How important is the USB cable for connecting external HDD/SSD?

[CuteStudio]

A USB disk cable with either work or give intermittent failures.

 

One properly working cable will pull the data off disk and into memory with exactly the same speed, accuracy and power consumption as any other properly working cable.

 

The disk will be caching and bursting the data into memory buffers etc anyway, but remember the data rate of a computer disk is FAR bigger than audio requires, which is itself a very low bandwidth option.

 

For instance to copy a 30 minute album from a disk - even in uncompressed WAV format will take a few seconds, and on really slow systems perhaps almost a minute. That right there indicates the disk subsystem is at least 30 to 100 times the required speed, so there is plenty of system time to do other stuff.

 

Many audio buffs are unaccustomed to computer hardware and still regard the CD as advanced, whereas in computer terms it has extremely poor error correction, data rate and capacity that's considerably bettered by the cheapest USB stick these days. And about that error correction: if you get the data off a computer disk/stick AT ALL you know it's the correct data, unlike a CD where you hope it's right but you're never quite sure.

 

So as long as it works, a $10 USB cable to your disk is EXACTLY the same as a $1,000,000 USB cable as far as the disk, computer and laws of physics are concerned.

 

Therefore it cannot and will not have any effect on the actual sound, although it may still affect the observed sound.

 

This is because psychological influences affect what we hear and our mood and expectations colour our listening experience, and a nice shiny, expensive USB cable we've just invested in will sound better to the buyer than the old cheap grey cable that's now in the cupboard: despite the fact that the sound coming out of the speakers is exactly the same as before.

 

 

[CuteStudio]

I tend to use optical as my final digital link to the DAC, it's just so good at isolating everything and to be honest USB/Optical links are pretty cheap these days too, some even built into the PC.

 

My ethernet and USB cables are all good quality $10 items that do the job as intended, going optical for the DAC may seem extreme but I must say it works like a charm and the $5 optical leads is faultless too.

[CuteStudio]

28 minutes ago, davide256 said:

 I couldn't hear any meaningful difference between good Toslink cable vs AQ diamond coax.

 

Some people hear differences but the actual sound will be the same.

The problem with Coax is that it's a metal wire connecting one box to another so giving a possible path for RF and ground noise to travel.

 

I'm not sure how you'd implement a poor toslink, isn't it digital?

I suppose there is always the old timing jitter, but that's an issue for any digital link to an extent and easily cleared up by re-clocking at the receiving end. We're always re-clocking anyway: all those bits on the CD in the shelf or on the HDD have to be clocked out by something, if it's a computer then it'll be a variable speed clock on the CPU and a super cheap tiny clock on the sound chip :). 

 

BTW what makes you think USB cable quality makes any difference connected to a DAC?

Are you concerned with current pulses down the power wires of the cable?

[CuteStudio]

No I don't think the specs do matter that much for optical, the steadiness of the clock is key, but beyond that it's digital, there's no physical impediment to perfection than I can imagine: is there a mechanism that I am missing?

If I use semaphore signalling to a friend on the opposite bank of a river, does it matter if my paddles are orange or white?

 

If your USB cables are influencing the sound of your DAC we need to look at the physics of the situation. Your transmitter is the same, your receiver is the same, just the cable differs.

 

Short of a digital error caused by a mismatched cable the only other thing is a partly mismatched cable causing reflections that may upset the timing slightly (by their levels being superimposed on the data levels).

 

But you say all cables sound different, so this discounts the digital domain because we'd expect a gaussian distribution of tolerances so we'd expect the majority to sound the same and the outliers to sound wrong.

 

We can also infer from your tests that all of the different sounds are wrong. It is unlikely that one particular cable is the 'right' one, therefore it points to a systemic flaw within the DAC itself. 

 

If we discount the digital and think about the mechanism of the error we're forced to concede that the analog section of the DAC is being influenced by the power or/and ground rail of the USB cable, insufficient filtering and/or isolation is therefore the likely problem with your poorly designed DAC.

 

Solution: buy another DAC, because that one's never going to be right.

[CuteStudio]

33 minutes ago, plissken said:

 

Being one to always interject some critical thinking into a thread:

 

Say you take two DAC's

 

Say you take two USB cables. One 'high end' and one built to spec but inexpensive.

 

Say both DAC's sound great with the 'high end' cable. Say one suddenly doesn't with the inexpensive but still built to spec. Do you see where this is going?

 

Your example is not self consistent.

 

What do you think a 'high-end' cable actually IS?

It's a cable hopefully (but not necessarily) built to spec that has a high markup.

 

A USB cable has two components, a) power supply, b) a transmission line for data.

 

If you are going to power your DAC from the USB cable it's not really HiFi then is it?
If you are not drawing power then it's a transmission line. Transmission line theory has no idea of the cost of the line, it just knows if it is within spec. This spec is achievable for pennies.

 

If you really think your USB cable is worth spending lots of money on you may also find some use for spatial containers and pingu wood.

 

 

In the meantime the rest of us will be solving real problems and making real improvements to the sound.