Not sure if any of you dummies have seen this, but it’s pretty great.
Innerclock has done a pretty extensive litmus test for many different modern (and not-so-modern) machines.
Interesting note: Rytm is among the best for both internal and external sync.
[http://www.innerclocksystems.com/new%20ics%20litmus.html](http://www.innerclocksystems.com/new ics litmus.html)
Been a while since I looked at this, it’s been heavily updated in the meantime.
OT fairs well as does the.A4. MD is all over the place lol
Trigger Riot & Z8000 are as tight as a gnats chuff 
Wow, the TR-8 is sloppier than the old machines it’s trying to model!
Where’s the monomachine?
:-{
Interesting how some cheaper devices performed so well- Volcas, MIDI Solutions Quadra Merge…
Word to that. Makes me happy that I route all my gear through a quadra thru and quadra merge.
A question I’ve never quite found an answer to:
Syncing between my Elektrons (midi thru to midi in), I’ve noticed fluctuations of up to 1.5 BPM following the on screen BPM, although I’ve never heard any audible jitter, and even syncing other gear from them things seem to stay rock solid. Is this normal midi weirdness that doesn’t have any real effect on the output? It seems to me that sending from a solid clock seems to correct for jitter, regardless of how quickly the receiving device displays it.
I believe it’s some sort quirk in the tempo display. I also haven’t heard any affect on the actual Audio. For me it happens between the Electrons and when I’ve got cubase running things.
Interesting that it may just be in the display…maybe Elektron are just hyperactive in presenting the possibility of jitter.
One thing I’ve noticed is in sending midi to my KP3, which is notorious for poor midi timing…from the solid clock of my Elektron machines, nothing sounds off beat when receiving clock. I’ve also sent some serious abstractions of clock (6+ instruments deep) via midi out/thru and never seen anything choke, jitter, or otherwise or off beat.
I’ve had the AK sending clock to the other 4 Elektrons and every screen except the Ak is doing the tempo dance.
When I say in the display I mean possibly a small bug in the software sending info to the display.
I wonder if the display is actually jumping between internal and external clock.
Might have to do an experiment.
EDIT…Ok I just did a test. It doesn’t seem to be a fight between int and ext clock. The jumping remains constant at around 1.5 bpm.
I did notice that the fluctuations were tempo dependent. ie Turning the Ak down to 40 bpm brought the fluctuations to a complete standstill on the AR and OT. The MD and MM still occasionally jumped but only by 0.1 bpm.
So the Screen jumping effect is scaling up as the tempo increases.
Minor tempo dance on the display of the AK/A4 and AR when it’s slaving to midi clock is nothing to worry about. (Major dancing up and down 2-3 BPM is another story.)
I have an Doepfer Dark Time sequencer that has very steady midi clock output and slaving my A4 and AR the tempo display values are dancing slighty anyways.
We are actually seeing at least 1.5 BPM jump. But it doesn’t seem to have any effect on the midi timing or the audio. I wonder if there is any latency compensation going on inside the machines and the display is reflecting this. 
I find the AR very bad when I use it as master clock to sync my MD & MnM, it doesn’t start exactly on the beat, and every 4 or 5 start I have a very high offset !
When I use MD/MnM as master clock to sync AR, it is very tight.
i’m using an ERM midiclock as my master. i’d like to see how it measures up on that test
Let’s think about the process:
Master sequencer has an internal timebase, controlled by a timer of some sort (probably super accurate) and the internal resolution of steps (very accurate).
MIDI output of the master sequencer: if the firmware is well written, clock signals will take priority over note data and other regular MIDI messages (so the Innerclock data shows that some gear has sloppy output), but MIDI has timing resolution (can only send a byte every 0.32 ms).
Transmission to slave: delay in transmission down a cable is negligibly small.
Sync to timebase of the slave sequencer: should be no problem if the slave’s firmware is well-written.
Calculation of tempo (bpm) on slave: the slave has to calculate the tempo based on the recent history of timed receipt of MIDI clock ticks, by calculating the average time between ticks and converting to bpm. So various manufacturers may use various methods of calculating a running average over either a specific number of ticks or over a fixed time period. If there is any variability at all in the output of the master, the running average tempo calculated by the slave will vary. The slave’s programmer has to optimize the averaging period to try to be long enough to smooth out any quirks in the master signal and to short enough to be responsive to smooth (intended) tempo changes.
Trigger the slave sequencer at the correct tempo: again, shouldn’t be a problem with good firmware.
Display the tempo on the slave: the slave device has to display the tempo with an update rate that’s fast enough to indicate true changes of tempo but slow enough to be readable when changes occur.
Finally, but most importantly, the listener: Colin Fraser at Sequentix (home of the Cirklon) did some tests and decided that he couldn’t hear jitter of more than 1 ms (IIRC), so machines that can achieve this are likely to be perceived as in sync. Minor variations in displayed tempo may be more like a distraction to the user than a useful tool.
Seems to be the easiest solution for nice clock, but how do you manage pattern change between several machine on same time ? And how do you plug any other controller ? Midi merge ?
[quote="“twyce”"]
Seems to be the easiest solution for nice clock, but how do you manage pattern change between several machine on same time ? And how do you plug any other controller ? Midi merge ? [/quote]
i’m merging my akai MPD with the midiclock for control, which works fine
the MD and A4 are slaved to the OT, which is set to receive master clock from the ERM. pattern change works fine, no problems, whether set to send transport, or receive it from the midiclock
Seems to be the easiest solution for nice clock, but how do you manage pattern change between several machine on same time ? And how do you plug any other controller ? Midi merge ? [/quote]
i’m merging my akai MPD with the midiclock for control, which works fine
the MD and A4 are slaved to the OT, which is set to receive master clock from the ERM. pattern change works fine, no problems, whether set to send transport, or receive it from the midiclock[/quote]
Very nice ! I think I will go this way for live situation.
For the studio, I spare money to buy ACME or Syncgen. But before I will wait to have more infos about the new product in development in ERM ! He made a prototype of ACME like solution. He said it will be more affordable. Wait and see !
it’s great having a dedicated master tempo knob, and LED showing the bpm at all times.
the other great thing (and main reason i bought the midiclock) is the restart feature. it restarts all downstream midi devices seamlessly, which means you can change projects on any of the boxes and have them restart in perfect sync. amazing!
i’ve been trying to slave my multiclock with an audio track from my MPC2500, but i’m not sure how to spit out 24ppq using ERM’s sync sample? any ideas? cheers