n0nx0x2

Discuss mods, hacks, tweaks, etc.

Moderators: altitude, adafruit_support_bill, adafruit, phono, hamburgers

Please be positive and constructive with your questions and comments.
3phase
 
Posts: 203
Joined: Wed Apr 22, 2009 2:06 pm

Re: n0nx0x2

Post by 3phase »

So i did some tests.. sofar most of the things o my list have disapeared , I was actually wrong that the back button dont does anything when beeing in pitchwrite and it works on te nonxox too..just no audio output..but now wit optical feedback..not bad .Timing on the midiclock is much better now.. before it was over 8 ms late in relation to a din sync clocked 303.. what was ok for a late feeling ..but different to the 303. sometimes the midiclock misses a start.. with ableton as master clock source..when you directly jump on stat again without having stopped before..1 out of 4 times..

din sync on kenton interface is still not working.. interesting is that it happens also in a 1 out of 4 times ratio that it gets off.. so that might be related.. dont looks like that the nonxox likes ableton live much as a master.. and kenton interfaces as din sync converter..

however.. the real 303 sticks perfectly to that combination without the slightest glitch.. 808 and 606 too..

whatever the reason is..roland machines are immune to that. i guess it has to do how the running voltage is handled..probably the rolands just start when it arrives aslong the clock voltage is positive..they dont wait for another clocktick..therfore they are immune to a tiny delay of the start flank that is probably caused by ableton.. but thats just an idea..

One other problem with the start handling on dinsync on the nonxox..when the dinsync clock drops for whatever reasons.. you cant go to the settings menu and change the sync mode to internal or midi..it just stucks up to the moment you give i a clock again.. only than you can change the clocksource..with midi that works fine.. so i would say the nonxox is better to be synced with midi..at least that works pretty fine for me right now..

3phase
 
Posts: 203
Joined: Wed Apr 22, 2009 2:06 pm

Re: n0nx0x2

Post by 3phase »

with the 303 when you set the pattern length you erase the time info totally and have to rewrite it again..

on the nonxox you can alter the length at any time.. and only when you make a pattern longer the additional steps get a rest command.. its not bad that you dont have to rewrite the time data..but the gap at the end of a pattern you made longer dont sounds musical and you need to rewrite or add time steps at the end immideatly..seeing this i would suggest to extend the pattern with tie commands instead of rest commands.. or triggers .. ties are probably better ..

another option would be just to keep the old timing info and be able to make the pattern longer or shorter as you wish..maybe even to teh extend that you extend over the 16 step limit and start using the time and pitch info of the following patter, performing an automated chain selectin this way....but ok..that would be probably a bit wild.. anyway..just a suggestion

User avatar
rv0
 
Posts: 395
Joined: Tue Jul 14, 2009 4:50 pm

Re: n0nx0x2

Post by rv0 »

3phase wrote: 1) while being in pitch write mode the back button could just do the going a step back when beeing pressed while the tap button is held.. this way you advance thru the sequence with the tap/next button..but can jump back when you accidently missed an edit .. that safes time and dont interfears with the original 303 conventions.. it appears rather as a bug on the real 303 since the back button just does nothing in pattern write mode.. probably it was originally intended to use it in such a way..
actually, on the 303 it does work in pattern write mode.
And in n0nx0x2 it works too.
Could be further improved by playing the note or allowing to write DUAS while holding back.
3phase wrote: 2) when writing pitches the non xox could accept the accent slide and transpose markers ..on the original 303 you need to do that in a second path..but when you know where you want the accent ..why not?
also works in n0nx0x2, but you have to press accent/slide/transpose first, before pressing the note.
could also be improved by allowing any order.
3phase wrote: 3) in realtime tap time writing only the tap button works..but maybe one could allow all keys plus the tap button to be used.. again a time safer. But probably too difficult, since the accent and slide info should stick to theire step number as on the real 303.. only the info of the button you use should be over written..
this stuff will be possible in the (now not yet implemented/empty) "pattern record" mode, though this is still a TODO, first the other stuff has to work perfect
3phase wrote:just a bit bad for nonxox users that touch the first time a real 303 than.. would feel a bit limited than *lol*
I already feel like that now when working on my 303s ;)
3phase wrote:Timing on the midiclock is much better now.. before it was over 8 ms late in relation to a din sync clocked 303.. what was ok for a late feeling ..but different to the 303. sometimes the midiclock misses a start.. with ableton as master clock source..when you directly jump on stat again without having stopped before..1 out of 4 times..

