Linking morphs across different body parts would be a different sort of problem than any of my scripts try to approach. There are various ways of doing that within Poser, as well as methods that require the editing of a CR2 file. There should be threads out there. Keywords would include PBM (partial body morphs), FBM (full body morphs), and various combinations of the words "morph" and "link". Hopefully there are still experts in that area posting here at Rosity.

Thanks. Those things are only an issue with extreme head morphs, and usually not for just normal human facial features. This one is pretty good. Here they are wearing each other's head shapes:
I'm going to revise what I said about the eyelashes: a cutoff of .003 is too short, it has to be at least .005. (For whoever is reading this and wanting to try it, to keep them from having to waste time on stuff that won't work.) They're still not perfect.

If I could program (but I can't) then an add-on I would set out to make is something that allows the unmatched parts to "follow" the matched parts when they are morphed. For example, the eyelashes would remain stuck to the eyelids, and would move with them, even without being matched to the other figure. Or, if one figure's teeth and gums are part of the head, they would maintain their relative position to the top and bottom of the mouth. This would be different from restoring the whole morph because it would work only on the vertices that are not included in the morph, telling them to come back into their approximate relative position to the other vertices before they were morphed.

Is there already a way to do this, and I'm missing it?

That can be done now, with areas like a brow overlay or lashes, but it does require a comparison to a source object. Basically, the overlay areas are set up to receive the same deltas as the underlying faceskin material. I've been doing this since P8 days, with Antonia. Because of how TDMT was designed -- the limited scope of the project -- there is no setup to allow one material to receive deltas from another material on the same object. A workaround, however, would be to use two instances of the same object to achieve that same goal. The source object should have the faceskin equivalent deltas and the second object's overlying brow/lashes/whatever materials should be correlated to the faceskin. I've used that trick a few times while refining face morphs.

I have done that also, but sometimes I wish for something stronger. The mouth is usually not a problem if both figures are human and the morphs are not extreme, but if, for example, I am trying to morph the Millennium Big Cat into Nergal then I would need all the parts of his mouth to scale with his snout, and without warping too much. (Yes I know there are easier ways to create such a character but just hypothetically, for the sake of illustrating the problem.) One of the problems with Satyress was that she has horns. The base of the horns is part of the head, and there is a chain of horn parts parented to it. If I apply a morph that changes the shape or size of her head, there is no way to tell the horns to move gracefully along with it, while maintaining their own shape. It makes logical sense to me that what I'm talking about could be done, but I'm not a programmer, so I don't really know how hard it would be.

The big trick to any more extreme deformations and correlations would be matching the surfaces of the source and target objects prior to comparison. That was always the part I struggled with, when setting up actors for use with TDMT. At the time of the last wave of script development, one user had great success aligning surfaces using ZBrush. Possibly the new surface-matching tools of Poser's own Morph Brush would be helpful for that sort of thing, now. The trick would always be minimizing distortion of the meshes prior to the comparison. That entire preparatory step is necessarily left up to the end user, with TDMT. To match the surfaces via scripting before the final comparison, one would need a comparison so the script could correlate the meshes for shape-matching. The quality of that initial comparison would vary, just like the final one, based on how well the surfaces correlated. That sort of problem is handled by programs like FaceShop, but I never had much luck with such utilities.

For the horns, I did something like that at one point, but I really had to hack the process to make it work. I used TDMT to grab the morph deltas for the underlying areas -- the bases of your horns, basically -- and wrote a separate one-use script to extend the deformations along the entire overlying object. It worked well enough in my case, but the motion of the deltas I needed to handle was simple, along only one axis. I can imagine a workaround approach for your horns, in which they are scaled down or flattened against the underlying surface, after which a TDMT closest vertex comparison might transfer the motion that you want. I could also imagine just using magnets to generate the required morphs, perhaps using weight-mapped magzone influence to keep the motion regular. Another possibility would involve manually translating the horns to their final intended position, then exporting as .obj and loading the export onto the original horn as a morph.

