Support
Looper's Delight
|
Support |
|
|
> > > even if your midilooper pairs note-on to note-off, > > the receiving devices generally won't be able to do this. > > huh? doesn't a midi-off event have a note number as well? Maybe I'm > missing something here. > yeah, your missing that words are not my thing... the problem comes when you get more than one note-on with the same note number (via overdub) before the first note-off. this is what i refer to below as overlapping note pairs. the looper can tell when these result from overdubbing and keep track of them, but the receiving device has no way of distinguishing them. > > when you overdub and play a note that is already in > > the loop, the next note off will stop it, even if you > > are still holding it. > > not incredibly hard to compensate for in code. If its build as a midi > looper there are many graceful ways to fix this problem. If > we're just > talking about using a midi-codec with a regular looper this > is definitely an > issue. > yeah, you can work around some of this, but it will be weird to get used to using it, and in some cases will defeat, however selectively, the very function you are implementing (ex. if overdubbing cc's becomes replace or record, it is no longer overdub) > > if you have feedback set to less than 100% and a note in > > the loop is fading, if you play that note again, the new > > notes velocity will be hijacked by the next note that was > > fading. (this is kind of hard to describe) > > I am totally confused by this one. I'm very interested, if > you have the > chance to explain this one further! > need some way of communicating music notation via text to do this easily... ok. i'll use overly simple one note example. 1. create a loop with a staccato c4 vel.127 on 1 of a 4 beat loop. 2. set feedback at 80%. 3. let fade a few times, say three times - vel=65. 4. click overdub. 5. play 1/2 note c4 vel.127 on beat 1. if your new note goes out after the original note, you get note-on.c4.v.65 followed immediately by note-on.c4.v.127, then say 100ms later note-off.c4 (or note-on.c4.v.0) you'll get your staccato note, but at the new velocity. (this simple example is a lot like the overlapping note pairs problem, it becomes a different thing when the notes pairs are part of two different phrases) if your new note goes out before the original note, you get note-on.c4.v.127 followed immediately by note-on.c4.v.65, then say 100ms later note-off.c4 (or note-on.c4.v.0)the new velocity will be overridden by the old decaying velocity. note that in both cases you lose your 1/2 note. (you may even hear a strange blip at the start) like i said, when the notes are part of two different phrases, this could be quite disconcerting... ok too many words, and probably even murkier... > > overdubbing continuous controllers will make a mess as each new > > controller move would interleave with the moves in the loop, > > causing toggling effects. (as cool as they may sound, they w > > Once again, there are graceful ways around this - the looper > could ignore > incoming CC commands if it knows that there are already CC > commands on that > channel/number in the loop or in that section of the loop (or > alternatively, > there could be an option to allow/disallow CC overdubs - so > the user can at > least be conscious to this problem). > i don't see that as all that graceful. if you ignore cc commands while overdubbing, you are hardly overdubbing. notes have similar problems, though with less frequency, and if you apply the same logic, you would disallow note overdubs leaving no overdubbing. if you make it selective and only dissallow same note overdubbing, it is going to seem mighty strange when you overdub a phrase and 2 or three notes (from either an earlier pass or the new pass) disappear. > All good points Jim, now let me bring up a couple of neat > ideas that make a > midi looper a cool idea: > > *no bump at the loop point - great for ambient loopers -smile- there might be a bump depending on how you handle an unreleased note at record end. but it wouldn't be the same as the bump you get on the repeater. assuming you end record with overdub, the first note in the loop for each held note would chop off that held note. > *ability to do a live quantize function yes. > *It could receive all internal control commands on its own > midi channel > (these wouldn't get looped) this is more of a requirement than a feature... > *Much easier to synch a device like this to midi time clock > *pitch, tempo, and volume changes are easy to implement and > are totally > smooth > *global search and replace notes (maybe is totally > specialized and obscure, > but I think it could be really cool!) yes, yes, yes. > *could configure a midi-filter before the looper, so that only certain > things are looped, but all things are passed through (this > means that it can > be placed late in the midi chain and easily switched to > looping whatever > devices you want) now i get to be confused. if it is late in the chain, how is controlling devices that precede it? maybe you mean using a midi patchbay? > *wet/dry mix easy to do in software hmmm. how? seems like this has the same problems as overdub. > *if this is done in hardware, it could be very cheap and simple > > Jon > hmmm... thorsten's midibox as a looper... you'd definitely need more memory, esp. for cc stuff. but you could use the mf module for control of 8 params/channels. could be pretty cool. for those unfamiliar with midibox check it out: http://www.ucapps.de/index.html did anybody make it this far down?