MIDI cord delay

  • Thread starter Thread starter midi
  • Start date Start date
M

midi

New member
Here's an interesting question: How do I calculate the latency introduced by a MIDI cord? That is how long does it take for the MIDI signal to travel per foot/meter of MIDI cable?
 
Here's an interesting question: How do I calculate the latency introduced by a MIDI cord? That is how long does it take for the MIDI signal to travel per foot/meter of MIDI cable?

You don't.

Unless you are considering really long lengths, the latency of a cable is negligible.
 
Electricity moves 18inches per BILLIONTH of a second.

Unless you're dealing with miles and miles of cable, it's irrelevant....
 
...according to monster cable its not :D

According to Monster Cable, the speed of light is a show stopper :rolleyes:

I've been working with MIDI for 20 years and I've never has to deal with MIDI latency introduced by healthy cabling.
 
Thanks everyone. I'm not running miles of cables (or anything like it!). It's good to know I don't have to worry about the cables introducing latency.
 
Yah, it ALWAYS takes a finite amount of time for electricity or
light to move from point A to point B....

Back when I was a beginning programmer a couple of decades ago,
my prof told us the reason that Cray Supercomputers were built in
an arc was to keep the longest wire in the machine down to less than
18inches in length.... that kept timing errors from creeping in and
messing up the calculations as the data was passed from cpu to cpu.

cray_Y-MP8_supercomputer.102627717.sm.jpg
 
Haha yeah I know, I'm just joking around. 18 inches :eek: Now those same circuit paths are microscopic, it's amazing how far we've come..
 
Midi is a 31.2 kilobaud serial protocol.

This means the 'what note, how long, how hard' information, as well as everything else midi is capable of expressing is transmitted single file @ a speed not greater than 31.2k.

It also means that there will always be SOME delay between pressing a command on a controller and the information arriving at the midi receiver. One thing that adds very little if anything at all to the midi delay is using a 20 foot midi cable instead of a 1 foot midi cable - once it's in the pipe traveling an few more feet is not going to slow the serial digital stream much.


.
 
If you're talking about just the cable, then forget it. Even for a 15 metre length the delay is so minute it's faster than we can think, let alone hear.

but ... and it's a big BUT ...

If you're talking about a MIDI to USB adapter, then there's latency and jitter in the conversion process, embedding the MIDI messages into USB data packets.
Martin Walker of SoS wrote in depth about his experiments in this area many years ago.

SysExJohn.
 
Something I thought of that I didnt mention the 1st time - Cables themselves add no latency. (ok negligible, the time it takes the speed of light to travel 6 feet - trillionths of a second, whatever...) Long cables are a problem because of induction and capacitance causing interference with the signal, and resistance weakening the signal. If there's latency, it's the protocol like ssscientist says, not that you could hear a 1/31200 second delay anyway. My trusty time-keepy drummer isn't that accurate! :D The latency you hear is ALL in your computer. SysExJohn says it all. If you had a hardware midi controller, a midi cable, and a hardware synth, you couldn't detect latency even if the cable was a mile long.

It is NOT the cable!
 
Yep! And it's the inductance and capacitance that limited the maximum length, specified by the MMA (MIDI Manufacturers Association MIDI Manufacturers Association - The official source of information about MIDI), of a MIDI cable to around 15 metres or 50 feet. Although using Cat 5 cable it's possible, I believe, to exceed this length. (No mile long cables, please!) :facepalm: ;)

The MIDI latency can be in the MIDI to USB converter, often contained in a little box in the middle of the cable, especially if that box is driven by generic Windows drivers. It will also give jitter, a considerable variation in the the latency.

I find the MIDI and audio latency using the interfaces on an Emu 1616m both PCI and using PCMCIA to be quite acceptable.

SysExJohn.
 
Bandwidth congestion or "choking" due to sending too many messages through daisy-chained devices is a far greater concern than cable delays per se.
 
Back
Top