I read a quote somewhere that said-
"One may walk even over the highest mountain one step at a time", so clutching a honey sandwich and wearing my old battered hat and best pair of boots, I shall embark on my own personal epic.

I will post all my tests as they happen, so all can point and laugh mockingly. I would obviously be gratefull for any scraps of direction, hints of promise and morsels of critiques that may filter down as I stumble along.

This is the first sequence - hereafter known as 02_07_65a.
I blocked it out roughly, pose to pose, using holds only. I quite like the popping feel to it - it seems to suit Ku Klips mood for me. I would quite like to make him "pop up" between the characters as he cuts fills and glues.
Of course - I could be going in the wrong direction, but I will depend on the forum to reset my compass.

In this sequence I think he shouldbe brandishing a pair of tin snips, Does anyone know if there is such a prop?

Thanks all,
Chris

Look in 2_07_064. Also, that's the location he starts his song in.

Thanks Martin, I shall get right to it. See already I am wandering off the path

The first test should have started in this position.

Is KK going to be doing some jumps in this song? Haved you thought about the style of the Oompa Loompas in the original Willy Wonka? ie the song is task oriented.

Cheers

Yes David,
As long as you all think its OK
I Want him to moove in a snappy sort of way when he's speaking, then dance like and smooth on the little chorus parts.
I'll look on you tube see if I can get some Oompa Loompas

I like your poses and I like the snap in the motion you are doing a great job go for it and keep posting!

Right, starting again-
this time saving the actions so I wont overwrite them by accident.
I have divided the whole song into sequences, and blocked them out with poses - very roughly.
The first sequence is "A little bit of snip, snip"
I will sort out the shears as an action object, giving him a snip, snip action to matchthe words.

Thanks Mark,
I know this is really rough, but Martin asked me to keep posting as I progree(and I havn't done much of that lately).
I wanted to keep it all to myself until I at least had some recogniseable animation to crit.
For now, I am going to post on until the whole sequence is blocked out.

Last one for tonight, more tomorrow.
I want to include some other props in this sequence - A glue pot - something he looks at, that sort of thing. Is this alright?

I can see this has lots of changes to be made, but heres the first for today:-
"A little glue to make it fit"

I shall be using Jason Ospias system rather than the phenome dicttionary, andadding the lip-synch as an action file over the top, so I guess theres no reason not to start that as well, since I intend to animate the "head up" seperately to the body.

Just wondering if I should post my "workings-out" for the lip synch - I mean show how I analise the sentance, then the stages involved in producing it - a sort of work through example with a clip for each bit.
This is the next bit - Please stop me if I'me rambling on(and on and on.
Also - check out this Oompa Loompa skit!

http://www.youtube.com/watch?v=bosmWANdw6o...ted&search=

And the next- f- Thats why they call me Klu -Klip

Thought we could have some chicken dance in the music bit.

With the lip-synch - I use Jason Ospias method, so checking out the first line of
"A little little snip, snip" I divide it into mouth closed and open, and mouth shapes or visimes.
A (visimes)> becomes a combination of EH and EE, shortened to EE or (wide)
A(open /close) > becomes closed, open, semi-open.

Little(visimes)>are IH-D-L>shortened to L-IH-L(narrower, wide,narrower)
Little(open /close)>semi-open. open, semi-open
Snip(visimes)>S-N-IH-P > IH-IH-IH-P > IH-P(Wider, narrower)
Snip(open /close) Semi-open, Open, Close

Before animating each phrase I key all the mouth shapes I am going to use, then animate the Visimes, then do the open close.

It takes longer to type this than to animate it!
I will post when its rendered.

Course it doesnt look too good yet - but I spent less than 3 minitues animating it and moe time typing the last post.
Heres just the wide/narrow pass, then the wide narrow/open close.
Next its time to define the mouth shapes better, and sort out the timing. I am actually doing this for all the lines of speech at the same time, but I wont post them at this stage. But the point is that its starting to take shape - from a distance in 3 mins- leaving plenty of time for finnesse
Chris

So now I've finished doing open/shut and wide narrow( shame those two wernt one pose) all through the sequence. Now its time to define the mouth poses a little better.
At least its starting to look like proper animation now, althoughI feel like he's shouting. Once I've done that I'll start on the tilts and noddy shots.

