PMD Hell... Again!



  • Ops, sorry, that's right, Use External Binary Morph Target; I think that creates a PMD file with the binary data, but I may be wrong.

    I'm with Poser Pro 11; my file has very old toons that I've reused many times - Judy, V3, Terai Yuki, M3 and the guy from Poser 7 (forgot the original name, it's now heavily modified); all of them have several pieces of conforming clothing merged into single figures, and most have been poly reduced; some 12 characters in the end (but if one counts the original conforming clothing that was merged, that makes some... 60 original figures, perhaps).

    I can do anything with these guys with no problems - merge conforming clothes, poly reduce, dynamic clothing, a few soft body (for the hair, haha, I found that works nicely with hair), fit clothing, change fitted clothing to conforming... not a single error in some 4 years or so, which I'm very much surprised as in the past dynamic clothes tended to fail quite a bit.

    Now I just turn on the external binary thingie... and there it goes, in less than a day! External binary morph targets is evil!


  • Poser Ambassadors

    @fbs7 said in PMD Hell... Again!:

    all of them have several pieces of conforming clothing merged into single figures, and most have been poly reduced

    I'm wondering if that is the issue.

    You've merged geometries together, and then reduced the count in the combined object. Do the morphs still work? Did you create morphs for them?



  • @Deecey said in PMD Hell... Again!:

    @fbs7 said in PMD Hell... Again!:

    all of them have several pieces of conforming clothing merged into single figures, and most have been poly reduced

    I'm wondering if that is the issue.

    You've merged geometries together, and then reduced the count in the combined object. Do the morphs still work? Did you create morphs for them?

    Yeah, the morphs work fine, as far as the morphs I use. I noticed that if I play with that long enough some morphs break (yuck), but I discard those broken toons as soon as I find them, as I've found the problem only gets worse.

    But then, with 300 morphs in each toon I don't check every morph around.



  • Btw, the morph operations that I do are all internal in Poser: fit with morph tool, fit with fitting room, merge characters and poly reduce. I use no external tools for morphing.

    I imagine that one or more of those are breaking the PMD file, I guess; it's anyway odd how I've been using PZZ for months after these operations are done (it's been long time since I last did them) with no problems, and yet the very first PMD, bang...



  • This is pet peeve of mine, so please feel free to ignore the following. Screaming that something is evil when you have an error doesn't help anybody. I understand you want to vent. I have obscenity laced tirades towards my computer at least every other week. Also, I had to replace my mouse due to repeated poundings. Ok, rant over.

    Lets work the problem. I have an idea for a quick test to help identify the issue.

    1. Create a scene as usual.
    2. Save the scene with binary morphs off (Scene 1).
    3. Close Poser. Then reopen Scene 1.
    4. Save the scene as a new scene with binary morphs on (Scene 2).
    5. Close Poser. Then reopen Scene 2.

    If you get the error, that means that a morph in the scene is incompatible with binary morphs. It would be a pain in the butt find what figure causes the problem.

    If there is no error, that leads me to think that an internal operation could be causing the issue. When a figure with PMD is added to the scene, Poser 11 will try to reference the original figure PMD instead of adding the morphs to its own PMD.

    If we can find what operations cause the PMD error, it will be easier for Smith Micro to fix. Thanks for letting me rant.



  • @willdial said in PMD Hell... Again!:

    This is pet peeve of mine, so please feel free to ignore the following. Screaming that something is evil when you have an error doesn't help anybody. I understand you want to vent. I have obscenity laced tirades towards my computer at least every other week. Also, I had to replace my mouse due to repeated poundings. Ok, rant over.

    Haha, no problem. Problem is, I don't now what's broken in the document, so I don't know which toons or objects to redo to clear that up. I only use PZZ and that worked perfectly well several years. Then I take a chance to save something with binary morph targets enabled, and it bombs on me in just a day.

    I almost always reuse work, so it's actually rare that I redo something from scratch. So maybe the original problem was from PP 7 or the like, but I have no idea. As the error message is very generic and doesn't tell what it was doing when it failed, the developers didn't help themselves to solving that either. It would be better to have something like "Tell developers Error Code 933874" so that someone can find inside the code which function or procedure failed.



