... So I saved two different ALLs under two different Perf Numbers 4 and 6 to try if I can switch/load between them via CC. So I'd send CC32 with #4 or #6. Did not work at all. I made sure to check each time the global Midi settings (Bnk2Perf ON)....
Oh, I'm a god-damned idiot. You need to send CC0 for bank change (there are 2 bank change MIDI CC's). Sending CC32 won't do anything. I haven't thought about that in a while since my octatrack only sends bank change CC0.
I don't know the QY700, you might want to ask someone who has experience with it. A MIDI message is 2 or 3 bytes. Program Changes are 2 bytes. The first half of the first byte identifies it as a PC message, the second half is the channel, the second byte is the program number to change to (pattern # for the LXR). I don't know what those 3 numbers in the QY are meant to represent.
It might help assigning the voices to different channels at first to make sure each channel is doing what you want, and sending one bank/program change signal at a time.
One huge thing for me, (I think I read right) is you can now triger via the drum select button while in edit mode. So you can hear what your doing while designing a sound without playing the sequence.
One huge thing for me, (I think I read right) is you can now triger via the drum select button while in edit mode. So you can hear what your doing while designing a sound without playing the sequence.
This just made me want to try this firmware. My number 1 bugbear with the LXR.
Still trying to figure out everything this firmware offers through this thread and the video. I'm really enjoying it, thanks Brendan !
However, I'm experiencing some midi issues. I would like to know if anyone has the same problem before going back to the stock firmware and check if it's also there.
I have the LXR synced to Logic through midi and when I press play on the daw the LXR starts. The problem is that after some bars it will go completely out of sync sometimes it's when I move a parameter on the LXR and other times it's just randomly even if I'm not touching it.
Oh, I'm a god-damned idiot. You need to send CC0 for bank change (there are 2 bank change MIDI CC's). Sending CC32 won't do anything. I haven't thought about that in a while since my octatrack only sends bank change CC0.
oh, I'm a god-damned idiot, too haha finally I figured it out.
so there have been at 2 mistakes on my side, even though I have to double check one of it. So the very stupid mistake was that I didn't know (or forgot cause I never used it before) that I have to select or reselect a pattern manually in order to change to a new loaded pattern/performance.
the other part - and there you comment helped me - was assigning voices/global midi channels.
[......]
Ok. While I'm writing this I just checked with the midi assignment. So in order for correct PC and BC it is necessary to set Global AND Voices to the desired Midi Channel you are sending the commands to (at least at my end). I thought because the commands for PC and BC are being global, one just have to set the Global Midi channel. If I just assigned the Global Midi channel there are changes but resulting in just weird behavior.
[.......]
And this is really great. Sending a BankChange Command in the middle of an 8 bar loop for the loading time of the performance followed up by an ProgChange for re-triggering a pattern and boooom there's the new performance. This was just what I needed in order to built a seamless live-set. Thank you so much for doing this.
However I have to send the PC on the first of the 8th (first step of the 8th measure) that the new performance/pattern changes at the first of the one (first step of the first measure), but if I send another PC on the first of the 2nd measure it changes immediately to the desired pattern. So it seems that for the first change of pattern it takes a whole measure (1-2-3-4) for the change to take place and after that it changes in time. Is this a bug or normal?
(I hope to be clear about the musical terms here, neither a native speaker nor a musician ; ) )
Oh and I do can confirm the pattern chaining bug, too.
...'m experiencing some midi issues. I would like to know if anyone has the same problem before going back to the stock firmware and check if it's also there.
I have the LXR synced to Logic through midi and when I press play on the daw the LXR starts. The problem is that after some bars it will go completely out of sync ...
Did you ever get a chance to check back to stock on this? I generally have my LXR synced to MIDI (octatrack) and haven't had a problem like this come up. Are you sending large amounts of MIDI CC or mod wheel by any chance?
It's been a while since I tracked from a DAW, but the general consensus used to be to sync the other way around - to slave logic to whatever hardware MIDI device was being tracked from.
