NAS + Ethernet DAC = lossless transport ?


since ethernet is lossless (error corrected), if you stream lossless digital music (eg FLAC) from the NAS to the DAC over the network (eg my Marantz AV7005 is such a dac with an ethernet port (its a preamp) - then isnt the transport lossless to the DAC? And wouldnt that be fundamentally better than any cd player transport (assuming the DAC was the same)?
chrysos
I believe that this is the future for sure. The folks at Linn stopped making CD transports some time ago.
Chryos,

Your original question was if network DAC would be FUNDAMENTALLY better. It has chance to be little better but jitter, as Steve said, would be predominant factor. Interpolation might be non-audible if infrequent but jitter converts to noise all the time. Designing low jitter clock involves also low noise environment (inside of cabinet) and that is not easy thing to do. D/A chip and analog electronics has also big effect on the sound.

Network DAC requires computer with bunch of cables that radiates EMI and injects noise to power lines. This noise might not only affect audio electronics but also increase jitter (noise on oscillator's supplies, PCB connections, D/A IC).
So, the actual bitstream is better with a NAS. Seems that approximations and interpolations to account for CD imperfections is something that can only hurt the sound quality.

For the jitter, then, why isn't the right product a DAC with a network interface that is internally designed for low clock jitter. The CDP has a read/transport problem to solve, and a jitter problem to solve, two places where it can lose audio quality. The Network DAC would have only one: jitter. The content would arrive losslessly.
Better is the question. You will have bit-perfect data from the network device, but as Kijanki says, the jitter from the CD transport may actually be lower, resulting in superior sound. Depends entirely on the master clock implementation in the devices, the Marantz and the Transport.

Jitter is actually more important than being bit-perfect.

Steve N.
Empirical Audio
Your data is stored on hard drive and will be always played identical while CDP has to approximate data in certain cases. When you read data CD in your computer and get checksum error computer will read same sector multiple times. You can do the same with many ripping programs (instruct them to read music as data). CDP cannot do that working in real time. It has error correction code that works up to 4mm scratches along the track but above it (4-8mm) it has to interpolate samples. When scratches are longer than 8mm (along the track) there will be gaps.

Another issue is jitter. Normally jitter affects asynchronous transmission like Spdif and is converted by DAC to noise, but Ethernet case is different since information is transferred as data packets without timing hence jitter doesn't apply. Clock for the DAC has to come from somewhere and it is recreated on the other side of the bridge. Again, jitter of this clock will convert to noise.