Improving sound can be understood in two ways : the audiophile way, supported by Bob Katz (digido.com goto “articles/Jitter”, and “FAQ/Jitter”) supposes that the playback is bit perfect, and that the sound is affected by the incoming jitter into the DAC. And the “engineer” way that states that the jitter is completely eliminated by the DAC, but that if read errors occur, the digital values are interpolated, thus different than the original ones.
Yamaha’s position about Advanced Audio Master is unclear (from http://www.yamahamultimedia.com/yec/tech/aam_01.asp# )
Maximize audio fidelity
Increase the lifespan of audio CDs
Increase the lifespan of important backups
Reduce skips and pops on mobile CD players
Increase compatibility with older CD players
Protect your valuable data (wedding photos, family videos)
Only the “audio fidelity” argument is on the audiophile side, but it can be understood in the two ways. Maybe, consious of the arguments against the audiophile position, they deliberately chose an ambiguous statement, so as to please audiophiles, but not upset engineers.
Oliver Friedman and I have checked that the digital playback of our CDRs in our CD Players (a Sony and a Yamaha CDX860) was error free, comparing the SPDIF output to a secure audio rip done with EAC. I checked a 24x burned CDR (an “audiophile nightmare”) for 4 minutes of playback. This makes impossible the slightest difference in frequency response.
The Philips paper (I could never reach it since it is in a restricted aera of their website) was quoted in EAC’s mailing list by Mike Richter, which lead me to perform the same experiment.
According to the measurments I made, recording sines from the analog output of a Yamaha CDX860 CD Player into the 24 bits 96 kHz analog input of a Marian Marc 2 soundcard, then performing analysis with SoundForge 4.5, and taking into accound all sources of error, such as quantization noise, size of the samples, and especially alias effects, I came to the conclusion that the frequency response of the analog playback of a CDR deviates no more than 0.5 db from a flat line until 20,000 Hz, which completely dismisses Philips statement about -2.5 db at 16,000 Hz.
This result was to be expected. As no digital error occur, the curve reconstructed by the DAC is exactly the same as long as the digital data is the same, the jitter having no significant effect on it according to various sources. Though the only number I could get about jitter was 10ps for high end audiophile transports, with the warning that it is only the average value; and 1000 ps for DAT decks, that would make them “worse” than CD Players (infos from digido.com).
All the details and calculus about my measurments here : http://pageperso.aol.fr/lyonpio2001/cdr/cdr.htm