this probleem has indeed been reported with the latest OS version. I havent been in my studio for 3 weeks, I'll hope to find some time this weekend to look into this. If you feel adventurous, you can downgrade to 2.08, which should not have this issue
3phase wrote: One other problem with the start handling on dinsync on the nonxox..when the dinsync clock drops for whatever reasons.. you cant go to the settings menu and change the sync mode to internal or midi..it just stucks up to the moment you give i a clock again.. only than you can change the clocksource..with midi that works fine..
just press stop on the x0x and it should be reverting to internal clock.
3phase wrote:so i would say the nonxox is better to be synced with midi..at least that works pretty fine for me right now..
When you use a DAW, you should change the clock setting on the x0x to "fixed width"
http://antonsavov.net/cms/projects/n0nx ... .html#sync
"MIDI-to-DinSync CONVERSION METHOD"

This is because a DAW never sends a constant clock (unless its Renoise), so the prediction algo can't work correctly.
Definitely read the manual, there's some cool stuff in there that make it more interesting that a 303 sequencer (imho)

3phase
 
Posts: 203
Joined: Wed Apr 22, 2009 2:06 pm

Re: n0nx0x2

Post by 3phase »

Did some more testing..the actual version seems to be an allmost perfect 303 clone os wise.. At least in all things i tried it seems to do everything just like the 303 would do it..


darffader wrote: actually, on the 303 it does work in pattern write mode.
And in n0nx0x2 it works too.
Could be further improved by playing the note or allowing to write DUAS while holding back.
I know.. had forgotten that it does this blind back stepping. but realised it on the nonxox again since it gives an optical indication that makes it much more usable..
3phase wrote: 2) when writing pitches the non xox could accept the accent slide and transpose markers ..on the original 303 you need to do that in a second path..but when you know where you want the accent ..why not?
darffader wrote: also works in n0nx0x2, but you have to press accent/slide/transpose first, before pressing the note.
could also be improved by allowing any order.
ah cool, thanks ... thats good.

3phase wrote:just a bit bad for nonxox users that touch the first time a real 303 than.. would feel a bit limited than *lol*
darffader wrote: I already feel like that now when working on my 303s ;)
The advancements of the sequencer grow fast to you..especialy pattern copy is gold ..
3phase wrote:Timing on the midiclock is much better now.. before it was over 8 ms late in relation to a din sync clocked 303.. what was ok for a late feeling ..but different to the 303. sometimes the midiclock misses a start.. with ableton as master clock source..when you directly jump on stat again without having stopped before..1 out of 4 times..
darffader wrote: this probleem has indeed been reported with the latest OS version. I havent been in my studio for 3 weeks, I'll hope to find some time this weekend to look into this. If you feel adventurous, you can downgrade to 2.08, which should not have this issue
which of the problems? that it sometimes misses a start? I really think that this goes together with the dinsync problem..which points to a problem how the start command is handled with the xox. fixing the one problem will probably fix the other with it.
Sofar one can state that the nonxox dinsync in is incompatible with kenton interfaces..but.. is that the only brand it cant handle? In the moment it reacts to incoming start commands just like the roland machines its likely that all devices on the market will work with the xox since they was developed to feed roland machines.

just press stop on the x0x and it should be reverting to internal clock.
That was the first i did..not working when the din sync clock was stoped before you had that chance to hit the stop button.
And the xox is automatical reverting to internal sync when no external clock is available? my doesnt do that at all..would be nice.. Is that an option that has to be activated?

