"Then I pressed one low note - then waited for a sec and then started playing the same high note 16th notes.

Soon, I heard the low note come back around and sounded "split" by the 16th notes. Not layered - but as if it was strangely gated by the 16th notes. Not playing anything at all the system kept playing something that sounded regenerative evolving over time. The interface was locked up. I couldn't press buttons or the touch screen. I left the keyboard alone and when I came back after an hour or so I saw an error message "Song/Pattern Data Overload" (or similar)."

Just in case it's not more widely known:

Have seen daughter hit a similarly impactful freeze bug with the Pattern Recorder, many times, and need to restart. It's somewhat difficult to Clear (Jobs menu) this Track/Scene portion of a Pattern Sequence, after it's recorded. Requires deleting the entire Pattern Sequence. Mind you, I'm forcefully restarting rather than waiting like Jason is, for the system to resolve itself, so there might be something missing from the Pattern's Scene/Track that makes it impossible for it to clear the note events in it. Btw, these tracks appear empty when you play them, but they're definitely full of something.

She's also seen these sort of bugs fill the Pattern Sequencer, giving these types of warnings about being full.

The missing first note that Dragos is talking about, seen that myself, many times, when doing Arp recordings into the Pattern Sequencer. It's quite difficult to get arps to loop nicely in the Pattern Sequencer, even if you do everything right, and have a perfectly matched Arp with right length to the Pattern Length. Either the trailing note hang, effect flush or missing first note issue will mess it up. So I've taken to using StepPolyArp on the iPad for this kind of thing, with much more luck, and rendering to audio in DAW. Not how I'd like to do this.

I read three completely different issues. The original poster is working with both the MIDI Song and the Pattern Sequencer (which do not work similarly concerning the LOOP settings). And setting the Loop Playback in the Linear Song, cannot be set frivolously (meaning - you cannot set it just anywhere).

You must heed measure lines, if you set a LOOP point so that it cuts a note or chord in the middle of it sounding, you will have selected poorly on where to place the Loop Start and Loop End Points. If you are not properly quantized (respecting measure lines) your result may well sound like someone stepped on the toes of the hanging notes.

Jason seems to have discovered an entirely unique problem but it doesn’t sound like it’s related. On the right track, however, looking for hanging notes (notes played too early perhaps, hanging over into the looped region). It could not have been in MIDI Song, because the setting described by the original poster do not exist in MIDI Song (again, because they are different… so whatever is happening is not happening in both.
(I think an Overload message has an entirely different cause).

And certainly anything to do with connecting to an external DAW is a completely different set of issues.

Without knowing the entire situation (what’s playing, what controllers are in motion, whether or not the selected Looped region can musically be looped, what effects are active, what side-chaining, motion, or interactivity is going on, or if the Arp Phrase is being offset by Arp PLAY FX, etc., etc) it’s really impossible to just venture a guess.

Extra Credit:
I’ll have to spend some time with the MODX seeing what happens if you set your loop region so that the data does NOT fit inside the measure lines — I expect it to not work, but I will be looking for what anomalies occur when/if used without attention to the actual region settings.

Known User Errors
__ If you create an Arp Phrase, again, setting the region is critical. I’ve found people short their User Arp by a measure, which causes strange results as the Phrase tries to playback when triggered
__ Play FX can offset the timing causing some notes to extend past the measure line… this can cause issues if the data is not Normalized.
__ Loop Start/End region must be able to musically loop

I'm guessing it's possible my note entry had a note ON with a truncated note OFF at the end of the measure - and perhaps this can be related to the cause of the overload.

Thanks for looking more into various recording scenarios trying to uncover corner cases.

Without a listing of note on/off & CC events with MIDI message precision in editing/removing - there's more of a burden on the recording feature to make attempts to "smooth out" artifacts during the recording process.

And certainly anything to do with connecting to an external DAW is a completely different set of issues.

Without knowing the entire situation (what’s playing, what controllers are in motion, whether or not the selected Looped region can musically be looped, what effects are active, what side-chaining, motion, or interactivity is going on, or if the Arp Phrase is being offset by Arp PLAY FX, etc., etc) it’s really impossible to just venture a guess.