Still havnt finished the blocking out of the sequence yet, but I have noticed I was keying on the sentances rather than the beat of the tune. I hope this wont be a problem later on.........

Erm....they are on the FACE interface. And erm again.....that's not Ku Klip's voice singing.

Not his voice?
OOPS!
Can someone point me in the right direction - Quick!
I know its not in songs - I can only find a wma version of the tune in my data file.

Yes Ken, sorry they are on the face interface, but I was just using the pose sliders at the time

I think there is no recording of that song with Ku-Klip voice. The situation is that Ku-Klip does not sing the song. He is only acting and doing some rhytmic choreographied sort of dance during the song. At least that is the way I see it.

Actually, we plan to get the voice actor to record it, but you can work with what you have, Chris. We'll have the voice actor match it.

Well, that is good news.

Chris, I'm really interested in reading your account of your use of Osipa approach of open - closed, narrow - wide visimes. Please continue to report on your progress, battles and learning.

Thanks for replying guys.
I will just press on with what I have then. We can always fine tune it later.
More after breakfast!

Looong breakfast.
during which I got to grips with the FACE interface.
I hope the sound track gets sorted out.

Working from start to end, I started refining the shapes for the mouth, curling the lips in for mbp and out for oo that sort of thing. Just roughly whacking them in there. Also moving the lower and upper lip down and up. The FACE interface makes this very easy. The upper lip looks very flat in this to me.

Once this was done I started adding head movement to emphasise the accent or volume of the speech. When Ku-Klip says "a little BRAISE and SOLDER WITH" I thought BRAISE and SOLDER had more emphasis, so moved the head up and down correspondingly.
When I added another for WITH, he started looking like a noddy dog, so less is more with this I think. WHen this is added to the body movement I have to make sure it matches, but its nice to start getting some life in.

Am I the only one to think that KuKlips nose is very smooth? Is ther a bump map missing?

Not too bad. I would say that the mouth shouldn't open that much....maybe half as much. And I'm not really seeing B or Sss sounds in the mouth shapes. Particularly the top lip isn't moving up or down. In face, neither is the lower lip much.....so the teeth aren't being exposed at all.

Thanks Ken - I had already noticed that he seemed like he was shouting, and that top lip did look very flat, even with the moustache in place.
I think thats the beauty of Osipa's system, Get a lot of the basic stuff done quick, and then correct and refine later.

I shall do the eyelids next, first the height, then the blinks.

The last lipsync looks quite convincing.

I see that you are putting the head motions into the lipsync action. That is not a good idea IMO. Adding the facial expression into the lipsync makes sense but head motion is so much linked to the rest of the body motion, the environment and the camera POV that you may find it hard to synchronize the head motion with the body motion if one is done in an action while the other is done in choreography once you get there. Consider those:

1) The amplitude of head motion (as well as body motion) is directly dependent on the closeness of the shot. Close shots requires very subtle head motions while long shots requires much larger head (and body) motions. And then medium shot, such as the one you use in your takes, requires medium head motions. Unless you plan to use the same medium shot throughout the whole sequence, you will need different head motion amplitudes from scene to scene.

2) While doing his singning and dancing, you might want him to look at different objects of characters around him. Trying to do that in a separate action can be difficult. And adding further head motions to look around over the head motions in the action file can be a fight.

3) In order to have varieties in the whole sequence, you will want to have either Ku-Klip to move and turn around or the camera to take different point of vues or move around as well or even a combination of Ku-Klip and camera motions. Some head motions may read well from one angle but not so well from another angle.

Thanks Yves.
I thought I would be able to "go back" and correct the head motion once I start blending the two actions together, but at the very least it now seems to add extra work, just to be re-done later. On that basis I will probably delete the head motion and leave it to do in the body animation.

My one concern is that I was going to use the eyebrows and eyes to anticipate the head motion. I suuppose since I have already done the noddy pass, I could leave off deleting it until the end.

I have added a very widening of the eyes just before a head lift eg for "with" and "glue", then after added a blink or two.

This is from my limited experience in animating yet. But there are a few things I learned.

I know that putting the head motion in the lip sync is usually not a good idea unless the shot is already well planned and you already have a very good idea of the shot angle and closeness.

