|
|
Extended IK/FK Handles
About Truespace Archives
These pages are a copy of the official truespace forums prior to their removal somewhere around 2011.
They are retained here for archive purposes only.
Extended IK/FK Handles // Tech Forum
Post by Stem // Aug 22, 2008, 4:12pm
|
Stem
Total Posts: 199
|
I have not seen mention of this problem so will post to see if any info.
Looking at Crazy bob (again) I was just moving the character around and noticed different behavior between the FK on the hands/waist/ankles and the IK on the feet.
If for example I over extend the handle on the hand and keyframe, the reulting anim is OK, all the model is how expected (pic 1), but if I over extend the feet (pic 2) and keyframe, when I move the keyframe in the dopesheet I end up with pic 3
Now yes, I have over exaggerated the extents to show, but should not the behavior be the same for the entire rid and the foot just stay where it is as the hand does?
Is the rigging on Crazy bob incorrect? or is this a TS problem? |
Post by Jack Edwards // Aug 22, 2008, 5:45pm
|
Jack Edwards
Total Posts: 4062
|
Yeowch that looks painful and definitely not good... :( |
Post by Stem // Aug 22, 2008, 6:52pm
|
Stem
Total Posts: 199
|
Hi Jack,
Yeowch that looks painfulYes, poor bob is getting into some strange positions these days :D
If you can repeat this on your setup, could you please forward to the devs,
TIA,
- Stem |
Post by Igor K Handel // Aug 22, 2008, 8:44pm
|
Igor K Handel
Total Posts: 411
|
Found this same behaviour only yesterday.
To emulate... grab Wizard's crazybob rig gizmo (walkatron 3000) copy paste the action clip along the track a few times. All works fine apart from 3 funky frames in the middle of the cycle.
But if you now delete say the first action clip. When you play starting at second action clip(a copy of the first) the clip itself works fine, but try scrubbing backwards (in theory where there should be no movement now)from the start of the now first action clip and you get same result as shown in your pic (if not worse).
IK
Wizards Walkatron 3000 gizmo scene -> |
Post by W!ZARD // Aug 23, 2008, 2:02am
|
W!ZARD
Total Posts: 2603
|
There is something funky going on with bobs legs - it's almost as if there aren't enough vertices in his mesh for the bones to get a good hold on.
There are a couple of other problems with poor old Bob - his skeleton is not symmetrical - one shoulder bone is longer than the other and his foot bones are not symmetrically connected to the mesh round each foot.
Bob is fine for general animation experiments but I advise using meshes you have built yourself - that way you can build them to bend the way you want them to - and if they don't well that will teach you how to do it better next time. |
Post by W!ZARD // Aug 23, 2008, 2:13am
|
W!ZARD
Total Posts: 2603
|
Sorry Stem - just re read your original post.
Bobs feet and ankles are actually rigged quite differently to his hands. Each hand has a single IK handle and no further constraints until you get to the full lock on the chest bone.
His feet each have an IK Handle (the white rounded rectangle), a full Lock (The small purple cube), Another IK handle at the ankle (the white plane) and a rotation lock at ankle and thigh ending with the full lock at the waist (the cube) and the IK handle at the waist (The white circle).
Thus the rigging for the IK chains is far more complex for the legs and feet than it is for the arms and hands therefore different behaviour is only to be expected.
Hope this helps:D |
Post by Leif // Aug 23, 2008, 4:52am
|
Leif
Total Posts: 276
|
For some reason I laughed out loud at that screenshot! |
Post by Stem // Aug 23, 2008, 7:10am
|
Stem
Total Posts: 199
|
Hi IK,
Found this same behaviour only yesterday.I did see this a while ago but was unsure if that was what should happen or not, but now I dont think it should react that way so made post.
I have thought about animating bob with hand stands and walking on his hands, as the hands stay in place better than the feet :D
I have been trying to completely re-rig bob, but as that is on a (slow for TS) laptop while at my mothers looking after her, it is a pain.
- Stem |
Post by Igor K Handel // Aug 23, 2008, 7:50am
|
Igor K Handel
Total Posts: 411
|
Hey Stem I reckon if it's a Irish dancer you are trying to animate you have the perfect pose!
Copy of your pic
IK |
Post by tomasb // Aug 24, 2008, 4:53am
|
tomasb
Total Posts: 261
|
this one is a confirmed bug.
