JBarrett
-
Posts
267 -
Joined
-
Last visited
Content Type
Profiles
Forums
Events
Posts posted by JBarrett
-
-
Many moons ago, there was a discussion on another mailing list about how to generate ideas for animation tests and whatnot. One list member posted a solution that involved randomly pulling from lists of words, which led me to create some Excel spreadsheets that I called the "Idea Generators."
Recently I decided to push the idea farther and break away from the Excel barrier. The end result is what I'm calling Brainstorm Online!
http://www.justinanimator.com/brainstorm.php
In short, the word lists are now an online database, and anyone can contribute to the lists. I've started populating the lists with some sample words, but with others contributing as well, my hope is to get hundreds of words in each list, which would mean a lot more variety in the results. This new version also allows for some words to trigger longer phrases, creating another level of variety and complexity.
I encourage everyone to stop by, check it out, and add to the word lists. While some of the results are pure nonsense, you never know when the right combination will spark an idea for a great project.
Cheers!
-
Wow...I'm starting to get....chills.....and stuff.......
-
Correct. There must be a key on all sliders at frame 0. Your earlier reference to 0% was a bit misleading.
-
You're right on the transition idea...the speed of the transition will be driven by the speed of the sound changes in the audio.
Re: what to do on frame 0, don't key them all a 0%. Just set the sliders so that the mouth looks appropriate for what's happening in the scene...which means you need to have a solid grasp of just what's happening in the scene. Is the character relaxed? Create a relaxed mouth. Can you hear in the audio that the character's lips are together, and then they open them at the beginning of their first spoken word? Then match the mouth accordingly.
There really is no generic pose you can use every time when beginning lip sync. It all depends on the emotional state of the speaker and what they're doing in the scene.
-
This is largely irrelevant, but my eyes skimmed the thread subject and it came to through to my mind as being about "exporting a shade at 7am"....
I need to go home...
-
Nice work on the update, Sam! Definitely lookin' better.
On a specific note, I like what you've done so far with the head tilt toward the end, but I'm not sure my original comment was too clear. It looks like he goes into a quick tilt for some parts, but then comes out of it back into his vertical orientation again. What I meant in my earlier notes is a level of tilting that continues through an entire pose, not just for a small portion of it. The angle you've got on the initial part of "dead right" looks good, but it still needs to be carried through to the end of that phrase for the full impact to set in.
Keep at it!
-
Whoops...forgot that the other reference I made to Keith's tutes was elsewhere.
Anyhoo, Keith Lango has some nifty new tutorials up on his site, some of which could help with certain issues in this clip.
-
Looks pretty good, Sam! Here are some things that could still be punched up a notch...
- The overall body poses don't really sell his tension as well as they could. The line of action is very vertical all the way through, even when he blows up and yells.
- Lip sync. Overall, it's not bad. Some shapes aren't as strong as they should be, though. Some examples...the "r" sound in "get in here" needs a stronger purse of the lips. The jaw should close a bit more for the "t" at the end of "I know that," and his relaxed mouth right after that, before the next phrase begins, is open a bit too much. It doesn't give the jaw any place to go for the first sound on the next word. Nice work on closing his lips during that frustrated part, but they need to open a little sooner...listen really closely to the audio and you can hear when the lips separate again.
- When he looks down, lower the upper lids more. A lot of the time they're still pretty far open, and it feels kinda odd. Try doing that yourself...at least for me, I find that even when trying to force my eyes "wide open," the upper lids are at about half-mast when I look down an extreme amount.
- Try adding some side-to-side tilt to his head on some poses. This would be especially helpful at the end when he goes ballistic, to contrast with his attempt to rationally explain his point to his audience in the earlier part of the clip. The head pose on "dead right" seems to be an ideal place for some tilt to really help sell his frustration.
- His brows could be used more effectively to sell all that inner turmoil he's feeling. Keeping them so generically wide doesn't say much about his emotional state. I picture a bit of a sad tilt to them in the earlier section, then as *his* patience is getting tried while he talks to them about trying *their* patience, start to introduce some downward poses in there. He's fighting within himself to keep from blowing up. Eventually it doesn't work, and he belts out his final line, where the brows should really dig in to sell that anger and frustration he's feeling as he tries to make his point.
