itsjustme

A couple of things I found in my testing so far might be helpful to others, so I thought I'd post them. First, if you want to script a compensated "orient like" constraint that has "store roll" turned off, the scripting looks like: Second, it appears that hiding and unhiding a bone in a Pose doesn't work in an on/off Pose, but it does in a percentage Pose. Here is a portion of a percentage Pose to illustrate (it's actually an on/off in the Squetch Rig, I just change it to an on/off in A:M afterward): The "hide" and "show" affect the bone when the Pose is at "100%", so, "hide" will unhide a bone when the Pose is at "0%". I went through several Diff/Merge programs, but none of them would merge files the way I wanted. I figured out that I can get the job done using jEdit with the JDiff plugin and recording a macro. It's kind of a low-tech hack, but it appears to do what I need. I still have a few things to work out, but I'm making progress...still a lot of testing and assessing to do.

When I've crashed TSM2 it's always been because my script referenced a bone that didnt' exist. Typically because I slightly mispelled the bone name. When TSM2 has not crashed but not completed all of a script it has been because I had a stray character or mispelling in some place other than a bone name. Thanks for the info, Robert! I'll do some more experimenting tonight.

I was concerned with altering it because I still am not sure what TSM is looking for in all instances...I've had it decide it didn't like a few things. I'm still feeling my way around. I completely understand, everyone has their preferences. I'm pretty happy with the latest Squetch Rig spine myself (of course, I'm biased), but I realize there are just as many reasons someone would want TSM's spine...it's very nice. this converting back to V12 stuff I don't see the need for? TSM Rigger (the last TSM step) doesn't read any model files at all. It operates on what is in your model window which means it could have been a V13 model by that point. the only Pre V13 things TSM needs is "components" You can run Builder in V13 and save its result in V13, close A:M, reopen A:M, reload that V13 model and TSM FLipper and TSM Rigger work just fine on it. I've had TSM disagree with me plenty, it wasn't made to do a few things I've attempted. I don't remember what I was trying the times I crashed it, but it could very well have been something silly. Breaking things is a good way to figure them out though. I'll have to illustrate what I'm doing, I guess...it may only make sense to the voices in my head. I'm sure there are several equally valid solutions and hundreds of individual preferences. I'm just working on one, but that doesn't mean it won't open up other possibilities.

That's what my app proposal above was about. I can almost see the code in front of me, but I haven't done any real programming in about 3 years so there'd be a certain time spent getting back up to speed on the little details. If anyone out there is enough of a programmer to read and write text files and do search/slice/insert type operations on strings, I can tell them what they need to do. My thinking was that if you have a v13 file that has the nulls in it and a v13 file (after running TSM and then saving using v13) that has every bone with the same name except for the added compensate bones, a program like KDiff3 could be used to set to over-ride the "" or "" tags with "" and "" tags from the original file that had the nulls. I think I'm close to getting that done...I'll do some more documenation reading for KDiff3 tonight. As I see it the only things that are hard-coded in TSM2 are the names of those component and Script files. Inside those it's wide open. The TSM spine has five segments at the standard setting...I forget if it can be set to three, it probably can. Also, there would have to be a bunch of markers added for the FACE controls...personally, I like the posable installation placement of bones (of course, I would though)...it would definitely make the face rig easier to install. That could be added pretty easily as an imported model, but, once it got exported, it would have to be converted to a v12 model to run the last step of TSM. I haven't quite deduced the proper workflow for creating them but the TSM2 rig has poses in it that contain information on both the On and Off positions, so I suspect this can be done. Just a theory. I do expect that some things for Squetch will have to be dragged into the model after TSM is done but that will still be a lot faster than having to manuallyreset all those compensates. I'm thinking that the only Poses that TSM would need to be able to make for the Squetch Rig would be the ones that need compensates reset, the others can be imported. That would save a lot of time on the development end of things. It will all get worked out, it's still early in the process.

You can build that rig using TSM. Just use a bone where there was a null TSM doesn't have to handle nulls. Nulls are just bones with a different onscreen appearance. There's nothing a Null does in a rig that a bone can't do also. I have a rig that has lots of nulls in it. They were all created as bones, placed as bones , constrained as bones... all that. When I'm done I just change their tags to NULLOBJECT, and then they appear as Nulls. I realize that it would only take a change of tag, Robert. I'm not sure of the number, but, there are a lot of nulls that would need to be changed...manually, that would take longer than just resetting compensates. Do you have an automated way of doing this? There are enough differences between TSM and Squetch that make installing the Squetch Rig using TSM more difficult. TSM looks for specific things, which also causes problems depending on what you're doing. What I'm doing is only using TSM to reset compensates. I'm not sure if TSM can make the FK/IK switch or reset compensates in both the on and off position of a Pose...I haven't gotten to that yet. However, since TSM can't make Expressions, there would have to be an import anyway. There are plenty of limitations.

