headphone cables, what's important and what's not.

[jonathane40]

Aftermarket USB cables seem to be all the rage with people paying as much as $15,000 for a short run of it!!! Lots of subjective performance reports are out there but none back with any objective measurements that demonstrates any difference. On the other side of the fence many believe there isn't or can't be any difference in USB cables ("they are all digital"). I thought it would be good to add some data to the conversation.

For my testing, I used the Sonore microRendu as my source host. This is a networked audio renderer and comes with improved USB implementation. See my review and measureents of the 1.4 version here: https://audiosciencereview.com/foru...-measurements-of-sonore-microrendu-v1-4.1867/

For DAC, I used our perennial favorite, the Schiit Modi 2 DAC. Of course I say that in jest as this is by far the worst DAC I have tested. It seems to be highly sensitive to power and USB conditions. I figured if there is one DAC that can show differences in USB cables, this would be it.

The first test was with a 28 gauge decent but generic quality 6-foot or so USB cable between the Sonore microRendu and the DAC in question. I then compared that to a very short, 6 inch or so generic USB cable:

View attachment 8509

What do you know! We have the world exclusive, first and only objective "proof" that USB cables make a difference in the analog output of the DAC!!!

With long cable (in yellow) we have all of those jitter/distortion spikes around our main and only tone at 12 Khz. None should be there.

With much shorter USB cable, their levels is much diminished leaving just a couple of notable spikes.

I also have on loan a short audiophile USB cable from Triode Wire Labs called uR-link. Unfortunately I can'f find the product specific page for it. It is a shorter version of this on their web site:
View attachment 8512

Here is how it did compared to similar length generic cable:

View attachment 8513

As I noted, it seems to actually do worse than my generic cable. I ran the test again and the output was different so not conclusive. But I say don't waste your money on this cable. The mere fact that the cable is short is bringing the benefit, not any magic inside it.

Figuring shorter was the key to better performance, I used a generic "stub" USB gender converter. The one I tested came with one of my many audio tweaks/boxes and looks kind of like this:

​

View attachment 8510​

Here are the measurements, comparing with the short USB cable:

View attachment 8511

We see better performance yet!

For final test, I actually have an audiophile version of such stub from Uptone which came with ISO Regen (see test of that here: https://audiosciencereview.com/foru...ptone-iso-regen-review-and-measurements.1829/)

View attachment 8514

The Uptone stub does even better, producing the cleanest output I have seen from Schiit Modi 2.

Conclusion So Far:
So it seems that USB cables do make measurable differences in the output of DACs with the shortest being the best. And carefully made one from Uptone being even better.

Should you run and buy these short cables/adapters? No, you shouldn't . You knew that was coming, did you not?

Let's continue our testing, this time using another cheap DAC: the Behringer UMC204 HD which at $79 is even cheaper than the Schiit Modi2. Here I compare the long cable to generic USB stub converter:

View attachment 8515

As we see, there is no difference between the worst performing USB cable (long and generic) and the shortest one (USB stub).

High performance DACs are liable to be even more immune to such vagaries, completely neutering any electrical advantage to shorter, better USB connections.

Graphs Can Lie
Note the vertical scale in my graph. You see that the top of the graph is at negative -120 dbV. This means I have highly exaggerated the differences. If we show that full scale and compare the worst (long USB cable) to best (UpTone adapter), we get this:

View attachment 8516

While differences still show, they are so, so down in level as to not ever be an audible concern. BTW, the Schiit normally does much better if fed with clean USB power. Here I am powering the Sonore microRendu with my high quality lab supply which helps a ton in cleaning its output.

So as usual, we can show measurements that bother the eye but not the ear.

Summary
For the first time anywhere, we have shown that we can change the analog output of a DAC by changing how we feed it in digital domain in USB. Shorter cables provided improvements that were visible. Alas, two other factors conspire to make this non-news:

1. Schiit Modi 2 that I have tested is horribly designed. It is sensitive to both USB power and data noise. Even a cheaper DAC like Behringer shows that such sensitivity is due to poor DAC design than anything to do with USB cables. The Behringer performance is better than Schiit with any cable and shows no need for tweaking the USB cable.

2. I have hugely magnified these differences. Without this magnification, the differences even with lousy DACs like Schiit are literally in the noise.

So nothing here explains the fantastical subjective differences people report between USB cables. I suggest those people conduct some blind test to get to the truth of whether they are hearing any differences especially if they are using better DACs than Schiit Modi 2.

Comments and feedback as always are welcome.

Thanks for the great test of usb cable!

Have you ever done a test with different headphone cables?

 

[artnoi]

Thanks for the great test of usb cable!

Have you ever done a test with different headphone cables?

Headphone cable measurement is very hard due to how we measure headphones themselves (the parameters like positioning etc are very hard to control. Plus there is always some background noise). Aside, you will have to convert analog sound back to digital to analyze it.

 

[amirm]

Thanks for the great test of usb cable!

Have you ever done a test with different headphone cables?

I have not. We will have to think about how to test them as noted above.

 

[jackenhack]

I have not. We will have to think about how to test them as noted above.

Isn't just resistance and capacitance measurement enough to know if headphone cables are ok?

 

[trl]

[...] Have you ever done a test with different headphone cables?

Source: http://www.mediafire.com/file/d9kb768up61eh3b/HD650+modifications.pdf - page 13, main page: https://diyaudioheaven.wordpress.com/headphones/measurements/brands-s-se/hd650/.

The brown plot is copper, the silver plot is the silver cable. Seems that there's absolutely no difference. Test conducted by Frans on HD650 cans (this is the only test I've seen on the web...yet).