- Even though the feet are off-screen, don't totally ignore them. True, you don't need to go whole-hog and key foot and heel rolls, but there should still be some feeling that the feet are planted on solid ground when they're supposed to be planted on solid ground. For the bulk of the clip it feels like he's floating, including his feet. You can just tell by looking at the legs that there's no bend happening during all those accents and steps.
- After his "I know that" accent, it feels like the hips and shoulders are going in opposite directions...the hips slightly left, the shoulders slightly right...leaving him in a bit of an odd stance.
- The arms feel a tad stiff. He's tense, so they should be stiff a little, but there's something un-tense-ly stiff about them that's hard to put my finger on.
- The completely-straight fingers during his "crazy" swing are a problem. Part of the problem is also the hand's angle to the camera. It could be turned a bit more so that the back of the hand is facing the camera. Another problem is the nearly-horizontal position of the forearm. Get the wrist higher, with the arm at more of an angle, and do something more dramatic with the fingers.
- Another tutorial to check out on Keith's site is his recent blurb on forces, which will help another issue with that "crazy" swing. The hips and arm both come out of their extreme poses at the same time, so the power behind the move doesn't feel quite right. If the hips are the driving force, then the arm should have a little delay to it.
That should keep you busy for a while.
- The overall body poses don't really sell his tension as well as they could. The line of action is very vertical all the way through, even when he blows up and yells.
-
Looks like a cool beginning, Sam! I love how you incorporated some of Jimmy Stewart's mannerisms into this guy.
The lip sync is pretty good, but could still use some polish in a few places. The last line especially feels like it needs more work.
Keep at it!
-
These are the things that stuck out to me...
- His nose/snout feels just a hair too large...not much, but enough to feel "off"
- His mouth also feels a bit big, partly due to the snout size issue, but also partly due to the angle. Looking at some Eeyore pics online, his mouth has more of a downward curve to it, where yours is more flat and leans more toward a horizontal orientation
- The tail needs to be wider, thicker, and more "blocky," like it's cut out of thick cloth.
- On a similar note, the head of the nail/tack holding his tail on needs to be a bit larger
Here's a pic I found that illustrates most of these items fairly clearly
http://members.lycos.nl/yoarra/color/coloring3/eeyore.gif
Other than that, it looks really good so far. Keep at it!
- His nose/snout feels just a hair too large...not much, but enough to feel "off"
-
Unlike Justin I'm partial to a less-is-more approach to lip-synch. The human mouth is lazy and will blend shapes and omit shapes wherever possible.
Actually, Victor, I think we both work along the same lines. Pretty much everything you said -- from the impact of attitude and context, to arcs on the mouth corners, to blending shapes -- is a part of the way I work, and I tried to make those points on the "Animate a Face" material as well. Apparently I'm not as successful in print, though...
-
I've seen lots of this kind of work -- even in high-quality productions like Pixar films -- and for me it leaves something to be desired. It's like they polished all the rest of the animation, but left the mouth in a barely-polished pop-through stage.
Because I live in a french part of the world, I'd like to share another POV.
Could it be that some latitude in the mouth shapes is left for later translations? Pixar films are being translated to several different languages after all. And by tweaking lip-sync loose enough, the translation is further simplified later on.
I'm not saying that all Pixar animators do that. The quality varies from shot to shot. There are some animators who work in this hit-the-peaks-only style, and others who go for a more polished approach. If you watch their films closely, you can see the difference between animators.
But you have a good point. The only thing that I'm not certain about is how much Pixar (or any animation producer) actually re-animates the lip sync of their characters for foreign releases. I know that often they'll get lucky with their DVD releases and have a non-English audio track that matches the English lip sync pretty closely in certain areas, but some of that comes from similarities between languages. However, if you mute the sound and just watch the mouth, it's definitely the English lip sync.
Let's see if we can get Victor in on this conversation for some more solid inside information...
-
An even more specific one: Exactly how is "ng" formed on the lips and by the jaw? It isn't. It's a sound whose appearence depends on the sounds surrounding it. Why animate that as a specific phoneme?
