Dhar
Mar 29 2007, 06:52 PM
Very first-getting-my-feet-wet take.
I didn't block...I...forgot how

So, I went ahead with the way I usually do it.
Somebody put a link to a blocking turtorial using A:M now I can't find it. Would that someone please post the link.
Thanks.
Xtaz
Mar 29 2007, 06:58 PM
nice start Dhar !!!!!
maybe if you work more with shoulders.. they seems statics ..
robcat2075
Mar 29 2007, 08:18 PM
QUOTE
I didn't block...I...forgot how So, I went ahead with the way I usually do it.
Ho-o-o-o-o-o-o-ld on there, You're making it too hard. Just block first.
QUOTE(Dhar @ Mar 29 2007, 08:52 PM)

Somebody put a link to a blocking turtorial using A:M now I can't find it. Would that someone please post the link.
http://www.hash.com/two/RCHolmen/PoseToPoseH.movDo it like in the tut.
Just the poses, held by paired keys. We don't care (too much) what the transitions looks like yet, but we care very much about seeing those poses so we can compare them to each other.
Watch the pitch video starting at 1:52 and check out the four poses I identify.
start
anticipation
**Leash held up
Looking to sky
Now here's some thing to always watch in video reference:
what are the hips doing? It's pretty rare for the body to do anything and not move the hips. Tinman's hips are moving a bit but not much.
Where do they move from the "start" to the "anticipation"? Was it the same direction as the hand?
Where do they move from the "anticipation" to "Leash held up"
Where do they move from "Leash held up" to "looking at the Sky"?
I clipped out just the poses here, you can scrub thru it to compare one pose to another.
Click to view attachment**there's really another important pose here, the arm swung to screen right on "lawn" but I didn't redline it, it comes later, when we do break downs.One other bit... you definitely want to do his left arm in FK, maybe even both arms. Even though we're just making poses now, the raw transitions will be closer to the right motion if they are moving in FK. Swinging arms are almost always better in FK.
Thanks Dhar!
Dhar
Mar 31 2007, 01:23 PM
Gracias, Marcos. And thank you Robert for the tute. I've never seen that one before. I tweaked the key poses and changed arm to IK but left the transitions alone since now I can see how easy they are to move, tweak or delete. It took me longer than normal but I'm sure I'll get quicker with time.
robcat2075
Apr 1 2007, 11:21 AM
QUOTE(Dhar @ Mar 31 2007, 04:23 PM)

Gracias, Marcos. And thank you Robert for the tute. I've never seen that one before. .
It's at the top of our "We're Off, We're Gone" thread.
I should have left it on the copy machine. Then everyone would have looked at it. (Old office workplace joke)
QUOTE
I tweaked the key poses and changed arm to IK but left the transitions alone since now I can see how easy they are to move, tweak or delete. It took me longer than normal but I'm sure I'll get quicker with time
Hey Dhar, could you check to see if your PRJ and CHOR are committed. I'm still getting just the bare layout I put up.
thanks!
Dhar
Apr 1 2007, 12:56 PM
Committed!
robcat2075
Apr 3 2007, 08:48 PM
This one's a
little bit long... it comes in two episodes.
2_01_56_take1commentsEpisode1MP4.mov2_01_56_take1commentsEpisode2MP4.movBe sure to update your SVN, I've committed a revised Chor.
Sorry this took so long, Dhar.
thanks!
Dhar
Apr 4 2007, 09:10 PM
Change never comes easy, but the resolve is there to learn this methodology.
I decided to start over. There were far too many keyframes to work with it. I also had to catch myself adding follow throughs and minor anticipations. I need to take a deep breath and slow down to allow this way of animating time to sink in.
Thanks for your tutorials Robert. They're tremendously helpful.
Jeetman
Apr 5 2007, 09:18 AM
QUOTE(Dhar @ Apr 5 2007, 01:10 AM)

