Bill_Y
-
Posts
260 -
Joined
-
Last visited
Content Type
Profiles
Forums
Events
Posts posted by Bill_Y
-
-
Some comments
(Do I sound like I have a cold? No, I sound like I have allergies)
I just wanted to say that this was an excellent critique with a lot of great information in it. As I watched this I kept thinking he should mention such and such, and then a few moments later he would. One thing that was done here that I think was particularly useful was the video taping Rob did of him self. I've found many times when I've hit a roadblock in how something isn't looking just the way I want it to that it helped to video tape the performance. This allows me to see exactly what my body is doing as I act out. Then, as Rob pointed out, I could spot missing actions that my body does, and add them into the performance. I think some people may be afraid of doing this as it feels like they are cheating in some way, but know that all professional character animators have either taped them selves, or watched another person act out a performance at one point in their careers. Great job Robert!
-
Just watched the latest version of this at lunch today, and wanted to say that the motion on the two characters looks realy good with the possible exception of the lip sync. It looks as if the charater is trying to make mouth movements on the exact frame that he makes each sound. Typicaly mouth movements lead sound by 2 frames at 24 fps because light moves faster than sound. Try moving the mouth movements back a few frames to alow the visual movement of the mouth to register with the viewer.
Another sugestion I would make on this is instead of trying to act out every sylable for the words, try empohasizing the key sylables. This will help smooth out out some of the words. For example the word, "today" just needs the OO shape and the AE shape to read. By simplifying the amount of movement shapes it will help sofeten the look of the lip sync. The T sound and the D sound are irelevent as they can be made from prety much any shape of mouth. Hope this helps.
-
I'm popping out of my little shell and checking some things out while we recover from siggraph. I watched through several of the animations for this shot, and I see the desired effect is a natural look. My number one observation is that though the current effect of his head snapping forward would indeed cause some ones hat to shift, I'm not sure that this movement it self feels natural to me. I'm trying to visualize any time that I would make a movement that sudden and sharp while talking. I'm guessing that the purpose of the shot is to have his oversized hat drop into his face as he talks. I personaly would try animating the shot as if the hat weren't going to drop into his face first, and realy capture his delivery. After that I would look for a moment in his movements where his head would tilt forward enough for the hat to just slip off his brow. To me the thought process area where he says, "Umm" is a point where I expect his head to make a movement like this. When this occurse gravity would kick in and pull the hat down over his face. This would probably cause him to switch from search mode to fix the anoying hat mode, and then drift back into search mode again. I hope this makes sence.
-
Wow, I just saw this hair topic for the first time, and must say your implamentation of the muhhair looks great.
As for grooming methods have you every herd of hair cells/cages? They are basicaly distortion cages that have hair that grows from one end to another. They are pretty cool because of how easily they sculpt lage quantities of hair.
-
he he well I soppose I might be able to swing in and offer some advice if people want it. if there is a spacific thing some one is having issues with, they could message me, and I'll take a look.
-
Uh oh... did I open a can of worms by posting.
We're busily working our butts off over here at the studio, but I do enjoy helping people if I can. How may I be of service? (Note: I'm saying that with an air of hesitancy as much of my time is very occupied currently) -
For example here: the Fabric_Denim material is used for the blue hat as well as the green jacket.
How spiffy! I always wondered if any one found my DT-Denim material to be usefull for something...
-
Wow, I can't believe that I've been out of the loop for so long thatI missed out on so much of this thread. I just wanted to say I think it's great you guys are doing this and that there is so much support for using it. Having use multi dimensional slider interfaces now on at least 7 or 8 projects it's easy to say there is no going back. I have made an adjustment to one of my beliefs though. Originally I felt that having more than 2 potential axes of movement would cause confusion, and for the most part I believe this to be the case, but when using a multi dimensional slider for eyes having the rotation of the slider control the dilation of the pupils is actually very useful. The main reasons for this is it eliminates another slider from cluttering up the interface, and it puts the eye dilation in a logical and easy to find place with out interfering with general intuitiveness of the interface. Anyhow, I've not had a chance to look at strohbehn's implementation yet, but hope to with in the next few weeks as perhaps I can offer some advice based off my experiences.