... And this is really great. Sending a BankChange Command in the middle of an 8 bar loop for the loading time of the performance followed up by an ProgChange for re-triggering a pattern and boooom there's the new performance. This was just what I needed in order to built a seamless live-set. Thank you so much for doing this. ...
Thanks That's the idea - I use the octatrack to coordinate my performance and patch changes, tends to work well.
... However I have to send the PC on the first of the 8th (first step of the 8th measure) that the new performance/pattern changes at the first of the one (first step of the first measure), but if I send another PC on the first of the 2nd measure it changes immediately to the desired pattern. So it seems that for the first change of pattern it takes a whole measure (1-2-3-4) for the change to take place and after that it changes in time. Is this a bug or normal?
(I hope to be clear about the musical terms here, neither a native speaker nor a musician ; ) )
Oh and I do can confirm the pattern chaining bug, too...
hm. I have to take a look at that, there might be something odd there. In general, the pattern should change based on the "pattern change instant" global parameter in all cases, regardless of whether the PC message is from the buttons or external MIDI, and whether the pattern change follows a pattern load or not. I know for a fact that the buttons will change the pattern instantly after a load. If it's not consistent, I'll change it.
I just got back from a conference, so hopefully over the next few weeks I can shake out a few of these bugs. Thanks to everyone getting back to me about this!
I just finished a new project using manual pattern switching per midi program change for the first time and it absolutely rocks. Together with the mute option for individual voices it allows one to create nice build-ups. Thank you for this great option!
Also I would like to play around more with your new features, but I downgraded to an earlier version of your firmware because of the sequencer record bug and the pattern chaining bug. Both are essential features for me. Could you fix this in the meanwhile?
yes, I have plans to fix a few things in the next couple weeks. Can you clarify what you mean by these bugs? Is it this?
... I can confitm, that something is strange here. I had some patterns with different notes and everything is fine when I select them manually. But when I try to chain them only the first pattern is played. Although the sequencer steps between the patterns.
Another issue I discovered is that transposed notes are no longer recorded to a pattern while record is on. I hope this helps!
about recording transposes - the old method of recording the transpose was causing some pattens to behave strangely, so I simplified things a bit. To 'write' the transpose to the current pattern, hold the record button (it will flash), then release.
If what you're after is altering just a few of the notes 'live' without affecting the position/velocity, this is more easily done now with the 'roll modes' in the shift-rec menu.
yes, pattern chaining stops working for me with your new version. I had a simple chain of two patterns, but only the first one was played. After downgrading to your last firmware everything was fine again.
I just noticed, that transposed notes are not longer recorded by using the old method. Thanks for your clarification. I use the transpose option a lot!
I didn't had the time to dive into your new roll features yet. But this is definitely on my todo list. I know your time is limited, but a video would be nice here focussing on the realtime variations you added (loops, rolls, custom-knobs, retrigger LFOs, etc.) and how you use them for live performances.
Thank you for sharing! You bring the LXR to a new level!
...(there still might be a thingy where chaining doesn't load properly from patterns sometimes while the sequencer is playing - i don't think that's me, I think that's a general LXR bug. I'll look into this, but any reports how/when/why this happens would be appreciated)
- audio output assignments load properly again
more stuff over easter break since I'm locked out of the lab anyway! :-B
One thing that I noticed over the weeks is that it is not only the audio output assignment that does not get recalled after a bootup, but also the note assignment on each voice as well as the macros assignments.
Just like with audio output assignment, to get them to load properly, after boot, I always have to reselect all of them indivisually on one preset. After that, all the preset will recall the assignment well.
A couple small bugfixes: - some timing issues on the looper fixed. Sounds much better and more consistent now. - fixed an issue with rolls occasionally taking a long time (1 bar+) to trigger. - releasing shift doesn't revert the PERF menu back to the top page - it will remember where you were. - the shift-rec menu is sticky now (ie, it doesn't go away when you release shift). You can escape the menu by pressing shift again, or a MODE button.
might try looking at putting a Korg-style "active step" mode on shift-step next
the latest firmware is always up at bredanclarke.com/files/LXR/, on github, I keep branches that mirror Julian's, which are probably the old ones you're seeing, but the branches with custom features will always be called custom-'something'. The latest one right now is custom-develop-voice-morph custom-develop.