Change never comes easy, but the resolve is there to learn this methodology.
I decided to start over. There were far too many keyframes to work with it. I also had to catch myself adding follow throughs and minor anticipations. I need to take a deep breath and slow down to allow this way of animating time to sink in.
Thanks for your tutorials Robert. They're tremendously helpful.
I know exactly what you mean Dhar. This is a somewhat new experience for me too but man I'm really finally learning pose to pose the way it's really done and so far I'm really liking it. Robert's video that shows the inbetweens is so inspiring to me.
I too have to fight of the urge to add the inbetweens. I had to add a couple to my shot I just committed because it didn't flow clear enough I thought to see the transition from the 2 poses. Other than that though, I'm staying as close to Robert's vision as I can.
George
Dhar
Apr 8 2007, 08:50 AM
"Slow and steady, steady and slow, that's the way I always go" Goofy.
Here's my next attempt.
Dhar
Apr 9 2007, 05:51 PM
Added some transitions and exaggerations.
robcat2075
Apr 9 2007, 08:05 PM
Here you go Dhar! Sorry This took so long.
This crit is brought to you by the word "simplify"

Simplifying the motion, simplifying the poses.
The challenge is that many times the simple-looking pose is not the one the rig presents to us as the easiest to get to. Those "simple" C curves sometimes require fine tuning more bones. The audience doesn't know how many bones we had to move to get a shape, but they do see the end result. So we have to get that end result in our minds eye and work towards that.
2_01_56_take4commentsMP4.movthnaksThanks, Dhar!
robcat2075
Apr 10 2007, 09:57 AM
Hi Dhar,
Here's the vid about the tinservant pose I promised. This is something I do a lot when i'm trying to find a pose or when I want to test an idea out. Once I've seen the shape in the mirror it's a lot easier to pose the character than if I'm only imagining what it might be. And I'll go back and check over and over. Was my elbow in or out? was my leg bent or straight? which way was my foot pointing? I don't have a photographic memory so I just keep gong back to the mirror and getting another look.
Neither you nor I have the exact proportions of the TinServant, so there's a limit to what this can get us, but except for the most mis-shapen beast this will get us a lot of good, pretty close answers quickly.
errata: At one point I say I "look at the Tinservant" but I really meant Tin Woodsman.
IMAG0015_TinServantPosing.ASF
Dhar
Apr 10 2007, 05:31 PM
Robert, which way do you want Tinservant bent to look at the swan? In your original comments, Tinservant's arc is parallel to TW's. In this latest video they're opposite.
Is it ) ) or ) (
robcat2075
Apr 10 2007, 07:33 PM
)(
robcat2075
Apr 11 2007, 07:47 AM
well, here's the story board drawing

and a frame from the pitch video
Click to view attachmentDid I do something later that made it seem different?
Dhar
Apr 11 2007, 06:09 PM
My mistake, I thought they had parallel arcs.
Here's the corrections. Tinservant's last anticipation to the final pose is a bit too exaggerated, but I'll leave it for now to see what your response will be.
robcat2075
Apr 13 2007, 11:50 PM
Ok, I didn't quite make the friday deadline... but here it is none-the-less
2_01_56_take03commentsMP4.movthanks, Dhar!
Dhar
Apr 14 2007, 07:20 PM
That much attention to detail can only come from many years of experience! Thank you, Robert, for your thorough explanation. This is what I've done so far, and I'm trying to apply the same priciples to Tinservant.
Dhar
Apr 15 2007, 08:46 PM
Pluggin' away at it.
Something is bothering with Tinservant's act, I can't put my finger on it.
robcat2075
Apr 15 2007, 09:27 PM
QUOTE(Dhar @ Apr 15 2007, 11:46 PM)