However, my own opinion: if your existing headphones cable has >1 Ohm/meter resistance, then probably worth "upgrading" it, but otherwise...I see no reason too spend anymore money.

L.E.: I'm sure same applies to speakers cables, if good quality cooper is used:
- Copper - 1mm thick @1m long: 0.021365 Ohms
- Silver - 1mm thick @1m long: 0.0201936 Ohms
Calculator: https://chemandy.com/calculators/round-wire-resistance-calculator.htm.

 

[Veri]

View attachment 19354

Source: http://www.mediafire.com/file/d9kb768up61eh3b/HD650+modifications.pdf - page 13, main page: https://diyaudioheaven.wordpress.com/headphones/measurements/brands-s-se/hd650/.

The brown plot is copper, the silver plot is the silver cable. Seems that there's absolutely no difference. Test conducted by Frans on HD650 cans (this is the only test I've seen on the web...yet).

However, my own opinion: if your existing headphones cable has >1 Ohm/meter resistance, then probably worth "upgrading" it, but otherwise...I see no reason too spend anymore money.

L.E.: I'm sure same applies to speakers cables, if good quality cooper is used:
- Copper - 1mm thick @1m long: 0.021365 Ohms
- Silver - 1mm thick @1m long: 0.0201936 Ohms
Calculator: https://chemandy.com/calculators/round-wire-resistance-calculator.htm.

This post (the last 'cable') section is also enlightening:
https://diyaudioheaven.wordpress.com/headphones/measurements/audio-quest/

 

[solderdude]

AQ has 2 measurably different cables coming along with their NightHawk.
These are so different that audible differences exist.
It is trickery on purpose by AQ to 'prove' differences exist.
In the HD650 above the same resistance difference (that's the only difference that counts) does not matter as the headphone is high impedance.
It CAN make a difference when the impedance of the headphone itself is very low as is the case here.

edit: Veri beat me to it by a few seconds.

 

[trl]

"Silver cable, Gold cable (Right channel only)", right?

 

[Blumlein 88]

So loudness and frequency response are audible. Just one more time we see 85% of high fi is frequency response.

While all good info, headphones cables aren't USB.

 

[Veri]

So loudness and frequency response are audible. Just one more time we see 85% of high fi is frequency response.

While all good info, headphones cables aren't USB.

Yes sorry a little off-topic

 

[trl]

So loudness and frequency response are audible. Just one more time we see 85% of high fi is frequency response.[...]

We can all try square-wave response, but also impulse response too; it would be interesting to see the diff. between more headphones cables. Of course, a new thread might be needed for that.

 

[Blumlein 88]

We can all try square-wave response, but also impulse response too; it would be interesting to see the diff. between more headphones cables. Of course, a new thread might be needed for that.

Okay so start another thread. FR will effect square and impulse response. So you'll learn the same thing in the end.

 

[solderdude]

3 wire cables, 4 wire cables, 'balanced' or not, thick, thin, supple or stiff, cloth covered or smooth, coiled or straight, flat or round, screened, braided, copper, silver, alloys, short, long, heavy or lightweight, coloured or plain black, tangle free, fixed cables or replaceable, plugs, connectors, split cables, single entry, aftermarket or original, generic or custom made, high or low priced ?

What's important and what's not... and to who.

 

[Patrick1958]

Balanced if possible.

 

[JJB70]

For the most part I find the cables that come with cables are perfectly good. Headphone cables can have a hard life and so good stress relief and sturdy construction are important but that doesn't have to mean expensive. Length is important, it can be a pain in the backside if they are too long (and also tends to result in damage as they wrap around chair rollers etc) but too short is just as bad hence why I do think there is a need for different cables for home and mobile use (long but not too long, and short). I don't use any balanced outputs so have no balanced cables. I prefer 1/4" jacks in terms of being more robust but in terms of SQ I see no advantage over 3.5mm. There is something nice about a nicely made cable with nicely made connectors and the tactile feel , but I don't see it as an SQ issue and even then the cables supplied with the headphones I've bought have for the most part had a suitably luxurious feel.

 

[JJB70]

I was interested to read of the naughty trick used to Audioquest to suck people into the whole expensive cable myth. Whilst I have never had much time for Audioquest and hate the way they peddle all those cable myths I did have a lot of time for the headphones they made, seems they even had to use their headphones to promote their snake oil ideas on cables.

 

[maxxevv]

I'm not sure if it makes a difference but the aftermarket TRRS cable I got for my HD6XX actually measures at half the resistance of the original Sennheiser one !

Would resistance at such levels make a difference ? I honestly can't tell from listening at all.

 

[sergeauckland]

I'm not sure if it makes a difference but the aftermarket TRRS cable I got for my HD6XX actually measures at half the resistance of the original Sennheiser one !

Would resistance at such levels make a difference ? I honestly can't tell from listening at all.

Rather depends on what the resistance was, and how that compares with the resistance of the headphone drivers. An ohm or even two won't make any difference to any sensible headphones, especially if they're driven from anything more than an essentially zero ohms output. If the resistance is more than that, then there's something seriously weird about the cable.

S.

 

[Maki]

I like a thick cable with a sturdy Y splitter that I can attach a shirt clip to underneath. Shirt clips are an amazing thing and should be included with every headphone. Luckily they're easy to 3D print.

 

[pwjazz]

I'm quite partial to 3.5mm connectors that support screw-on 6.35mm adapters. These are the sort that come with a lot of the cheapish NewFantasia cables and similar. When I built my own cable, I used these:

They've got a decent heft, the threaded post is narrow enough that the 3.5mm end fits through most phone cases, and having the 6.35mm adapter screw on rather than snap on means that I don't have to worry about it coming off when I unplug the headphones from my desktop amp.