I never said that center tap would be better for noise rejection (though it absolutely could be).
It would only be truly differential if you center tapped it and connected the shield to the center tap, such thatThe parenthetical remark renders the first statement false. A center tap decreases noise rejection, at least at audio frequencies.
The second quote is certainly suspect. There is no attempt to make the connection differential, but to receive the signal properly differential techniques should be used if a high CMRR is to be expected.
Ooooh, I can see that one is going to be controversial. You seem to be saying SQ differences between different cables arise from running single-ended, as most audiophile systems do, and not from the cables themselves.Yes, cable manufacturers and inexplicably, many audiophiles, don't like to hear this stuff. But there is so much product out there that does not support the standard that there will be plenty of need for exotic balanced interconnect for a long time.
Is it going too far to draw some further conclusions from that statement:
1. All well-designed cables sound (much) the same in a balanced set-up.
2. Money spent on expensive cables is (largely) wasted in a balanced set-up. Therefore:
3. There are SQ differences due to cables in a single-ended set-up so expenditure on cables can be justified. But if you choose a balanced system, this will gain at least the same SQ improvements as buying expensive cables, but without the extra cost. Pace - I accept in principle it is more costly to build balanced amplifiers than single-ended.
C. "your equipment must support the standard, AES48; I can tell you that hardly any high end audio gear supports the standard "
Are you saying that some amplifier hardware fitted with XLRs and said by the manufacturer to be 'fully balanced', is not fully balanced and does not support AES48? Presumably XLRs could be put on for show without the correct wiring behind? If this is correct, someone suitably qualified should start naming names.
To be clear, I am saying that SQ differences between different cables are audible if there is no termination standard and if there are signal return currents present in the shield. So this can mean both single-ended and balanced, if the latter is improperly executed. Since that happens a lot in high end audio, we have a high end cable industry making expensive balanced line cables! Quite the opposite of how it should work.
Regarding your numbered points:
1), 2), 3) all Yes, if AES48 is observed.
C. Quite a lot of high end audio hardware has XLR connectors and are not balanced at all; the connectors are there for convenience only. Some VTLs and Lamm products are set up this way (examples I have had contact with did not have pin 3 connected to anything in the amplifier; this will introduce a buzz if the source driving that amp meets AES48 as the circuit is incomplete). Other high end audio products have XLR connectors and are fully balanced (such as ARC and mbl) but don't support the standard regardless. The mbl amp I had contact with was balanced because when it was monostrapped, you could then use the XLR input (meaning you needed two of these amps for stereo). This approach causes signal return currents to be present in the shield and results in a very low CMRR, making the connection susceptible to noise pickup and cable construction. The ARC stuff I've had contact with has had fairly high output impedances, preventing the preamp from driving anything less than about 30,000 ohms. The output was two single-ended stages that were driven out of phase with each other. This causes signal return currents to be present in the shield, and the relatively high impedance of the resulting system causes the setup to be sensitive to cable construction.