Free DAW/Midi sequencer with can be used in sync as midi slave with Midi Clock ?

[tatanka]

Hello evrybody,

new tool, new problems. I was just fine with Tracktion as main DAW and Mulab Free for fast, easy trying out of compositions, now next problem which both dont solve.

I have a great new composition tool which only works in sync as midi master with MidiClock protocol.

Both Tracktion (no MidiClock) and Mulab (no slave at all) dont support that . So the search starts again....

By the way: as third DAW I have Mixtrack, no midi sync either....

Anybody has any ideas (I have of course looked for LMMS, Studio one free etc.) for a free DAW/Midisequencer on windows 7 which works as slave with MidiClock ? It is only to have the possibility to listen to the raw version of a composition and then change parts "live" in the composer program and then save the result as midi. Mixing etc. I will do in Tracktion.

Thanks for evry, just thought after years I finally found what I need...

Kind regards,

Richard

[lfm]

There is a reason that midi clock is not supported for sync due to how it works - not reliable for proper sync.

You have separate clocks on the different system you try to syncronize, and you send clock pulses from one system to the other, but the clock frequencies may differ a bit. That is how I understood it. So the longer the clock runs, there will be tick differences in between.

So there is MTC for proper sync.

Ask makers of your tool why they don't support MTC.

[JoseC.]

Midi Clock and Midi Time Code are two different things. BOTH are reliable for proper sync, as long as they are well implemented. MTC uses absolute time and for that reason is used mainly for video and film sync. For syncing sequencers, drum boxes and such, midi clock is the way to go (I will leave out things like Din Sync and other analog ways of making machines work together). The problem with midi clock is not that "the clock frequencies may differ a bit", since the very reason of establishing a master clock is precisely that the slave does not use its own clock anymore. In a nutshell, the problem is the way computers and DAWs handle it. Since software sequencers started sequencing audio, too, realtime midi synchronization depends also on the audio stream processing, with all the latency compensation issues that rise in this forum from time to time. This tends to introduce inconsistencies in the midi clock stream (jitter).

Anyway, the problem with syncing audio and midi means that most apps will not process audio while slaved to an external sequencer, or if they do, they will not do it too well and sync will fall apart if pushed any hard. This is the reason why Rewire was invented. If the OP's composing tool does not support Rewire, I would suggest either using the old way (some hardware module as scratch sound source connected directly to the main composing program, and then export midi to the main DAW), or exploring the use of a virtual VST rack like Cantabile. I cannot think of a free solution now, sorry.

[lfm]

The problem with midi clock is not that "the clock frequencies may differ a bit", since the very reason of establishing a master clock is precisely that the slave does not use its own clock anymore.

I see it this way:

midi clock - 24 ticks each quarternote, that is what is sent.

The actuall timebase for unit sending it and unit receiving will be slightly different. Not by a mile, but still different. So a movement by 1 tick on slave system will not be exactly the same as 1 tick movement on master system.

So the movement in time will vary after x clocks. All systems will not be in the same sample spot, if to use the final outcome - audio.

That is why midi clock is less reliable as an sync all thingy. And tempo settings on daws and external master for clock also need to be the same - or 1 tick will make very different advance in time.

Settings for sync in daws are usually audio on a certain input, and when switching to MTC that is used instead as sync.

Midi clock is not useless for all puposes - or it would not exist. But still if to choose a reliable sync - MTC is the way to go.

So I understand why daws do not implement anything but MTC which also easily translate to SMPTE code.

[JoseC.]

But the differences between master and slave are not related to the sync format, but to the way each process their own data. Even if you have a dedicated master clock device, which tend to be pretty expensive, BTW, it is up to each slave to react to the clock stream consistently. This would not be any different with MTC. The thing with MTC is that it uses absolute time, so it is perfect to sync sound to film, where there is direct correspondence in Hours Minutes Seconds Frames, but to sync midi sequencers, drum machines, synths arpeggiators and such, HMSF is meaningless. A steady pulse like Midi Clock is enough or rather, should be enough. The problem is that within modern computers the midi stream process is asynchronous, and on top of that modern DAWs make it also dependent on the audio stream. That is the reason why some people are still using Ataris and Amigas for midi sequencing, and there seems to be a comeback of analog ways of syncing external equipment, for applications where things really need to be tight.

[lfm]

This would not be any different with MTC.

I think it is different - at least depending on if daw can set projects to musical orientation as primary, like Cubase, if that means that time is truly secondary to orienting itself through project.

I know main ruler can be set this way, and linear or not, but still think time is the bottom line. The choices are most probably visual to user, and in occations about anchoring clips to timeline or to musical time - which is a common setting in all daws. I don't think that it is possible to sync audio, plugins with certain samples delay - all needs to be translated to samples/time somehow - and then you have midi clock becoming an obstacle since host timeline is different from external devices - you never receive external now positions, kind of.

Midi can send start position and stuff, to a certain bar and quaternote, tick etc - and run from there. So that is not the major issue as I see it, not implementing midi clock as external sync. But then needing to resend position in between, I guess, if clocks are lost in between - so there are disadvantages with midi clock that would create problems. So unless this protocol is fully developed there is little reason for host to implement midi clock sync.

So the crucial point I believe to be that mtc sends all timing - nothing to translate for host already orienting itself in time. And all is there in protocol to make it work.

Please correct me if I'm missing something here.

If any host is truly orienting itself(or can be set to) in musical time - then midi clock might be as good as anything else but with some ifs.
- if resending current position in between as well if a clock pulse was lost
- and most of all having the same tempo maps in external system as in host, perfectly mirrored.

That is OP's question in this topic - but never saw this in the 10 daws I 've tested - and I understand why. Nor in the 4 portastudios I had. Tape sync also translated to mtc for sequencers.

[JoseC.]

MTC does not send any info about tempo changes, so it is not really a valid solution for syncing sequencers. Anyway, of the DAWs I use, Ableton Live can be both master and slave for clock, and slave only for MTC. Sonar can also be master and slave for midi clock, but it will not play audio as slave, only midi. OTOH, Sonar can slave to MTC, but it can also be master and send MTC to several ports at once.