[ZappaMan]

If you replace usb power line by a clean battery power, clean the usb data line, then there should not be much difference between the usb transports, as clocking will be done by the dac anyway in each case. So what tweaks of expensive usb transport will improve it over laptop ?

The usb signal is transmitted in analogue form, by creating square waves, so the ability to create the most accurate square wave is important, allegedly.
But if you see usb data as zeros and ones then, you’ll never see why everything matters

 

[naynay]

If you replace usb power line by a clean battery power, clean the usb data line, then there should not be much difference between the usb transports, as clocking will be done by the dac anyway in each case. So what tweaks of expensive usb transport will improve it over laptop ?

We audiophiles so wish that was the case as we would save a few pounds but it just isn't unfortunately.
Once you audition one you will hear but only do so if your prepared to pay.

Just to add the best sound comes from using the internal SSD as is with most high end streamers.

 

[iDesign]

I’m finding that with digital, it all comes down to power and ground quality, and clock stability/precision. Closer you are to the DAC, the bigger the impact, but I contonue to be shocked how much things upstream continue to have an impact. The quest for the mythical digital moat goes on, but everything (still) matters.

Its the uncertainty about what’s upstream that is the question. TIDAL and Qobuz allow for artists/labels to submit albums through third party services like Indigoboom, Record Union, DistroKid or Tunecore and that sloppy process invites serious questions about the quality and provenance of some recordings. How was the file processed? How was the file modified to add headers? We can’t know. Now let’s use MQA as another example. Beginning in February 2019, major record labels such as Atlantic, Warner, and numerous small labels began distributing most new releases only in MQA 48kHZ 24 bit to TIDAL and Qobuz (yes, Qobuz is publishing MQA formatted albums— see Audiophile Style for more on that). Each week more and more albums are appearing only in MQA format and that is what’s concerning about digital and streaming— the digital quest might be in vein if the files from TIDAL and Qobuz are inferior to begin with. For me the CD is the source of truth and it just sounds better.

 

[JaZZ]

The usb signal is transmitted in analogue form, by creating square waves, so the ability to create the most accurate square wave is important, allegedly.

I doubt this. Otherwise the transmission via Toslink with its bandwidth limitation (→ rounded squares) would sound worse than every other connection. Also, given jitter immunity on the DAC's side signal-shape accuracy shouldn't matter – theoretically –, as long as the digital info is losslessly transmitted. So rkt31's question is justified: What kind of tweaks are there supposed to be at work in said expensive servers? Why not just use a battery-driven audio player as source instead?

 

[ZappaMan]

I doubt this. Otherwise the transmission via Toslink with its bandwidth limitation (→ rounded squares) would sound worse than every other connection. Also, given jitter immunity on the DAC's side signal-shape accuracy shouldn't matter – theoretically –, as long as the digital info is losslessly transmitted. So rkt31's question is justified: What kind of tweaks are there supposed to be at work in said expensive servers? Why not just use a battery-driven audio player as source instead?

I have no skin in the game.
I find it hard to imagine even how the checksum process is performed on an analogue signal, it must be within a certain tolerance?
If the square waves are the same from all sources to all downstream components then I’d be surprised. And if the variation in the reproduction of the wave makes no difference then I’d be surprised, but I’m just an onlooker. I’d like to understand it better.

 

[ray-dude]

Generally speaking, better power, better clocks, dedicated OSs that minimize other processes

By better power, I mean power supplies with more stable reference voltages (low ripple), fast transient response (big and fast current reserves), and robust/stable ground. If you have a perfect reference voltage, that doesn’t vary with instantaneous demand, with an ultra low impedance connection to the digital circuit, the power supply basically crushes/absorbes and switching or transient noise in the system. That then does not get transmited downstream (by cable, by power lines, by air)

the better/more stable clock, the better as well

at the end of the day what matters is having the best reference voltage and ground as possible for the DAC, and the best clock stability. Anything upstream or Downstream or in the air that takes away from that, takes away from what the DAC is capable

what is amazing to me is how something 3 steps away from the DAC can still be audible. Non sensical and madness inducing.

 

[JaZZ]

I have no skin in the game.
I find it hard to imagine even how the checksum process is performed on an analogue signal, it must be within a certain tolerance?
If the square waves are the same from all sources to all downstream components then I’d be surprised. And if the variation in the reproduction of the wave makes no difference then I’d be surprised, but I’m just an onlooker. I’d like to understand it better.