There may be a difference between keyframing and manual moving IK handles due to order in which handles are solved (from limbs toward skeleton root) but that stretching of bobby's hip bone is wrong. Solver did not converged and stopped prematurely. This happens also in 7.1 and it will be fixed in a patch. |
Post by Stem // Aug 24, 2008, 4:15pm
|
Stem
Total Posts: 199
|
this one is a confirmed bug.................. and it will be fixed in a patch.Thanks tomasb,
At least it is not just me going "crazy", and I think "bob" will be pleased to have this fixed :D
- Stem |
Post by Electric Jim // Aug 24, 2008, 4:29pm
|
Electric Jim
Total Posts: 98
|
Regarding this part of the statement...
There may be a difference between keyframing and manual moving IK handles due to order in which handles are solved (from limbs toward skeleton root)
I'm wondering whether this impacts on the issues raised in the "Skeleton Animation Question" thread. There, several of us discussed the unexpected/undesirable playback behavior often seen for skeletons in the frames between keyframes, although the skeleton correctly aligns at the keyframes themselves. In such cases, the skeleton can seem to "float" or otherwise "break" IK locks at the in-between frames of the playback, even though the locks appear to function properly while the skeleton is being posed.
I guess my question is this: Does the difference between keyframing and manually moving IK handles, due to the order in which handles are solved in the two situations, mean that we must always expect such undesirable playback behavior in the in-between frames? If so, this tremendously reduces our ability to properly predict the way keyframed behavior will play back, and, therefore, the usefulness of skeletons for character animation in general.
I really hope I'm just misunderstanding the implications of the original statement... |
Post by Igor K Handel // Aug 24, 2008, 11:01pm
|
Igor K Handel
Total Posts: 411
|
LOL Bad Timing, I just spent 8 hours yesterday try to pin down this issue.
Few things I can confirm ref stretching
Stretching seems to only start approx 6 to 8 minutes after the first keyframe is set.
At times windows Page file size shoots up to 1 .2 gig once this has initiated.
CPU and graphics card loads, temps remain at same levels as pre keyframes, not sure if this has any relevance but there ya go.
100% cpu load is the case during scrubbing, with or without keyframes present. Don't think this is an issue.
The gremlin always appears in joints high in the skeletons heirarchy (the root, or possibly next down, not yet confirmed 100%)
The gremlin is present both locally and in shared space the timing of first stretch appearing is still approx 6 -8 mins from setting first keyframe.
In all cases the stretching can be repaired usually by rotating (I have used a pair of legs for testing) the hips, which appear to get a twist added. However, after further scrubbing or altering of Keyframes the keyframe either before or after the altered one gains the stretch again.
If I set keys on frame 1 2 and 3 and now I scrub to frame 1. Now from 1 I go to frame 2 and alter frame 2 keyframe, it seems that the preceeding frame (1) is warped. However if I go to frame 3 then scrub back 1 frame and redo frame 2, it is now frame 3 that is effected. IE the direction of scrub alters which frame is subsequently warped.. Still to be confirmed, but in the main this, on the surface, appears to be the case.
Jim even with keys set on frames 1, 2 and 3, this behaviour kicks in, so not just inbetweens, it's appears on already set keyframes as well. BUT as it's now confrimed as a bug, and being addressed in a patch (as per Tomasb's reply) I guess we just have to wait.
Continued and repeated testing on the same rig confirms that it is always the same joints that have the unwanted stretch/rotation applied.
Am VERY close to a fully working full IK leg rig. So looking forward to the patch with much eagerness.
Thanks for confirming bug Tom, have been spending hours trying to pin this down, yesterday was just 1 of several test sessions.
If any further test rigs or video examples needed I have em saved and ready to go (was planning on spending today looking at this again, but as now already being addressed, no need?)
I realise this isn't the bug thread but thought I should add these results, as many of those who have been trying to pin down this problem are currently following this thread.
Hope this helps. (personally I am optomistic this is about to be resolved, the future of TS character animation is moving forward which is where we all want to go:D )
IK |
Post by Stem // Sep 28, 2008, 6:14pm
|
Stem
Total Posts: 199
|
Hello,
this one is a confirmed bug.
This happens also in 7.1 and it will be fixed in a patch.
