News for hackers. Stuff that matters.

A comment on Terminal latency (2017). Parent comment

Re:Terminal latency (2017) by yellowapple on Wednesday March 20, 2019 @02:34PM (#19445334)
> Organ players have to deal with that kind of latency on their instrument[1], but it's a hard skill to learn, and limits the music that can be performed on the instrument. Say, following a drummer is out of the question.

Field musicians (marching band, drum & bugle corps, etc.) have a similar issue because of the sheer size of the "stage"; the distances between performers is large enough that the speed of sound actually causes notes to be off-beat to the audience if they're "on-beat" to the performer.

Instead, each performer has to basically ignore one's own ears (at least for rhythm; one should by all means use one's ears for matching pitch, though in theory a sufficiently-fast-moving performer might need to pitch up/down relative to what one hears to account for Doppler shifts) and watch someone up front (e.g. a drum major). Even in non-marching settings (e.g. your typical orchestra or symphonic band or even rock/jazz/etc. band), the percussion section is typically positioned rear-most on the stage for this exact reason (and in this case, it is okay to listen for the beat, since the sound of said beat would be coming from behind you).

So for an organ player (organist?), your best bet would be to place the drummer behind the organ, then give the organist a mirror to watch the drummer's sticks and keep time with the visual beat. That way, by the time the organ's sound actually comes out, the drummer's sound will have arrived and they'll be in sync (assuming you have the distance right between the drums and organ; this is something you'll probably have to tune for every stage, given air-pressure-related differences in the speed of sound).

[ Parent | on HN ]
Re:Terminal latency (2017) by bluGill on Wednesday March 20, 2019 @03:15PM (#19445756)
Organs are even harder because it isn't just speed of sound. There is often mechanical delay before the pipe even gets the air after the key is pressed. In some organs the pipes are "slow" to speak meaning that rank of pipes won't make a sound for some time after the air arrives. Skilled pipe manufactures can minimize this (called voicing, though voicing covers other areas of pipe fine tuning), but in some cases that involves other compromises that aren't desired.

Organs have gotten better and worse in this regard several times over the hundreds of years since they have been known.

[ Parent | on HN ]
Re:Terminal latency (2017) by yellowapple on Wednesday March 20, 2019 @07:30PM (#19448023)
My point is that if you put the drummer far enough behind the organist and have the organist track the drummer by eye instead of by ear, the speed-of-sound-induced delay will match the organ's mechanically-induced delay and everything will sound right.

Another option would be to place the organist next to the drummer, but keep the organ itself far enough in front of the drummer such that by the time the sound comes out it's in time with the drummer. I don't know if there are any organs that work this way, though (I'd imagine it'd have to be either mechanically complex or electronically controlled; the latter, though, gives an option for a synthesized "monitor" that the organist can hear, which should hopefully make the delayed sound of the actual organ less of a mindscrew).

This all assumes you have a large enough stage to be able to pull this off, and (in the case of fixed organs) have actually built the organ such that the sound does emerge from in front of the drums.

[ Parent | on HN ]