I like the name F.A.C.E. though in theory it could be used for other things as well. Something I didn't catch an answer to while quickly skimming this topic was the question to using dopesheets with these. The answer is if the multi dimensional sliders are set up correctly then you can control them with in another pose. Thus you make the poses with the controls set into the shapes you want, and let the dope sheet control these. Another problem I think some one asked, or if not might eventually is getting the interface camera to export with a model. If you drag the camera into another model it gets recognized as a regular bone, and thus you can make a pose that scales it with the interface for hiding the whole thing when it's not in use. If anyone has any other questions I can potential help a little.
For what it's worth at this point I've made some neat discoveries on using weights for facial influence that have some really interesting potential for the future, but I've not had the chance to test it fully yet. When I do, I'll post the project file here.
-
I've not commented on TWO since back in August, but thought I might try and offer a few sugestions. When I look at the mothers face in the concept drawing and look at the face in the modeled version I see two diferent characters. Niether are bad, it's just one has a diferent feel from the other. The concept drawing has a very Chuck Jones look to the face thats not comming through in the model. I'll start with the eyes. When I look at the concept drawing it uses what I like to call the bulgy pyramid eyes. By this I mean that the general shape of the eye is triangular. The base of the eye is defined the a joly cheek that is so plump and round that it creates a foundation for the eye to sit on, and helps indicate that this is a generaly jubulous character all the time. From there if we trace the corners of the eyes up towards the arch of the top of the eye you'll notice that it isn't perfectly rounded. There is a change in the slop from one side to the other and where they meet almost creates a peek right above the pupil. The only thing missing from the concept drawings and a Chuck Jones drawing would be the pointed lash off each of the eyes; anyhow, if we look at the modeled version a lot of this kind of character is lost as the eyes are much more rounded with with a baggy lower lid. Though this isn't nessicarily bad, it is diferent. I personaly don't know which style appeals more the individuals involved with the project, but it is something to be aware of.
The mouth area also suffers from a deviation from the concept art. In the concept art the nasolabial folds arrow at the corners of the mouth. The top edge of the folds curve in such a way that if we were to continue the flow of their lines, they would end right at the top of the nostral wings. When we look at the modeled version the nasolabial splines run below the nose. This can limit how far you can push the smile when it comes time to animate as it can only become so defined before it starts curving in a potentialy undesirable fashion. If you are not planning on going for extream smiles then this might not be a big deal, but it is something to be aware of.
The last thing I would point out is the jowels on the modeled version of the character. It seem as if the modeled version of the character was given bull dog jowels that are very comon in some characterizations of old people. Again though this is a perfectly acceptable look for an old person, the concept art shows a much higher cheeked individual than the model presents. Both work, but are diferent.
-
To close out this topic for those who might only read this forum, Duel was released Today. Here is the official announcement.
-
Here is another sample of how things are coming together.
-
Thought people might be interested in how the trees are going to look animated. This is a test of wind effecting a tree.
-
Thought I'd show one more image of the temple area. Again it's not finished, but will give you a basic idea.
-
And just so people know that is a 3D set.
-
I thought people might like to have a look at the scenic island where the Duel transpires. This is missing some of the vegetation and other such nurnies, but should give you an idea of what the setting is going to be.
-
Ahh the picture helps, and indeed the problem is a splinemanship issue. You are using dangling splines to close spots that should be 5 point patches. Near the nose of the character run the spline from the forehead down to the one thats coming up the edge of the nose. Here is a suggestion. When linking splines together look at any given CP and ask your self, "Are there only 2 splines crossing here?" If you answer, "no" then 9 times out of 10 that is a bad modeling choice, and should be reworked. There are 3 such problems on each side of your current face that I can spot just from the image.
-
wax
in (2003-2004)
Well technically translucency with depth occlusions would do this with out having to need real subsurface scattering; however, we don't have depth occlusion for translucency in A:M currently. So to achieve this effect you might try using a spherical combiner that adjust the ambiance of a surface staring from the wicks position as being the most "self illuminated" or ambient and then gradient it to complete occlusion or no ambiance.
-
I feel so.... special....
I think Morgan does as well...
-
I couldn't watch this one. The link says the account with the video needs to be upgraded to be watched.
-
I'll have to bite and ask this question..... why are you guys doing this.... this way?
Why would pose sliders not handle all of this...
seems like adding an extra layer of controls.
What is the bottom line advantage?
Mike Fitz
www.3dartz.com
That is a very good question Mike, and here is the answer: Simplicity. Can you do this with just pose sliders: yes! However, when it comes to animating a face an intuitive multi directional interface can speed up your workflow exponentially over traditional sliders. Fewer controls that do more actual make posing the face much faster and gives better feedback. I know for a fact that NO ONE at our studio will go back to traditional sliders again unless we are forced to. That is how much this kind of interface improves workflow.