Pluggin' away at it.
Something is bothering with Tinservant's act, I can't put my finger on it.
Two things I notice immediately...
-The arc on Tinwoodsman's arm for "mow the lawn" has become rather triangular. It may be dipping too much.
-TinServant seems to be holding his tablet out too far.
-How about a "checkmark" motion rather than just a stab down?
-It may be necessary to point his right foot more right at the outset, to enable the last pose he goes to.
-the wide camera angle makes it seem like he's leaning too forward at about the midpoint. It may be necessary to un-lean him.
oops, that was more than two
Dhar
Apr 16 2007, 06:57 PM
Thanks Robert. How does this look ?
Tunames
Apr 17 2007, 05:31 AM
If I can comment on this I'd say that TW's movement in to "feed the giant swan" doesn't look right. As he starts his turn his head cocked way to the side and as he looks up he leads the motion with his chest his head should look up first then the chest should follow.......just my observations
Dhar
Apr 17 2007, 06:39 AM
Thank you for the critique, Steve. Everyone's welcomed to comment.
Dhar
Apr 17 2007, 06:15 PM
Tweaked TW to lead with the head for the swan, and Tinservant checks off the list instead of simply putiing a line thru it.
Quick Edit: Updated .mov since I replaced TW proxy with the final.
robcat2075
Apr 18 2007, 04:39 PM
QUOTE(Dhar @ Apr 17 2007, 09:15 PM)

Tweaked TW to lead with the head for the swan, and Tinservant checks off the list instead of simply putiing a line thru it.
Quick Edit: Updated .mov since I replaced TW proxy with the final.
Hi Dhar, did you ever try "scooping" into the second pose like I mentioned?
robcat2075
Apr 21 2007, 02:43 PM
We're Done! (With this shot) Dhar is the first to finish a shot for WOWG!
comments..
2_01_56_Take8CommentsMP4.movI couldnt' stop myself, I had one more idea...
2_01_56_Take8CommentsBMP4.movand the shot...
2_01_56-take.movThanks for your patience Dhar, good work!
martin
Apr 21 2007, 03:32 PM
Whoohoo!
HomeSlice
Apr 22 2007, 03:58 PM
Way to go Dhar!
robcat2075
Jun 24 2007, 08:44 PM
Hey Dhar, here some housekeeping for you... the dog leash!
replace the leash in your chor with
DogCollar_proxy2.mdl from the same prop folder and see if you can get it to behave well.
here's a test movie showing it working in a test.
Click to view attachmenti found I have to hide the "Handle" group before doing any simulating or the leash will behave very weirdly. I've also found I have to turn "Dynamic Options" OFF or it won't simulate at all. That's in V13. I'm told it's fixed in V14 so you'll find out which way that goes now.
thanks!
Dhar
Jun 25 2007, 06:04 PM
I looked at this model, Robert. How is it supposed to work?
Update your file, I had to save it as in v14.
robcat2075
Jun 25 2007, 06:19 PM
QUOTE(Dhar @ Jun 25 2007, 09:04 PM)

I looked at this model, Robert. How is it supposed to work?
Update your file, I had to save it as in v14.
Hmmm, I was hoping
you knew.
Here's my theory... somehow you use the preroll time from 00:00 to 04:04 to get the leash into the resting position we want it to be in at the start of the shot, sort of like in this storyboard drawing:
http://www.hash.com/forums/index.php?s=&showtopic=27249&view=findpost&p=226555 ://http://www.hash.com/forums/index.ph...t&p=226555 maybe have the TWM's arm (holding the leash) raised way up at 00:00 so ithe collar isn't touching the ground, then lower it into position and leave enough time for the leash to come to rest.
That's my theory.
Dhar
Jun 25 2007, 08:39 PM
Sorry Robert. I'm not very good at guessing what a rig is supposed to do. Maybe whoever built this model can clue me in as to how it's supposed to work? I don't even know how you did it in your test.
In the meantime, I added some facial animation to Tinservant.
2_01_56_take
robcat2075
Jun 25 2007, 09:12 PM
QUOTE(Dhar @ Jun 25 2007, 11:39 PM)

