were pitching towards. That's not an SMS problem.

And conversely, SMS has done _beautifully_ with all its other eFrontier purchases. Manga Artists especially seems to have taken the top of the comic illustration field, and Moho/Anime Artist is second only to Adobe Animate/Flash. I come across people talking about them all the time just by looking up stuff about art and illustration. Only Poser keeps failing, and has pretty much no visibility outside of our community. But only Poser keeps adding features that only make sense for the 3D professional customers they don't have and ignoring the dozens of very easy, very low hanging fruit features they could add that would make creating stills _so_ much easier and more enjoyable for the customers they do have.