Jitter, CDPs, Transports and Streaming


In my personal, digital audio journey, I have found that I prefer the sound of CD players over streaming through my computer to a DAC. I have tried 4-5 different steaming/transport configurations and found CD players to sound more natural with less digital glare and cause less listener fatigue in each comparison I made. I attribute this to jitter and the increased levels caused by noisy computer environments and the additional circuitry and wiring between a source/transport and DAC. I am sure component quality plays a role here and I’m sure there are CD transport and DAC combinations that sound better than some standalone CD players.

I got to thinking that DACs have buffers that they read from and realized that the upstream source shouldn’t matter, but they apparently do. Why doesn’t the buffer completely eliminate the relevance of the quality of the source? Are there types of DACs, like asynchronous DACs, that make the CD transport or computer source quality irrelevant?
128x128mkgus
I got to thinking that DACs have buffers that they read from and realized that the upstream source shouldn’t matter, but they apparently do. Why doesn’t the buffer completely eliminate the relevance of the quality of the source? Are there types of DACs, like asynchronous DACs, that make the CD transport or computer source quality irrelevant?

It depends on the way the buffer works. The buffer in the Benchmark DAC3 for instance is not a resampler or a PLL receiver. I have not personally heard it, but customers have told me that putting my Synchro-Mesh reclocker ahead of it has no effect on it, indicating that it may actually reject most jitter. I have no idea how good these customer systems are however, so I may still hear a difference in my own system.

Making the source jitter irrelevant is a good idea, but only if the internal DAC clocks and circuits used deliver really low jitter also. This is in no way guaranteed. Those circuits can easily take a 1psec jitter spec oscillator and convert it to 300psec and they usually do.

This is exactly why I prefer DAC’s that have no reclocking on the S/PDIF inputs. This way you can drive those inputs with lower and lower jitter sources and reap the benefits.

Steve N.

Empirical Audio

This is exactly why I prefer DAC’s that have no reclocking on the S/PDIF inputs. This way you can drive those inputs with lower and lower jitter sources and reap the benefits.

Interesting. It makes a lot of sense. Does that mean it doesn’t have a buffer it reads from - the DAC just takes the data stream and converts it to analog in real time?

There is always a small buffer of a few words, but it is clocked synchronously with the clock recovered from the incoming stream by the "Receiver" chip. The clock is contained in the S/PDIF data stream and it is recovered by this chip.  Some of the newer Receiver chips can reduce jitter even more if they are implemented on the circuit board optimally.

All DAC's work basically in "real-time", even if they have asynchronous buffers.  The buffer delay is much smaller than anything audible.

Steve N.

Empirical Audio

Thanks, that makes things more clear. So even though everything has a buffer, it sounds like it is operating off the clock of the source, unless it is asynchronous which reclocks the signal with its own (hopefully) high precision clock. Is it safe to say that all CD transports will sound the same if one uses an asynchronous DAC?