I hesitated for a long time but I’m back to hardware… I made my wallet explode with A4 + AR purchases!
But I still want to use some VST and don’t really want to get rid of them (U-he synths mostly), but refuse to fight with midi clock problems, and I know it’s quite always the case with DAW’s clocks.
I had this idea: getting a MM or an OT and sequence the VST’s with its sequencer without syncing the DAW through Midi clock, just setting the right BPM in it in order to get, for example, the delay times right. All the automations would come from Elektron’s sequencer. I think (I hope) enveloppes would be ok (gate triggered). The DAW would not start or stop. It would only an passive environment to play the VST’s.
My question is:is it really doable without hassle? Should I expect some major problems?
Thanks in advance for your advices and best regards,
Take a look at the result. Do they line up perfectly? No they don’t. Can you hear it? Well yeah you can. Does it matter? Well probably not, but it’s there.
Ok, so now do the same test but this time use a DAW as midi clock master for both the AR and A4.
Is the result the same as the first test or is it worse? For me using my DAW it’s pretty much the same.
The points I’m trying to make here is that:
Using a DAW doesn’t necessarily mean that you will get problems with midi jitter and delay. You need to measure and evaluate how big “problem” you have with your setup.
Using hardware does not magically “fix” the shortcomings of midi jitter/delay. There’s plenty of hardware synths, drum machines that are terrible at midi.
I suspect people who come from a pure DAW/plugin world get a bit freaked out when they buy “real” gear and start using midi clock and sending notes back and forth since they are used to always having sample accurate timing in their DAW.
I’ve never found any midi jitter with Live slaved to my OT (other than the initial second or two after hitting play). Works really well for VST control and the like keeping it all syncd up. Would you not be better looking into this route a little further before the “messier” alternatives?
By all accounts Live’s midi clock is pretty stinking as a master but it seems to behave pretty well when slaved. I’ve had a lot of fun with it. I’m little more than a hobbyist so perhaps I’m not hearing or noticing midi jitter but it seems fine to me and very stable after that initial couple of seconds sync thing that happens after pressing play.
For scenes and the likes I really don’t know. The above suggestion of midi channels should work but further OT/Live integration is on my to do list - waiting on my Octa Capture to arrive to set up and things before attemping it!
Yeah trusting your ears is the most important. There may be no need to worry at all.
Regarding MIDI jitter. There is some really interesting info out there, specifically the Innerclock Systems litmus test is one good resource. However there is no need to worry about potential issues. Focus on your setup and goals first then seek workarounds if needed.
Imagine just using your U-HE plugs and other VSTs as another sound module. You will only have a minor amount of latency from the audio interface you are using. You can probably tweak and get your latency down to the smallest buffer size without any issues.
slightly off topic, just love to discuss this, forgive me.
about freaked out: well, DAWs do not have sample accurate clock generation. what u hear when using DAWs is not tight. u have clock jitter all the time, u just dont notice it because u have sample accurate timing “inside”, eg. a delay plugin locked to bpm, drummachine seq, rewire sync, etc. if u dont believe me load some cpu hungry plugins and do a litmus test while loading. its fascinating this hasnt got better since 30 years, rather worse because software and OS is more complicated - so I suppose. mind dazzling there seems to be a tradeoff between functionality and clock tightness problem u cant just iron out with faster cpus or more memory.
just compare how much tighter a MPC3000 or a ASQ-10 is compared to a MPC1000.
AR is a complex computer too. it would be silly to expect the tightness of a simple analog sequencer from a machine that does so much.
that said of course there is simple gear that has terrible timing.
maybe its possible to have a computer hardware design that ensures correct timing to the outside world while keeping sample accurate timing inside. von neumann architecture does not seem to scale in this regard.
at same time it comes down to the programmer. Roger Linn probably could write a VST DAW that generates tight midi out clock.
Here’s the two Litmus tests I’ve done with my DAW using a plugin. One audio mixdown (aka Render) and one where I recorded the playback output from the DAW.
Cubase Audio Mixdown
This Test: Cubase 7.5.20 Build 174 using Kontakt 3.1.0.4 Audio Output: 16th Clave
Number of samples [48 kHz] between consecutive Sixteenth Events:
6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/
Maximum variation between any two consecutive Sixteenth Event intervals:
0 Samples [0.00ms]
Litmus Ranking: 5
Cubase Playback (recorded using RME Digicheck v5.71)
This Test: Cubase 7.5.20 Build 174 using Kontakt 3.1.0.4 Audio Output: 16th Clave
Number of samples [48 kHz] between consecutive Sixteenth Events:
6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/ 6000/
Maximum variation between any two consecutive Sixteenth Event intervals:
0 Samples [0.00ms]
Litmus Ranking: 5
Synopsis of the two tests:
It doesn’t get any tighter then this. No difference between playback or an audio mixdown. Doesn’t matter what plugin I use. I can add effect plugins as I wish and as long as my processor can handle it.
If your DAW software isn’t as tight as this with plugin instruments and effects you really should consider getting a new one.
I don’t really understand the logic in starting the playback in your DAW before cpu hungry plugins have finished loading? What does that prove?
When I buy hardware with a sequencer made in 2014 I expect/demand tight timing. I don’t think that’s silly at all.
With dedicated hardware boxes like the AR and A4 Elektron have full control over both the hardware and the software. No third party USB drivers, other processes running or anything like that.
Having said that I’m quite happy with the tightness of the AR and the A4.
Regarding Roger Linn; He might be the Chuck Norris of hardware drum machines but he is human. I strongly doubt he’s any better at making a Windows and/or MacOs based DAW then the people at Steinberg, Avid, Apple or Ableton.
just saw ur reply now, lars.
thanks for your input.
of course DAWs are sample accurately tight. thats what I meant with “inside”.
if the cpu comes to its limits, a litmus test just recording audio out (mix down is always fine) does not get good marks.
of course u are right I am responsible to keep the cpu load sane. but thats not always obvious, eg. u might be messing with unison voice amount in a fxpansion DCAM synth - well, that is obvious
however, DAWs do not behave well when it comes to produce a tight midi clock or, even worse, if they have to slave to an external midi clock.
thats why the innerclock boxes sell well.
AR syncs well. however we still have to be aware some user operations can fuck up the timing. to keep the cpu load always sane is basically the programmers job but the user will always find a way to do something unexpected in a complex machine.
its nothing to get crazy about. just have to start&stop a DAW-midi-gear-environment once in a while. and be aware sync might break if we do nasty things.
