MIDI rhythms sound weird, slightly off

[douglas stringer]

I've been noticing that some of my midi loops (sent from PT5 to my drum module) sound a little bit "off." sometime. It's almost as if the beat skips ahead (a few milliseconds) or something. Others listening to my system have heard it too. It's like every once in a while (erratically) the beat stumbles.

The beats are 1 bar looped and fully quantized.
I'm using a MOTU Midi Express USB. I'm currently using the USB port but I have also tried it with the serial. I also am using two active USB extensions when I'm using USB and two 9 ft serial male to female when I'm using serial.

Has anyone else seen/heard of this??

Thanks!

Doug Stringer
[email protected]

[Dr. Marius]

We have made intensive measurements on (almost) all available Midi interfaces. Even though the Midi Express USB provides steady timing, it surely has some delay on its outputs (aprox. 5 msec longer than Midex 8/Unitor etc). Some drum machines (like the Korg Electribe series), has a rather slow response time. Combined with the delay of Midi Express, you will get VERY bad timing -especially when using beat clock. My advice is to get: 1) A Steinberg Midex 8 (which provides the best available USB/OMS timing, and absolutely trouble-free operation). If you haven't yet upgraded to OS 9.2.2, you must do so (USB drivers has been improved a lot)
2) A Griffin "G-Port" adapter + an old serial interface (Like Opcode Studio 4/128X etc). Those interfaces are still the most reliable solution for OMS/OS9. Be aware that the new Midi engine in OSX, might not support serial ports at all.

Best regards

Dr. Marius

[jackruston]

There are issues with the midi timing in pro tools. There have been quite a few threads about it in the past. I have found that when doing drums etc, it's better just to import/record the loops as audio and work with that instead. Or you could use logic to build up the midi rhythm section and import it from there as audio. Running plugs etc seems to make the timing worse.
Jack

[Dr. Marius]

According to our intensive measurements, the Midi-timing in Pro Tools, is just as good as in ANY other application (Logic/Cubase etc.). Emagic/Steinberg just have a better marketing dept.......!

With all respect.....

Dr. Marius

[MikeC]

I've experienced the same midi timing problems/sloppiness using the same Motu USB interface. You can really hear the timing problems with complex arrangements. I understand that this can occur with USB interfaces and that a serial connection to your computer will help reduce the problem.

I'm using Digital Performer 3.0 for all midi work where the midi timing is rock solid. Their proprietary Midi Time Stamping (MTS) tecnology seems to do the trick. I ought to try PT with a serial interface as it can get tedious re-recording/re-importing the audio.

Cheers,

Mike

[tld]

Regarding the midi timing of Protools itself, I have to agree with Dr. Marius. I did a test and tried everything I could imagine to create midi slop in Protools, testing with an SCII card to rule out midi interface delay etc. This was on a stock 9600/350, PT 5.1.1, OS 9.0.4. While recording the original midi track, I recorded audio from the SCII to disk as a reference. I also added several other midi tracks playing the Access Virus plug that were simply loaded with streams of controllers, pitch bends, you name it, just to keep Protools midi real busy. Then I duplicated the original midi track several times, recording the resulting audio from them to disk. Remember, this is a stock 9600/350.

The worst midi slop I got was just over 1 ms, which by any standards is good.

Tom

[Barnaby Bristol]

With a G4/533, OS9.1, PT5.1.1, MTP-AV, G-port connected with 4 MIDI ports to a Micro Express to GigaStudio, we've exhibited glitchie timing as expressed here. Not slop like one encounters in loaded MIDI, but hiccups and micro jumps in timing. I haven't spent any time looking into it yet.

[douglas stringer]

The problem is very undermining to using Pro Tools for midi. Has anyone been able to confirm that it is a problem exclusive to Pro Tools? Or to USB midi boxes?

[img]images/icons/confused.gif[/img]

[PacMan]

Let's face it. Midi timing has always sucked. Regardless if you have a usb interface or stealthport, etc. I don't believe that midi timing is any worse on PT, than on any other platform. That sounds like a load of propaganda to me. [img]images/icons/smile.gif[/img]

To some people, 1 ms is not that bad. For others, its unacceptable. The reality is that most outboard midi gear has delays that are more than 1 ms. (especially older gear)

Personally, for me it all depends upon what instrument I'm using.

If you're so worried about timing, then use a software sampler, which will give you sample accurate timing. Can't get much tighter than that.

I still use plenty of outboard midi gear, for a variety of sounds, but if there's any part that requires Rock solid timing, than a software sampler is the only way to go for me.

My akai just sits there gathering dust nowadays [img]images/icons/smile.gif[/img]

My 2 cents

[Robin Hughes]

Its worth noting that some timing skips in drums can be the fault of the midi spec and its serial nature.

It is impossible for a unit to receive 2 midi comands simultaneously , including controller data , aftertouch , sysstem commands and note offs.

Working with samlers it can be advantageous to add stepped amounts of silence to the beginning of a sample and adjusting the midi position accordingly. Working in the list editor can pinpoint these details.

I read an article on the net where the author described how the Haas effect can dramatically effect perception of rythms.

Tapping a pencil along to a metrenome it is possible to get the timing so close that either the metronome or the pencil click dissapears. This subtle effect will be in place in a rythm track. As long as everything is repeated identically perception will not change . However if one note is randomly prioritized to occur before another ( due to midi information being simultaneously transmitted and the system having to make a choice ) , then a sudden perception of a change in the rythm will occur.