Midi Cable

[NiceKid]

Well, me and my friends need to sync two laptops, it's easy when we do it at home. Just plug the midi cable, slave one of the laptop, and done. I wonder, if we do it on stage, and we use longer midi cable (well, actually, never seen a long midi cable) is it any problem with that? Latency etc?

[HammondToby]

MIDI works with cablelengths up to 6 meters, then the signal must be amplified to prevent signalloss. I've got some 6-meter-cables for my studiorig and never had problems with them. But keep in mind, that the latency increases with the length of the cable.

Best regards
HammondToby

[Athox]

Just remember to use the same setup every time (same cables for each function), so you'll get used to said latency, and work around it by pressing keys/pedals just a few milliseconds ahead of normal.

[Nico]

Why the latency increases? The signal in the cable is at lightspeed so it wouldn't matter how long your cable is.

Nico

[Grey Loki]

Actually, it's very far from lightspeed indeed. In large gigs (usually open air gigs, though the recent Diana concert in Wembley stadium used them too), to reinforce the sound from the stage to the back of the audience, some companies install delay towers (basically extra speaker stacks, but off in the middle of the crowd). However, if you just 'let rip' with all these speakers, there'd be all sorts of chorus and flanging and phasing going on between the sounds from the two sets of speakers, because the delay tower's signal has got further to travel. To compensate for this, you get a delay unit (not unlike a tape delay of yesteryear) and install it into the line for your main FoH speakers, so that they produce sound 'in time' with your delay towers.

At least, that's how I think it works :D

[Baddox]

Grey Loki: That's because the sound coming FROM the speakers goes at, well, the speed of sound, which is a LOT slower than light AND the current in a MIDI cable. That said, the SIGNAL would theoretically travel at or near the speed of light, since it's basically pushing an electron out the other side the instant a new electron gets pushed in one side. With a digital signal such as MIDI, you'd be a lot more concerned with signal loss, which increases as the length of the cable.

EDIT: Just after posting, I was thinking that most likely, the impedance of the copper in the cable would impose a delay that if I remember correctly would increase with the length of the cable. However, I think this would still be minimal.

[Deceit]

Grey Loki's got a point.

An electric signal (which is a wave) propagates in real media (cables) with attenuation and delay as well; this is due to impedance, too. A longer cable means higher attenuation and delay, too.

And, we're rather far from lightspeed.
An ideal media without losses (which is impossible due to complex impedance Z=R+jX) would mean a wavelength=lightspeed/frequency, but this is way far from our real situation.

The real deal with digital signals is timing and error avoidance. Errors are due to noise. A lowpass analog signal must avoid distortion for an ideal transmission, which is rather unlikely to achieve.

We can allow distortion on digital signals but we cannot allow errors. This is why, instead of the so-called Signal to Noise Ratio, the quality meter for a digital transmission is the Bit Error Rate.

As for the minimal delay, we should check the data sheets concerning cables and the line attenuation coefficient.
Right now I'm translating my mother language subjects to English, I hope it's clear anyway.

Oh and yes, different speaker configurations do change the perception of sound. Theoretically, you could even find a situation where you couldn't hear, in certain spots, a single sine wave (speakers do interfere both in phase and amplitude of sound).
Anyway as many of you know any sound is a sum of infinite sine waves (Fourier's Theorem) so it's actually rather hard to occur in such a situation.

And I guess the speed of sound has its own effect too.
There are also echoes, which are basically waves bouncing back to us. A chorus in the end is a sum of rather short delay lines, right?

Reality is a (chaotic!) sum of these phenomenas, the deal is understanding which one is most frequent and has the greatest effect; getting back on topic, if you had to cope with >6m cables you'd better get a ->POWERED<- midi patchbay of some sort; I've been using a MidiMerge 2x2 for a while (the pink one -_-), which is the cheapest solution to the problem anyway. You have to give power (amplify) the signal in order to avoid problems due to the length of the line.
The Merge usually operates passively with the power coming from the midi devices. That's why there's a power jack whenever you're using long cables.

Deceit.