As for eyebrows motions, a good deal of eyebrow motions comes from the expression of the emotion. So eyebrows would fit well along with the lip sync since the emotion and the talking (or signing) goes well together. There is, however, some eyebrow motions that are related to anticipation of eyes motions and also to the gazing direction which themselves are linked to the scene. What I found work well is to overload the eyebrow motions in that case. That is do the emotion related eyebrow motions along with the lipsync but add the eyebrow motions related to the anticipation and gazing direction, and thus the environment in the choreography

As for eyes, there are two types of eye motions. 1) eyes motions that are directed by the environment, like when the character is looking at something and 2) eye motions that are related to internal thinking like when the character is recalling something, someone or some event or just thinking or lost in thoughts. Internal thinking eye motions can (and should IMO) be done in the lipsync action. But environment directed eye motions should be done in the choreography. And for that, I found that using the global eye target works the best for me. Not the eye target bone but the eye target null.

My interpretation of what Osipa says on head tilt is a bit vague, so I re-read some more. He says to do head-tilt second after synch,the order being:-
1 Sync
2 Head tilt
3 Eyes
4 Eyelids
5 Brows
6 Finnesse
......but with a catch as to what second means - in that with full body acting, do that first, then the head tilt.

Currently I have worked through the song, completing(more or less) steps 1-4,

So.......
should I post up the rest of the sync without steps 2-6 (by deleting the unwanted stages 2-4) so people can crit it,
Then go back to the body animation and complete that, finishing up stages 2-6
Or......
finish the facial(1-5) stuff first? I think the former may be the best option. I will try to take on board everything you say about the eyebrow motions and eyes.

Chris

You can post at whatever stage you want. It's just easier to do 2-6 in a chor where you have a context (camera and props). I doubt Osipa is using actions at all.

I agree with Ken. Post whatever you have whenever you feel.

I have the feeling that Osipa, even though he is an A:M user, does not bother with doing the lipsync in a separate action. His workflow is like he is doing everything in the same module which for A:M is the choreography. We could also do the same thing but A:M allow us to separate the steps and I find that immensely helpfull.

I havn't spent long on Two this week,being a bit tied up in other things, but I promised Martin that I would post.

I love the way the bones change to blue in keys on 14. I think that makes animating so much easier.
I think I skewed things up in my transfer of files today, but I will be able to sort it out.

Thanks for the helpfull comments, Ken and Yves.
Since as far as I know there is no shot planning yet for this scene we will have to pick it up as it goes.
For now this is a pass one step up from blocking out, so I know its not much good - basicly just moving the hips and limbs a bit
What I really want to know is if if I can use this sort of movement for KuKlip.
I read your post from the song thread, Yves..........
" I felt the quick tempo fitted more a ballerina than Ku-Klip. Ku-Klip, being large would not move that fast. "
.......and thought well what if he could?

Also seem to be having some sound distortion.

I think the sound problem was due to having a wav file in both the body action file and the lipsynch action file.
OOps

Chris

Things havnt gone too well for me regarding output over the last 2 weeks. I lost the ability to render as mov files, then the sound went all screwy then finaly my PC gave up the ghost. Still I'me back, and I intend to move on, hopefully with daily WIPs .
This is the first

Not too shabby, Chris. Looks pretty good Nice flowing movements.

The shoulders need animating though. They look stiff.

Thanks Dhar, I missed that.I will add some movement in that department today
C

Todays update:-
More shoulder movement and some cleaning up.
I added some head movement on the snip snip to emphasise it. Also I wanted to see it with the hair, so we got that as well.

Almost ashamed to put this WIP up.But a daily update it is.
Although the animation is nowhere near finished, I am starting to think that this will be close to how I envisioned it. I want KuKlip dancing around. interacting with Woot, so I feel I ought to include Woot in the next WIP as well, because its about to get a little more "intimate" .

I love where you're going with this. Can't wait to see more.

A disapointing end to my promised update. I started a "test render"this morning of KuKlip and Woot together in the workshop, but with 40 hrs still to go I am afraid I stopped the clock. I cant believe it took that long!
I'me rendering a quick take of the continued action file, wich if its ready before bed I will post.
C

A shaded render is taking 40 hrs to render? Maybe you should turn off hair. That would speed things up.