You're right, it wouldn't be represented by a specific shape on the lips and jaw. And that actually goes right along with my preferred way of animating, which is to keep animation controls for the lips, jaw, and tongue completely independent of each other. The "build a full mouth target for specific sounds"-method of facial setup is very limiting and cumbersome IMO, and whenever I have any input on the facial rig of a character, I strongly discourage that path.
To clarify, by "animating all sounds" I'm not talking about animating all parts of the body that are involved in the creation of sound (you left out lungs, BTW
). I stick with the three most visible elements -- jaw, lips, and tongue. When I animate, I try to get these items as complete and accurate as I can based on what I know of real speech and the way our mouths move, which includes all the blending that goes on for sounds that aren't as enunciated or articulate as their neighbors.Most often when I hear people talking about skipping shapes for lip sync, they're talking about hitting only the most clear/accented sounds with accurate shapes, and sliding between these key shapes without any care for the sounds in the middle (and I apologize if I mis-interpreted your comments as being along this line). I've seen lots of this kind of work -- even in high-quality productions like Pixar films -- and for me it leaves something to be desired. It's like they polished all the rest of the animation, but left the mouth in a barely-polished pop-through stage.
Call me crazy, but for me it's important to believe that the character is actually saying what they're allegedly saying, and that means having those three key mouth elements -- jaw, lips, and tongue -- move through all the appropriate contortions to accurately represent the sounds being created as the character speaks. If the accuracy is only 50% there 'cause animator only hit the primary accents and extremes, it's very distracting, and makes it hard to believe that the character is truly alive.
The effort expended toward creating the illusion of life in an animated character shouldn't take a back seat once the body motion is done and it's time to animate the character's lip sync. The movement of the mouth has the potential to contribute to the life of the character if animated with care and attention, or to detract from it if it's neglected. I'm just trying to encourage the care-and-attention route.
-
Do NOT animate every sound you hear just the ones that matter. What matters? There's the art to it... learn this by thinking, trying, experiencing, learning, doing again.
Hit the initial poses to phrases a little BEFORE the sound never on it. In real life we're used to seeing people form phonemes without sound and then the sound begins to come out. It's how people talk all the time. In fact with some strong phonemes we're quite used to seeing them form on the lips while an entirely different sound is being uttered. Like in "I love dogs with big ears." The "with" "w" will be formed on the lips solidly AS the person is saying "zzz" at the end of "dogs."
G'luck.
FWIW, you can animate every sound you hear, and it will look just fine if you know how to do it properly. When we speak, our mouth forms every single sound. It's the way that the sounds are formed that is the tricky part to recreate when it comes to lip sync.
I've frequently heard of the suggestion to skip shapes when doing lip sync, but I'm personally against it. When you skip shapes, the end result will look pretty good, but it will look like you skipped shapes...it will feel like something's missing. When we speak, we don't skip anything. Instead of skipping shapes, take a little time to observe speech very closely and learn about how the mouth shape blends between the creation of different sounds. The art of accurate lip sync is all in the blending. The final product will be much better as a result, and your animation skills will have grown in the process.
-
Take a well-deserved rest, Sam. Really sweet work on this little clip!
3...
2...
1...
WOOOOOOHOOOOOOOOOOO!
-
Looks pretty cool! Can't wait to see this bugger MOVE!
-
-
Looking better, esp. on the hip movement. A few more notes...
- The two holds on his "surprise" reaction are still very stiff. Loosen/soften them up a touch.
- During his opening walk, his left arm looks like it's going through the left leg.
- Speaking of arms, the looseness of the arms during the opening walk is getting better. Some other parts still have the arms feeling too stiff, though.
- Another arm note. As he's cautiously advancing toward the gap, his right arm is getting lost behind his body, with only the fingers showing up below his butt a few times. Try cheating the arm a little more into view for a little better silhouette during that segment.
- As that right arm comes back into view before he leans over the edge, the wrist and fingers snap into position a bit too solidly. Ease into those positions more smoothly.
This is really turning into a sharp piece, Sam. Add another "Woohoo!" to your collection.