-
Must be something wrong with my AM install, something like that would have made it a pretty pointless suggestion.
Ok so you constrained the controls to the main camera. Now what happens when you add 4 or more characters to the scene to animate? Are you going to position each characters controls in different locations of the main camera? Also the control rig for this was a simple one, but when you add in all the controls for a full face the realestate for the controls is going to obscure more and more of the camera. The idea of having the second control camera was to eliminate the interface interfering with the main camera or any other view you may choose to animate from and to operate more like a slider interface window; however, all these things can be over come if you real wanted to have the sliders in the main camera view. As I originally said I personally wouldn't do it.
-
I personally wouldn't advise using a single camera for both the controls and the character. The main reason is that you almost always want to animate the face relative to the camera. It's one thing to do the initial rough of the lipsync from a face cam, but ultimately if the sync looks funny from the camera, then you're probably going to want to tweak it. Now by making the controls viewable in a single view I would suspect the view would primarily be the camera. If this is the case then the controls will grow, shrink and twist away from the camera as the character moves. having a dedicated control window allows you so view the character from any angle and still have full acess to it's controls.
-Bill Y.
-
This might sound silly, but because you said the pug-ins started working in 11.1 when you copied them to the 11.0t plug-in folder I suspected the 11.1 options for the plug-in folder might have been pointing at the old 11.0t's folder. So as a test I checked on my system and strangely enough the plug-ins folder for my 11.1 is pointing at the 11.0t's directory. This is probably why people are having the problem.
-
Awww... Time to add in my 2 cents. The big advantage to working in a visim fashion over phonemes is the speed in wich you can convey emotion and speech. With phonemes you use several sliders to vary the degree that the mouth moves into specific and often limited shapes. Yes thes shapes can say the words, but often the words are sterile because there is no good way to add emotion too these shapes with out adding more sliders on top of what you already have. With this approach you can soon end up with dozens of sliders and counter sliders to convey emotion as a character talks. Beyond wich you have to make sure that as thes shapes blend together they don't cause the mouth to overshoot your original lip sync as they are conveying the emotion. In other words this can be a real pain.
Lets look at this from a more technical stand point. with an "O" sound the mouth opens in an up down fashion, with an "EE" sound the mouth opens in an up down fashion as well. If we use sliders to convey the "oy" sound in the word boy we can see some ridiculous overlap. Basically as the "O" slider lowers the lips the "E" slider moves the lips back up. So long as both of these happen at exactly the same time the transition should be relatively smooth; however, We're now juggling 2 sets of key frames that have to be alined just to deal with opening and closing the mouth properly. If one of the sliders keys gets offset even by a tiny amount it can cause the blending to look odd. Now lets toss a sneer on top of this wich pulls the lips up even further. We're up to 3 sets of keys that are all effecting the up down movement of the lips! This is ridiculous because the more sliders that effect that upper lip the more potential there is for things to go wacky due to a missing or misplaced key, or even worse having one of the sliders overdrive the lips into a very ugly pose requiring more key frames to try and compensate for the problem.
With visimes the "O" and "EE" mouth open slider are one and the same. Thus the keys can't get offset or over driven from the other sound. True when we add a sneer on top of this it is done with a separate slider, but the great thing is we are only dealing with 2 sliders instead of 3 or more (depending on potential counter sliders). Because of this simplification of the face sliders it is then easy to create multidimensional controls to even further simplify the process. One control can handle both the opening and closing of the mouth and the width of the mouth at the same time. This reduces 2 controls for the mouth down to one for even less hassle while animating. If you want to test out a simple visimes interface in A:M you can down load a project I made a while back and see how easy and fun it is to control. You can find the link and peoples opinions of it here: Visimes interface. You can also try doing a little lipsync with it.
-Bill Y.
www.anzovin.com
1_05_12 Tinwoodsman
in TWO Animating
Posted
I'm sorry if I'm a bit late discussing this shot, but I would like to make a suggestion. As I watched this shot my eye was drawn to Scarecrow moving out of the shot instead of Tin mans performance. I'm not sure if this upstaging is the goal of the shot or not; however, if the performance is supposed to be about people feeling sorry for the Tin man, then I would suggest freezing the Scarecrows position on screen after he oils him. Let him listen with compassion to his dear old friend. By turning and walking away the Scarecrow looks like he got the Tin man think about his love as a ploy to sneak off.