[Steve Roberts]

Consider Color Correction for Digital Video by Hullfish & Fowler.

Steve

[Steve Roberts]

Hmm … maybe we’re all talking about the same thing. Which motions are too smooth to be stop motion?

I agree that everything after the snowflake cutting looks like it was animated with a computer. The cutting and folding was probably stop-motion done on a blue or black BG, then comped over the background. The flake floating away was probably the last frame of the stop-motion animated in the computer. Shooting the folding on a keyable BG would have allowed repositioning and rotation in an animation app.

Steve

[Steve Roberts]

Folding an airplane or snowflake is infinitely easier using stop-motion photography than using AE. Go ahead — try it:

Lay out a piece of paper and fold the airplane. Open it up and draw lines where the folds were. Scan it. Bring it into AE. Using masks on 3D solids, create one shape for every unfolded segment of paper. (For instance, a piece of paper folded into quarters would need four solids, one mask each.) Now, because you’re going to need to rotate each piece as it folds, move the anchor point for each solid to a fold line. Which one in the case of pieces with more than one fold? Probably the one closer to the center of the “sheet” of paper. When done, go back and fold the paper again, carefully notating the order in which things fold. Now set up the parenting so, basically, the ancestors are closer to the center. I suppose you should start from the outside, and as you move in, parent each piece to the next one closer to center. Fold the thing in proper order by rotating each solid, and make sure your layers are set to “alpha add” mode. The tricky part can be rotating a piece on a non-orthogonal axis, that is, on a fold which is not straight up/down or across. Let us know how it goes.

Or, you could just shoot yourself as you do the first step. 🙂 Or you could shoot yourself later. 😉

If you want to be a creative motion graphic artist, use any method available to you. Scan theatre tickets. Make crappy photocopies. Pull pages as you scan them. Make little models. Belief does it. Shoot stills. Shoot cutout paper letters. Set up a stills camera, shoot, then import the numbered shots as a sequence. Instant animation.

The computer is a poor tool for inspiration, in my opinion. Its job comes later in the process. It is just one tool in your creative process. Don’t assume you need to use it all the time.

my 2 cents (Canadian)
Steve

[Steve Roberts]

Nice piece.
Analyze it, step-by-step, and it won’t become so intimidating.

And when you analyze it, don’t think “wow” or “Geez, I could never do this” or “my brain just doesn’t see the world this way”. Think “okay, what happens next?” and “was this part done with a computer or not?”.

Also:
“I can do this with some animated masks”
“I can do this with a gradient wipe.”
“I can do this by cutting the background out of a photo.”
“I can do this with a particle generator.”
“I can do this one by hand, while the rest are handled by a particle generator.”
“I can put these two shots in the same comp, but these two shots have to be separate.”
And so on.

That’s the HOW of it. As for the WHY of it, the story or sequence, break down the shots as if from a brief or a script. “This appears, this turns into a bunch of these, as we pan these appear, smoothly, crudely …” Or “what are the elements of winter? Home, neighbourhood, trees, snow, cards …” Write this down on paper.
And storyboard it by saving stills from the video so you can see how the sequence develops.

Break it down. The sooner you get started, the sooner you can make your own original works which you’ll like better.

Hey, you said you’d love to do stuff like this … 🙂
Steve

[Steve Roberts]

I’ve seen similar mixed results on an aerial flyover of a Mayan temple in a forext. Twixtor did a beautiful job on the trees, but not on the temple. The details on the temple gave it trouble, probably because they were lines about 1 or 2 pixels apart, so Twixtor wasn’t sure if the temple was moving or not from frame to frame.

It’s similar to the spoked wheel effect, where the number of spokes and the rotational speed conspire to make the wheel appear static.

Of course, there may have been a way around my problem, but I was in a rush. 😉

Steve

[Steve Roberts]

How silly of me. I forgot the COW’s HDV forum.

[Steve Roberts]

Here’s a link:
https://www.adamwilt.com/HDV/

I haven’t worked with HDV, but anecdotal evidence suggests that as soon as you capture the footage, you get it out of the HDV codec as soon as possible and never go back. The HDV codec is very lossy and can’t handle multiple compression cycles. I’m not sure how that relates to the intermediate codec, though.

Hope that helped,
Steve

[Steve Roberts]

[Aharon Rabinowitz] “If your client failed to mention that they wanted it widescreen, you *might* be able to diplomatically convince them that its their fault for not telling you this, so that you can get paid for the time it will take to do this.”

All excellent advice as usual, Aharon.

I’d add that, with regard to this point, it’s important to indicate the deliverable format (4:3) on your next estimate. That way you can charge for the extra work if the client decides to change the format to 16:9 later.

Steve

[Steve Roberts]

I’ve heard good things about Sony Vegas, but I’m using FCP right now.

Steve

[Steve Roberts]

Unless they’ve updated ReTimer, I’ve come to prefer Twixtor 4.5, at least on a shot of two kids running through high grass. I recommend you download demos to see which plugin works better on your specific footage.

Keep in mind that not all footage can be slowed down if the motion can’t be interpolated because the information just isn’t there. For instance, the spokes of a blurred mag wheel of a fast-moving car just don’t exist on the film.

If you shoot 60i, you have a better chance of getting the temporal information you need.
Try to make the action go as slowly as possible as well. You can always speed it up in post.

I haven’t tried this, but you might want to shoot with a high shutter speed to avoid motion blur and get sharper (motion-wise) images that are easier for the software to interploate.

I recommend you do some tests with your own footage.

Anybody else?
Steve