[Baddox]

There is no error checking or avoidance with MIDI. It's a one-way signal with no handshaking (the synthesizer can't say to the keyboard, hey, the parity on the last signal didn't match the data, send it again). Therefore, I would assume latency on the actual cable would be minimal, even negligible. The MIDI latency we usually notice is due to the computer or synthesizer we're using to output MIDI data as music. Anyway, when you're talking about delay modules for speakers, that is only because the speed of sound, not the speed of the signal in the cable. Think about it, if you're playing live, your monitors are being fed from a cable that goes from the instruments on stage all the way back to the sound board and all the way back again, and there's no noticable delay introduced there.

[Deceit]

There is no error checking or avoidance with MIDI.
Midi latency is a fact, it is proven to exist in many cases. It is not as frequent as malfunctioning cables though ;), and it occurs most commonly in complex chains of signals (the so-called midi thru daisy chains).
There is no handshaking of course, I was talking about the principles of lowpass digital signal transmissions in general.
The latency on the actual cable is something you could measure but I guess it's a minor issue, the point is long midi cables (>6m as HammondToby pointed out) are subject to attenuation to the point you have to amplify the signal by spending electrical power on it. I could take a picture of the interior of my midi merge unit. I am not assuming MIDI latency is audio processing latency.
Anyway latency on cable connections may be negligible in this case, yes. But we're confusing two phenomenas and two totally different types of connections. Midi is a digital signal, while audio is analog.
The topic is about midi cables, and the correct, engineered answer to the question
we use longer midi cable (well, actually, never seen a long midi cable) is it any problem with that? Latency etc?
Is NO, latency is a negligible effect in MIDI transmissions, as they are still in a close range. The signal does not reach the speed of light as transmission lines are subject to capacitive, inductive and resistive - conductive effects. It is anyway fast enough for all the delay effects to be almost null.

The ONLY issue with midi cables is the one HammondToby clearly pointed out. After about 6m of cable, the midi signal needs amplification to be received correctly by another midi device, which must be supplied by a midi merging or routing unit.

Baddox, I tried to analyze the problems against each evidence, by considering some of the "negligible" and "non-noticeable" effects, which are anyway a part of the whole reality.

Grey Loki, in the case of really long wires and analog transmissions that delay could occur. But, it's true, when you add speakers to the environment it's probably to enhance the effect rather than correcting delay issues (well, you could also avoid crosstalk and cancellations but that's a long one to write about). The acoustic analysis of an environment is rather complex and involves sound sources as well as geometry of the room, damping or reflecting materials as well as density and pressure of the air itself.

As for the case of Wembley, I guess that may be a distance - yet most likely quality - issue.
speed of sound at sea level = 340.29 m / s means you'd hear sounds exactly one second after they are being produced at 340 m from the sound source.
In just about 150m you'd hear 500ms of latency which is still a noticeable delay.
If the source is moving (which is hopefully not our case) you also experience Doppler effects...our world is damn cool.
Deceit.

[hanton]

It maybe somewhat off-topic, but isn't it the doppler effect which is being utilised in leslies?

[Baddox]

It sure is.

[Omega Monkey]

Just to comment, Grey Loki, I think you got the delay tower thing backwards. What they do is delay the rear speakers by some number of ms (which you can figure out if you know the distance they are from the mains and the speed of sound) so that they dont cause phase interference and whatnot with the sound arriving at the back of the "venue" (usually a bigass field) from the mains.

So in other words, if your rear speakers are 50 ms worth of sound distance from the mains, you add 50 ms of delay to the rears (or just slightly less), so that you are getting the same "sound moment" (for instance a snare hit) at close to or precisely at the same time for both sets of speakers. That is because the "cable time" is negligible, but the audio delay can be very significant. Really, anything beyond about 20 ms or so is going to be noticeable.

The only thing Im not sure about is how the delay would affect listening positions between the 2 (or however many you had) sets of speakers. But I think the attenuation from the natural dispersion pattern of the rear speakers (ie most of the sound would be going "forwards, so any delayed sound would be volume masked by the mains) would minimize that.

[NiceKid]

why i asked this question is, i watch a keyboard player, he take a laptop and play some softsynth with it. The my friend asked him, how about the sequencer? the backing track is played with another laptop, downstage. :redface:

so after 6 metres we had to use "midi amplification". I'd never saw this "thing". Something like rack or stompbox? *sorry, really" blind about this kind of stuff :redface: *

edit : oops forgot to thanks for all the replies :)