Anything on this yet? A possible date of a patch? I know you like to keep all info internal, but roman did state a patch in a few weeks, and that was nearly 2 months ago.
regards, |
Post by Stem // Jan 31, 2009, 7:33am
|
Stem
Total Posts: 199
|
Anybody there? |
Post by frootee // Jan 31, 2009, 8:35am
|
frootee
Total Posts: 2667
|
hi Stem. The patch is still in development. |
Post by Stem // Jan 31, 2009, 9:30am
|
Stem
Total Posts: 199
|
The patch is still in development. and the sun still burns and the earth still revolves :D
Sorry, could not resist.
I was hoping that bug fixes for bugs from 7.1 would have a little urgency in being made /released. |
Post by Igor K Handel // Feb 1, 2009, 7:58am
|
Igor K Handel
Total Posts: 411
|
I struggle to stay positive so will keep it short.
14 months since purchase of software that includes IK
6 months since submitting examples and bug reports
2 months until I will finally give up and "defect"
Patience running out... 14 months wait just to to get proper working IK is just crazy.
IK Handel
(a somewhat inappropriate name in the circumstances) |
Post by Electric Jim // Feb 1, 2009, 9:42am
|
Electric Jim
Total Posts: 98
|
I know what ya mean, IK! I've pretty much put my 3D hobby-related interests on hold 'till a solid fix for this stuff is available. (Fortunately, I've got enough other distractions in my life to "fill in" for the time I used to play with 3D.)
I know there's a ton of "working" functionality available in tS, so I could theoretically choose other types of scenes/projects to work on. But the role tS always used to play in my life was as a fun way to let my mind wander freely and see what I could come up with. So having to work on scenes other than those I want to work on simply because the necessary functionality isn't "really" there is far more frustrating than fun. (Picture this scenario: I work on a scene, painstakingly constructing my little character's skeleton -- which requires various work-arounds on its own -- then posing and keyframing him the way I want. Only to see that his animation doesn't play back the way it's supposed to. I spend numerous days trying to figure out what I'm doing wrong, and how to fix it, only to conclude that it's a problem with the program, not with me, and all those weeks I had expended are now wasted, because I simply can't accomplish the basic, dinky little animation I want to (and should be able to), given the bugs in the program. That's not fun; it's simply frustrating, sometimes to the point of being infuriating.)
Add to this the difficulties with what should be simple texturing tasks, including those related to textures being lost between Workspace and Model side, and I finally had to admit that I was spending more time getting annoyed/angry at tS than I was enjoying it. Not a good thing for what was supposed to be (for me) an enjoyable, relaxing hobby. (It's for these sorts of reasons that I've long been expressing the wish that Caligari would just fix what's there before worrying about adding enhanced functionality. But, as I eventually admitted to myself, there's no point in continuing to flog that dead horse.)
I recognize the very real possibility that by the time the necessary basic functionality does get fixed, I will have been away from 3D for so long that I simply won't be able to "get back into" the hobby, mentally. That'll be a shame for me, but c'est la vie. (Having invested so much time -- and, previously, money -- in the various versions of tS over the years, I simply can't see myself switching to another program. Especially now that tS is free, and for me, 3D is just a hobby. SUBSEQUENT EDIT: A possible exception would be Blender, with which I've played and by which I've been greatly impressed, function-wise. Though I'd probably wait at least a couple versions, for some GUI-related enhancements.)
What strikes me as more of a shame is the thought that there are probably many others like me who love the potential that the tS7.n architecture clearly presents, but who are getting really tired of waiting for the "basics" to be made conveniently and robustly usable. I love the great work that programmers and script writers like trueBlue, ClintonMan, and Dele (and others) are putting out. But for me, that work can't compensate for the fundamental shortcomings of a 3D program with IK that doesn't work properly, texturing that is problematical, and so forth. I actually get sad thinking of all of tS's potential very possibly ending up wasted because of the way these issues are being addressed -- or, more precisely, the way they're not.
Just my two cents' worth, for whatever it's worth. (Besides the two cents, I mean.) :) |
Post by Stem // Feb 2, 2009, 3:43am
|
Stem
Total Posts: 199
|
Hi IK,
Patience running out... 14 months wait just to to get proper working IK is just crazy.
It has been admitted that this has been a problem since 7.1,.. so we are actually looking at bugs going back closer to 3 years.
It does make me wonder as to what actual internal testing and private beta testing is actually made before a release, certainly very little for the character animation as for as I can see, certainly as locks dont lock correctly between keyframes and characters are ripped apart.
