MarOS1.5.0 - Beta Version Moderators: altitude, adafruit_support_bill, adafruit, hamburgers, phono

[Paradigm X]

yeah, massive thanks, its much appreciated.

my computer is in bits at the moment and i havent had a chance to fix it and try this. i will really try to get it done this weekend.

all the best

[ioth]

I've got another bug to report: The sync mode in User C sometimes resets itself (I've got it in classic) to the new sync mode.

[Nordcore]

Bug confirmed - doubly used EEPROM address. Give me a minute to fix that ...

[Nordcore]

Here is the fixed version.

Sorry for the delay, had an important call from my pillow ...

• removed EEPROM address overlap in user C settings (C# and D#) with randomizer "FAV" settings.
• changed default behavior for settings if "magic value" not in EEPROM (= used after software update from non MarOs firmware): now *all* settings (even C#) are set to their default. Defaults are (slightly) changed

As the bug might have corrupted the USER C settings, it is recommended to check them.

List of USER C settings

C# Settings:

1. disable high octave shift (two octaves up, the blinking one)
2. don't store tempo changes immediately
3. live edit sets edited pattern in chain
4. -
5. preserve selected pattern when changing mode
6. -
7. auto increment for step mode
8. -

D# Settings:

1. classic midi sync (start/stop) behavior. If off: new mode, start on first note
2. disable clock Fix (fakes a clock if starting midi note comes immediately after the real clock)
3. preserve swing & gate len when loading parameter preset
4. help LEDs (Dark light on LED 1,5,9,13 for orientation)
5. less accurate, but faster MIDI sync tempo measure (if jitter is too high to get any results)
6. MIDI CONTINUE behaves like MIDI START (resets to first note in pattern). If off, it tries to kick in proper to current song position (SPP)
7. diminished 2nd for phrygian and arabic scales
8. major 7th in minor for harmonic minor

F# Settings:

1. -
2. -
3. -
4. -
5. -
6. tempo encoder has click on every 2nd edge (not 4th as standard)
7. -
8. -

bold=default setting=ON. Everything else is OFF.
Default is set if the "magic byte" is missing, which is (only) the case if there never was a MarOS on the box before

EDIT: Firmware File deleted, see viewtopic.php?f=13&t=33914&p=588826#p588826 for latest Version!

[Nordcore]

Yesterday I worked a little on the upcoming manual.
If you look at the example page (... single page - sneak preview...) for the randomizer mode, you'll see why not everybody runs into this issue above: the double permanent memory usage does only happen when you use the "E" (re do) function of the randomizer. (As that function stores the used algorithm in EEPROM, so it is retained over a power cycle. )

================

Regarding the key assignment for the RANDOMIZER scale:
I would prefer to have:
C = major (our root note on the x0x)
C# = pentatonic (just the black keys = pentatonic scale)
D = chromatic
D# = minor (=Eb =the minor third to C )
... at least that would be better memorable for *me*.

[ioth]

Hey Nordcore,
thx again for the fix!
Your scale suggestions sound fine to me, although I don't get why you say all black keys is a pentatonic scale.
I'd suggest the C minor pentatonic (C Eb F G Bb) so you always have C as your root note when randomizing stuff.
And with the manual, I thought the oct. down (A key) transposes the highest note one octave down and not the lowest one.

best, ioth

[ioth]

Major bug discovered:
When you press slide in Play mode it always slides to the highest C, when it should be just sliding the notes in the pattern.
Also a minor bug, when you go to user C while your seq is still playing, the leds don't light up and you can't change anything.
Not too bad though, just wanted to let you know.

[Nordcore]

Bug confirmed ... and fixed.

• Sliding notes always sliding to C2 when pressing "SLIDE" in play mode - fixed
• lowest notes not silent when pressing "REST" in play mode - fixed

(... and thanks for proof reading the manual page - that's also fixed...)

=============

Turning the mode knob "outside" the PATTERN PLAY to EDIT range whilst the sequencer is running is ignored intentionally.
It should end the current mode (stop the sequencer...) and go to the new mode.
But that would require some fancy (=lengthy code) debouncing of the switch.
I'm not happy how it is now, as it feels broken - but it is very cheap on the code byte count side.

[ioth]

kewl gonna let you know if I find sth else.
As I said, I also don't think the mode knob behavior is a problem. It's not beautiful, but if the essential part works its fine to me.
by the way, what language is the xox os written in?
Must be really tough to always have an eye on the byte count when updating stuff...

[Nordcore]

It is written in C.
And the lack of code space has one benefit: no feature creep.

Completely different thing:
The RUN/STOP key acts when pressed in the stock OS, but on release since Sokkos.
Is there any reason for that?
I would prefer the "on press" version, if there is no reason, why it has been changed.

[ioth]

Found another bug, when transposing down (semitone as well as octave) in randomizer mode, the end step gets deleted, I guess that's not supposed to happen.
My suggestion would be that the end step stays the same, even if you completely randomize your pattern.

And for the Run/stop button, of course on press would make more sense.

best, ioth

[Nordcore]

Confirmed, thanks for pointing that out ... that was really a mess.

The randomizer ignores the "end of pattern" mark (used on patterns which are shorter than 16) steps nearly everywhere.

• RUN/STOP key action on button down (was on release )
• fixed settings 1 (D#) key 6 (had also "Encode Q2" = double speed for tempo encoder on it ...)
• fixed RANDOM mode (... now it plays again, start_tempo() was missing)
• added synchronized RUN/STOP for MIDI_SYNC mode, MIDI_CONTINUE starts x0x on next bar
• fixed Octave DOWN/UP for "C"- Key in pattern edit
• Randomizer works with pattern length shorter than 16 (obeys "end of pattern" mark)
• Randomizer sometimes writing out of (pattern) array bounds fixed.

The pattern length thing works much better now, but I expect to find some more quirks there.

Edit: seems something really went wrong with this version ... I have to check that, until than: not recommended...

Edit: Outdated file removed, see viewtopic.php?f=13&t=33914&p=592644#p592644 for next version

[ioth]

Will let you know when I find sth else.
Say, are you by any chance affiliated with the Nordcore Label? That'd be funny! :)

[Nordcore]

No, nothing to do with the label.
Just a nick that came into my mind, one evening, some years ago, for registration in another forum ...

[ioth]

yo, I just uploaded the new version and its suuper buggy :(
There are a few things:
Scale randomizers also randomize the end step,
When a pattern is randomized its (sometimes?) impossible to enter step edit mode, it just stays on the first step and doesnt move, this is probably the meanest one.
when I transpose a pattern down (semitone and octave) it still deletes the end step,
loop random just sets the end step at 5 everytime (when it shouldn't change the end step, just the pattern, right?)
randomize end step makes a 16 notes pattern most of the time, sometimes it works.
And the first time i uploaded the os my xox started to randomly freeze, though that didnt occur after i updated a second time.

Sorry for the bad news