Elektron-Users

again - in 2004 - if you are designing/manufacturing a world class drum machine - regardless of the fancy voice modeling and DSP synthesis you build into it - if compromises in the OS mean it can't fire on time properly and more significantly if it falls behind on much earlier products and designs then it needs to be addressed - don?t let 'em off the hook!

Regards - David


Actually i think this has to do with listening experiance...the MD sounds and behaves timingwise like a computer sequenced system of the 80?s..we have the same lag here ..same groovy circlyng that happend by squeezing the note data thru a midi cable..you always had wobble around 2 ms...similar groove
My guess is that the Elektrons didnt had acces to real drummachines and was living rather in a computer geek world...

The Sidstation indicates that...
Some design flaws in the user interface of the MD indicate that aswell... Its made by programmers that do music..
while the mpc3000 was done by Linn..a musican that learned to program...

Linn started with clocking and sequencing the thing... sound engine later..
Elektron started with the soundengine...clocking and sequencing later...

You can see that in a lot of details like that mute and pattern access need double keystrokes...
no problem for a geeky point of view...but something the direct access used musican have to adapt to... on an old mpc each function has a dedicated button. The mpc is simple in relation to the MD..but all sync issues are 100%

The MD works well when you work in a single pattern...
Or you use songmode..but only geeks would do that

there we come to another point good point about the midi alike timing of the MD

The MD with it limitetd multibar abilitys is ment to be arranged in a DAW sequencer.. so with the highspeed midiinterface you can record all your moves in a DAW and after that run the drumsequencing from the DAW using the MD as expander without loosing the timing...

When the MD would be absolutly straight you couldnt do that without loosing the tightness..

So in general wobbely timing is bad..but as with the MD you can choose internal or external sequencing without loosing the feel...

quite unique for drum machine...

again we see that the MD is rather an soundexpander that later got a sequencer than a drummachine that is based on a sequencer...

Its quite a lot of sound they squeeze out of 2 dsps... when i would use my 8 dsps on the G2 i wouldnt be able to simulate a MD...

dreg wrote:
just try to finish a song is my take :-o

Word!


however...its a forum to discuss such issues... but in the end of the day the question is ..
can you do music with it...and the answer to that question is that are rather to many MD records out there than to few..so YEs...it works out...
Timing is not precise..but within the range where it still sounds good and natural. Not irritated as you can have it with sloppy DAW/soundcard combinations.

Toni wrote:
I find this thread very interesting and I'll add my little counter-voice (not that it means to anybody anyway).

............
ps. nevermind the way you feel, you have got to see this as an act of love.


some true in your article aswell..its not about measurments its about the feel of a thing. And we are sensetiv enough to feel frequency related phenomens well above 50k...and timing is a frequency related phenomen..but..
its like water...when its meandering in nice ways it feels good...everything on a 100% straight clock wouldnt sound nice aswell..something jumping around in an edgy fashion is the worst...circular motion...like water..

what pushed against your nerves by listening to the mpc was that there is absolutely no movement..its so straight that it hurts..

in this sense the MD is better...but the wide range of swing causes editing problems...

I ve circular timing on my nord modulars and the 808 without variating loop length

In ideal it would be possible to set one track into clock precission mode and there you place a sound event as you like that gives you the loop marker.. the other 15 tracks can cycle as they do now... or you swith that precission track mode off and all 16 tracks do as they do now.

rudebop wrote:
.but..
its like water...when its meandering in nice ways it feels good...everything on a 100% straight clock wouldnt sound nice aswell..something jumping around in an edgy fashion is the worst...circular motion...like water..

what pushed against your nerves by listening to the mpc was that there is absolutely no movement..its so straight that it hurts..

in this sense the MD is better...but the wide range of swing causes editing problems...



with all due respect we've gone over this already,
features like SWING exist so you can humanise your grooves. Maybe you dont like the mpc3k because you dont know how to program swing yourself, and you prefer the md because by the fact its timing is already off 'it does the swing for you' .....personally, if I'm not using swing I want the steps to be straight. I've had problems recording straight loops into ableton live, from the MD...and just assumed the problem laid elsewhere...but the push-pull thing is making sense.

I'm not down with the 'let it be sloppy, cos thats like human swing' rhetoric. Also, at first I thought this timing issue was down to midi & sync, but learning of the INTERNAL timing issues sucks, also explains alot of the nagging quirks I've had in the past. I just totally assumed the internal timing was spot on, and the midi could be improved.

anyways, good to know. Keep up the good work guys

LOL @ toni ^^

there we come to another point good point about the midi alike timing of the MD

The MD with it limitetd multibar abilitys is ment to be arranged in a DAW sequencer.. so with the highspeed midiinterface you can record all your moves in a DAW and after that run the drumsequencing from the DAW using the MD as expander without loosing the timing...