Sorry Robert. I'm not very good at guessing what a rig is supposed to do. Maybe whoever built this model can clue me in as to how it's supposed to work? I don't even know how you did it in your test.
I built it. The leash has a dynamic constraint on a chain of bones to make it floppy.
note this stuff again
QUOTE
i found I have to hide the "Handle" group before doing any simulating or the leash will behave very weirdly. I've also found I have to turn "Dynamic Options" OFF or it won't simulate at all. That's in V13. I'm told it's fixed in V14 so you'll find out which way that goes now.
put the leash in the chor. animate it moving around to test it. find the settings that make it correctly floppy.
here's the test prj
Click to view attachmentIt's pretty much trial and error in every new chor until you get it to work. You may be able to scout up some input from someone onthe forum who is more knowledgable about dynamic chains also.
Dhar
Jun 29 2007, 09:27 AM
I'm still experimenting with this. here's dynamic constraints with stiffness set at 20%
2_01_56_take
robcat2075
Jun 29 2007, 11:42 AM
Well, now we know that 20% isn't it.

Only 99 more to go!
the collar seems to be moving on it's own. It really should just hang from the leash in a casual fashion.
a thought: Make sure that none of the keyframes from the previous leash rig are still there.
Dhar
Jun 29 2007, 06:53 PM
I did this by hand.
Sorry Robert, I'm just not a programmer. I don't think I'm cut to experiment with dynamic cnstraints. They're too unpredictable for me.
Check the link in my last post and see if it works for you.
robcat2075
Jul 1 2007, 09:14 AM
The great thing about the drawing board is you can always go back to it.
Just think of the pride you'll feel when you finally nail this!
Dhar
Jul 1 2007, 03:21 PM
Back to the drawing board
robcat2075
Nov 17 2007, 04:31 PM
QUOTE(Dhar @ Jul 1 2007, 05:21 PM)

Back to the drawing board

Hey, Dhar... I haven't rustled up a new shot for you yet, but we still need to get the dogleash working on this shot, so lets finish this one up.
If you haven't figured the dynamic constraint out yet, the next step would be to find some one on the forum who can fill in the gaps and tell you what you are missing.
thanks!
mtpeak2
Nov 18 2007, 07:43 AM
Well, I took a stab at getting the dynamics to work, not sure if this is what you're looking for. The stiffness could be upped a little. Also the collision is working. Here's the results.
Dhar
Nov 18 2007, 03:43 PM
That looks great to me, but Robert's the one with the final call on this. Thanks for taking a stab at it, Mark - I'm all thumbs when it comes to technical aspects of the program.
robcat2075
Nov 18 2007, 06:49 PM
That's an improvement. is there a way to get the collar to not get yanked so high in the air? If this thing could just hang limply from Tinmans' hand that would be about all we need.
One thing i found in testing the leash long ago was that i'd get a much calmer simulation if i hid the part of the leash in the TMs hand. the 90° bend where he holds it seemed to cause alot of unneeded collision activity.
mtpeak2
Nov 18 2007, 07:52 PM
Is this better?
robcat2075
Nov 18 2007, 09:00 PM
that's even closer. What is causing that weird pop in the leash starting around the 30th frame, where the middle gets pulled into a sharp angle?
Is there a way to make the collar heavier? it moves a bit like a balloon on the end of a string would.
mtpeak2
Nov 19 2007, 03:02 PM
Well, I tried to add more mass to it, but now collision no longer works. So I tried redoing the whole setup from scratch with my original setting and collision detection still does not work. So I don't know what to do at this point.
robcat2075
Nov 19 2007, 03:17 PM
QUOTE(mtpeak2 @ Nov 19 2007, 05:02 PM)

Well, I tried to add more mass to it, but now collision no longer works. So I tried redoing the whole setup from scratch with my original setting and collision detection still does not work. So I don't know what to do at this point.
Doe this mean anything, where the settings are the opposite of what they should be...
http://www.hash.com/reports/view.php?id=4472
mtpeak2
Nov 19 2007, 03:36 PM
No. Martin straighten me out, I need to simulate from a front view. Yesterday I simulated from the camera view, which worked, but that has issues.
mtpeak2
Nov 19 2007, 05:13 PM
Well, I don't think this is much better and I don't know what else I can do.
robcat2075
Nov 25 2007, 05:44 PM
I'll try to take a look at it. Could you commit the last version?
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please
click here.