When you use a DAW, you should change the clock setting on the x0x to "fixed width"
http://antonsavov.net/cms/projects/n0nx ... .html#sync
"MIDI-to-DinSync CONVERSION METHOD"
have had a look in the manual..but is that fixed pulse width not only for transmitting din sync?
anyway..i ve a mac on a rme midi out te clock is pretty stable.. around 0,1 ms clock jitter..thats pretty good
but i will set it to fixed..
Last edited by 3phase on Fri Apr 11, 2014 3:23 pm, edited 1 time in total.

3phase
 
Posts: 203
Joined: Wed Apr 22, 2009 2:06 pm

Re: n0nx0x2

Post by 3phase »

One thing that gets on my nerves on the real 303 often..and i wonder if live wouldnt be better without that "feature"..
pitch write.. as soon you activate it new pitches get written you cant play the keys without overwriting.
Thats a little to quick..often i overwrite important pitches by accident.. never havethis problem with time mode..but pitchmode invites to press the keyboard just to hear the 303..or xox..


I personally would prefer it when just pressing the pitchmode button only enables the keyboard and writing starts on pressing the start button...as soon you press the function button recording ends and the 303 starts directly playing ..

Thats the way i write on a 303 anyway.. The final recording path ..when i ve found my line...i stop and restart the 303 in pitch write mode..enter the piches.. jump direcly into play..hit the pattern clear button and change the time info if necessary exactly on the next one..all out of one move without getting interupted.

Thats a workflow that dont works on the nonxox like on the 303.. on the nonxox you have to start it by hand again after you have ended the recording by pressing the function button.

Not a big thing.. but..together with a write protection that the 303 dont has..it would give a good workflow and more comfort..where you can rehearse the pitches without overwriting anything.. ...start recording by pressing start..and have it automatical playing the moment you end the recording. If you dont want it to directly start playing..you just use the stop button to end the recording... it just jumps to function mode anyway..

when you want to edit in pitchmode you anyway need to press the Write/next button.. so only when the write button is pressed the pitch info on that step can be overridden.. but when you dont touch that write button the pitch info in the memory stays unaltred regardless what you play..

So everything stays the same..just you need to presst run to really start writing the pitches.. butz get repaid for tha extra press with the possibilty to get directly into a running sequence..

303 owners can try that.. do you like it when it starts playing as soon you end the recordin by pressing function again?
And do you sometimes accidently overwrite good pitch info by accidently pressing a note button?

Since you often enter pitch mode just to alter acc and slide such a protection routine for the pitch data is usefull..and when the nonxox would allow to activate the start while beeing in pitch write mode like on the 303 we get a dynamic write feeling.
As i say..since years i always finaly write the pitches with activated run button.. Thats my personal routine to initiate the real run..and it feels good when it just runs when i am thru with the sequence..feels alive and makes things faster somehow,.

3phase
 
Posts: 203
Joined: Wed Apr 22, 2009 2:06 pm

Re: n0nx0x2

Post by 3phase »

That you cant engage run while you are in pitchwrite mode is a 303 feature missing.. would be nice if that little detail would be added..


but on top of that the above mentioned pitchwrite protection could be an option in the settings.. that would be nice comfort..

User avatar
rv0
 
Posts: 395
Joined: Tue Jul 14, 2009 4:50 pm

Re: n0nx0x2

Post by rv0 »

3phase wrote:
3phase wrote:Timing on the midiclock is much better now.. before it was over 8 ms late in relation to a din sync clocked 303.. what was ok for a late feeling ..but different to the 303. sometimes the midiclock misses a start.. with ableton as master clock source..when you directly jump on stat again without having stopped before..1 out of 4 times..
darffader wrote: this probleem has indeed been reported with the latest OS version. I havent been in my studio for 3 weeks, I'll hope to find some time this weekend to look into this. If you feel adventurous, you can downgrade to 2.08, which should not have this issue
which of the problems? that it sometimes misses a start? I really think that this goes together with the dinsync problem..which points to a problem how the start command is handled with the xox. fixing the one problem will probably fix the other with it.
Sofar one can state that the nonxox dinsync in is incompatible with kenton interfaces..but.. is that the only brand it cant handle? In the moment it reacts to incoming start commands just like the roland machines its likely that all devices on the market will work with the xox since they was developed to feed roland machines.
missing the start is something that has been reported with latest 2.09 version. I have not used the final 2.09 build, some things were changed last minute and I was not around to test them before release. 2.09 changed a lot of things about how internal clocks are handled (see release notes if you're interested in the technical details)

