XLR connectors are readily available with 3, 4, 5, 6 and even 7 pins and I have done stereo audio on 6 pin XLRs. The reason you don't see it being done on 3 pin XLRs is probably because XLRs are used primarily in commercial or professional applications which also typically utilize balanced audio.
XLR is the go to cable/connector for balanced audio which requires three conductors: positive polarity, negative polarity and ground. We could just as easily use a TRS for balanced audio transport, but for various reasons the XLR has become the King. If you were to send two audio signals down a single XLR then the signal would be unbalanced and would be susceptible to interference, wouldn't have any where near the same distance available, and would have much lower signal to noise ratio. It's doable, but not worth the sacrifices to save a little bit of cable.
History of the XLR.
The two inside conductors are not individually shielded and will Crosstalk a Lot if used for stereo.
XLR connectors are readily available with 3, 4, 5, 6 and even 7 pins and I have done stereo audio on 6 pin XLRs. The reason you don't see it being done on 3 pin XLRs is probably because XLRs are used primarily in commercial or professional applications which also typically utilize balanced audio.
I'm curious, is is sum-and-difference stereo like we use in stereo FM, or is it two discrete full-bandwidth channels?
Technically, AES3 or AES/EBU is two channels of PCM digital audio, not necessarily stereo. So it is indeed two discrete channels. I have used AES3 between a number of different devices in order to keep the signal in the digital domain and avoid format conversions or the additional latency you noted.That was one of my first thoughts. I'm curious, is is sum-and-difference stereo like we use in stereo FM, or is it two discrete full-bandwidth channels? I had thought of using AES for interconnection between two digital consoles on a show I'm working on now, to send two unrelated channels from console 1 to console 2, and bringing back four unrelated channels from console 2 to console 1. It never came to pass since console 2 doesn't have AES outputs like I thought it would, so we are incurring the excess A/D and D/A conversions I hoped to avoid.
It's 1's and 0's.
Well yes, but I was thinking about the data. I could see, for stereo uses, using sum-and-difference technique in the data to encode and reconstruct Left and Right, under the presumption they're related, that most of the information would be common to both channels. If the two channels were unrelated, there would be more crosstalk and artifacts since each one would have to be built from their sum and difference, the difference signal probably being lower-fidelity than the main channel if they took their cues from multiplex stereo.
It doesn't make much sense to do this with digital audio since there isn't, as far as I know, a bandwidth problem to overcome as there was with stereo radio transmission.
And edit: I clearly typed this too slowly, as Brad provided the answer while I typed.
We use essential cookies to make this site work, and optional cookies to enhance your experience.