XLR vs RCA ... and which cables are you using?

Hi all appreciate your advice here

At the moment I use RCA>XLR from Naim NAC272 into ATC19As (Mogami 2579 terminated with Neutrik) and happy with the clarity these bring

I assume balanced XLR>XLR is the recommended connection for my new Bartok when it arrives

If so I need to acquire new cables

Online I’ve seen scepticism about audiophile grade cables, and a lot of love for Mogami

The obvious answer is to try it myself … but I’m interested in your opinion and recommendation


Cable choice is always a contentious subject. I have long believed ( and it is belief as scientific justification seems hard to come by) that cables can change the sound sometimes subjectively positively, sometimes negatively.

Aside from subjective tastes it also appears that a cable that sounds well in one system may not in another. This makes recommendations difficult.

I have also been unable to find any correlation between price and performance.

So, with that preamble out of the way and with over 20 years experience of using XLR/XLR analogue connection the best sounding cable that I personally have found is Sommer Cable, Vintage balanced . This is generally not sold in the retail market but can be bought properly terminated from a custom cable supplier who mainly deal with the pro market. They are Designacable in Nottingham. I assume that you are in the UK but if not buying from elsewhere like the EU before October 31st should be straightforward.

They should cost circa 17gbp/terminated metre. I have replaced thousands of pounds worth of cable from top end well known hifi brands with them to connect my Vivaldi or Vivaldi/EAR 868 to my ATC 50ASL’s and with the single ended version of the cable in other applications.

Another good alternative is Van Damme Starquad ( same supplier).

However, as always with this kind of thing YMMV.

1 Like

I can add a bit of an explanation here as to what’s going on.

Cables have, among others, three electrical properties which are commonly referred to as LCR:

  • L – Inductance
  • C – Capacitance
  • R – Resistance

The magnitude of these properties is determined by a number of factors with the most notable being conductor material, dielectric material, the geometric relationship between conductors, and overall cable length.

L, C and R have an impact on the electrical signal travelling through the cable and in many cases that impact is not the same across all frequencies. Furthermore, since there’s a source of the signal and a load acting upon it the electrical characteristics of those components play a role as well.

By manipulating L, C, and R a cable can become an effective filter through which certain signal frequencies come out looking a bit different than when they went in. In short, cables can be filters if they are engineered a certain way…

Rather than go through all of the ways in which L, C, and R can impact a signal I’ll give a couple of examples of other electrical devices that manipulate those properties in order to modify a signal:

  • Tone controls
  • Crossovers

So, through the associative property of internet forum logic it’s safe to say that cables ARE tone controls. A corollary is that really bad cables are crossovers. The exercise is left to the reader :wink:


Thanks Pete, I will check them out

Thanks Andrew, are there preferred cables in use on the dCS shop floor?


This one is really a matter of behavior within a given system / room combined with personal preference. You’re going to have to find something that is well-built and has the right sonic seasoning for you.

I’m also not aware of any negative interactions of analogue signal cables with any of our components so as long as they aren’t some really wacky design there shouldn’t be any issues.

1 Like

Should probably try to answer the original question…

If I recall correctly @reeve_mike has ATC loudspeakers in his Rossini system and he may be able to offer some insight into what has and hasn’t worked for him.

All things being equal it doesn’t really matter which output you use. Both are excellent. The choice comes down to the downstream components and their designs / implementations. If the downstream components are truly balanced (and not all components with XLR connectors are) then the balanced output on the Bartók is preferred.

In your case I’d recommend the balanced outputs as they will provide the best interface with the electronics in your speakers and also give you the most flexibility in placement of the Bartók within the room.

1 Like

Thanks Andrew. However the naysayers always assert that the measurable differences in L, C or R values between cables typically used for audio transmission purposes when given their comparatively short lengths are too small to result in audible differences. Further that frequency dependent factors such as skin effect are irrelevant at audio frequencies .

Well that’s the opinion they hold. I am not qualified in such matters to comment further .

Indeed, but this is based on the assumption that nothing has been done to purposefully manipulate L, C, and R nor does it take into account the interaction between source and load.

There’s a certain manufacturer of electronics that prefers very long speaker cables and many of their customers absolutely agree with this assessment even though it runs counter to logic. After all, cables should only be as long as needed to get from point A to point B, right? Turns out the added inductance of the very long cables is beneficial due to the design of the amplifier’s output stage.

Some cable manufacturers have adopted a similar premise and used cable construction techniques to emulate a longer cable in terms of inductance but without the added increase in resistance and capacitance.

Yes, the differences are very small, but so are the perceived sonic differences that people report upon comparison between different designs and grades of cables. No one is reporting massive changes in sonic characteristics, just very subtle nudges to different parts of the presentation.

No where here have I mentioned anything relating to cost / benefit of high-end cables. It is possible to get extremely high-performance cables at any price point and there’s a wide variation of behavior of those cables in any given system.

More money doesn’t always mean better. It can, but that’s far from universal.

Thanks Andrew

Thanks @Andrew for dragging me into a cable discussion … :crazy_face:

I have always used the balanced outputs on all the dCS units that I have owned.

@Andrew is correct - the balanced outputs of my Rossini feed a pair of SCM50ASLs [via a preamp and an active crossover that blends in two subwoofers].

My personal preference is to not use cables as tone controls - I have units that I can insert if I feel the need for such, e.g. Maselec MEA-2 and Dangerous Music BAX EQ.

I thus go for cables that I believe are neutral in my system context from an LCR perspective and sound so to my ears.

I have used Apogee Wyde-Eye cables (sadly no longer made) for many years as both my digital and analog interconnects.

I have a large stock of Wyde-Eye but if I need a cable that I don’t have I make up something from Belden or LC-1 which Belden makes up for Blue Jeans Cables.