as for dinsync interfaces, my x0xes work perfectly (like all my other dinsync devices) with at least 4 dinsync sources I tested with (none of them were kenton, but I don't think the problem is related to the kenton interface). I won't go into this any further until I can test 2.09 myself (tomorrow)
3phase wrote:
When you use a DAW, you should change the clock setting on the x0x to "fixed width"
http://antonsavov.net/cms/projects/n0nx ... .html#sync
"MIDI-to-DinSync CONVERSION METHOD"
have had a look in the manual..but is that fixed pulse width not only for transmitting din sync?
anyway..i ve a mac on a rme midi out te clock is pretty stable.. around 0,1 ms clock jitter..thats pretty good
but i will set it to fixed..
The x0x is a midi dinsync converter in itself.. the sequencer gets driven by dinsync, when you use midi it gets converted to dinsync first.

so the conversion is both for the internal sequencer (as I just mentioned above) as for transmitting din sync.
Definitely set it to fixed for perfect results when syncing from a DAW.

3phase
 
Posts: 203
Joined: Wed Apr 22, 2009 2:06 pm

Re: n0nx0x2

Post by 3phase »

darffader wrote: but I don't think the problem is related to the kenton interface)
But what else could it be? ok..when something goes wrong theese days and since a few years it has allways something to d with ableton live.. Soo lets assume that live does something wrong.. but that can only be in the event timing.. like sending the mid start command after the first clock bit. Or the other way around.. not much else could have an fx on the kenton interface here..

But whatever it is.. Its not effecting roland machines.. So something makes the xox not to start one the right clock tick..
Since it behaves rateher random in that regard i assume that this is event timing related too.. sometimes the running status vltage goes high befor the first clock flanc..and sometimes it gets high shortly after.. That dont irritates the rolands.. since the dont wait for a flanc.. the trigger when the voltage is high.. so when its high when the start BANNED arrives they see that as positive flanc and dont need to wait for the next clock pulse to trigger. I assume thats what the xox box does..and so its more dependent to the correct evet order than the roland machines..
so the conversion is both for the internal sequencer (as I just mentioned above) as for transmitting din sync.
Definitely set it to fixed for perfect results when syncing from a DAW.
logically.. outside of teh cpu it needs real exsisting clock pulses... but that are not the same signal as it sends to its dinsync port? interesting.. one could do with the dinsync output different things as the internal clock . but how does that prediction thing effects the timing of the sequencer? isnt it trigering on the positive flank anyway and ignores the width of the pulse?
But thanks for the tip.will set it to fixed..

3phase
 
Posts: 203
Joined: Wed Apr 22, 2009 2:06 pm

Re: n0nx0x2

Post by 3phase »

One Feature request..

Its a great idea to have this wrap master counter for restarting the xox..but.. it could be brought to even better use when you could apply this to pattern changes as well.

For example you have a 7/16 long pattern running..and want to go back to a normal 16/16 pater..how do you do that? allmost impossible task to manage this by realtime counting and waiting for the right cycle.. most of teh times you need to stop the sequencer..

Now there is the restart function, very good..but that dont safes the new pattern from starting in at the wrong point.. you always have bad sounding notes before you perform the restart..

Wouldnt it be better when patterns could be changed acording to the master counter and not to their length? either as a permanent option..or by pressing function while changing the pattern.. maybe booth? Pressing function while changing the pattern would match the start logic..

when you press function on the pattern change the 7/16 pattern would repeat until the master counter reaches a 4/4 one and starts the new selected 16/16 pattern right in time..

that would allow to combine uneven counting times with regular 16/16 patterns way more smoothly and musical as on the 303.
Last edited by 3phase on Sat Apr 12, 2014 4:35 am, edited 1 time in total.