I tried to document the DAW-to-sequencer issue as detailed as possible in its thread (linked in my post above).
Please let me know if there's any info that I can add or if you can reproduce it on your side based on the information in my post.

Since we've got a list of Known User Issues, an in summary of known product issues pertinent to the Pattern Sequencer:

Antony has provided a reproducible bug report. I think that should be shown a degree of respect, and that he should be thanked for doing that, as he's articulated it quite well.

I've confirmed I've experienced this previously, and previously found (and now described to him) a workaround, and seen others not only describe and use this workaround, but also seemingly acknowledge why it's needed - folks that have been given or lent units from Yamaha and don't have a 1/100th of the experience I've had with the Pattern Sequencer such that I was able to find this workaround - so I presume they were told by Yamaha about the issues of edges of loops and the need for the "clean" way of working.

Antony attempted to admonish me for not filing a bug report - which might be a point, if I believed it would have done any good and hadn't already seen that the issues were known about at the very release of the Pattern Sequencer.

What's super interesting, to me, is that Antony then went on to describe a scenario in which the workaround I've described no longer works.

After that, two more pointed examples of issues with the Pattern Sequencer (one each from Dragos and Jason) regarding timing seem to indicate there might be general problems with its implementations at edges of timings of its looping (both starting and finishing).

I have more generally confirmed that there are many issues with the Pattern Sequencer, most of them too odd for me to bother chasing down, instead I've been preferring to think in terms of how to actually get things done, with the way the thing in front of me actually is, and what else can help me get to where we want to go. I'm leaning ever more towards an Akai MPC Live II, or heavily customising Reaper, or both.

I had a reprieve from work.

I had a look at this again.

I deleted all User Arps, Patterns, MIDI songs from storage, and deleted every version of my Performances (Teardrop... see tutorial I submitted).

Armed with the experience of writing that tutorial, and able to jump through the hoops in quick succession I started from scratch.

This time I had NO issues.

This is good news, the recorders can work properly.

The bad news is I cannot figure out why it was happening before.

My guesses are any of the following:-

- Bad playing/timing
- Multiple aborted attempts
- "New" load-ins of MIDI Songs or Patterns
- Multiple counts of Use of "Undo Media Record"
- Quantization changes (I tried various)
- Some unnoticed parameter setting
- Multiple incremented Stores (v1, v2, v3 etc)

Although In the aftermath I feel a bit stupid now, I was definitely experiencing it, to the extent I was consciously trying to figure out how to stop it or workaround it for many hours.

-

I found the problem. REPEATABLE

My Scenario
4 Part Performance. Part 1 Piano (Note Shifted -12), Part 2 - Piano, Part 3 - Octave Harpsichord (Note Shifted - 12)

Key board Ranges Set and Matched at Part and Element Level -> Part1&2 (D1 to A1), Part3 (A3 to E4)

Part 4 = Drum D'Elo Kit

Keyboard Control OFF for Parts 1,2 & 3.

The intention is to Record a MIDI Song (MIDI Recorder), from which to create a 1 bar User Arp.

I am using a 1 Meas Count in, BPM set to 72.

The Drum Score is in the attached link.

1st CLEAN Run Through
----------------------------------------
From Home Press Play/Rec in left hand tab.
Select MIDI.
Press Record (Arm)
Press Play (Record)
Record Many Bars.
Press Stop
Press Play, no loop, it sounds OK, Stop
Press Loop ON, Start 001:01, End 002:01
Press Play. The Loop plays and sounds OK

Assume you made a mistake so you decide to "erase" and start again.
There is a feature for this "Undo Media Record".

2nd "BUGGY" Run Through
‐------------------------------------------
Press UNDO MEDIA RECORD <<------------ THIS IS THE BUG
Repeat Record, Loop and Play as per 1st Attempt. The first beat is now double-tracked.

The buggy first beat will be carried through into the User Arp if you create one. When I first heard this (Saturday, Sunday) I assumed there was some kind of delay switched on... there wasn't. Then I thought there must be some unknown Arp feature switched on... there wasn't.