Those are good suggestions, although I don't completely understand all of them. Do you still have that script? The other thing about the horns, and this is probably just a limitation of my understanding of Poser, is that what I do to the base doesn't affect the rest of the horn. So if I move them they split apart. I checked to make sure that the horn is in fact parented to the base, and it says it is, and they move when I pose her like they're supposed to but they won't stay stuck together when I morph her. Probably your suggestion about creating a morph and importing it would work but 1. it bugs me that they won't just do what they're supposed to do in the first place, and 2. with my TDMT matches between figures, I've really been thinking in terms of the match as a whole, instead of just for a specific morph. Making a match where the morphs hardly ever need restoring seems like an art in itself. My final match between Apollo and M4 was really very good, except the eyelashes weren't perfect, and I've barely even used it except to harvest the ear morphs (that's why my ear matches are getting better, because I've been collecting them from every match I've done) but just having made it is satisfying in itself somehow.

I think that one-shot script was handled using my temp.py, which I rewrite each time I need a one time use script. That is perhaps a flaw in that workflow. Hmm.

You might want to try a morphing-as-translation test for your horns, to make sure it will actually do what you hope. If I understand what you want, you might be better off using ERC or Poser's built-in dial-linking process (don't recall the name of it right now?) to link the actual translation of the horn props to the shaping morphs for the underlying head. I seem to recall that using morphs for translation can lead to some distortion of shape at settings between 0 and 1. Or perhaps I'm thinking of rotation, and translation is fine? Hmm. Not really sure. At any rate, dial-linking may be a preferable option.

I get the sense that my answers are becoming frustrating for you. Sorry about that. :( My Poser workflow tends to involve a lot of workarounds and ad hoc processes, as I experiment to make the most of the sadly limited tools I've been able to script with Python. TDMT never reached the point of refinement where it could do what I wanted, right out of the box, as it were. Spanki, a fairly skilled programmer, was primarily concerned with making the core process work, perhaps as a proof of concept as much as anything. My expansions on his foundation saw me learning as I went, making it up along the way, and involved a lot of trial and error. I frequently hit my limitations as a programmer and had to switch tracks. If you look at the tutorials I cobbled together, they show a lot of complex workarounds that I had found useful to make the most of TDMT's fairly constrained toolset. None of it, sadly, ever worked as well as I had hoped. I've managed to do a few things, but I don't think anything I ever tried was as ambitious as what you seem to want to do.

When I was matching Antonia with V1 and V3, the process involved several passes as I gradually refined the surface matching for a better final correlation. A lot of magnet-morphing, followed by a TDMT correlation run and a shape transfer. Then cleanup, more magnet-morphing, lather, rinse & repeat. Even then, the best I managed still required significant cleanup of most of the mouth morphs. That was the most ambitious I got, with TDMT. Since then I've largely used it to clean up morphed topologies using tricks like the eyebrow-eyelash thing I mentioned above. Things like your Nergal and your horns are uncharted territory, where I'm left guessing about what might work.

I'm sorry if my tone comes off like I'm complaining. I'm actually charmed and delighted by the whole thing, and I'm usually laughing when I type that stuff, and I forget that people can't see that. Poser is very complicated to learn, I bought it thinking it would be _easier _than drawing, but it is much harder, and I sometimes get impatient because I generally know what I want to do but it becomes very complicated to pull it off. Your scripts are awesome. TDMT changes the whole game, it opens up all morphs for all figures. I use that morph restoring one all the time too. It would be ridiculous for me to be frustrated because the Millennium Big Cat's teeth stick out of his head when I try to turn him into Nergal, when I know fully well I could just chop off his head and give him a new one, or make his head invisible and position a man there and make the man's body invisible, or photoshop it, or make the cat's teeth and tongue invisible, or just draw a picture of Nergal, and I don't actually need a picture of Nergal for anything anyway. I really appreciate your taking the time to discuss this stuff, it's all interesting to me. Sometimes I don't process it right away because I get information overload when I'm hearing a bunch of Poser stuff that I don't fully understand, but I store it away and come back to it later.