Optimal USB Cable Length?

At the bottom of page 186 of the 5th edition of Robert Harley's book The Complete Guide to High-End Audio, he says this: "Finally, here's a hint regarding USB cable length. Unlike speaker cables and analog interconnects, where shorter is always better, the best-sounding USB cables are 1.5 meters long."

Now, why would that be so, I wonder? It hardly seems believable to me on the face of it, but perhaps there is a reason. Leaving aside a discussion of whether any digital cables can possibly sound different, which has been a point of controversy in other threads, I'd like to know if there is any rational to this, assuming for the sake of arguent that signal transmission and reception could be affected. Why 1.5 meters in particular?

 

I've no idea but is shorter always better for speaker cables? Pierre Sprey (Mapleshade) doesn't seem to think so. From the Mapleshade website:

"NEVER use speaker cables shorter than 8'. Amazingly, 4' sounds much worse than 8'. Contrary to common belief, shorter interconnects (2 m or less) and longer speaker cables always sound WAY BETTER than the opposite—based on extensive head-to-head tests."

Source: http://mapleshadestore.com/upgrades_cable.php

 

That's bizarre. So far, these statements have been offered without supporting evidence or plausible explanation. Also, I suspect high system dependence if indeed there are audible changes due to length of cable.

 

I'm not going to say the usual suspects like resistance/capacitance/inductance/etc actually validate the Mapleshade statements, but one could at least come up with some theory around it. The USB length is a head-scratcher. The best USB cable would be no cable at all. And from all the friendly discussions on USB cables, we're told they're The Great Problem, plagued with noise and badness. I couldn't possibly see why as short as possible wouldn't be the goal.

 

I sense I high degree of Making Stuff Up.

With USB cables, it could make sense to make them long under the premise that keeping your noise-sensitive analog gain stages as far as possible from the electrically noisy PC -- or that the extra length increases inductance and helps reject RF noise (in which case, a ferrule would work even better). But to specify an optimal length?

With speaker cable, additional impedance can interact with the speaker crossover and woofer tuning to change the sound. I stuck 2 Ohm power resistors in series with my Infinity Reference-3 speakers back in college to see if they would behave better with my amp with an 8 Ohm minimum load spec. It totally changed the sound and fattened up the bass. On those speakers, it was an improvement. So I could believe that with particular speakers, extra long speaker wires could have a good or a bad effect, but to proclaim longer is always better with all speakers?

What are these people smoking?

 

High end weed?

 

Mapleshade is....special.

 

The only explanation I could possibly muster is that it is easiest to keep a USB cable at 90 ohms with a 1.5m cable length. I have absolutely nothing to back that up, just throwing **** at the wall.

 

Yes, and you are destined to burn in flames, damned in hell forever, if you don't believe Robert Harley's truth. Obviously, 1.5 m is exactly 1/100.000.000.000 of the distance between the earth and the sun, who can deny that? What can sound better? Mind you, even Bach was obliged to to resort to numbers like this to well temper his clavier.

Of course, some ill-minded guy could come along and suggest that measuring the distance between Andromeda and our galaxy could hint for an ideal length of 6ft instead of 1.5m. Who knows? Of course, the last judge is always our ears. Anybody swear about his fine ones, no guarantee on a collective scale.

And, since we still hold to this funny sort of things, I have some late news, astonishing in their content, but from a reliable source. You won't believe it, but Santa is not real stuff, it's just your parents' excuse to drop you toys for Christmas. Trust me, sad but true.

Sad but true as the **** we music lovers are all time supposed to swallow in the name of the "big names", who are just poor souls trying to support their families through silly statements like "1.5 m is the perfect length for a USB cable". Wow, how could they say that without even knowing my zodiac sign???

I am taurus. suggest me my correct USB cable length accordingly, please. Otherwise, your measures are too commonplace for my taste. Keep in mind my own peculiar listening habits: I like tight bass, delightful mids and natural treble. Not like anybody else, of course.

Just having fun, you know. I am here for that....

 

Why is 90 ohms important? Impedance matching can supposedly minimize reflections within cables, like with 75 ohm coax. Is there some sort of computer interface standard which specifies 90 ohms for USB?

 

I'd recommend going to 2 meters, or perhaps longer for recordings that are prone to edginess in the treble range. The extra length will give the over excited electrons time to cohere into a better aligned image without smearing the upper transients.

On the other hand of course I'd recommend a USB cable as short as .25 meters to brighten up an old sludgy recording, assuming the dielectric capacitance is a multiple of the fundamental harmonics in the source recording.

That is if I was in the habit of recommending complete ********.

 

I wish I had your answer. I am just (regurgitating) that which I have read on a few blogs and forum posts kicking around the 'net. For this reason (the 90 ohm spec), Supra USB cables from ebay are in high demand because they aren't enormously expensive (about $50 for this length) and are 90 ohm compliant. Personally, I've not actually compared USB cables but if I did and heard a significant difference, I'd put my gear down and go outside.

 

I remember reading claims like this from some small "high-end" manufacturer guy a number of years ago who was all cryptic about his experiments that somehow prove that there's an optimal length. As usual, no evidence provided.

