Yes, that's what I have. The start point in relative time happens several seconds after the absolute zero. This is shown in the second picture of my recent post.

[And one result is that if I now rewind the tape to its furthest possible limit (absolute zero), I will see a negative number on the AKAI counters, which are both showing relative time]

Okay, so do the musically significant events start when you expect?
Is the Bar and Beat counter the same as the counter of the XF when things are running?

In SONG mode, yes, but with exceptions.

If I start the AKAI playing from the relative zero point, or any point later than that, the Motif plays with it, in time, and the bar:beat:clock counters match up.

With the AKAI stopped, I can move to any bar at or after the relative zero point and the Motif will also move to the same bar. However, the Motif time display will always show an extra 120 ticks.

[If I move the AKAI transport anywhere earlier than the relative zero point, for example to the absolute zero point, I start seeing more troubling things such as the Motif location going to bar 1021 as described above]

Is the Bar and Beat counter the same as the counter of the XF when things are running? If yes, then that is all that matters — who cares what it says when Not running? In both “exceptions” the transport is NOT running!

Bad Mister, thank you for sticking with me through all this. There is a new chapter to this story, but first to close the last chapter so you know where things were.

AKAI DPS

At the end of the trials using offset,
- the AKAI DPS played well with the Motif in SONG mode if and only if starting play on the AKAI machine from zero relative time or later.
- they still didn't play in sync when in PATTERN mode.
- if started from zero absolute time, the AKAI always triggered the Motif playback immediately, rather than after the pre-roll had completed

New chapter: ROLAND VS2000 as master

I got the idea to retry the Roland VS as clock master, incorporating the ideas about offset I have gotten from this exchange. So I setup the SYNC OFFSET TIME in the Roland settings, using a time in H:M:S:F that was equivalent to 2 bars, much the way you described in the AW4416 synchronization guide. Results: a great relief!

- When I rewound the VS2000 to absolute zero, it showed nothing on its Bar/Beat/Clock counter (only stars). Only after it played through the two bars (equivalent) of pre-roll time did musical time 001:01:000 first appear on its counter.
- Even better, the Motif slave (in SONG mode) only began playing after the pre-roll (unlike when the AKAI was master)
- Even better, the bar counts and metronomes of the two devices were in sync
- There were no funny 120 tick offsets appearing at any time
- Even better, when I switched into PATTERN mode on the Motif, the pre-roll period worked again
- Even better, the metronomes of the two devices were again in sync
- Even better, the good luck extended to recording a PATTERN CHAIN. The pre-roll worked, the metronomes were in sync, manual pattern changes were quantized to the bar perfectly!

So, in the end and in short, I got everything I could hope for.

THANK YOU

We really did get bogged down with the AKAI machine. For some reason, it just doesn't seem to work like the Roland or like your description of the AW4416. I know it would have been very frustrating for you to try and troubleshoot a setup with this device that you've never seen, which appears to act in a non-standard way, and with me as the conduit of information.

But I greatly appreciate you staying in the conversation and helping me get to a good resolution. I can't wait to get back to actually making music now!

Awesome! You’re very welcome!

And it will not work right in Pattern mode... as the article says, you’re free-wheeling... there is no intelligence matching exact location. That is handled by Song Position Pointer... which as the name states, works on linear structures, “Song”

There is no such thing as Pattern Position Pointer(sic) — it simply does not exist.

I understand that getting the two devices playing correctly together in PATTERN mode relies on me rewinding the Motif to the top every time before I initiate play from the master device.

But once playing, I am finding that changes of pattern in both PATTERN MODE and PATTERN CHAIN mode are staying right "on the grid". By that I mean that I can hit the button to trigger the next pattern just in time or a long way early, but the actual triggering is always happening right on a bar line.

I'm happy to get a video to demonstrate this if it would be helpful in any way. Either way, it's all good news as far as I see it.

Thanks again.

Free wheeling does not mean it will not sync. It will sync and behave absolutely properly but getting it there requires your effort everytime. You must start from the top and reset/rearm it each time.

The clock works like a set of gears — once they interlock they run fine. But be off by a silly 120 clock ticks and it will remain (faithfully) OFF by that amount. It will not drift (it can’t) it will simply remain OFF by the particular amount.

I used to run a Motif sequencer in Song mode in sync with the AW4416, via MTC, and simultaneously run my RS7000 in Pattern mode via standard MIDI clock (free-wheeling). The Motif was playing a fixed composed sequence, the RS7000 was in Pattern mode, which means I would manually bring it in and out when appropriate. You could ‘perform’ on the RS and with it... free-wheeling was the way to go. It was like working without a net, but when you understand that once you drop it in on the beat, you are in... (Turntablists take bigger risks) so you can use Pattern mode just know that you are manually addressing the sync’ing (start point) issue!

Please, no videos. I know exactly what it is doing, and how it is behaving. I’ve been here, done this, and own the t-shirt!! 🙂

Sounds like fun with the RS7000, Bad Mister. I imagine you could get some interesting choppy quick-trigger effects if you were so inclined. I probably would be -- but maybe you have better tastes!

I now have a quick follow-up to get some closure on the AKAI vs Roland MIDI behaviour mystery. Maybe this will help somebody else somehow.

I decided to use MIDI-OX to get a look at the SPP and realtime messages generated by the two audio recorders.

For each machine, I did the following short routine while logging MIDI output in MIDI-OX:

1) Locate the recorder's transport to the start of the first musical bar.
2) Press play
3) Press stop after a short time (a bit less than a bar)

The Roland VS2000CD machine's output

Note: I filtered out the MIDI clock messages here as the Roland is constantly sending clock, even when stopped, and it gets too much to view easily.

The AKAI DPS16 machine's output:

My layman's analysis

From what I can glean from the messages,

- the Roland sends a SPP value of 00 00, equivalent to the start of the first bar. When play is pressed it sends a START message. All very straightforward.

- the AKAI sends an SPP value of 01 00 that is later than the start of the first bar (I'm guessing by a 16th). When play is pressed it sends the same message again, followed by 5 clocks and finally a CONTINUE message.

So this seems to explain the source of problems I was experiencing: the AKAI approach is fine in SONG mode but falls down in PATTERN mode because the user (me) manually returns the slave machine to 0001:01:000, rather than 0001:01:120, resulting in the master and slave playing from different positions.

I don’t think so.

I’d reach out to Akai for support, you’re guessing!

Ok, I'll leave this thread here. Thanks Bad Mister.