It is incredible how even minuscule delays can be heard. I first laughed about calibrating sound systems with microphones to compensate for different cable lengths until a sound engineer let me listen to the differences.

He could hear if a box had a delay about 2-3ms. Quite hard requirements.

Note that delays can also be an issue when singing in real life. In a large church choir, people in the back have to start singing a tiny bit earlier than people in the front because of the speed of sound in air. And then there is the pipe organist too, often at an ever greater distance.

https://music.stackexchange.com/questions/30454/how-does-a-p...

You also see this in marching band, where playing in time with the conductor's hands results in messy sound for the audience, because of how far apart different band members might be. The conductor also can't listen to the sound and hear if the band is in time - so you have to watch their feet instead. There's some good videos about these effects on the Drum Major Academy website: https://drummajor.org/marching-band-sound-delay

> I first laughed about calibrating sound systems with microphones to compensate for different cable lengths until a sound engineer let me listen to the differences.

Seriously? Even if electricity has a drift velocity of 50% the speed of light, that's still 150,000 meters in 1 ms - how long is this cable???? There must be something else at play.

It's not the transmission through the cable that matters, it's the fact that you're using a long cable to put things far enough apart that the speed of sound matters.

There's something weird here though. I reckon more 4 or 5 ms of latenecy is enough to put me off playing something in time (this is about what you aim for using DAWs to record music) and yet if you just think about playing a piano, the player's ear is about 1m from where the sound is generated. The speed of sound in air is ~300m/s so there's 3 ms of delay right there, even if the hammer mechanism is instantaneous, and I've never heard a pianist complain about delay.

Maybe it's just familiarity but even that feels too simple an explanation

I'm an organist. The organ console is often placed quite some distance from the pipes—especially in the instance of remote divisions such as antiphonal stops (placed at the back of the sanctuary). In addition to distance, the mechanism of the instrument itself introduces a bit of lag. It's something that you just get used to. Choral and orchestral musicians also often deal with these sorts of issues: sometimes they have to play entirely on sight (watching the conductor's baton) rather than listening to the (delayed) sound.

I used to play Dance Dance Revolution quite regularly. One of our controllers had a lag that made it unplayable for most people. But I found my brain adjusting to it extremely quickly, and I think it's because I was accustomed to doing that already during organ playing.

Somehow the audience also manages to adjust too, trained or not. The violins and cellos are several metres apart yet I think the audience would agree whether they come off together or not, regardless of where they're sitting in the auditorium

I wonder if actually it's the delays plus reverb we manage to adjust to. I remember playing church organ too, a while back, and I found the delays much easier to get used to than say a 4 or 5 ms delay in a DAW, even though the delay was probably much bigger. It's like it's training + room acoustics that we use to adjust, rather than the raw numbers of the physical delay

My wife is a musician and sang in a highschool/college choir. She hated when they combined with the orchestra and the orchestra conductor conducted a song. Said it felt like the conductor was off a bit because he compensates for different delays.

"Maybe it's just familiarity but even that feels too simple an explanation"

I think it is. Because it has always been like this for anyone since he played the first time. You press the key and (a bit later) music comes. I doubt anyone consciously notice that it is not instant, but your mind notices and adjusts to it.

But I could play either Jerry Lee Lewis style, arms length from the keyboard with my head back, or Glenn Gould style, hunched over the keyboard and my head moves a metre between the two. So that's around the 3ms threshold where I (and the sound engineer above) claim to be able to detect delays, and yet I don't think I'd notice a delay between those two playing positions

Your brain is very good at compensating for noise and delay in a feedback loop it controls so it's hard to say one way or another. Maybe it feels like a difference in expression instead of a change in delay? A better experiment would be to, like, hm...

let's say you're standing a few meters from a metronome and a musician playing to it. If the musician is close to the metronome, you will hear their playing on beat. The further the musician plays from the metronome, the more delay between you hearing the metronome and their playing, unless the musician stands between you and the metronome. When would you notice the delay?

Yes, something like that. As I mentioned elsewhere, I wonder if the result might be affected by the acoustics of the room as well as the physical proximity of the sound sources

It depends on what scenario you're talking about. One scenario where delay is often used is when you have speakers at different locations to provide even coverage for a crowd.

For example, at a really large outdoor concert, you might have speakers near the stage, and another set of speakers further back. If you don't add any delay, then the people sitting far back will hear the signal once from the speakers closest to them, and then they will hear it again from the speakers that are further away (by the stage). The effect will sound like an echo.

So you delay the sound coming out of the further back speakers just enough so that as sound arrives from the stage, the speakers further back produce it at the same time.

And it's not just large venues. in small venues, you sometimes cannot cover the whole area evenly with one set of speakers, so you might have small "fill" speakers to deal with coverage gaps.

That’s not as bad as the nerve conduction latency.

Also, on a grand piano, the hammer generally strikes before bottoming out, unless you play sufficiently loud, the timing dependent on how loud you’re playing. You have to begin moving your fingers or hands downward into the keys farther in advance when playing quietly, which is a whole thing when your hands are at different volumes. The mass and mechanical advantage of the finger affect the timing and volume, too.

Yeah, that's correct, but saying "cable length" and not "acoustic time distance cause by microphone placement" seems like it would confuse folks who might not have tried to solve phase alignment problems into thinking that the difference is wholly in the "cable length".

But that could just be my reading of the thread.

Ah, so it's the soundstage, not the mic that's the problem?

Absolutely minuscule delays due to the distance between your ears are a guide to your stereo perception.

Human hearing is quite extraordinary.

Electricity travels at a big fraction of the speed of light, he couldn't hear a cable length delay. Sound travels about a meter at 3ms in air, that's more likely what microphone calibration was for.

You're right you can't all hear each other. Some choirs use instructions like https://www.makingmusic.org.uk/resource/zoom-online-rehearsa... where everyone is on mute and can just hear themselves and the choirmaster.

I think a technological solution would exist: A and B hears choirmaster, A and B's sound is sent back to server which syncs it with the choirmasters sound before sending to C, D, E and F. C, D, E and F's sound is sent back to the server which syncs it and sends to G, H, I... etc.

So most people can sing along hearing some of the choir and the latency overall shouldn't be too bad. Scales with log(n) * max(latency)

I've found that the lag impacts even day to day speech. Turn taking cues are all out of whack, and I notice that e.g. my teammates interrupt each other significantly more. It makes conversations in groups larger than 4-5 people very awkward.

https://endlesss.fm/ gives you a shared loop pedal. It's not real time, but just as social and the looping means you don't have to worry about lag.

I saw someone on a facebook group for choral directors who got 10ms latency over 30km. It can be done but requires somewhat specialized hardware (ethernet cable to the router, external audio interface and mic), network setup and software

ninjam should work, although I have not tried it myself, anyone else?

https://www.cockos.com/ninjam/

I haven't tried it, but maybe singing a round would work? The whole point there is the delay.

In a round, the delay is under your control. If you start singing the first line when you hear me sing the second one, you're actually starting to sing at (second line ts + $latency). And then I hear it another $latency later. A round could work if your call software compensates for this and delays your singing for me until the next round "entry point".

That is, you hear my "Gently" and join in with "Row," but it's not played back for me until I start singing "Merrily." This could work, but as you can see, it would need some clever delay management.

Then how is Smule possible?

The parts are recorded at separate times