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

[mario1089]

MarOS 1.5.0

UPDATED FOR VERSION 1.5.0

Here is my new alternative firmware for the x0xb0x called MarOS1.5.0 It`s based on sokkOS2.0 and adds quite a bunch of fresh features.

This latest version contains various improvements on features, workflows, fixes bugs etc. Storing and Restoring Runtime Paramter Presets may be the most obvious improvement, but please read the new documentation to get an overview. (06/05/2016)

User nordcore has rewritten significant parts of the code, and thereby freed a lot of memory space. He also improved the timing and stability of the machine. Nevertheless, as of now, DIN Sync, and MIDI Beat Clock Output, as much as several other features, are sort of untested. I'll be here for a while to fix severe bugs, if I can. After the biggest problems are fixed (if any), I'll post the source code, let others continue to work on this.

So this version is a "beta".

Features:

-- MIDI Sync Mode significantly changed: Start/Stop, trigger patterns, transposing patterns including automatic scale correction – all by MIDI notes from your DAW / Midi keyboard.
-- Smart randomizer and macro functions to create cool patterns without any step editing. 20(!) macro / randomizer commands available in edit mode.
-- New Swing: Now allowing for fine adjustable swing 0%...100%.
-- Swing applied to both MIDI clock and Din Sync output.
-- Adjustable Gate Length allowing for shorter notes or almost gapless patterns.
-- Automatic Random Variations of current pattern to have the xox be kinda "jamming" - adjustable for accents, slides, octaves, rests and where and how often variations should happen.
-- Parameter Adjustment Mode - all new sequencer parameters editable with the tempo knob and visible on the bank LED.

Hex-file and documentation you find here:

[pangrus]

Hi Mario,
first of all I must say that the new randomizer is fantastic. I'll love to have the chance to randomize slides, instead of the da-da-da-da thing, but it's up to you, take it as a suggestion.
I've found some bugs:
1) Note F2 is supposed to select pattern 4 but it select pattern 8.
2) When selecting a new pattern with midi notes and latch on, before going to the right pattern a piece of the pattern in bank 1 is played.
3) Latch D#2 key works as latch when pressed together with pattern select keys , but works as play on-off if pressed alone.
Hope this may help.

[mario1089]

Thanx! Hope you have fun! I`m currently doing my first track with the xox, and I got a really nice groovy pattern here, perfectly matching my track - and not having touched a single step:)

I also found something wrong with pattern select (that strange bank1-piece, and more), and I`m happy you report that other bugs. I`ll have a deeper look during the week and release a fix soon. Guess won`t be hard to fix..

What I also found is that it is very important make the DAW send the starting noteOn a little before the related beat clock pulse.. which means a note right on the downbeat will be randomly sometimes after or before the clock, so half the times the xox is a 1/192 or something late.. so I delayed the triggering noteOns by 5 ms, by shifting the MIDI track in the track settings (I`m on banned-base) -- since that I got perfect sync.

Yeah, and slide-rand -- I did not do it in the first version, as slides are a bit special, no point doing slides after a rest for example.. to have a good slide-rand I think it needs some more smartness than messing the bits. Currently I got one button left ( CHAIN-PREV I now used to have a random repeat/loop-rand to make an existing pattern be more repetive in one of a couple (random!) ways ), so I think I do some slide rand on CHAIN-NEXT.

[pangrus]

Mario, it was a pleasure!
Would be great to have the chance to even use the midi clock based start-stop, perhaps using the USER positions of the selector.
This willl make your firmware usable by everyboby, even the "hardware only" guys.
Do you think that's possible?

[antto]

no point in doing slide before a rest you mean ;]

iirc, slide before rest works in ReBirth and ABL as a Tie kind of, extends the slid note onto the Rest step, with the same note attributes (ignoring the note attribs of the "rest" itself, except for the slide value)

so, you can actually have a c# with slide, followed by 3 Rests (with slide) and that'll make a 5-step long c# note

another thing, i hope you changed the original randomizer in the firmware, it's very very cyclic
a good idea is to grab a 16bit PRNG instead.. i can point you out some very nice ones if you want

haven't tested your firmware yet, hopefully i'll check it out soon :wink:

[mario1089]

well, I didn`t change the randomize, I kinda hoped it`s good enough, but I had the plan to improve it. what I actually thought would be enough is to count the milliseconds since start of the OS and use that value to change the "seed", but if you come up with something better, I`ll put it in.

