[Ben G unguren]

Briefly, you can do a pretty nice imitation with levels, blurs, and time remapping — maybe a little gradient. The “camera flash” comes from a film camera when it is turned off — the film doesn’t immediately stop — a few of the frames slow down, causing (1) overexposure, (2) increased motion blurriness, and (3) the transferred footage to be sped up (that is, if the film is slowed down while shooting, it appears to speed up with projection).

So as the camera “turns off” you use time remapping to take, say, 10 frames (of “real time”) and crush them into 2 or 3 (it speeds up). At the same time, apply an increasing blur. Then take your levels effect’s input white and animate it all the way to 0 (so it animates over the 3 frames). Mess with the gamma levels of red green and blue, just for fun. Applying a SLIGHT gradient is handy for the “light leak” look (one of the sides lightens up before the others). When the camera “turns back on” (if it ever does) you do everything the opposite, and if it’s the same footage, make sure that you take out at least a second or two in between. Here’s something I made a long time ago; I’m not sure where the actual file that made this is:

-ben

[Ben G unguren]

“I assume AE interpolates the motion without loss of quality and so saves me time re-timing all the separate elements.”

Correct.

[Ben G unguren]

You need to render whatever works (looks right) with your editing system. You can render either upper field first or lower field first. DV-based editing systems, for instance, are lower field dominant, so if you’re working in Final Cut Pro with miniDV footage, or Avid DV, or Premiere, you should render lower field first.

If you are unsure whether to render upper or lower, try this: create a composition with the NTSC-DV or NTSC-D1 preset, depending on your system (if you don’t know which one you should use, try doing both and this might help you figure that out as well). Create a quick animation of some little circles or text moving quickly across the screen (so it takes, say, two seconds to cross), over and over (a horizontal hail storm), for about 10 seconds. Render this out upper field first (label it upperFirst.mov, or upperFirstDV.mov if you’re doing DV and D1). Render it again lower field first. (You set the field order in the Render settings of your render queue. You don’t need to make any changes to the composition itself.) Import these movies into your editing program, drop them in a timeline, hook it up to a TELEVISION (not a computer monitor) and see which looks better. If you do this right, it should be extremely obvious. It may take a few attempts to get the right codecs and sizes, but that’s all a part of the fun. Good luck,
ben

[Ben G unguren]

Rob,

Are you vertically scaling your footage at all? That could be bad for interlaced footage. If you are, you should correctly interpret your footage (select the footage in your project window and press command+f) and then render to fields. In fact, maybe you should be doing this anyway.

ben

[Ben G unguren]

This is kind of obvious, but just in case…. Are you choosing “Best Settings” in your Render Queue? Maybe it’s stuck at “Current Settings” and your text isn’t already anti-aliased.

[Ben G unguren]

Holly,

Place your moving background layer UNDERNEATH your text layer (if there is more than one layer making up your background animation, you will need to precomp it, or get all those background layers in a new composition and nest it). Next to your layer’s blend modes pop-down menus are the track matte menus. (Press F4 to toggle your timeline’s view in order to see these if you can’t already.) Change the background layer’s track matte from “none” to “Alpha Matte [text layer]”. The text layer’s visibility will turn off, and the background will borrow the transparency information of the text. Voila.

That’s kind of a muddled description. But I think if you try it, it will make sense.

ben

[Ben G unguren]

[Naveen Mallikarjuna] “my RAM preview could look significanly different from what I end up with as my final render output”

If you have all your layers set to best quality, motion blur and frame blending turned on, and if you aren’t doing anything fancy in your render settings like overriding shutter angle, and if your comp window is set to full resolution, what you see in your RAM Preview should match up with the final render. RAM preview frames, from what I understand, are essentially uncompressed “RAW” frames, so you won’t be able to anticipate any kind of compression. Generally, you should render with a high-quality compressor and then mess with compression setting after that (thus avoiding the need to keep rendering over and over).

A major exception to this are fields, which isn’t a limitation of AE so much as a limitation of yoru computer monitor (or your TV, depending on how you look at it). Your computer monitor doesn’t present video with fields; it is progressive. Your TV shows video with interlacing. This means that when you are rendering with fields you can’t see what the moving image will look like on your monitor (the more you do it, the better you get at guessing, however). That’s why it is a good idea to render out a part of your moving section, import it into a program like Final Cut Pro, and hook up a television monitor. Then see how things look.

ben

[Ben G unguren]

Both DVD and DV/DVCPRO in NTSC-land are 720×480 pixels (non-square). Therefore, the size — and framerate of 29.97 — are correct for a video DVD. However if you render with DV/DVCPRO you will be compressing the video in a different way than MPEG2 compression. Thus you are better off to use a lossless (or virtually lossless) codec like Animation (or MotionJPG) set to best quality. As far as I understand it, these compressed-at-a-higher-quality movies have a better chance of good MPEG2 compression for your DVD.

-ben

[Ben G unguren]

You need to render with a codec that has an alpha channel. A common one is the Animation codec, using RGB + Alpha (or “Millions of colors+” — the ‘+’ means alpha). Animation can also render WITHOUT an alpha channel, so make sure it is set correctly.

The easiest way to do this is to use the Output Module preset (in the Render Queue): “Lossless with Alpha” (which means “Animation set to full quality with Alpha”). You may want to toggle your checkerboard background in your comp window to make sure your transparency is working correctly (that is, make sure your text is genuinely over “nothing” and not just a black solid).

ben

[Ben G unguren]

You need to render with a codec that has an alpha channel. A common one is the Animation codec, using RGB + Alpha (or “Millions of colors+” — the ‘+’ means alpha). Animation can also render WITHOUT an alpha channel, so make sure it is set correctly.

The easiest way to do this is to use the Output Module preset (in the Render Queue): “Lossless with Alpha” (which means “Animation set to full quality with Alpha”). You may want to toggle your checkerboard background in your comp window to make sure your transparency is working correctly (that is, make sure your text is genuinely over “nothing” and not just a black solid).

ben