MPC X, MPC Live, MPC One & MPC Key 61 Forum: Support and discussion for the MPC X, MPC Live, MPC Live II, MPC One & MPC Key 61; Akai's current generation of standalone MPCs.
By Yorgos Arabatzis Fri Dec 15, 2017 11:52 am
The problem starts when Live is in slave mode so you’re bound to use it as the Master and i need Tempest to be the master..Never had this problem when i was using the Octatrack..
Off timing is noticeable when you change bpm in real-time and you have a warped Loop locked to the tempo..We have a song in our band that in a section i have to drastically change bpm from 90 to 30 and back and the transition is not that smooth...
AKAI must fix it it’s critical!
By soultrane28 Fri Dec 15, 2017 3:23 pm
analoguebubblebath wrote:Ok I have some fairly valuable research to share after extensively testing the brand new MPC Live unit Akai sent me (after sending back my old one due to apparent hardware issues).

On both 2.0.3, and the latest firmware 2.0.5 I found the exact same timing results when testing at 120bpm:

-The standalone internal sequencer timing is rock solid. Not even 1ms of clock jitter. (The previous internal clock jitter I was experiencing must have been from a bad unit, hence why Akai replaced mine).

-This is the big one: When slaving the MPC to an external clock (I used a Cirklon), the clock jitter is HORRIBLE. I measured up to 15ms clock jitter, which anyone can notice. This renders the MPC completely unusable in musical production for me when slaving it to an external clock. You can easily hear when comparing metronome clicks with other gear the MPC is all over the place. Very disappointing performance.

-Slaving the Cirklon to the MPC, however, the timing is absolutely rock solid. I tested the metronome clicks on my Analog RYTM which were sync'd to the Cirklon, and they were dead accurate with not even 1ms of clock jitter. **Funny note here is that the BPM display readings on the RYTM and Cirk were constantly shuffling, of course making you think the timing is not accurate. When I recorded and measured the clock jitter here, the displays were wrong (they must be rounding up or something). The bpm displays would shuffle back and forth from 119.9 to 120.1, while the performance timing was perfect.

So overall its incredibly disappointing to see the MPC's timing instability when slaving to an external clock, which i'm sure is how a lot of others setups are arranged. It's slightly redeeming that slaving other gear to the MPC works properly, though I will have to change a bit of my setup now in order to make this work.


Sorry about your frustrations but as I'm understanding this, if the MPC X (or Live) is the timing master, the new MPCs actually perform better and with less jitter than even the 3000 (which I believe had 0.1ms of a jitter as master or some minuscule number)?

That's the way I want to use it, and then bounce the audio, but I'm waiting for confirmation that the internal clock is as solid as the 3k or 60
By smellypants Fri Dec 15, 2017 5:57 pm
soultrane28 wrote:
analoguebubblebath wrote:Ok I have some fairly valuable research to share after extensively testing the brand new MPC Live unit Akai sent me (after sending back my old one due to apparent hardware issues).

On both 2.0.3, and the latest firmware 2.0.5 I found the exact same timing results when testing at 120bpm:

-The standalone internal sequencer timing is rock solid. Not even 1ms of clock jitter. (The previous internal clock jitter I was experiencing must have been from a bad unit, hence why Akai replaced mine).

-This is the big one: When slaving the MPC to an external clock (I used a Cirklon), the clock jitter is HORRIBLE. I measured up to 15ms clock jitter, which anyone can notice. This renders the MPC completely unusable in musical production for me when slaving it to an external clock. You can easily hear when comparing metronome clicks with other gear the MPC is all over the place. Very disappointing performance.

-Slaving the Cirklon to the MPC, however, the timing is absolutely rock solid. I tested the metronome clicks on my Analog RYTM which were sync'd to the Cirklon, and they were dead accurate with not even 1ms of clock jitter. **Funny note here is that the BPM display readings on the RYTM and Cirk were constantly shuffling, of course making you think the timing is not accurate. When I recorded and measured the clock jitter here, the displays were wrong (they must be rounding up or something). The bpm displays would shuffle back and forth from 119.9 to 120.1, while the performance timing was perfect.

So overall its incredibly disappointing to see the MPC's timing instability when slaving to an external clock, which i'm sure is how a lot of others setups are arranged. It's slightly redeeming that slaving other gear to the MPC works properly, though I will have to change a bit of my setup now in order to make this work.


Sorry about your frustrations but as I'm understanding this, if the MPC X (or Live) is the timing master, the new MPCs actually perform better and with less jitter than even the 3000 (which I believe had 0.1ms of a jitter as master or some minuscule number)?

That's the way I want to use it, and then bounce the audio, but I'm waiting for confirmation that the internal clock is as solid as the 3k or 60



Where did you hear this may I ask?

Its been reported in this thread that the Live acting as the master has less than 1ms of jitter both externally and internally which is very good.

The 3000 though is legendary... About 4 samples of jitter max sequencing externally, and about 20 samples with its internal engine.

At 44.1khz 1ms is about 44 samples. :smoker:
User avatar
By MPCWeapon1 Sat Dec 16, 2017 1:56 am
analoguebubblebath wrote:I've had my MPC Live for 2 weeks, over the weekend I ran some timing tests and found the MPC Live's sequencer timing to be VERY inconsistent. Here's my post straight from the Akai Professional community forums:

"I've been integrating my new MPC Live into my analog hardware setup, where I am syncing the MPC to other sequencers and drum machines. On first adding the MPC to the setup, I started to hear a bit of timing discrepancy between the MPC's drum patterns I had made and other hardware.

