The concept behind lossless compression is sound. It should generate output identical to the CD original. A relatively full explanation can be found here: http://en.wikipedia.org/wiki/Lossless_compression
Shorthand example: if a digital signal is 111111111100000011111111110000001111111111000000, that takes 48 characters to describe. I could also describe it as 3x((10x1)&(6x0)) which takes 16 characters to describe. In terms of data file size, I've achieved 3:1 compression. What's obvious is that you need to be able to decompress (translate the formula) in realtime. Compression formulas are much more complicated than this little example, but for computer chips these days, this is a very easy task.
Having said that, I've NEVER listened or compared or tested Apple's output, but I expect that feeding to a good D/A converter would yield CD quality.
Shorthand example: if a digital signal is 111111111100000011111111110000001111111111000000, that takes 48 characters to describe. I could also describe it as 3x((10x1)&(6x0)) which takes 16 characters to describe. In terms of data file size, I've achieved 3:1 compression. What's obvious is that you need to be able to decompress (translate the formula) in realtime. Compression formulas are much more complicated than this little example, but for computer chips these days, this is a very easy task.
Having said that, I've NEVER listened or compared or tested Apple's output, but I expect that feeding to a good D/A converter would yield CD quality.