My feeling is that it is an example of a phenomenon that happens frequently in high end audio. An advantage that a certain design approach or technology may have, which at best is significant only marginally and occasionally, is viewed without proper quantitative perspective, and as a result is over-hyped.
Nearly all cd transports and cd players read each bit, byte, and word exactly once as the disk is being played. If an error occurs in that reading, error correcting processes that are provided in the format usually allow on-the-fly correction of the error, with bit perfect accuracy. But occasionally an uncorrectable error will be encountered, particularly if the disk is scratched or otherwise not in good condition, in which case either interpolation (essentially guesswork by the machine as to what the sample value is likely to be) or muting will occur.
Data from a hard drive, on the other hand, can be read multiple times, if necessary, by the associated computer-based hardware, so that bit-perfect playback is more likely. The PS Audio Perfect Wave Transport emulates that kind of process, and is able to do that by virtue of the fact that it contains a very large buffer memory, into which data can be written at much faster than real-time rates.
That's the theory. I don't have meaningful statistics readily at hand about how often uncorrectable errors typically occur when doing single-pass reads of cd's, but for disks in good condition my impression is that it is either zero or a small handful of samples AT MOST (44,100 samples occurring per second for redbook cd's). So under typical circumstances that theoretical advantage of hard drive playback will be insignificant.
As always, of course, specific practical implementations of either approach may vary significantly in quality and performance.
Regards,
-- Al
Nearly all cd transports and cd players read each bit, byte, and word exactly once as the disk is being played. If an error occurs in that reading, error correcting processes that are provided in the format usually allow on-the-fly correction of the error, with bit perfect accuracy. But occasionally an uncorrectable error will be encountered, particularly if the disk is scratched or otherwise not in good condition, in which case either interpolation (essentially guesswork by the machine as to what the sample value is likely to be) or muting will occur.
Data from a hard drive, on the other hand, can be read multiple times, if necessary, by the associated computer-based hardware, so that bit-perfect playback is more likely. The PS Audio Perfect Wave Transport emulates that kind of process, and is able to do that by virtue of the fact that it contains a very large buffer memory, into which data can be written at much faster than real-time rates.
That's the theory. I don't have meaningful statistics readily at hand about how often uncorrectable errors typically occur when doing single-pass reads of cd's, but for disks in good condition my impression is that it is either zero or a small handful of samples AT MOST (44,100 samples occurring per second for redbook cd's). So under typical circumstances that theoretical advantage of hard drive playback will be insignificant.
As always, of course, specific practical implementations of either approach may vary significantly in quality and performance.
Regards,
-- Al