We need to support all the standard AAC and MP3 frequencies. AAC(from 8 to 96 kHz) MP3 (16 to 48 kHz) Our data is in the audio frequency range. Let's forget about what we are inserting because it is irrelavent to what we are looking for. The issue is not whether our encoding survives the compression because it does survive any standard AAC and MP3 compression. The issue is audio degredation from decompressing and recompressing. We are trying to reduce the amount of audio degradation. Every time you compress you add more loss. At a fundumental level both AAC and MP3 both do the following. 1. The signal is converted from time-domain to frequency-domain 2. Psychoacoustic quanatization (remove or reduce content that is unperceivable to the human ear). 3. Redundancy elimination (compression) We are looking for ways to reduce the audio degredation. Since we know the audio has already been compressed step 2 is redundent and may cause further degradation. What we are looking for is the ability to adjust or remove step 2 through the AAC/MP3 creation parameters. Alternatively, Does TI know of any open source AAC/MP3 encoders that we could modify? I am assuming TI's encoder libraries are in binary format that can not be modified.
We need to support all the standard AAC and MP3 frequencies.
AAC(from 8 to 96 kHz) MP3 (16 to 48 kHz)
Our data is in the audio frequency range. Let's forget about what we are inserting because it is irrelavent to what we are looking for. The issue is not whether our encoding survives the compression because it does survive any standard AAC and MP3 compression. The issue is audio degredation from decompressing and recompressing. We are trying to reduce the amount of audio degradation. Every time you compress you add more loss. At a fundumental level both AAC and MP3 both do the following.
1. The signal is converted from time-domain to frequency-domain 2. Psychoacoustic quanatization (remove or reduce content that is unperceivable to the human ear).
3. Redundancy elimination (compression)
We are looking for ways to reduce the audio degredation. Since we know the audio has already been compressed step 2 is redundent and may cause further degradation.
What we are looking for is the ability to adjust or remove step 2 through the AAC/MP3 creation parameters. Alternatively, Does TI know of any open source AAC/MP3 encoders that we could modify? I am assuming TI's encoder libraries are in binary format that can not be modified.
With lossy compression techniques like AAC/MP3, there will almost always be some amount of audio degradation, so your posting doesn't make sense to me. If you want a lossless comrpession method, you should look at something like FLAC. Or, alternately, you can simply zip your audio for a basic form of compression.
So to summarize, if you want to "remove" step 2, what you are really saying is you want lossless audio compression. See FLAC, or just zip the audio. If you want to reduce the degradation from step 2, don't compress as hard. For example, you can compress 8 kHz mono audio at 16 kbps, and it sounds good to me. I hear almost no distortion at 12 kbps either. At 8 kbps, I hear serious distortion.