When the MD would be absolutly straight you couldnt do that without loosing the tightness..

So in general wobbely timing is bad..but as with the MD you can choose internal or external sequencing without loosing the feel...

quite unique for drum machine...

again we see that the MD is rather an soundexpander that later got a sequencer than a drummachine that is based on a sequencer...
i have never ever experienced the MD as being ment to be sequenced from DAW... that's really your own take on it

divi wrote:

i have never ever experienced the MD as being ment to be sequenced from DAW... that's really your own take on it


Its not my take..its in the design of the machine drum itself.. i just tested it and was astonished that it didnt fucked up the timing..later i learned that this is because of an allready "fucked" timing....
so sequencing the MD over midi dont sounds worse than internal sequencing...
which is in the end of the day rather prctical because you can choose each way of working without destroying your established listening experiance.

Khazul wrote:
Interesting thread.

Some additional observations:

1. Myself and Tarekith have both observed agravation of arpegiator timing issues with a Virus TI (in old TI OS versions) when using an MD as the master clock.



Hi Khazul - thanks for the reply and detail - all talk of Sync aside which is very important I know - but getting the internal clock/event tempo stability straight needs to be nailed first which is what these test show as being flawed.

REgards - David.

Please! Please! PLEASE! - I know all these cross threads about Midi and Sync and external control are all valid issues in their own right but they dilute focus on the core issue - I agree that personal timing and feel are subjective but what I think must be stressed is getting away from the 'Mystic' mysterious 'black art' theory of sequencer timing. There is nothing mysterious about it - one of the reasons this never gets resolved is because there is a tendency to reduce what is absolute science and logic into the realm of magic and voodoo. Event/Step/Clock precision is science. Period. Time is not a variable. OK in a Black Hole maybe but not in music. One second is a bullet proof absolute. In the same way a quarter note time division at any given tempo is an absolute. This is NOT OPEN to debate. By all means choose where you want notes to fall along that absolute timeline to create any mystic feel you like but don't fuck with the timekeeper. Lock it down, keep it running, dont interupt it's flow - all things that need a reference or trigger from that time clock must get it WITHOUT compromising the solid and continuous flow of the clock itself.

Regards - David
www.innerclocksystems.com

For those interested - a quick history lesson that relates to my previous post - before CPUs and DSP - sequencers were built around simple voltage gates, switches and timers. The master tempo clock was either an internal analogue VCO that generated regular Square/Pulse waves or; an externally generated Square/Pulse wave running at a particular tempo. Just like today, there were tight and sloppy Square Wave clock generators - BUT (and this is the important bit in all this) - the way the sequencer designs of the day responded to the incoming Square Wave Sync Pulse was fundamentally different to the way most/all modern sequencer devices are designed today.
In a vintage (say 1978) step sequencer - the Tempo Clock voltage Pulses hit a combination of gates and switches - and then - depending on the gate and switch positions - a voltage at the precise sync timing clock pulse is then fired off to generate a note or pulse etc. Because these designs are all voltages (real electrons) - they are all running around at light speed (or close to it depending on what you read) so pulses in and pulses out are effectively instantaneous. The critical thing in all of this is that the switches and gates are all passive to the Sync Pulse Tempo Clock stream. They do nothing to it in order to do their job. There is no possible way for any component or operation state of the sequencer to pause, stall, rush, drag, jitter or halt the tempo clock pulses. When designers moved away from this proven method of handling event synchronisation using discreet and passive components and instead towards CPUs and EPROMS running instruction sets to handle many different tasks - this is when things started to go deep south on the sync front. And it?s easy to understand why: The CPU and OS all-in-one design philosophy means that tempo and sync are just two aspect the machine has to manage now governed by the speed of the CPU and the number of tasks it has to achieve overall. Of course we now have CPUs running at very high speeds but they are also running incredibly deep and complex instruction sets. Designing a modern sequencer that has all the features we have come to expect while still maintaining the sort of razor-sharp tempo/clock performance we had in 1978 is not easy - sure - but not impossible - Roger Linn proved that hands down with the MPC-60 and the MPC-3000 and the later only runs a 16MHz processor at its core! Its not about speed - it's about smart design that uncompromisingly puts clock/sync/tempo/event scheduling at the absolute top of the feature list and building everything else around it. Add features - sure - but if they fuck with the clock - find another way to get there. And if it can't be done - leave it out!

Regards - David.