At least it is not just me that thinks this is a complete and utter farce.
Well, I think it is getting close to just removing TS/Vray from my PC as it is currently just taking up HD space.
Regards,
- Steve |
Post by frootee // Feb 2, 2009, 5:36am
|
frootee
Total Posts: 2667
|
It does make me wonder as to what actual internal testing and private beta testing is actually made before a release, certainly very little for the character animation as for as I can see
Incorrect. |
Post by Stem // Feb 2, 2009, 5:59am
|
Stem
Total Posts: 199
|
Incorrect.
mmmm,
well that implies that the bugs have been known about and just a case that the devs have never been bothered to fix them. |
Post by Stem // Feb 2, 2009, 6:26am
|
Stem
Total Posts: 199
|
Hi frootee,
Why did you delete your reply of "Wrong again. whats your second guess" |
Post by frootee // Feb 2, 2009, 7:12am
|
frootee
Total Posts: 2667
|
Sorry Stem. I didn't want to sound antagonistic, so I chose
to edit myself.
However, I took issue with the statement about IK not being thoroughly tested. While I cannot state specific information, I can say the all the beta testers and developers are working very hard to improve trueSpace, in all areas, IK included. We hammer, hammer, and hammer as much as we can. We're not sitting around doing nothing.
Sure it's taking a long time. But saying 'the devs overlooked this' and 'very little testing has been done for character animation' is not only grating, but simply incorrect. |
Post by Stem // Feb 2, 2009, 7:55am
|
Stem
Total Posts: 199
|
Hi frootee,
However, I took issue with the statement about IK not being thoroughly tested. While I cannot state specific information, I can say the all the beta testers and developers are working very hard to improve trueSpace, in all areas, IK included. We hammer, hammer, and hammer as much as we can. We're not sitting around doing nothing.
You may take issue, but I have issue with the bugs.
As for as the testing go`s, well, I think of it as this.
These bugs have been in TS since at least version7.1 (as admitted on thread), so that was one release made with the bugs,.. since then there as been another release and then a patch file release. So that is 3 releases over a period of around 2-3 years all with these bugs.
Now for character animation. One of the first checks I make is if locks work correctly between keyframes and they dont. So was this basic function missed in testing?, if yes, then I would say that the testing was lacking. Again, with the pulling apart of the character, was this not seen? as this shows up quite obviously on attempting character animation.
Now if these bugs where seen, then why where they not fixed.?
regards,
- Steve |
Post by trueBlue // Feb 2, 2009, 8:10am
|
trueBlue
Total Posts: 1761
|
Steve
Can you supply a data set/steps/scene/example of the bug in question? I'm fairly sure this would be forwarded to the developers by one of the Beta testers and or developers sometimes drop by here as well. Perhaps you could create an example and post it in the Bug section? |
Post by Stem // Feb 2, 2009, 8:48am
|
Stem
Total Posts: 199
|
Hi trueBlue,
For the problem with the character being pulled apart, tomasb did reply to thread confirming the bug (http://forums1.caligari.com/truespace/showpost.php?p=78552&postcount=10) so did not think examples would be needed.
As for the bugs with the locks, there where a number of threads/posts made back in aug last year (including post to bug thread), including post from myself of this issue ,.. but there was nothing from the devs in reply to these posts. Are you saying examples are now wanted ? it would be strange examples needed 6 months later after the reports
regards,
- Steve |
Post by trueBlue // Feb 2, 2009, 9:01am
|
trueBlue
Total Posts: 1761
|
Yes it would be helpful. |
Post by TomG // Feb 2, 2009, 9:06am
|
TomG
Total Posts: 3397
|
An example could then be used in the current beta version to see if the problem is remedied, improved but not complete, still the same, etc. Always good to have an example of your particular problem, rather than a general description, as then specific tests and comparisons can be done.
HTH!
Tom |
Post by Stem // Feb 2, 2009, 9:12am
|
Stem
Total Posts: 199
|
Hi trueblue,.. Hi TomG,
I did post a simple example of where locks do not work. (http://forums1.caligari.com/truespace/showpost.php?p=77939&postcount=9) you could start there.
I will find some time to put other examples together if needed.
EDIT: on the same thread I did post another simple example of lock problems (http://forums1.caligari.com/truespace/showpost.php?p=78022&postcount=10). The very simple example should be easy to test
Regards,
- Steve |
|