This isn't a solution for Kdiff (a text editor?) but my notion for a Flash app was to write a loop that would -take a list of bone names, -search thru the mdl file for an occurance until it found one -backtrack to replace it's SEGMENT tag -from there search count thru the and tags to find the correct closing tag -start over on the next bone name in the list First, I'm taking the original model (v13), using a macro in jEdit to convert the bones section of the file to a v12 set of bones, pasting the v12 bones into a dummy v12 model, opening the v12 model in v13, running a basic installation setup for TSM (with an altered script that includes the constraints that need compensating in the Squetch Rig) so that it won't error out, save the resulting file as a v13 model, drag the new Poses to the original model, delete the unnecessary TSM bones, cut out the updated bones section and then use KDiff3 to compare the original bones to the updated bones (all that would need to be done is to put back the nulls), then paste the fixed bones into the final model. That's what I'm presently doing...I just have to get the nulls back using KDiff3, which I think will only take a little documentation reading on my part to figure out. The bones have to be converted to v12 because TSM can't see v13 bones. I could have built the Squetch Rig using TSM, but, TSM can't handle nulls...the way I'm presently doing it appeared to be the easiest solution given all of the variables.

I ran out of time again this evening, so I don't have the example ready to post yet. I have one more thing to overcome...and it's a very simple thing, I think. Dragging each compensate bone made by TSM into the Squetch portion would be just as tedious as resetting compensates, so, I decided to use KDiff3 to just put back in the and references. Where I'm hung up is doing the actual replacing using KDiff3...it could either be the "Preprocessor command" or the "Auto merge regular expression". I've never had to do that kind of thing in KDiff3, so I'm still figuring that part out. All I need is to have "NULLOBJECT", where it occurs in the original file (just a text file of the "BONES" section before using TSM), replace "SEGMENT" in the post TSM file (a text file of the "BONES" section after TSM). If anyone has KDiff3 pointers, or can recommend another cross-platform free tool for text file comparison (with instructions that will help me in this), I would be very grateful. If not, I should be able to figure it out eventually...I'm pretty stubborn.

After tonight's testing, I think I have a decent handle on it. I've run out of time for today, but I think I can show my work sometime late tonight (which is my tomorrow). I had to convert the Squetch Rig into a v12 set of bones without nulls (I made a macro for jEdit and used a dummy v12 model) and saved that out as a separate model since I needed TSM to be able to reset the compensates. From there, I figure I can just drag the Poses and added compensate bones that TSM adds to the v13 model.

You're dragging those in from another model now anyway, so you don't really need to write them in TSM, right? Yes, they are presently being imported. No, I mentioned it as an added caveat for anyone that might want to add Expressions. It can be worked around, it just adds another step in the process. The easiest way I can see is to make Poses that only had the constraints that need compensating, then dragging or importing them into the rigged model. Using TSM to build the Squetch Rig would take more work than just doing a hybrid installation, I'm thinking. All that would be necessary would be to add the compensated constraints to one of the scripts and actually have that part of TSM in the rig temporarily (since TSM looks for each system). I'll run some tests tonight.

well, that would describe rigging in general anyway. Absolutely. The biggest hurdles are the inability to make nulls (there are a ton of them in the Squetch Rig) and I haven't been able to make more than one sub folder for Poses so far (lots of sub folders in the Squetch Rig). I have some things I'm going to try tonight. Thanks of the tips, Robert!

The ability to add nulls using the SDK was added after TSM was made...I think in v13. If the components are altered, they have to be done using v12 or earlier since the file format changed in v13. I haven't been able to figure out a way to make more than one sub folder for the constraints...anyone know how? ---------------------- EDIT ---------------------- I forgot to mention that Expressions aren't scriptable either.

Yessir, that's where I'm at, Dhar. I've been modeling a character for my personal project that I'm going to use in the tutorial...I have been pretty obsessive, I realize. Also, I want to figure out the possibilities of using TSM in the installation process, I don't want the tutorial to be outdated immediately after I finish it. I figure the character will be finished late this week and I'm hoping to have the TSM integration figured out by then as well. I'll start a thread for my character when he's finished and then I'll get started on the tutorial. Sorry it has taken so long, I've been extremely busy.

