MIDI turns 30

[Eanna]

Cheers, I'll get thru that thread tomorrow.
A quick scan of the thread shows that there's no silver bullet, no guaranteed solution - except perhaps the Silent Way and one of their modules (ADAT -> Midi or SPDIF -> Midi).

Real World facts about Jitter is always interesting too.
TCP/IP networking has tools to discover where slowdowns are taking place from hop to hop, from client to server. No such facility with Midi, unfortunately.

[jhulk]

bud

i have built several shruthi synths

all the different filter boards and the shruthi controller

built a few midipals but olivier now sells them as a complete unit as some were findng it hard to do the surface mount chip

built an ambika six voice

i can use my transwaves in the shruthi olivier wrote me a python script called convert.py

it takes my 128 single cycle wavetable and turns it into a 16 single cycles of 128 samples+1sample each the 1 sample is for interpolation

it has 2osc +sub + noise and a pure analog filter

my fav is the matrix 12 version called mission filter it has leds for the wolfs eyes and they light up

i create sounds then sample them so that its polyphonic as its a monosynth

[Cochise]

Unluckily I have no suitable external MIDI clock source ATM.
Never had the Atari, my first MIDI sequencer in '92 was C64 based, but... not the Sequential Circuit interface here



Can't post any screnshot since my C64 has gone long time ago (even the 5" floppy looks no longer usable); No graphic interface, all alphanumerical; you can imagine the event editor screen looking similar to the MIDI monitor in SFP, plus the event timing columns.

I used to link it to this tech toy (good lo-fi sounds capable though, not velocity sensitive keys, but note velocity recognized from MIDI input)






The other piece of hardware in the MIDI chain was at the time a Yamaha RY-30



I remember almost tight timings from that rig (...I'm planning to check some recording)
RY-30 sequencer though is 24 ppq only and it doesn't sound very tight by itself to my ears nowadays, especially using layered sounds, if I remember right...
[ADDED] was remembering wrong, see my next post [/ADDED]

Its clock could even be tight, but what if clock distortions happen once the signal entered the DAW? Will MIDI events sent from a keyboard to the sequencer reflect the same distortion?

[dante]

Depends a) on the DAWs MIDI resolution and b) if it remains unquantized.

Those two factors allowing, then 'yes' I would guess

[Eanna]

For me, once a stable clock has made it into the DAW, then the likelihood of the DAW losing time is very low. You really should not see jitter in that case. It'd be the equivalent of the DAW losing time to its own clock, which would make that DAW pretty useless!

If you have a load of tracks playing back, then jitter may occur I guess - and I would guess that the jitter variance from correct clock would be small. But a stable clock available to your DAW, whether that's sourced internally or externally, should be enough to keep time during playback.

Others are talking here about Parallel Port interfaces. I've no experience of them - but this is no doubt a 'lighter' protocol than USB, and would not have to share parallel-port-time with other devices in the way USB devices on a USB host controller might... But then you're still at the mercy of the DAW's master clock thread, and the threads that send to the external Midi Ports set up for sync, which do exhibit jitter behaviour. You're just taking the USB out of the equation, which might be enough for a good midi DAW...

As for Midi triggered from a keyboard - the keyboard isn't sync'ed to the clock - it will just be sending midi note events out to the DAW... only reason that might cause mistimings is with a bottleneck in midi where midi event collisions aren't handled gracefully by the midi subsystem in your DAW, which again would be sloppy coding on the part of the DAW programmers. As before, I would guess that jitter deviation from standard clock would be small / not noticable.

I am happy to be wrong about anything I'm saying about this. My experience with this is more like thought-experiments than real expeience.... I've experienced shocking jitter from Ableton sending clock out to a little drum synth I had via an Emu usb interface, so I've tried to understand the base issues, enumerate the variables, see what variables have the greatest impact, and minimise the set of variables... that's the theory anyway!

[Cochise]

...I remember almost tight timings from that rig (...I'm planning to check some recording)
RY-30 sequencer though is 24 ppq only and it doesn't sound very tight by itself to my ears nowadays, especially using layered sounds, if I remember right...

Just checked some recordings and... I was wrong.
Beats from my old MIDI rig actually weren't sounding so tight.
I'm gonna post some examples below; these were mixed home in early 90ies using hi-fi amp and speakers for monitoring, no room treatment, recorded from mixer to a stereo cassette deck; unluckily the tape is no longer in good conditions.

http://dl.dropbox.com/u/71859916/92/C64 ... attern.wav

http://dl.dropbox.com/u/71859916/92/C64 ... attern.wav

http://dl.dropbox.com/u/71859916/92/Kal ... 2BRY30.wav


This other example had been recorded in '92 using the Yamaha RY-30 alone (plus an Ibanez delay pedal in the send/return mixer chain, probably), no MIDI; and it beats tighter to me...

http://dl.dropbox.com/u/71859916/92/par ... ng%231.wav


...good times when I was concretely studying/composing, kind of too self made probably

[Cochise]

Actually the Ibanez Digitally Controlled Pedal model PDM-1 (in picture above) had been used in the send/return chain especially in tracks involving the whole rig...



Probably in some track the delay speed didn't exactly fit the tempo, making it harder to understand what isn't tight

[Cochise]

Here some Super Track screenshots

http://www.richard-aicher.de/html/c-lab_supertrack.html

[Bud Weiser]

..., no guaranteed solution - ...

Real World facts about Jitter is always interesting too.

There isn´t any guaranteed solution for MIDI jitter w/ computers.
MIDI-jitter, like in the audio world too, is not only a matter of drivers but also the clock.
Since all the connections are on the motherboard, the motherboard´ss clock is responsible as well as the drivers are.
In addition, some MIDI interfaces use a different clock than the others,- there´s TGT and QPT (or is it QPC ?).
I´m actualy not sure if is called Query Performance Timer or -Counter ...

As examples, Steini MIDEX 3 and 8 use QPT, Emagic AMT the other one,- RME too.
The interfaces coming w/ DirectMusic drivers and using QPT work best w/ "use system time stamp" checked in the sequencer,- others want it unchecked.
Matching hardware is definitely a factor and you might have luck w/ a motherboard but not with everyone.
Other factors like MIDI handling of the sequencer/host come in addition.

Bud

[Bud Weiser]

Here some Super Track screenshots

http://www.richard-aicher.de/html/c-lab_supertrack.html

I remember we worked w/ that in pro studios, DX7 as masterkeyboard and Obie-8, Prophet 5 and Oberheim DMX and Linn LM-1 connected.
It worked great at that time and before ATARI came.

The SCI C64 sequencer was just only the beginning.

Bud

[Cochise]

Yep, screnshot shows 1985 as release date. I got it later in '92, paying the equivalent amount of almost 40 Euros for interface and software; I remember I had to go to Milan (1250 Km from here) to purchase.
I worked some month in a pub (not as a musician) to buy the RY-30.

In '85 (probably) the only sequencer I had the chance to handle was the one featured in a Poly-800 Korg, which I had here for some week; MIDI was a misterious word for me at that time.

[Eanna]

Hi, the good news is PizMidi site is back online:

http://www.thepiz.org/plugins/

The many utility Midi plugins are here:
http://www.thepiz.org/plugins/?p=pizmidi

[Cochise]

http://www.atari.com/arcade/asteroids#! ... roids/play

Something reminds me that in the 90es I was probably convinced Atari was for videogames. (The fox and the grapes inspired behaviour?).

[Eanna]

Level 1 complete.

Screw music, this is the nuts!

Cheers Dude...