So you've recorded your some fucking, you've even imported it to the computer, and done a little editing. Now you just need to re-encode it, and upload to yuvutu.

I don't know about the rest of you, but I'm still a bit unclear on what the best options are for uploading to yuvutu. What I do know is there's a 100MB per file limitation. The resolution the video player on yuvutu seems to be 480x380 (someone please correct me if this is wrong), which is a rather odd resolution, but we can still work with it. And finally, as most (all?) flash based video players do, the final files are .flv format.

Unless I've missed it, yuvutu doesn't seem to have a FAQ pertaining to ideal encoding settings, so I've just borrowed from another popular video site (vimeo), which suggests using h264 (mp4) compression. This is generally considered a good compression for web based cuz you get much better compression than mp2, while not greatly decreasing in quality. As for bit rate, 1300-1500kB/s is what's recommended, so I usually try to hit this range.

Because of yuvutu's seemingly non-standard resolution, scaling sometimes becomes a bit funny. I've noticed some of my video's will sometimes lose their proper aspect ratio, depending on what resolution I render at. One option is to actually make your video 480x380. This will most likely create some black bars, which makes the video a bit larger, but the upside is that you know the aspect ratio won't get distorted.

On one of my videos in particular, I noticed the video was a bit jittery, and this got me wondering about the frames per second (FPS) yuvutu uses. 25 is standard for PAL, 29.97 (sometimes called 30) for NTSC. Sometimes for lower resolution video, you can get away with reducing FPS, perhaps to something like 15, 18, or 20. This will reduce the file size, but at the expense of less smooth video. If your video is on a tripod, you might not even notice the difference. But with jerky camera work and the fast movements in some of these videos, it might come out looking pretty choppy. I don't know if yuvutu changes the FPS when they re-encode your video. This would be helpful to know, for setting an ideal default. But for now, best bet seems to be leave it at what it is.

And the last thing I can think of is the fact that when you upload your video, it gets re-encoded anyway. For one they need it to fit their resolution (althought I've noticed if your video is smaller than the yuvutu screen, it just gets zoomed to fit). So I wonder if there is a certain way to re-encode which would already meet the requirements of yuvutu, and thus avoid a RE-re-encode. Cuz this would certainly help on not losing quality. My guess tho is it doesnt matter, especially if you request the watermark, then I think they have to re-encode anyway.

This post is kind of a mixture of my own questions about encoding, and hopefully some tips to help others that might be confused. I'd love to get feedback from other's that have been dealing with this, and have an idea of what works best. If we get more comprehensive input, I'd be happy to create a more proper encoding guidelines for people to follow.

I say stick with mp4. It works well, fits everything I want into the 100MB file size, and shit, I guess you could later watch it on your iPod. I haven't had the scaling problems you mentioned, and as for the shutter speed / FPS - shoot with lot's o' light and a narrow aperture. The more data available in the original, the less compression your software has to do, therefore the better quality to the end user.

I don't guess anybody lets you upload .flv directly, do they? I'd think that'd be the way to do it, quality control wise.... But, of course, the computer nerds that run these sites wanna have an excuse to "review" and beat off to our stuff before publishing it. It's their version of a backstage pass.

(p.s.: - so, you're recommending a rate of 1300-1500 kb/s for mp4? I haven't tried changing any of the parameters - I just do what the Mac suggests.)


_________________
Sweet Jeebus! What in Hell is THAT!?!

Most encoding codecs use a GOP (group of pictures) structure which predicts the amount of encoding frame by frame based on previous and future frames processed. Using a tripod will drastically reduce the amount of encoding required since part of the frame is predictable from one frame to the next. The less randomness between frames the higher the picture quality after compression (and recompressiom after compression to flv). Using good lighting reduces the amount of grain (and randomness) so less compression is required and images are clearer. Hope this helps.