3phase
 
Posts: 203
Joined: Wed Apr 22, 2009 2:06 pm

Re: n0nx0x2

Post by 3phase »

And one unimportant bug report.. couldnt get the led dimming to work.. But.. while i tried to do that my xox reduced its volume drastical...
Is there a possibility that the cpu can effect the volume of the xox? I dont think so ..but?
Probably my xox needs to go back to the workshop.. :-/

Anyway..one other thing.. the swing settings are not all really funky.. the higher values seem to miss some steps in between.. when you go above 4 its quite big jumps but not really nice values for the speed i am right now working on..(128bpm)
A finer resolution would be nice


on my machine only values in between -3 to +6 are musically usefull..and +8 it goes into half time even on the shortest PW setting. But even when it would work it would be too strong than to be cool really.. I would suggest to have a a smaller general range but finer steps..and maybe selected steps . not just in even intervals.. handpicked..like on the 909. some shuffle % just sound better than others on the usualy tempo range from 110-135 bbm.
Anyway.. wouldnt hurt to have all 909 values available ..its just 6 .. and some more in between..

User avatar
rv0
 
Posts: 395
Joined: Tue Jul 14, 2009 4:50 pm

Re: n0nx0x2

Post by rv0 »

the pulse width setting is important for things like the shuffle to work well, though it should only be changed from defaults when using a DAW, with hardware it's all fine.
as for shuffle resolution: keep in mind the software gets only 48 ticks per quarter note to work with (24 rising edges, 24 falling edges)
the 909/707 shuffle should be perfectly possible with the current code version (an analyses was made of how these and other devices handle shuffle/swing before the implementation began). The more extreme settings are there because crazy people like me like to do crazy stuff with it, you don't have to use them if they are not useful for you ;) (extreme settings like the Sync Unit DC plugin did before with a 303: http://soundcloud.com/rv0/extremeshuffle ) .. Same for the negative swing, it's not useful in normal 4/4 type music, but it has it's uses, like antto demonstrates in this jam: http://www.youtube.com/watch?v=D0NKSTeHGAk

3phase
 
Posts: 203
Joined: Wed Apr 22, 2009 2:06 pm

Re: n0nx0x2

Post by 3phase »

darffader wrote:the pulse width setting is important for things like the shuffle to work well, though it should only be changed from defaults when using a DAW, with hardware it's all fine.
as for shuffle resolution: keep in mind the software gets only 48 ticks per quarter note to work with (24 rising edges, 24 falling edges)
the 909/707 shuffle should be perfectly possible with the current code version (an analyses was made of how these and other devices handle shuffle/swing before the implementation began). The more extreme settings are there because crazy people like me like to do crazy stuff with it, you don't have to use them if they are not useful for you ;) (extreme settings like the Sync Unit DC plugin did before with a 303: http://soundcloud.com/rv0/extremeshuffle ) .. Same for the negative swing, it's not useful in normal 4/4 type music, but it has it's uses, like antto demonstrates in this jam: http://www.youtube.com/watch?v=D0NKSTeHGAk
oh i wouldnt say that negative shuffle is unusefull.. just extreme settings are most of the time ..but funk is relative..

But wouldnt it be better than when the min/max range setting for the shuffle would alter the time base value instead of just limiting the range of the factor parameter input?

So you can adjust how far your swing should go but always have the maximal step resolution of up to 16 variations in each direction? So almost any possible swing factor, or shuffle, could be achieved and when you prefer smaller values you get a very fine resolution?

The whole clock shift thing is pretty good.. actually it would be great if positive and negative shuffle value could be set independently .. than you also can realize offsets with the whole machine with this and can make it fit precisely also with late machines like modern dsp based synths..or even software synths you migh want to run from a midi hardware sequencer but than have to cover a latency of > 10 ms..whatever.. wouldnt hurt to be able to use positive and negative swing settings simultaneously..

for the handling it wouldnt be so easy maybe.. but maybe when theh direction you move the tempo encoder first makes it jump to the positive or negative setting page. or pressing 3/4 and back simultaneously when pressing the tempo button gets you to the negative setting..while only 3/4 and tempo switch go to positive swing values?

when you than have a negative swing setting you only need to use the positive swing setting page to move from a negative to a positive swing.. but with an offset defined by the initial negative swing value.