Ok, I'm clearly missing something really important... I downloaded your FIRMWARE.BIN file and plonked it on my SD card, but when I fire up the LXR I don't get any of these fancy new features (can't preview voices, no extra pages in Perf, etc.) What have I screwed up?
Aha! Section 8 of the manual eh? I got as far as the Quickstart section, which says to just plop stuff into the card and get going. Glad to know some people are more patient than me. Thanks faktoid.
It sure is! Just spent 2 hours playing with the pattern scaling. So glad it's not a nice day outside. Only downside is my previous Performances won't load anymore, but the new one I just made loads fine, so I'll just have to get busy making more Thanks for doing this Brendan!
Comments
I don't know the QY700, you might want to ask someone who has experience with it. A MIDI message is 2 or 3 bytes. Program Changes are 2 bytes. The first half of the first byte identifies it as a PC message, the second half is the channel, the second byte is the program number to change to (pattern # for the LXR). I don't know what those 3 numbers in the QY are meant to represent.
It might help assigning the voices to different channels at first to make sure each channel is doing what you want, and sending one bank/program change signal at a time.
It is more an extended firmware version, which adds new features to the official firmware.
so there have been at 2 mistakes on my side, even though I have to double check one of it. So the very stupid mistake was that I didn't know (or forgot cause I never used it before) that I have to select or reselect a pattern manually in order to change to a new loaded pattern/performance.
Did you ever get a chance to check back to stock on this? I generally have my LXR synced to MIDI (octatrack) and haven't had a problem like this come up. Are you sending large amounts of MIDI CC or mod wheel by any chance?
It's been a while since I tracked from a DAW, but the general consensus used to be to sync the other way around - to slave logic to whatever hardware MIDI device was being tracked from. Thanks
I just got back from a conference, so hopefully over the next few weeks I can shake out a few of these bugs. Thanks to everyone getting back to me about this!
Together with the mute option for individual voices it allows one to create nice build-ups.
Thank you for this great option!
Also I would like to play around more with your new features, but I downgraded to an earlier version of your firmware because of the sequencer record bug and the pattern chaining bug. Both are essential features for me. Could you fix this in the meanwhile?
If what you're after is altering just a few of the notes 'live' without affecting the position/velocity, this is more easily done now with the 'roll modes' in the shift-rec menu.
yes, pattern chaining stops working for me with your new version. I had a simple chain of two patterns, but only the first one was played. After downgrading to your last firmware everything was fine again.
I just noticed, that transposed notes are not longer recorded by using the old method. Thanks for your clarification. I use the transpose option a lot!
I didn't had the time to dive into your new roll features yet. But this is definitely on my todo list. I know your time is limited, but a video would be nice here focussing on the realtime variations you added (loops, rolls, custom-knobs, retrigger LFOs, etc.) and how you use them for live performances.
Thank you for sharing!
You bring the LXR to a new level!
- pattern chaining works again
...(there still might be a thingy where chaining doesn't load properly from patterns sometimes while the sequencer is playing - i don't think that's me, I think that's a general LXR bug. I'll look into this, but any reports how/when/why this happens would be appreciated)
- audio output assignments load properly again
more stuff over easter break since I'm locked out of the lab anyway! :-B
- some timing issues on the looper fixed. Sounds much better and more consistent now.
- fixed an issue with rolls occasionally taking a long time (1 bar+) to trigger.
- releasing shift doesn't revert the PERF menu back to the top page - it will remember where you were.
- the shift-rec menu is sticky now (ie, it doesn't go away when you release shift). You can escape the menu by pressing shift again, or a MODE button.
might try looking at putting a Korg-style "active step" mode on shift-step next
where can i found the newest version of your work?
the newest version on github is 9 month old.
thanks in advance