Post
by Kai Stromler » Wed Jan 15, 2003 2:42 pm
AMV ideas happen by magic, usually listening to music, because I can be listening to music while I'm doing other things, while this is harder to do with anime. Sometimes the anime will suggest a song, but the overwhelming majority of the time it's the other way around.
Once this happens, I grab a pen and a piece of paper, or a napkin, or a receipt, or my arm if I'm walking somewhere and just having the song in my head, and write it down so I don't forget it. As soon as possible, it gets entered into the SH prelim-ideas list, and sort of into the production queue. As fast as my editing pace is, it runs horrifically behind by idea pace, so there are usually fifty entries in this file, which is constantly in flux.
Eventually, ideas rise to the top of the queue, usually within about six months after originating. After that, the longer the idea stays in the queue, the less likely it is to ever get made. Once the video immediately before it goes under production, the idea gets formally listed in the SH internal catalog with a working title and a provisional catalog number in addition to its artist, song, and anime, and I start thinking about how it's going to develop in the breaks when I'm not working on the video then under production.
Once that video has 'cleared', been finalized, mixed down to a distribution version, and archived in all its parts on CD, the next idea in the queue goes under production. All germane or possibly germane source is copied locally (if working from other people's encodes, a rare situation now), or captured en bloc (if working from VHS or DVD; a good, properly tuned DVD capture is practically indistinguishable from a rip, and a hell of a lot easier for me to work with), and separated into episodes or ~20-minute bricks if the anime doesn't naturally divide itself on such basis. These are cut through in sequence with a view of getting any and all cuts that may work in any way with the source music, in both hq MPEG2 and lq MPEG1 formats. The usual goal is "100% overage"; to have at least twice as many seconds of footage as seconds of music in the song.
If I've been preoccupied and have not done so yet I rip the music after the completion of clipping, and map out a timing chart on paper, pegging the starts of all lyrical lines, changes in the riff structure or anything else of note, to the second, and create a parallel feel chart with notes about what impression the music is giving, so that I can reflect it accurately with the visual. Of late I have been doing both synch and feel-match more and more by gut, but it's nice to have things plotted out on paper as a backup if you get uncertain or lost.
Then the video-making actually starts. My vids are not really planned to any great degree but real-time drafted in a trial-and-error-process using the lq MPEG1 clips and the playback capabilities of my editing environment. If a clip isn't working, I delete it off the end and swap in another. In this neolinear fashion I go from one end of the song to the other, making the video work with the music, and making sure that it's consistent from clip to clip rather than purely random. This is why I need overage; to make sure that I can get an optimal rather than a bad clip for every part of the video, just in case something really good was used in some other place earlier. If a clip needs some effect that I can only apply with AviSynth or VDub, I'll make a note of it and apply the effect to the hq version that's going to be used in the final video.
That's basically it. Pure naive barbarism. Most videos will have a 'key' shot or two, usually the image that inspired the idea, but that's about it with regard to planning. I'm eternally thankful that I lucked into the one equipment setup that would allow me to do this, during the extremely brief period that it was on the market, because otherwise I wouldn't be able to edit AMV at all.
Ofcourse, to some people perhaps that isn't a bad thing.......
--K