I just finished a stop-motion project (as DP and editor), albeit on a GL1. The process used would be exactly the same for an XL-2, though.

I simply ran the tape for 5-10 seconds for each set-up and then adjusted the length of each clip as needed in post. For stop-motion, you are not going to need 24 or 30 frames per second anyway, unless you're a complete masochist, as shooting that way would require an obscene number of movements for your subject. I probably ended up with something like 1-10 "frames" (short clips of static video) per second, and the results were more than satisfactory.

The drawback to shooting this way is that you have to manually adjust the length of each clip, and I'm sure it's not hard for any of you to imagine how repetitive that can get after a while. But the advantage is that it forces you to compose the amount of movement that's necessary--in post, when you can actually *see* it, rather than relying on guesswork during production. For example, if you're shooting frame-by-frame (like you would with film or single-frame video capture), you have to be able to visualize your models' movements while shooting. If the model is supposed to take a few steps and then pause, you have to account for that, or else you end up having to duplicate frames and whatnot. With the "short clip" method, you're basically building your movements from scratch in post, which makes you pay closer attention to the animation.

Short clips also produce a more natural-looking pattern of video grain, because the grain is allowed to "move" naturally, rather than laying there in a static pattern on each frame. The natural grain motion, in my opinion, helps bolster the illusion that the models are actually moving on their own.

I'm really happy with the way my project turned out. Here's the site (which has a clip) for the project I'm talking about in case anyone is curious:
http://anti-arktikos.com/

The picture mode which shoots for 5 seconds is perfect for this work... that way, every clip is consistent...



ash =o)

In theory, yes. Except for the fact that (as someone has already said) timed intervals would be hard to work with when you're moving models around. The amount of time needed to do that is wildly variable, and you'd either end up with lots of unusable clips of yourself making adjustments or else you'd be standing around waiting for the next 5-second grab.

Or did I misunderstand what you're getting at?

Doing Stop motion?

Use a rebel XT or other Digi SLR set up to a computer using a live capture...

Talk about HD.

Depends on what you're doing. The project I worked on included live-action and puppetry. Also, if you're outputting to anything more high-res than NTSC or PAL, you still have to have a really beefy editing setup or render times will kill you. I was looking at routine 16-hour renders at NTSC anyway, because of all the compositing and color correction this thing required. 64-hour renders just weren't an option in my case, nor was plunking down the $$$ for a faster computer.

And if you're spending that much money, why not shoot on 16mm?

Video is way cheaper.

I just mean if you're doing only stop motion, then you just take the picture sequence into AE and downres to SD... it's a nice way to do it. In production of one right now .Will post when done

The XL1s have a photo mode that only records for 5 or 6 seconds. This means you could just speed up a clip and every frame would be the same length... you can take as long as you like in between... I did pixilation with the XL1s all the time...


ash =o)

I kind of half-way thought that might be what you meant. My GL1 does that too. In retrospect, it might have been a good thing for me to use that. After a while though, when I became accustomed to it, my grabs became pretty uniform in length anyway.

I still kind of feel like the grain is probably more "natural" looking with clips of video, but it sounds like you got some good results.

DSLR worked fine for _Corpse Bride_ :)

I'm sure they also had better computers to work with than most of us do, too. And a lot more people working on it.

I'm not trying to be all contrarian, I'm just pointing out the potential workflow issues that come with processing such large image files into a video stream. It's probably not practical for a lot of situations.

Agreed, I'm sure they had a whole animation studio. And they used modified Canons, not stock rigs. OTOH not every photograph has to be a nine megabyte raw CCD file, DSLR does great JPEG's. And there's open source tools for streaming JPG's into an MPEG or whatever. Geez I gotta try this someday lol.

What your doing sounds cool, I'm just adding the SLR info in case someone else wants to try. Video will do animation too, people do it all the time. Actually the first short I remember making was a stop motion with my Video8. I sure wish I had those tapes now! :)

Yeah, you should definitely try it. I had a great time doing stop-motion. It was also frustrating and tedious at times, but on the whole, lots of fun.