- The two holds on his "surprise" reaction are still very stiff. Loosen/soften them up a touch.
-
VERY nice work! Really fun little clip there. As with many others, one of my favorite bits is his run in the distance. Really well executed! Woohoo!
-
The odd thing is that Williams' "belt line" comment contradicts the illustration he made just one page before (the weight shift example on page 146).
Re: the hip issue, all I know is from personal observation of my own walk, which meshes with the weight shift illustrations. It's a similar balance issue to when you're standing with your weight primarily shifted over one leg. The weight-bearing leg is straighter, with the pelvis tilted up on that side. The relaxed leg is bent, in part because the pelvis is lower on that side.
It all depends on the look you're going for, though. If you want to be realistic about it, then the pg 146 weight shift guide in Williams' book is a good example. If you want to break away from tradition and come up with something different, feel free to break the rules and switch the weight the other way. I was just suggesting the more realistic approach since (as stated before) a lot of the rest of your motion was fairly realistic in its approach
Anyway, it's lookin' better! Keep at it!
-
Looking great so far! A few things to think about...
- Not sure if you're intentionally going for a caricatured walk, but for a more natural walk, the hip tilt should be reversed, with the side over the planted foot tilting higher than the side over the passing foot.
- Consider reducing the gap between his two main reaction poses after he sees the gap. It feels a little too forced; i.e. "look at me, I'm reacting", especially compared to a lot of his other movement, which feels much more fluid and natural.
- The quick backward step after he reacts to the gap has an odd feeling to it. The bouce, pace, and final pose make it feel kinda feminine, whereas most of his other moves have a clearly masculine flavor. Try reducing the bounce level, and also varying the pace of that move. Perhaps he makes the steps a little quicker, but the body move into the final pose is more long and drawn out, with lots of ease on the end.
- After backing up, his wide back-and-forth look feels a tad redundant, considering he'd already done that when originally reacting to the gap. Consider something more subtle. Even though he doesn't have any eyes to indicate more specific direction, you might be able to give the *impression* of an eye dart by using a very subtle head turn. This could also blend in with his cautious advance toward the gap, which would *really* give it some nice flavor.
- Some of the holds are a too solid. Hopefully smoothing these out is on your list of thing to tidy up. If not, consider adding it. The arms could especially use a bit more fluidity and flexibility in many places. They feel a tad stiff in spots.
Really solid work so far, Sam! Keep at it! Woohoo!
- Not sure if you're intentionally going for a caricatured walk, but for a more natural walk, the hip tilt should be reversed, with the side over the planted foot tilting higher than the side over the passing foot.
-
Remember, this is cartoons...NOBODY should be trying to 'lip-read' your character.
I hate to take issue with this, but I must...
I've got a close friend who is mostly deaf, and who relies a lot on lip-reading. Despite his severe hearing problem, he still loves to go to the movies, and one of his biggest complaints about animated films is the poor lip sync on characters. The only way he knows what's going on is if he lip-reads. If the characters' mouths don't give him an accurate picture of what the characters are saying, then it can literally ruin the experience because his whole connection to the story is broken.
-
Looks very cool, Justin does some great work. How long is this going to be? With 2 disks I imagine it's pretty long, just curious. Sounds like something to add to the ole wishlist.
I handed over more than five hours of raw material. Chances are it'll be edited down to somewhere around four hours (wild wild guess).
Thanks for the pre-comments, guys.
I hope folks will find it useful. -
When creating lip sync, sync to sounds, not words. Listen to each sound being made, figure out the shape the mouth needs to be in to create that sound (including taking into account how strong the sound is in relation to surrounding sounds), and recreate that shape using the controls you've made for your character's mouth. Lather, rinse, and repeat.
I do lip sync pretty much straight-ahead, with the only pre-game analysis being on an emotional level; i.e. getting a bead on the emotional state of the person speaking.
Swashbuckler animation tests (flat color rendering
in Work In Progress / Sweatbox
Posted
The speed is one problem, but from what I can see most of the readability issue comes from his body blocking the actual cut. I was going to suggest a larger anticipation before the cut, but considering the quickness of his motion, even a larger anticipation might get lost. Tricky, tricky...