About MIDI sync vrs hardware guys - Always-start-on-MIDI-Start is for me totally annoying, a pointless behaviour. But I actually think the new MIDI SYNC mode should be optional (at one of the track modes dial position), so that the old behaviour is still there for people that need it. Unfort. I was a little too lazy to follow that approach, lets see if this changes someday:)

[antto]

nah, the seed doesn't help, the randomizer itself is of poor type, it has a fixed "length"
it's like a sequence of values, changing the seed only jumps to a given point in this sequence but doesn't change the values

can't find the article but iirc it was by Marsaglia and it involved a number of coefficients, which makes the randomizer with a very very long period

the simplest thing you can do is to use the same type of randomizer as the one in the firmware, but make it 16bit at least, this would make the cycle much longer

[Paradigm X]

Well i personally would find the midi start very useful as i dont use computers anymore, and just sync the patterns to my 909, so i would hope that it makes its way back in somehow. The randomize functions sound great.

Cheers
Ben

[mario1089]

MIDI Start: I see the point. I think I turn it on per default, but make F#1 toggle it on/off (the next black key after latch key). so anyone not interested in computers or the new SYNC has it working fine, while the others like me can quickly turn it off.

I guess I do some fixes today and come up with something later today. So, any other wishes?

[mario1089]

Ok, here we go - MarOS1.0.1 is attached.

SEE FIRST POST FOR CURRENT VERSION!!

-- MIDI Start: By default classic behaviour is back - starting on recieving a MIDI start command. To turn that off, which is strongly recommended for anyone who likes to use the new MIDI sync mode in a MIDI DAW environment, press F#1. Pressing it again, brings it back to classic mode.

-- Randomize: CHAIN-PREV: Simplify the pattern by looping parts of it.

-- Randomize: CHAIN-NEXT: Randomize Slides.

-- Some internal improvements in randomizer (@antto: Using some very simple 16bit algo now, that reliably did a good job for me in many projects. Scientificly it`s not a good one, but it is by far good enough for the needs ).

-- a couple of stupid-early-version-bugs fixed for pattern selection

Have fun!

-Mario

[Paradigm X]

Wow, that sounds fantastic, many thanks for including the 'original' midi sync mode too, will have to give it a try. :mrgreen:

Im well out of the loop, looks like antto has made a new uploading software, need to try that, i was lucky with the original software which generally worked.

Thanks again, cheers, Ben

[antto]

yup... viewtopic.php?f=7&t=29282
it's not just for uploading firmware, it is supposed to do everything that c0ntr0l does, and runs on win32/osx :mrgreen: .. it seems it's very unpopular even that a bunch of months passed since it was released :?

[mario1089]

c0nb0x is great. Nice. Stable. Cool. I`m new here and I ve never tried the other controller, but I`d assume there had been a reason for antto to rewrite it:) Unpopularity seems to be a matter of this forum having not exactly too much traffic. I wonder how much xoxbox` are actually delivered. Any gut feeling? 1000? 10000?

[pangrus]

Just loaded the new firmware. Thanks Mario, for the chance to slave it again to midi clock.
Unfortunatly, it seems something is not working properly.
When the xOx is in PATTERN MIDI SYNC mode, you start it from your external midi master and then switch to PATT EDIT to use the excellent randomizer.
But pressing the CHAIN buttons stops the xOx. When you release the CHAIN button the xOx start again, but obviously it run out of sync.
Anyone else noticed this?

[mario1089]

I`ll have a look if something is wrong. But in PATT EDIT mode the x0x is not synced to MIDI. While editing you have to sync "manually". Not the nicest thing, but for me it is kind of ok to work like this. Did it (MIDI sync while PATT EDIT) work in SokkOS1.9?