Even your Wave cables with their massive ferrite population will certainly round the signal to a significant degree (not like optical, given the higher bandwidth, but enough to deviate from the ideal).

Just so much for now, the kitchen is waiting...

 

[ecwl]

I think everyone’s mileage may vary and my listening abilities may be different than yours. But I can get my Blu2 USB to sound the same as CD playback but it requires careful setup. First, even though the USB is not galvanically isolated, the dual BNC output is. I can either play off my iPad via USB to Blu2. Or my main setup is via ultraRendu with LPS1 power supply but I find the main determinant to sound quality there is actually making sure the USB streamer, the network routers, my cable modem and the NAS are grounded (I had to add ground wires) as it takes away high impedance leakage current noise away from being transmitted to the Blu2 via those Ethernet cables. That’s my experience but your mileage may vary.

 

[Triode User]

Even your Wave cables with their massive ferrite population will certainly round the signal to a significant degree (not like optical, given the higher bandwidth, but enough to deviate from the ideal).

Just so much for now, the kitchen is waiting...

Just on that and not to divert the thread but the target frequency for the ferrites is well outside the signal frequency.

And I have some fermented red grape juice waiting. . . . .

 

[naynay]

What tweaks ?

If you have time read 6moons.com part 1 CX+EX solution.
Part 2 currently in progress to follow in a few weeks time.

 

[miketlse]

Just on that and not to divert the thread but the target frequency for the ferrites is well outside the signal frequency.

And I have some fermented red grape juice waiting. . . . .

I hope that it is some decent wine. Life is too short to drink vinegar.

 

[JaZZ]

Just on that and not to divert the thread but the target frequency for the ferrites is well outside the signal frequency.

And I have some fermented red grape juice waiting. . . . .

We were talking of the signal shape that's compromized by bandwidth limiting. You can't avoid a rounding of edges by any form of (low-pass) filtering. That's not to say the digital signal itself (or its content, resp.) be compromized.

 

[Triode User]

We were talking of the signal shape that's compromized by bandwidth limiting. You can't avoid a rounding of edges by any form of (low-pass) filtering. That's not to say the digital signal itself (or its content, resp.) be compromized.

Too late, the red grape juice has captured my attention . . . .

 

[tunes]

Come on, let's be serious, if someone wants the ultimate 'pc mode' experience with DAVE then something like Aurender W20 or similar quality digital transport would be required in the chain, not the cheap toys mentioned above (JitterBug, etc.).

I use a basic DAP via optical out with no issues and it sounds great. You are tethered by the cable but so are your ears.

 

[JaZZ]

Generally speaking, better power, better clocks, dedicated OSs that minimize other processes.

Just playing the devil's advocate...

By better power, I mean power supplies with more stable reference voltages (low ripple), fast transient response (big and fast current reserves), and robust/stable ground. If you have a perfect reference voltage, that doesn’t vary with instantaneous demand, with an ultra low impedance connection to the digital circuit, the power supply basically crushes/absorbes and switching or transient noise in the system. That then does not get transmited downstream (by cable, by power lines, by air)

It's somewhat plausible that every process that takes current introduces some interference and that it's dependent on the capabilities of the power supply how much this takes effect. Now we're still talking of a digital signal that's not directly affected by such irregularities. We just have to take care that the interferences don't reach the DAC and further downstream devices. Your scenario is that a well-designed server will eliminate much of the interferences by means of low-pass filtering and/or galvanic isolation (against the transmission of interferences via ground connection). Unfortunately the signal processing in the server will produce some interferences itself, thus is depending on a capable power supply for minimizing them. It reads like you would be better off by renouncing the server and just using a row of battery-driven iFi iSilencers or iDefenders instead as proposed by rkt31.

the better/more stable clock, the better as well

If you ask Rob, that's absolutely counterproductive, since any pre-DAC reclocking measures will only do harm to the sound or nothing in the best case, since the DAVE itself has a very capable clock on board – the cause of its immunity to signal jitter. How can double re-clocking be useful for sound quality?

What is amazing to me is how something 3 steps away from the DAC can still be audible. Non sensical and madness inducing.

Hard to believe, but I believe you. dmance reports the same. However, it looks like we will never get a signal clean enough for our demands (that can't be improved any further), and then there's the mechanical aspect and the noise of the tides...