Is a femto clock important if you aren't using asynch USB?
I am considering buying a used Wyred 4 Sound DAC2 DSD se for streaming from my Bluesound Node 2i, and the seller told me that this DAC2 se is one of the early ones that does NOT have a femto clock. My Bluesound does not have USB output, so I won't be using the USB input on the DAC2. My understanding is that the femto clock is mainly useful to sync up the digital stream from a computer; is it likely to make much difference if my source is a streamer and the input is SPDIF?
I have an old W4S dac2 and it is indeed a good dac if it works in your system. Since you're shopping in this price range, I doubt you'll hear the difference with or without a femto clock (this was a flagship dac over a decade ago but worth a couple hundred dollars today). Be aware that the Dac2 is not driverless. I got it work fine with my bluesound node but then Roon will not see it. Jerry |
Other than Asynchronous Rate Converter used in some DACs (like Benchmark), most of them use PLL (Phase Locked Loop) and have variable D/A converter clocks with S/Pdif transmission, to adjust conversion rate to average incoming words rate (otherwise samples would be lost). It used to be done with VCO (Voltage, Controlled Oscillator) adjusted by comparator comparing phase of incoming signal to phase of the VCO util they are even (same frequency). Today it is likely all in digital domain, but still, D/A clock has to follow incoming signal rate, hence it is variable. These "Femto Clocks" are very stable, I'm sure, but average frequency stability is not important since we cannot detect such tiny clock differences/changes. Jitter of the clock is important and likely audible when above about 50ps. It is "picoseconds" - not femtoseconds, making me suspicious of snake oil. With async USB samples come in packets (frames). Each frame delivered at some fixed rate (like 1kHz) contains multiple samples. DAC places them in the buffer and upon buffer over/underflow signals back to increase or decrease number of the samples in next frames. That way D/A clock is fixed, but samples are not lost. Unfortunately connecting computer also connects high frequency noise. There are extremely low noise switching power supplies (SMPS) used by companies like Rowland or Benchmark, but those used in computers are cheap and extremely noisy. This noise is finding any alternative return to ground, including one thru DAC grounds. Optical isolation with USB would be my first choice of action. |
Thank you all for your input, but I should have made my current setup more clear. I now have a Benchmark DAC1, which I like pretty well; I was not planning on using the Bluesound’s DAC, which I do not like as well. My goal with the W4S DAC is to improve on the Benchmark; what are your thoughts? PS, the rest of my system is a McCormack DNA1 Gold rebuild with gravity base jr (LOVE), a Modwright 9.0 SE (meh), and Usher MD2’s (like very much). |
@dirkjanl and @erik_squires I wish the two of you would get together and publish a book "DACs for Dummies". I love the insight you both provide on this forum, but sadly I can only follow what you're saying about half the time (my bad). However, I do appreciate Erik's attempt to break through the fog by closing with a warning against chasing a 10 year old DAC... The Modwright is a good unit, but it might make more sense to sell it and put the money towards a new (or at least <3 years old) DAC, and go directly to my DNA1 (using the Bluesound for volume control); any thoughts? |
