Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Compression of multi channel MF and HF data (Read 4067 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Compression of multi channel MF and HF data

Hi all,

I'm working with multi channel MF (medium frequency, 0 Hz - 50kHz) and HF (high frequency, 0 Hz - 2.5MHz) signals stored in raw .wav files with 12 and 5 channels, respectively. I want to compress the .wav files into another format, lossy or not. The signals in my raw files are across the channels typically very similar in amplitude, but not in phase [meaning that for example channel 2 is a delayed version of channel 1 etc]. The first problem is the high sampling rate of the HF files. I have tried to insert a fictive sampling rate of 20kHz into the wav headers so that the ogg/vorbis codec allows to encode. This seems to work, at least I experience that no frequency components are damped/removed. However, when I use the ogg/vorbis codec, it seems that this lossy codec destroys the phase information across the channels. Are there any advanced options within the ogg/vorbis framework which can be used to handle this? Are there any other codecs suitable for my files?

Sincerely,
Geir K Nilsen
Project Engineer
Bjørge Naxys AS

Compression of multi channel MF and HF data

Reply #1 –
I doubt that you will have much luck with lossy codecs in general. They are tuned to make the compressed file "sound like" the uncompressed data. If your data don't closely resemble something that matches human hearing then the results probably won't be useful to you.

OTOH, lossless compression will obviously preserve all of your information. If two channels are similar but phase-delayed then the lossless codec probably won't be able to utilize this similarity in improving compression though.

Compression of multi channel MF and HF data

Reply #2 –
WavPack seems like an obvious choice for this. I believe that it will handle the channel numbers and sampling rates you have, and you have a choice of lossless or lossy. The lossy mode is based on variable quantization, so it will not destroy any frequency or phase information.

http://www.wavpack.com/downloads.html

Since your files do not represent typical audio, I would suggest the using the -h and -x4 options to optimize the compression (the compression step will be fairly slow, but decompression will still be fast).

If your wav files to not have any channels assignments, I would suggest the -–channel-order=... option to make sure the channels are not grouped into stereo pairs, and if you try the lossy mode I would suggest a starting point of -b4 -s0 to target 4 bits per sample and turn off noise shaping (and be sure to keep the -h and -x4 options).

Good luck!
 

Compression of multi channel MF and HF data

Reply #3 –
Quote from: bryant on
WavPack seems like an obvious choice for this. I believe that it will handle the channel numbers and sampling rates you have, and you have a choice of lossless or lossy. The lossy mode is based on variable quantization, so it will not destroy any frequency or phase information.

http://www.wavpack.com/downloads.html

Since your files do not represent typical audio, I would suggest the using the -h and -x4 options to optimize the compression (the compression step will be fairly slow, but decompression will still be fast).

If your wav files to not have any channels assignments, I would suggest the -–channel-order=... option to make sure the channels are not grouped into stereo pairs, and if you try the lossy mode I would suggest a starting point of -b4 -s0 to target 4 bits per sample and turn off noise shaping (and be sure to keep the -h and -x4 options).

Good luck!


Thank you, that was quite of a good tip!

Thanks again,
Geir