I did a bit of research/testing and ended up recording clicks from all of the devices simultaneously. In my DAW I then went and measured the distance between clicks from each device in milliseconds. The 2 main pieces of gear I use are the Cirklon Sequencer, and the Analog RYTM. Both of these devices showed very consistent click distances. At 150bpm the distance between the start of the click waveforms was consistently 400ms with both of those machines. Very solid timing.

The distance between the MPC clicks however varied between 396ms and 408ms, without any sort of patten as to when the shorter or longer distances appeared.

12 milliseconds is VERY Noticeable in this case, which is why I was hearing my MPC drum patterns sounding a bit sloppy when played alongside some tight sequenced synth lines from the Cirklon.

To prove the accuracy of this test, I did the "waveform start measuring test" again recording audio of 3 different MPC Metronome types, and 2 different drum programs. The tests were done zooming in on each waveform start to make sure I was getting a highly accurate reading.

I first did this test with the MPC receiving MIDI clock data from the Cirklon, but I ended up getting identical results when I recorded and measured all the clicks again by turning of MIDI clock receive and running the MPC separately by itself."


Man, I was really thinking about copping a MPC X to use with my Motif...now I'm not so sure. Too much money for the sequencer not to be rock solid.
By soultrane28 Sat Dec 16, 2017 2:13 pm
Ok but at about 1ms we've reached the point of diminishing returns, imo. BTW, the poster merely says "less than 1 ms" but not how much less...it might be sample accurate.
User avatar
By Danoc Sat Dec 16, 2017 5:45 pm
Im thinking 0 ms :hmmm:

soultrane28 wrote:Ok but at about 1ms we've reached the point of diminishing returns, imo. BTW, the poster merely says "less than 1 ms" but not how much less...it might be sample accurate.
By renegadebliss Thu Dec 21, 2017 7:21 am
Frisbi wrote:I could say that every time I used midi clock to sync different machines i had some problems, issues and latencies.
I think it's something related to midi protocol, which is a really old technology.


I has nothing to do with really old technology. It has to do with bad programming. Whether it's the programming by Akai with MPC software , or on the Linux distribution they are using, it's bad programming.

I have a Sequentix P3, which has a MUCH slower CPU then any MPC, especially the Live/X and it has rock solid timing.

The MPC4000 slaved to my P3 is a great combo. And is rock solid. Sorry to hear that the P3 --> Live combo won't be a solid
User avatar
By MPCWeapon1 Fri Dec 22, 2017 11:17 pm
Lampdog wrote:
MPCWeapon1 wrote:Too much money for the sequencer not to be rock solid.

This is my WHOLE reason for not jumping in.


I decided to cop The Live. I love the MPC Touch workflow so much I had to upgrade to The Live. I'm not looking for it to replace my 4000 though.
User avatar
By MPCWeapon1 Sat Dec 23, 2017 12:49 am
Made my first beat with The Live and Motif together without the Desktop computer, and it tracked the audio from The Motif without much of a problem.

We’ll see how it goes.
User avatar
By Juxx Sun Jan 28, 2024 8:59 am
I don't know How I ended up in this thread in 2024... but on GearSmut I found this link to Jitter timings if this helps?

https://www.innerclocksystems.com/litmus

the X and 4k are rated..... (this is all greek to me....No offence to the actual greek MPC'ers)
--------------------------------------------------------------------------------------------------------------
Akai MPC-4000 / OS 1.71

Sequencer

Tempo-Slave Format > 5-Pin MIDI Clock

External Sync Source > GRIDLOCK II
Source Jitter > Zero

Maximum Audio Out Jitter >
58 samples / 0.604 ms

Maximum MIDI Note On TX Jitter >
58 samples / 0.604 ms

---------------------------------------------------------------------------------------------------------------------------------

Akai MPC-X / OS 2.7.2

Sequencer

Tempo-Slave Format > 5-Pin MIDI Clock

External Sync Source > GRIDLOCK II
Source Jitter > Zero

Maximum Audio Out Jitter >
53 samples / 0.552 ms

Maximum Gate Out Jitter >
194 samples / 2.021 ms

Maximum MIDI Note On TX Jitter >
202 samples / 2.104 ms
By HouseWithoutMouse Sun Jan 28, 2024 11:06 am
FWIW and semi-related. Here are my findings of a Force's MIDI Out and Gate Out jitter. Recorded from the physical electrical TRS MIDI output, into a computer audio interface's audio input, as 44100 Hz audio WAV file on a computer, and differences between times of starts of consecutive MIDI messages i.e. timing jitter measured from the WAV file with a Python script. Python code included in the opening post.
viewtopic.php?f=49&t=214354

In my tests, maximum MIDI note-on jitter on a 120 bpm 4ppq pattern was close to 3 ms. Surprisingly, MIDI Clock seemed to jitter slightly more, even though it should be a "realtime" message with maximum priority.

Gate Out jitter was over 4 milliseconds, and it's either -4, 0, or +4 milliseconds all the time, looking like the gate signal is driven by a 240 Hz timer interrupt. Maybe the Force is worse in this regard than an MPC. I don't have a physical MPC I could test with.

Particularly the MIDI Clock jitter is much worse than what I compared it with, Arturia Keystep 37.

Whether any of this matters at all for person X, is completely up to person X to decide.

A surprising finding I made during this testing was that it takes 10 ms for a MIDI message to simply pass through the Force. Set a track to monitor a MIDI input, reroute to a MIDI output. A Note-On message sits waiting inside the Force for 10 milliseconds to be sent out.

I don't know how the older MPCs operate, but with the current generation, _internal_ MIDI timing, when the MPC or Force is the clock master and its own sequencer is triggering its own samples, then the timing is sample-accurate, and there's no jitter. Things start to go slightly downhill when sending MIDI messages between separate devices. This seems to be common with all DAWs, and MPC/Force is just a DAW in a box.