3phase
 
Posts: 203
Joined: Wed Apr 22, 2009 2:06 pm

Re: n0nx0x2

Post by 3phase »

But that depends of cause what you see as a negative shuffle.. i was thinking that its a delay of the uneven 16th positions while a positive shuffle is a delay of the even 16th positions.. But.. maybe wrong interpretation and negative shuffle is shifting the even 16th early.. but that would be more difficult to achieve than just working with delays...

In any case i see use in a general offset since its sometimes nice to have the 303--or xox ..late..

In my setup i use a daw as clockmaster just for the reason that this allows me to have the 303 on a separate midi port i can delay just as the track needs it..and especially on the 303 i change that setting often..while on drummachines its usually set once and never be touched again. But with a synth like the 303 any possible timing/feel can be musically usfull since it can cover such a wide range from soft and hard sounds..

xlarge
 
Posts: 151
Joined: Wed Jul 11, 2007 2:02 pm

Re: n0nx0x2

Post by xlarge »

http://antonsavov.net/cms/projects/n0nx0x2.html

New version out!

What's new in v2.10:
New: Shuffled clock added. In addition to the tick-based shuffle from the previous version - shuffled clock is only available when the x0xb0x is Master. It can generate shuffled clock (which affects all sync outputs, DinSync / MIDISync / XSync).
Changed: Changes were made to the internal clock math: the digital oscillator (which generates the internal clock) now uses 64bit integer math to calculate the frequency, since shuffled-clock requires a bit more accurate coefficients.
New: Added a "Freeze Frame" feature. It's a sort of buffer-override / glitch / stutter effect. It's available in Pattern-Play and Track-Play mode.
Changed: Some small cosmetic changes were made: the Tempo LED blink is now even shorter.
Changed: The numeric LED display animation (for displaying numbers from 0 to 999) is no longer tied to the tempo. The animation is now at a fixed rate.
New: MIDIPLAY mode now uses the "midi-note-tracker" code to track up to 8 simultaneously held notes. Now one n0nx0x2 can drive another one with the MIDI notes. It's also possible to play the x0x via a MIDI keyboard in legato style.
New: Added a "PANIC" function to MIDIPLAY mode. It kills any stuck notes and resets the midi-note-tracker.
Changed: Added handling of MIDI_START to override any unhandled MIDI_STOP, this happens with some stupid DAWs which don't send MIDI_CLOCKs while stopped.
Bugfix: The note-gate-out (from XSync) was sometimes getting stuck "high" on sequencer stop.
Changed: Added code to set the note-gate-out to "low" on MIDI_STOP, otherwise it gets stuck (due to some DAWs which don't send MIDI_CLOCKs after a MIDI_STOP).
Changed: The "Dump Settings" function in Computer Control mode now prints the values more sanely specifically the midi channels were printed 0 to 15 while they had to be printed 1 to 16 instead (now fixed).
New: Added visual indication for the at-boot functions (those who didn't have any).
New: Added an option to disable the output of MIDI notes from the patterns (they are enabled by default).
New: Added an At-Boot keycombo to set SPI to 4MHz.
Changed: "MIDI-Trigger" mode (A) is now renamed to "Pattern-Trigger" mode.
Changed: Pattern-Trigger mode: small button optimization.
Changed: The test-tone (in Pattern-Play mode) now has a shorter retrigger time.
Changed: some small adjustments were made to the retrigger times in some places (they are shorter now).

3phase
 
Posts: 203
Joined: Wed Apr 22, 2009 2:06 pm

Re: n0nx0x2

Post by 3phase »

And a feature request. Its a pitty that nonxox cant transpose patterns via midi.. there are hardware 303 that can do that. How about other xox os versions? can sokkos transpose patterns via midi?

Locked
Please be positive and constructive with your questions and comments.

Return to “x0xm0dz”