Stuart Rogers

Robcat's image shows it but he didn't explicitly point it out - unlike you he had "Hold Last Frame" set to ON.

The best example of spline continuity is the case where two splines meet at a CP. Having the splines cross each other is better continuity for A:M's purposes than if the splines meet but don't cross (i.e. they approach each other, touch at the CP, and then turn back to the side they came from, rather than crossing). This kind of continuity problem often hides itself for peaked splines, as it's really not apparent then. If you make sure your splines cross you're less likely to end up in the position where you get a patch made up from just one spline, which will result in it not rendering (this is a feature by design, not a bug). That said, not so long ago I still got such a patch hole even though all my splines crossed properly. Also, you'll probably get better results if, where splines cross (or at T-junctions) they do so as square-on to each other as possible. This isn't always possible - you usually need to make compromises somewhere along the way.

I've been seeing this too. I found that if I hide everything in the model except the five points (and maybe some surrounding points) it seems to greatly improve the chance of getting the five-point button made available. I don't know if it helps, but I also usually rotate the model such that the five point patch is in the plane of the screen.

I quite agree - a last moment transition to a gallop-type motion would heighten the effect of the pounce.

It's a simple decal I made in Photoshop Elements - a few dabs of red over a transparent background, saved with an alpha channel (that is, saved as a 32-bit TGA).

You could add all manner of nurnies. Depending on the relative scale of your model, perhaps an image I made a couple of years ago can provide some inspiration... <Clickety-click>

You'll kick yourself when you see what the problem is... Half your dugout is under ground level. You have an object called Ground which, I'm assuming, is the default Ground model that A:M supplies when you create a new choreography. This Ground model is a simple square plane with no holes in it, so it obscures anything underneath it. Half of your dugout is, as you point out, underground so, naturally enough, the ground plane obscures the underground part of the dugout. You need to modify your Ground model so that it has a hole in it corresponding to where the dugout is. A:M doesn't dig holes for you! Can you say "D'oh!"?

I tried that, but all I got was a slightly moist and sticky mouse.

To duplicate a bone structure... In the PWS, drag and drop the root bone of the branch you want to duplicate while holding down the control(?) key (ALT key on the Mac).

That's a good start - keep at it!

The philtrum (the bit between nose and upper lip) needs reworking. I also think you have far more splines running down the nose (and philtrum) than you need. The eyelids look fine. And I like the shading.

refraction in a transparency will lengthen render times, but I can't detect that a transparency on a plane (not using refraction anyway) invokes any more overhead than any other decal, which is negligible.The last time I did this was with V11 for my Mattress movie, for which I had my movie on a plane, and several semi-transparent planes - whiteout panels, titles, and credits. Planes that weren't in use were keyed inactive. Each plane was at 100% ambience; there were no lights, no shadows. Rendering took several seconds per frame. This struck me as rather long. Then we have a fast solution!

Instead of modeling the hole in the plane you could use a transparency map ... with a soft edge and apply that to the plane.I find that using transparencies tends to increase render times considerably. If you use a modelled iris you could get softened edges by setting a depth of field for the camera and then placing the iris closer to the camera than the near focus. I usually do fades as a post process with my movie on a flat patch with ambience=100% and no lights. It's just occurred to me that a fade to black can then be attained by animating the ambience to 0% over time. This would be quicker than using a transparency.

This looks like a spline continuity problem. The spline that starts running up the edge of the blade takes a sudden turn and then runs around the blade. This is the smooth spline. Another spline then runs around the other side of the blade, takes a sudden turn and runs up the edge of the blade. This is the peaked spline. You should start breaking and reconnecting splines so that there's only one that runs up the blade edge.

Only seen on film, eh? So how come some of my digital photos have lens flare? Lens flare is a side effect of the lens, not the film. Now, we're using CGI to synthesize images of real world objects - who's to say that we can't use CGI to synthesize a real world photograph of real world objects? IMO the use of lens flare in CGI is still valid if the artist wants it - but I would thoroughly agree that lens flare should be subtle, an addition to suggest a real photo without being obvious. This was more distracting than any lens flare! Whatever, it's a nice picture - if it got oohs and ahs in the contest, you're definitely doing things right.

Because she's seen Shaggy having difficulty in "The Door's Stuck" and doesn't want to get caught out like that.

That's odd - it's always worked for me.

I agree with Martin - there's some good direction here. The one thing that didn't quite work for me was when you cut from looking down on the green guy (Alfie?) to what I guess is his point of view (to watch Peggy sliding down the pole). I would put the green guy in that second shot to confirm that this is indeed the case. Why? Because when he comes running, he doesn't end up in the same room in which the bell was rung, so it makes me wonder why he's stopped in the room he's in. I did like the shadows in the telescope room - that long agony you suffered over them has paid off. I liked the broomstick launch, although I wonder if a lot of squash and stretch would give it more impact.

When you select the CP, don't click directly on the CP but on the spline you wish to adjust. Make sure you're closer to the target CP than the next CP along.

Does your character's rig include a heel lifter in each foot? I'm guessing not, and that you're trying to rotate each foot bone around the toe end. I've never been able to do that without some slippage (at least, not without laboriously keying every frame). If that's what you're doing, I think the slippage you've got is probably the best you'' be able to get. If your rig does have heel lifters, then for this movement you should be able to keep the heel permanently lifted so that when you place each foot of the ground it stays put until it's time to lift off.

If you render separate sequences, one from each camera, you can then have each sequence overlap a little. This gives a bit more latitude for editing - you can adjust timing, use soft transitions, etc. If you translate/orient your main camera to other cameras, then you have to make fine-scale editing decisions earlier in the process.

Shhh! Don't disturb him - genius at work! I suspect most animators here, once they've gone past the tutorials, animate phonemes 'by hand'. Using the automated dopesheet that A:M provides is handy for a quick'n'dirty approach to block things out, but I find it's worth the effort to go through the tedious process of hand-tweaking mouth shapes for each phoneme. But then again, I've only lip-synched a talking mattress...

A:M ignores the DPI setting - it only looks at absolute pixel dimensions. Most image file formats don't include a DPI value. JPEG is a prime example of this.

Something like the Optimus keyboard, perhaps?

Because my iBook is ancient, and doesn't have the horsepower to handle most H.264-encoded QT movies (your movies use H.264).