  • @fbs7 said in PMD Hell... Again!:

    @willdial said in PMD Hell... Again!:

    This is pet peeve of mine, so please feel free to ignore the following. Screaming that something is evil when you have an error doesn't help anybody. I understand you want to vent. I have obscenity laced tirades towards my computer at least every other week. Also, I had to replace my mouse due to repeated poundings. Ok, rant over.

    Haha, no problem. Problem is, I don't now what's broken in the document, so I don't know which toons or objects to redo to clear that up. I only use PZZ and that worked perfectly well several years. Then I take a chance to save something with binary morph targets enabled, and it bombs on me in just a day.

    I almost always reuse work, so it's actually rare that I redo something from scratch. So maybe the original problem was from PP 7 or the like, but I have no idea. As the error message is very generic and doesn't tell what it was doing when it failed, the developers didn't help themselves to solving that either. It would be better to have something like "Tell developers Error Code 933874" so that someone can find inside the code which function or procedure failed.

    Oh, and I forgot to say... I'll never ever ever again enable Binary Morph Targets in any situation, ever again. I kinda promise that every 3-4 years, then take a peek on it, anyway, hahaha...



  • @fbs7 that's how I do it!



  • I guess as long as the PMD files are a sort of "Black Box" it won't change because no one is really able to determine what goes / went wrong.
    (I know there is a program out there to read it, but that one is just another Black Box because it's compiled and doesn't reveal it's secrets on how to read a PMD)

    Time for SM to finally release the code to read the PMD files!

    Until that happens I won't touch the PMD option with a ten-foot barge pole.

    • Sure, it seems to work for some. Emphasise "some"! Is this sort of Russian Roulette?
    • But for others it still doesn't and annihilates hours of work.
    • AND: we still can't find out WHY the Dickens!

    So is this DAZ with it's proprietary file format?
    Or is this Poser with it's widely-acclaimed open source format?
    ?

    With storage media as cheap as today I can well live with that bit of extra storage space needed to save my morphs "the safe way" without the PMD minefield.

    Come on SmithMicro!



  • @karina ask @anomalaus He just worked on it (he decyphered it and he's able to read it) from source from @adp

    http://www.sharecg.com/v/89075/gallery/8/Script/PMD-File-reader-writer-WIP



  • Never ever use Binary Morph files. That's what I've learnt since Poser 7. Morphs get lost using them. From time to time I test it with the same results: randomley loosing morphs, that are saved as pmd by using binary morph targets. It is devasting to hear, that this bug still exists.



  • @fbs7 do you have a broken scene file that you can share? With the associated PMD files.

    I'm currently in the self-imposed hell of trying to read .duf/.dsf figures and load their morphs (nearly done, yay!) and translate their rigging (getting quite depressed, now) into Poser, without the use of DSON importer, so I don't have a lot of Cerebral Processing Unit cycles left to spare. As @barnardino helpfully pointed out, @adp's PMD library has the necessary decoding routines, but needs some wrapper code to actually select and read/decode the PMD files which are problematic. I think that would be really helpful to get to the bottom of this problem you and many others have suffered, which Poser's own uninformative messages don't even begin to shed light on.

    If, as a collective organism, given adequate tools and knowledge, Poser users can analyse and determine where the problem lies, and present sufficiently detailed information to SMS, a final solution may be possible. Individually, we all struggle with insurmountable issues. Together, we can "Rule The World!" </Party Political Broadcast> ;-)

    Deciding to never use binary morphs, has been a pragmatic solution for many of us, who just want to get on and "Do Poser!", yet there are obviously plenty of experienced Poser users out there who use them every day. Given how much of a struggle it was to eliminate many of the esoteric bugs with PMD in the early days, it's absolutely not statistically impossible for others to linger, dormant, like viral spores, waiting for the moment to launch our anxiety and frustration levels into orbit, again. I've just wasted hours on a bug in my own python code to split whole body mesh morphs into per-actor delta morphs, only to find the error was an oversight in an indentation level which put code meant to be executed every loop iteration inside a preceding conditional (Ah, the joys of python wrangling!)



  • @anomalaus

    Yeah, thank you; let me find a file that's small enough (my files tend to be quite large), and doesn't a myriad of textures around (otherwise would have to send them too)



  • It's a little pricey, but the late Dimension3D has Binary Morph Editor at Renderosity. I've used it a few times to copy morphs between scenes.


Log in to reply
 

Looks like your connection to Graphics Forum was lost, please wait while we try to reconnect.