Despite their slightly higher (than Wyde-Eye / Belden 1800F) capacitance, I use Supra EFF-IXLR cables in a couple of places because their stiffness helps with the cable routing [and because the runs are 1 m or less and because I was given them for free :grinning:] and have not noticed any detrimental effects.

The ‘best’ connection method and cable are those that give one the sound that one prefers …

1 Like

With my ATC SCM 50 Active + SCA2 preamp I used (of those that I liked) Acoustic Revive PA or TripleC-FM, Siltech and Harmonix XLR IC. But also I always have a fallback - Canare L-2T2S XLR :slight_smile:

Earlier as source I used dCS Debussy + NB and now placed an order for Bartok.

Thanks for this Andrew. It’s unfortunate that many audiophiles aren’t aware of this information and most cable manufacturers don’t publish the L/C/R specs for their products. The last time I checked Nordost did publish L and C. I’m unsure which others elect to do so.

I have Soundlab speakers and stats “prefer” low “L” (Inductance), very low “C” (Capacitance) and medium “R” (Resistance/Impedance). I would love to prioritize shopping for speaker cables down selecting for applicable L/C/R specs before auditioning. But alas that would be too easy, would help the consumer and would tend to marginalize the mystery behind cable marketing; which may even dissuade some folks from spending more for cables than the equipment they’re connecting.:roll_eyes:

Auditioning components can be fun. On the other hand, I’ve never applied the word fun to cable auditioning. It’s more a necessary PITA!


Just installed my bridge connected to Luxman DA-06 DAC. For whatever reason, much better sound quality, especially presence ,sound stage, and lower frequencies with RCA vs. XLR(XLR is a high quality cable so do not think it is a cable quality issue-RCA cable is normal quality, nothing special. Wife notice the SQ difference immediately.

Be sure that your bridge is configured as follows:

  • Volume level set to maximum (0.0dB)
  • Dual AES sample rate set to 352.8K
  • AES downsampling set to 192K

If the Dual AES rate is set lower the Bridge will output dual AES for some content. When 1/2 of a Dual AES connection is fed to a DAC via a single AES input it can sound odd (to put it mildly).

" Just installed my bridge connected to Luxman DA-06 DAC. For whatever reason, much better sound quality, especially presence ,sound stage, and lower frequencies with RCA vs. XLR(XLR is a high quality cable so do not think it is a cable quality issue-RCA cable is normal quality, nothing special. Wife notice the SQ difference immediately."

Putting aside the basic theoretic electrical compatibility, around a quarter of a century of my practical experience using balanced connections shows that the practical execution of the interface is the important factor. Although I have complete faith in what dCS are doing at their end of things with a fully differential implementation that doesn’t necessarily extend uniformly to other manufacturers.

So if single ended connection sounds better with your Luxman than when the conditions outlined by Andrew are met for AES connection then single ended is what suits your setup best. Nothing wrong with that.

As a caveat however cables do make a difference and I wouldn’t assume that your " high quality cable" is necessarily comparable on all fours to the RCA one that you are using. I would try other AES cables before reaching any conclusion about the interface per se. BTW, the cable must be specified as AES/EBU and not simply be a cable terminated with XLR connectors. It must, for example, have a characteristic impedance of 110 ohms. Your posting does not make this clear.

Hi Andrew:

Thank you for your reply/advice.
Both my wife and I notice the “dramatic” sound difference AES vs. SPDIF 75 ohm RCA

I am using Roon.
Device Setup in Roon:
MQA Decoder Only
Volume Control Fixed
Resync delay 0ms
Max Sample Rate PCM 192K
Max Bits Per Sample Rate 24k
Max DSD Sample Rate Up to DSD 128
Clock Master Priority Default
Enable MQA Decoder YES
Multichannel Downmixing Downmix as needed
Headroom Management Disabled
Sample Rate Conversion Disabled
Filters Bypass All Filters

Seetings in Mosaic
Current settings Dual AES 352.8k
AES Downsampling 192k
SPDIF Downsampling 192K
DSD Downsampling off
I did not see a volume control option.

AES Cable is Audioquest Carbon 110ohm

Thanks for your help.

Best regards,


Hi Pete:

I am new to the forum and worked this weekend so apologies for the late acknowledgment of your response.
The AES cable is 110ohm audioquest carbon.
I use balanced cables output from the luxman dac to the luxman integrated amp and prior to that used balanced cables from MarantzSA11S1 SACD player to several integrated amps with great success.
Per you suggestion, I am going with what sounds best in my system, the inexpensive 75ohm RCA cable.

My DCS dealer has been great and wants me to bring in my AES cable so they can try it and also send me out with a loaner cable from a different brand (not trying to up-sell me).

Best regards,



My DCS dealer, Music Lovers Audio in Oakland and San Francisco (Spencer and Graham) have been great and walked me through the settings in Mosaic and Roon and also offered to send me home with a different loaner cable. I may try a loaner cable this weekend. They were surprised there was a significant difference in sound.



I have to say that I have had very good results ( although not the best) with AQ Carbon albeit for USB interfaces. However cables are unpredictable as to how they behave at least from an audiophile perspective . I recall reading a quite long piece in one of the German magazines which both measured and listened comparatively to a number of interconnect cables. They were able to measure differences ( not just L,R,C but impulse responses etc) but were unable to correlate any aspect of the cables measured performance or physical construction with what they actually heard.

So I would take up your dealer’s offer out of interest but there is no guarantee that it will be better than the Carbon - it may or may not in your particular circumstances and not forgetting your personal preferences in sound. Your single ended connection could be the one to use in the circumstances.

Just anecdotally, and not regarding a digital connection, one of my preamps definitely sounds best singled ended in and balanced out! That goes to show something even if I am not sure what :grinning:.