Percussa Home / Shop / Contact /

MIDI weirdness (really!)

#1

Hello!

Struggling to understand what is going on!

I copied a preset to a new slot. Did some slight changes, but mainly to the midi contrller number, as I wanted to map my Fadefox EC4 differently.

NOW – MTIN will receive, learn very nicely – BUT does not pass on the signal out to modules. I litterally can’t see any signal change on the right scope when I select the correct CV out.

However, when I go back to the previous preset, the learned CCs are gone from the MTIN, BUT will still send the correct cc to the correct place! How strange is that?!

And Iäve naturally rebooted a couple off times. I’ve removed and re.instated MTIN several times. Nothing happens. And in the old preset, if I instantiate a new MTIN, learn a new cc, again nothing is output!

I am completely flumoxed and imagine I’m making a really silly msltake somehow, but not matter how many times I check all settings, etc, nothing si diffreent and the “new” MTIN just don’t want to output a signal…

Any help or suggestions appreciated!

Thanks,
Girilal

MIDI Bindungs disappear after reconnecting a USB device
#2

ok, lets clarify something first, otherwise its not going to make sense :slight_smile:

so you do not use learn on MTIN.
the parameters you see CV_A to CV_H (?) , are the CC numbers for that output
i.e.
CV_A = 1 , means its receiving CC1, and will output that on CV_A

(if you did learn, then you’d actually be changing the CC number used for the output… very meta :wink: )

or rather, thats how it should work :slight_smile:

I seem to have broken it, (probably in the last refactor?) I cannot get it to output anything.

that said, I did take a quick look at the code, and it seemed to be fine…

what is odd, is the flow is very similar to how midi learn is being used on parameters, and that is working - and thats why I didn’t spot the break, as that is working fine.

I’ll need to take a deeper look to see whats going on…

for now, you can take a different approach, one that I often do and that is to use ATTN.
basically, ATTN has normalise voltage on its input, so if you midi learn the params on it, you’ll get a

as for the rest of your post, im not quite sure whats going on…
if you did learn, you shouldnt see output at all.
if you did it ‘correctly’ , I dont know how your getting output as Im not getting anything :wink:
do I dont know whats going on.

but honestly, I cannot even think about that, until I have MTIN working, and so can try to reproduce the steps you mention.

1 Like
#3

ok, this gets stranger :laughing: (… well, probably not, as you’ll see below)

so I built/ran it on mac, worked fine…

tested again, on SSP, still not working…
ok, I’ll build for SSP with some debug - ran into new compiler issues with juce (not uncommon!)
so did a minimal fix, and ran a full build…

whilst doing this, logged into SSP, killed synthor in prep for testing with my new debug version.
ran it up again… just to check something, now worked find !
rebooted, to run ‘as normal’
now works fine !!

I think in my case it was possibly ‘user error’ ?

for some reason it only works in usb port 2…
I could swear this did not used to be the case, on later firmware, but looks like I mis-remembered
(in fairness I usually have my network adapter in usb 1, so that could be what I was thinking of)

the other side, is there is no automatic ‘reconnect’, basically if you connect after preset load it will not detect it, you have to reload the patch (or delete/add the module)

finally, Im using an Erae 2, not the simplest controller so I could have done something wrong lol

I suspect I did a combo of the above… probably tried usb1 , didnt work so switch to usb2, but didn’t remember the no-reconnect…
then I instantiated a new module, and ofc, all was fine with that one…

so yeah… user error, but understandable :wink:

back to the OP…
so now was working, I tried dup-ing the patch, and switch between the two.
and this seemed to work perfectly… i.e. both patches continued to work, as I went back n’ forth…

so Id probably need a more concrete test case.

again, sorry for the confusion… Im working on something else thats pretty complex, so probably my brain, just overloaded… and forgot the details.

#4

Hi, Mark!

Speaking of user error – yup! I had indeed gone full meta by continuously changing the receiving CC value! :slight_smile:

After your first reply, I suddenly remembered that this is in fact the second time I have done this! I’d forgotten how it works!! Again!

So, at my end everything works now! I forget which batch of your software I installed, but it was more than a year ago, probably two. So, perhaps the bug you are looking at now hadn’t been manifested yet?

Anyway, sorry for my confusion and for the bother!
But mainly thanks for pitching in so quickly! Saved my gig yesterday afternoon!

Girilal

1 Like
#5

as above, I don’t think there is a bug, I think Id just forgotten how it worked, amusing considering I implemented it, and spent hours testing it :laughing:

I think there are two things really
a) unable to use USB 1
this is an OS thing, I cant do anything about, well except to reconfigure the OS.
I could look into this, as I think its possible, as my ethernet and wifi dongles work fine in both usb ports.

b) connect / reconnect logic.
this Im pretty sure I could improve, also, I could make it cleared on the midi page if it thought the device was connected/disconnected. together, this probably would probably fix the more confusing scenarios … save me some head scratching!

anyways, glad you got it working, thats the main thing.

1 Like
GRID Plugin Project - a better 8-channel sampler
#6

I think that would really be a great help! SSP is kinda “magic” in some aspects of how it works, so anything that gives a bit more info is very welcome! .-)

Looking forward to more of your brilliant work as always!

Girilal

1 Like