Hi Homeslice,
I tried to do a final quality render in the scene cho. It only had Kuklip and Woot there along with the workshop and I was very suprised it would take that long.
[System Description: 2*P4 3.2GHz 4GBRam 2*Nvidia GForce 7800 XP Ho,e SP2
P 4 1.8Ghz ,384mb ram, ]

For som reason it was only using 50% on the ram. I have no idea why it should take this long, it is probably my pc at fault.

Off to bed now, managed to squeeze in another few secs though.
Tomorrow- less gardening, more animating"

Todays update,
As usual not finished but I wasnted to start on Woots reaction.
C

I worked out that its the high details on the models themselves that takes the time rendering(.....DOH!)
So I've switched to using proxy Woot in the rendering - I was already using him in the action file.
Unfortuneately there is no(I think) proxy KuKlip, so I've just saved him as a proxy, and deleted
the decals and materials. HAIR stays off for now. There are some sound problems with this render as well, but its just a test, so I'me ignoring it.
At least this way I can see some renders in a reasonable time.
This is 03.11 to 08:00, Its a "Take 1", so its the first time I've seen it, with no corrections as
yet - quantity as opposed to quality.I think it would have been better if I had started like this in the first place!

Chris

I worked out that its the high details on the models themselves that takes the time rendering(.....DOH!)
So I've switched to using proxy Woot in the rendering - I was already using him in the action file.
Unfortuneately there is no(I think) proxy KuKlip, so I've just saved him as a proxy, and deleted
the decals and materials. HAIR stays off for now. There are some sound problems with this render as well, but its just a test, so I'me ignoring it.
At least this way I can see some renders in a reasonable time.
This is 03.11 to 08:00, Its a "Take 1", so its the first time I've seen it, with no corrections as
yet - quantity as opposed to quality.I think it would have been better if I had started like this in the first place!

Chris

OK
08:00 to 15:00
The action here doesnt match up with what I had in mind - but still take 1 and all that.
Its my intention to have him reach off screen for a sort of braising torch, swapping it for the snips. Then he puts it back and produces a pot of glue, sticking his hands slightly together.

C

edit Made this update too large! I will split it and repost

Hey, we record the real actor singing the song tomorrow!

Excellent News!

Click to view attachmentI'me still getting sound distortion problems on rendering out a .mov file, so the sound on this is a bit screwy. I had thought it was using the same sound file in three instances(Woots reaction, KuKlips motion and his lipsynch) but since I have deleted them this doesnt seem the case.

A shorter update tonight - one thing, the jump and spin- The timing will be on twos rather than 4s where it is now. I just had to stop early.

Messed up earlier updates by making last nights too large, so I have re-posted it. Looking forward to cleaning this up now.

Yesterdays update turned into a nightmare. The action file is OK, but when I tried to render I got into all sorts of problems.
The first time I tried- this is the same settings each time - It was going to take 8 mins. After rendering out I had 2 soundtracks, all mixed in a weird way. So I put it on again - 27 mins and it froze near the end. So I rendered it into tgas - still froze.Then I checked the settings - seemed to be ok, then rendered again.
This time got a message " Quicktime may not be installed". Well I knew it was, but did it work? No.
Then followed 2 hrs of trying to install an older version of QT - I had no idea that "You cant install this version of quicktime since you have a newer version on here already" would get so annoying.
Sodelete Quicktime, Itunes(how didthat get on here?, Ipod and anything else with an I in it, THEN go to system and delete the sneaky hidden bits, then clean the registry, then re-install older version......
AND IT STILL WONT WORK. but at least I can render files, even if the sound is rubbish...... OH,now the sound scrubbing feature on AMs stopped to.


This is the last update till tuesday. I am off to Brussels in Bel for the weekend!

Here it is, Chris.

Final Ku-Klip Song

Thanks Martin, just got back from Brussels today after a nice romantic break. Didn't get out much, but I did manage to check out this place- the world comic strip museum:-

http://www.awn.com/mag/issue4.03/4.03pages...nsbrussels.php3

Set in a beutifull art nouveau building it houses an outstanding collection of Comic Strips and BD, my personal favourate. Nothing by Stan Lee though!.

Nose to the grindstone tomorrow,
Chris