top | item 10485746

Mad Generation Loss

10 points| thisisparker | 10 years ago |parkerhiggins.net | reply

7 comments

order
[+] keenerd|10 years ago|reply
> The last seconds of the recording have been transcoded nearly 200 times. All together, the recording represents nearly 20,000 individual mp3 encodes.

Ew, O(n^2). They actually do chop it up into seconds and then run each second through the encoder multiple times and then stitch it back together. This also creates the stutters every one second.

If I have a moment today I'll put together a more reasonable bash script that repeatedly copies and re-encodes the input. 200 single passes and no stuttering when it is stitched together. (Though unless I chop the first second off of each pass, it won't be any faster.)

[+] asymptotally|10 years ago|reply
Ew, premature optimization
[+] kidsil|10 years ago|reply
Unlike audio where I'm very aware of Flac format, I'm not that familiar with raw formats in video that are still relatively small (like Flac still is compared to WAV). Any suggestions?