All I know is that the USB cable should not go beyond 16'5" (5m) as per the standards - you'll need active hub or extenders beyond that length. As for how short they can be, I actually have a 6" USB cable here connected to an external hard drive and I certainly haven't noticed any data loss or speed degradation . Don't see how this would be bad for a DAC...

 

It seems a bit odd to me that Bob Harley, who has a pretty high profile in audio publishing, would damage his credibilty by making claims like this. The rest of his book seems mostly reasonabl, comprehensive, and well-written. I can't vouch for all of it, not having read the whole thing. It's kind of a reference. But I suppose his commitment to subjectivism creeps in with offhand statements sometimes.

 

Sorry for the tangent, but I've seen many, many accounts of shutting down services, playing tracks from an SSD, optimized OS's, etc... I was just watching an HD broadcast of the NBA Finals through the same network cable (cat5e mind you!) as my music stream and I, unfortunately, noticed no extra texture on bass, or vastly improved treble extension after shutting down the HD stream. I won't say I don't believe things like analog cables and power conditioning benefit sound quality, but some of the lengths the computer audio folks go to are just incredible, even to me, a self-proclaimed audiophile tech geek. I am still open to try some things, but I am starting to come around to the notion that marketing's ability to create solutions for problems we didn't know we had is quite powerful.

 

I had a 22' Belken USB cable to transfer my music to my stereo for about 4 years. I loved the sound of it but I changed my room around and I am now using a much shorter USB cable. The 22' USB cable did not exhibit any odd characteristics.

 

Yes. 90 ohms is the USB standard impedance for signal transfer. Cables, connectors, traces should all have this characteristic from what I gather from the USB spec. It is actually +\- 15%, so 78 - 104 ohms is within spec, 90 being the target. As to "why" they chose that spec, I am not sure, only that once such a spec is decided upon, maintaining it allows for better signal transfer. Here, we are talking about power and data, over separate wire pairs, all housed within the same cables, so I don't think the same rules should apply as with analog audio transmission or those of SPDIF digital audio data transmission. It seems more compatible with power and noise rejection and then allowing the connected devices to control the operations.
-Bill

 

It seems to me that Robert had some confused idea of the requirements of analog and digital, presumably because digital behaves differently and is less well understood by most. His comments about 1.5m lengths being better than .5 or 1m are likely stemmed from the SPDIF specs, which are different from USB. In SPDIF transmission, specifically more important in optical versions, there are internal reflections due to cable construction and a longer cable allows for reflections to be far enough out of sync as to not interfere with the data receiver, at least that was what I was taught, and I think 6' or even 12' might be better there than 3'. Wadia engineers agreed back in the 90s when I was more curious about it. I just accept it now and use a 6' cable for most optical connection applications.

Pierre sounds like he is simply not making himself clear, although I don't know the context of any of his other statements made about cables. If you take the whole message in, to me it seems if he is saying that it is better to use shorter interconnects and longer speaker cables than to use long interconnects ans short speaker cables. That is typically true as interconnects carry a low level signal which is more susceptible to interference and losses than the high level signal of the speaker interface. It is also device dependent however as there could be a situation where a strong input signal is used with a low impedance output into a very high current amplifier, attached to very demanding speaker loads. In that case, a shorter speaker cable, at least a larger diameter cable, would allow greater current flow. That's all fine theory, but in practice the typical listener might be hard pressed to hear any differences based upon the cable lengths alone. Construction and materials play a more important role.
-Bill

 

3 to 5 meters is the spec governed on USB cables. DEC, IBM, Intel, Microsoft, NEC and Nortel all designed and worked on these specs.

If only they made a Litz wire USB cable all those little 1's and 0's would sound so much sweeter. Here's to dreaming.

 

Diplomatically and well worded.

 

Why are we even discussing this when we all know that taping a baggie full of pebbles to the connection points will solve all our problems regardless of the length of the cables?

 

0

 

Very nice post. Thanks. Audience is now marketing a USB cable with separate lines for power and signal, to keep them isolated. The wires are combined at the connectors. In the standard version, both ends. But they also have a version with separate connectors at the source end for power and signal. I wasn't aware that any computers separated the two connections. But maybe there are special interface cards for PCs that do that. I can see the advantage there, but haven't really seen any actual applications.

 

This is a situation where I can see the advantages in theory, but am not afflicted by the problems it claims to solve.

The Supra USB cables mentioned earlier http://www.ebay.com/itm/Original-SU...115?pt=LH_DefaultDomain_3&hash=item417b03f103 do seem to be the pinnacle of USB cable engineering at a decent price, with the power and data leads separate, but I'm still not quite sure this actually makes any difference.

 

How would you actually use the separate leads though? For instance, on my desktop system, I have a Mac mini and a power speaker/DAC combo. Since both these devices are independently powered by wall outlet current, I wonder if I could just hook up the signal connector and leave the USB power connector part unused. (In your example, the SUPRA doesn't have dual connectors, but Audience makes a cable that does.)