The Only way to remove this is to delete the MIDI Song from Data Utility. Store the Performance. And then start afresh (and don't make a mistake).

I believe the bug mechanics are the same for Pattern Record, i.e. Undo Media Record.

I will now hand over to the Forum Hive-Mind for further testing and reporting.

Thanks all for your help.

HOLY COW!!!

I remember getting yelled at on these forums for claiming there was a undo bug, ages ago.

Good detective work!

@Anthony
I've reenacted your situation above and think i know the answer.

Problem: In song mode there is no quantise. You record 4 bars of drums but only want to keep bar one. But, If you play the first beat of bar2 slightly early, it will always appear as a 'double' drum in bar 1, because thats where you played it. This is the source of the Double. And it will be there in both Loop or the User Arp you create.
If you quantise the arp 'afterwards' you just move the offending note closer until you get a phase effect

Solution Number1: Record your drum part in Pattern Mode with quantise on as you record, not after the fact. This will be clean as a whistle. And for this drum loop you know you want it quantised to 1/16th notes before you start.
And as BM pointed out if you set 'Key On Start ' to 'On' and "loop' to 'OFF', you'll find it hard to go wrong

Solution Number 2: And in cases where you know you want 1 bar, In both Pattern and Song mode just play 1 bar, and don't overrun

Good Luck

@ Graham... did you try it?

I won't be near my MODX now for at leasr a few weeks.

IIRC.... the Double First beat will appear on the track, after you hit Undo Media, after you rerecorded and played back.... before you loop it.

@anthony
Let me start off by saying i feel your frustration. I had something similar a few years back when trying to use DAW for the first time, where nothing seemed to work. (and i took time off work that ended in frustration!)

Yes my solutions work. yes you will get your Drum Arp done if you follow solution 1. I think the undo media thing is confusing matters. I could not get my Montage to recreate that issue.

Please follow my instructions above.. Let me know if you need a screenshot or 2, or more clarity

@Graham - this is part of a longish process, via which Antony has not only successfully created his desired Drum Arp, he's also documented exactly how he achieved that goal, in one of the best written examples of the problems inherent to Yamaha's inconsiderate manner of designing operations.

You should read that, here: https://www.yamahasynth.com/ask-a-question/tutorial-creating-your-own-user-drum-arp

And, generally speaking, if you can't reproduce something (good or bad) don't presume those having differing experiences (good or bad) aren't really having that real happenstance. In fact, presume it is, especially when it's this well documented and researched.

I can personally attest to having experienced similar outcomes from a multitude of different approaches, and many other peculiar and trifling issues with the Pattern Sequencer.

At a certain point you'll likely realise that the Pattern Sequencer can be considered a half baked response to the Fantom, just as the Montage can be thought of as a response to the Jupiter 80.

for Antony I have no relationship with musical equipment manufacturers, and never have. I've owned quite a few more synths than I'd care to admit, despite being no form of musician. I love arpeggios and sounds, so fiddle with these. I work with computers, so don't want to use plugins to make sounds. It's an interest more than a hobby. Recently begun more serious DSP coding, which is crossing over in interesting ways.

[...]despite being no form of musician. I love arpeggios and sounds, so fiddle with these. I work with computers, so don't want to use plugins to make sounds. It's an interest more than a hobby.

The typical narcissistic obsessed troll profile.
You can always find one on every professional / specialized forum on the internets and they're all very similar: outsiders of the field, zero real understanding of the field, not actually doing anything in that field but spending a huge part of their life explaining online to everybody (lots of them actual pros who use the tools in real life) how things should be, what needs to be done and why everythings sucks (hint: because nobody listens to their advice).
Like here, they have a phenomenal ability to drag down a forum, because they are willing to spend a ton of time just arguing. Andrew is here for the forum, not for the MODX.

@anthony
Please bookmark my replies for when you return to your keyboard and we can move on from there.

@Dragos - perhaps focus more on my shameless love of the analogue, tactile, dilettantism and two strokes in your next attacks.

@Graham - highlight the part about turning on live quantising, I think. It's a partial solver to some of these problems at the beginning and end of Patterns and something I think Antony has missed, probably due to being a good player.