I did a little messing with it...I'll do some more tonight. It looks like it will work, but I haven't done a really good test yet. Thanks for the info, Robert! -------------------- EDIT -------------------- I should have said "might" work...there are some hurdles that I'm not sure about yet. -------------------- EDIT -------------------- I should have gone to bed, but my curiosity got the better of me...if it's possible to get the Squetch Rig in there, it will be a lot of hoops to jump through. Maybe I'm not looking at it from the right angle, I'll give it another go tonight.

First, does the Pose work when you have the Pose open (while editing it)? Did you have any Actions open at the time you made the Pose? Sometimes the changes you make won't "take" if you have an Action open at the same time. Are you checking the Pose in an Action or Choreography? If you can post the model, or a test model, maybe we could point you in the right direction. If that's not possible, give us a step by step explanation of what you are doing. Hope that helps.

Very cool! Thanks to Raf and company!

Here is a rigged ball that Robert Holmen made...it's pretty cool.

Some things have changed since the beginner tutorials were put together...I don't think it's a "problem" with v13.

Added some FK finger controls (they are also on the Foot Gizmo version toes) to make it easier to add overlap to finger movement. There is a quick video attached here to give you an idea of how to use it. I can post the steps to update a previous installation if anyone needs it. If anyone finds a problem, let us know and we'll fix it as fast as we can. -------------------------------- EDIT -------------------------------- The next rig release is here...this version was deleted. FK_finger_controls_Sorenson3.mov

Very nice model, Scott! The possible problems I see are with the arms and legs...although it may be a style choice (like the leg length) as well. Arms and legs have at least one axis that is a straight line (for the bones), the axis that the limb bends on. I would move her knees outward (or move her feet inward) and her elbows up for that reason. Feel free to ignore the suggestion, you've done some fantastic work so far.

Very cool test, Stian!

I didn't scale it, I just moved it. The reason I moved the bones was that I sat in a chair, breathed in deeply, and noticed that not only did my chest expand (muscle movement), but I also slightly moved my upper back, neck, and head (bones). I can probably do the pose without moving the upper back, but this is a case for the 2001 rig where it is clear that because of dependencies, if you move higher-level bones (such as pelvis, back, back2, shoulder) after lower-level bones (hand, or hand target), you mess up the positions of the lower-level bones. This can be frustrating for anyone that is not intimately familiar with all of the ramifications of the rig. It can also be frustrating for anyone that has a need to go back and clean up animation. A quick look at the squetch rig ... it looks like when FK_IK is on, the hands stay where they are. I think the target in question is not directly below the geometry for the back and shoulders. This is how I was expecting (incorrectly) the 2001 rig to behave. Oh, sorry for the misunderstanding, Chris...I should have read more carefully. You could make a modified 2001 rig by changing the parenting of the IK hand targets.

I don't think you need to scale the "Back2" bone, Chris. The chest expansion using muscle movement should be enough for a "breathe" Pose. The other motion would be handled using spine and head rotation and translation. Hope that helps.

Another quick update. This adds separate left and right "Blink" Poses. It should make it easier to wink or offset blinks. If you have a previous installation that you want to update, copy the "blink_switch" null twice, move them to the left and right limits of the "Blink" box in the FACE interface and rename the nulls "blink_left_switch" and "blink_right_switch". Then delete the relationships for the "Animation_Controls/FACE Interface/Blink" Pose, the "Animation_Controls/FACE Interface/FACE off/Joint Controls/Split Controls" Pose and the "Animation_Controls/FACE Interface/Constraints (do not touch)/Face Control Constraints" Pose. Drag and drop those Pose Relationships from a version of Squetchy Sam to replace them with the updated Poses and drag the "Animation_Controls/FACE Interface/Blink_individual/Blink_right" and "Animation_Controls/FACE Interface/Blink_individual/Blink_left" Poses from Sam to your character as well. Then, drag and drop the Smartskins for the "blink_right_switch" and "blink_left_switch" from Squetchy Sam onto your character. If anyone finds a problem, let us know and we'll fix it as fast as we can. -------------------------- EDIT -------------------------- The next update is located in the next post in this thread.

Very cool, Lazlo! Good luck on the project!

Fantastic work, Rodger!