Hello, I am currently writing a combination of programs to do the following: - Transmitter: read an MP3 file from disk, send it in separate data chunks (still encoded, preferably smaller than 5 kB) through the network to the transformer; - Transformer: decode the received data chunk, re-encode it at a lower bitrate or as mono, send it to the receiver; - Receiver: decode & play the received data chunk.
I have a few questions on how to do this best: - The transformer operates on a single chunk at a time, not more. I believe it would be best if the chunks would start and end on frame borders. How do I locate frame borders in the transmitter without decoding the MP3 file? - I have read on this list that a single frame can reference many (up to 255?) previous frames. I can send multiple frames per data chunk, but how many should I include? I'd like the data chunks to be smaller than 5 kB each, if possible. The transformer program operates on one data chunk at a time. - Is there a way to let LAME do both the decoding and re-encoding of the received data chunks at once? (Perhaps this is what lame_set_decode_on_the_fly() does?) For those interested in the reason: I realize there are more efficient ways to do this, but the programs are part of a larger framework that adds information about the client device (e.g. whether it has only 1 speaker) to a proxy (the transformer), which will then decide if the audio should be re-encoded to reduce client bandwidth and CPU usage (think mobile devices like PDA's and cell phones). Kind regards, Tomas _______________________________________________ mp3encoder mailing list mp3encoder@minnie.tuhs.org https://minnie.tuhs.org/mailman/listinfo/mp3encoder
