Is Balanced or Single-Ended Better at Preserving Phase and Sound-stage Information?

[Ron Resnick]

In your experience is balanced or single-ended inputs, outputs and interconnects topology better at preserving phase and sound-stage information?

Have you done direct experiments to evaluate this question?

 

[vinyl_mike]

I know this was a cable question, but I believe that it relates more to circuit design than cables. Therefore, my comment is a wimpy “it all depends.” The immersive three-dimensional sound stage is a critical element for me. Because of that importance, I've spent a lot of time trying different combinations with different equipment. For example, my VTLs were designed as differentially fully balanced beginning to end. For these, it’s not even close, balanced wins hands down for soundstaging and this was confirmed several conversations with Luke Manley. Hervé from darTZeel says that while his Zeel cable is the best, single ended is second, followed by balanced. I agree and noticed a more natural sound stage and with greater depth with single-ended darTZeels, and the same with Goldmund amps. With my Tenor amps I've achieved an amazing soundstage with balanced even though the components are not inherently balanced. However, it is my understanding that Tenor uses extraordinarily good transformers for the balance inputs and outputs. I am certainly not an engineer, but just from experience, my conclusion is it really depends more upon the inherent design of the equipment – not the cables. Just my two cents here.

 

[Ron Resnick]

I know this was a cable question, but I believe that it relates more to circuit design than cables.

Thank you for your reply.

I agree that it relates more to circuit design than to cables, which is why I included inputs and outputs in the question.

If an inherently single-ended circuit transformer balances at its input a single-ended input to plug into an XLR, I used to wrongly label this as "pseudo-balanced." I now know this is called "transformer-balanced." But it still seems kind of bogus to me, because there is no true differential circuit between input and output.

The theory I was aiming at evaluating by this thread is "does a true differential circuit, balanced from input to output, preserve phase, timing and sound-staging information which may be lost in single-ended designs or in pseudo-balanced designs or in transformer-balanced designs?"

Presumably Vladimir Lamm (Lamm) believes the answer is no, Jeff Fischel (conrad-johnson) believes the answer is no, and Stavros Danos (Aries Cerat) believes the answer is no.

Presumably Luke Manley (VTL) believes the answer is yes, Audio Research believes the answer is yes, Ralph Karsten (Atma-Sphere) believes the answer is yes, and Ken Hayes (VAC) believe the answer is yes.

In answering a different but related question, Gary Koh advised me that preserving consistent electrical characteristics (loop inductance, conductor to shield capacitance and conductor to conductor capacitance) between left channel and right channel interconnects in a very long interconnect run contributes to preserving phase, timing and sound-staging information.

 

[leyenda]

Ime single ended wins even though the circuit is fully balanced. Also tried impedance matching in fully balanced system which may be slightly better but still, single ended it is for me.

 

[sbnx]

Someone can make a bad balanced cable or a bad single ended cable. For any given cable has the designer considered if phase is being shifted by the cable in some way. Has the designer considered how they will deal with mechanical noise? How does the design deal with RFI/EMI? IMO these considerations are more important than if the cable is XLR or RCA. The component (Preamp, DAC, Amp etc) generally determines if the cable should be XLR or RCA.

 

[Ron Resnick]

Someone can make a bad balanced cable or a bad single ended cable. For any given cable has the designer considered if phase is being shifted by the cable in some way. Has the designer considered how they will deal with mechanical noise? How does the design deal with RFI/EMI? IMO these considerations are more important than if the cable is XLR or RCA. The component (Preamp, DAC, Amp etc) generally determines if the cable should be XLR or RCA.

Do you have an opinion on the propriety of mixing and matching balanced sources with single-ended downstream preamps and amps, or vice versa?

 

[PeterA]

Ron, I think you wrote that you have both in a conduit going from your preamp to amp. In your system, which sounds better, and which preserves phase and sound stage information more accurately? And if that is difficult to determine because you don’t know what is more accurate to either live sound or the recording, which way presents a more convincing sound stage to you in your room? For phase, I would be listening for the naturalness of spoken voice, singing voice, and the impacts of things like drums and piano strikes. Also, whether or not you can distinguish clearly between the up and down strokes of a bow against strings.

 

[adrianywu]

I think phase shift is more of a function of the bandwidth. Since phase shift occurs as the frequencies are rolled off at either end, and this is cascaded from stage to stage, a bandwidth much wider than the audible range is desirable. SETs tend to have more limited bandwidth since air-gapped output transformers have limited HF extension, this can be a problem. On the other hand, NFB has the potential to cause phase shift as well if not implemented with care, and SETs usually eschew NFB as part of their design philosophy. Phase shift can happen in cables if there is impedance mismatch leading to HF roll off, esp. with coaxial cables. Another potential problem is group delay that some people believe can occur due to skin effect of thick conductors, but I don't know whether balanced vs. SE makes a difference. Using input transformers for signal balancing (usually to ameliorate noise issues) can also introduce phase shifts. Personally, I think a true balanced circuit design without NFB or transformers is most likely to preserve phase accuracy. Another advantage of a balanced circuit is the ease in switching phase, and I really appreciate this as some recordings are in reverse phase, or even out of phase.