boafzilla Posted June 23, 2012 Share Posted June 23, 2012 (edited) I'm having an issue where some samples loaded into Ultrabeat will only play on the second note event and while playing the hit twice. To be specific, I have a few hi-hats routed to the 5-6 channel of the Ultrabeat that will not play exactly when triggered. This channel does not have any effects or send routings. I have tried changing the hits routing to channel 3-4 and it works just fine. This issue happens regardless of how the hit is played (i.e. midi from a controller, midi event in a midi region, clicking the key next to the drum in ultrabeat, adding notes to a region [with the midi out switch enabled]). Any ideas? I'm about to dump UB and just use an EXS24 I'm running 9.1.7 I realize this sounds like a hazy issue, so please let me know if you need more information. EDIT: Clarified topic... Edited June 24, 2012 by boafzilla Quote Link to comment Share on other sites More sharing options...
David Nahmani Posted June 24, 2012 Share Posted June 24, 2012 EDIT: Clarified topic... Thanks a lot for doing that! Would you mind uploading your project to this thread? You could save a copy of your project and delete all tracks but the UB track and its auxes, then in the Finder, Control-click the .logic file and choose "Compres..." to get a .zip file, then at the bottom of the post box click "Upload attachment" and attach it here. Quote Link to comment Share on other sites More sharing options...
boafzilla Posted June 24, 2012 Author Share Posted June 24, 2012 Thanks for the quick reply, David! See attached for the compressed Logic project (it's not very large). 21 copy.zip Quote Link to comment Share on other sites More sharing options...
David Nahmani Posted June 24, 2012 Share Posted June 24, 2012 I'm afraid your project is corrupted. And that's one of the strangest corruption I've ever seen!!! I'm seeing all sorts of other problems as I'm playing with your project. Meters showing signals while the project is stopped, no input monitoring armed, no record armed, UB return channels showing signal when UB is not playing, some UB voices showing signal on the Aux, but the signal not being routed to the Stereo Output, etc etc. Quick Workaround: insert a Channel EQ on the Aux receiving UB 5-6, leave the EQ on but flat. For some reason, it solves the hit and miss double hat issue. I've tried reimporting the tracks in another project but unfortunately the issue was imported along with it. Quote Link to comment Share on other sites More sharing options...
Eriksimon Posted June 24, 2012 Share Posted June 24, 2012 I haven't seen all those extra problems, but the actual issue is baffling enough. As the OP described, you have to play the MIDI note (doesn't matter how) twice, to get two rapid hits of the hat (± 1/32 apart). Even bafflingererer is that if in Ultrabeat you turn off the sustain on ENV 4 of the offending hihat sample, the problem is that you now need to play the midi note three times and the third time the sample will play three times in rapid succession... :? :? :? And indeed it is tied to Ultrabeats' 5-6 out. Any other output and the behaviour is back to normal. So I'd say that's the workaround here. Do not use output 5-6 for hihat samples. I tried the other hihat samples, same behaviour. I tried a snare sound on that same track, and, to my surprise, no problem. Only the hihats appear to do this... Very weird issue indeed. Maybe a fresh instance of UB will not exhibit this issue. Create a new SI track, and move your HH midi regions to it, load Ultrabeat and configure it for the hihat(s) only. Quote Link to comment Share on other sites More sharing options...
David Nahmani Posted June 24, 2012 Share Posted June 24, 2012 Thanks for your help Erik. Did you try my quick workaround for "fixing" ouptuts 5-6? Wondering if it also works for you? Quick Workaround: insert a Channel EQ on the Aux receiving UB 5-6, leave the EQ on but flat. For some reason, it solves the hit and miss double hat issue. Quote Link to comment Share on other sites More sharing options...
Eriksimon Posted June 24, 2012 Share Posted June 24, 2012 Thanks for your help Erik. Did you try my quick workaround for "fixing" ouptuts 5-6? Wondering if it also works for you? Quick Workaround: insert a Channel EQ on the Aux receiving UB 5-6, leave the EQ on but flat. For some reason, it solves the hit and miss double hat issue. Yep, it does. In fact, any insert plugin will do. As long as you can dial it to just passing through the signal. I tried the Pedalboard (without any pedals) and the Gain, both solve the issue, so that's an even easier workaround. It's an odd bug. We need a software equivalent of dr. House to diagnose this one... Quote Link to comment Share on other sites More sharing options...
David Nahmani Posted June 24, 2012 Share Posted June 24, 2012 Isn't that weird? Never seen that before. I tried another plug-in yesterday and it did not work though. When I used the Channel EQ it worked. Haven't tried Pedalboard or Gain plug-ins but I trust you they work as well. Weirdest thing. Quote Link to comment Share on other sites More sharing options...
Eriksimon Posted June 24, 2012 Share Posted June 24, 2012 Futzed around with it some more...: Play with the decay time of env 4, with sustain off make it shorter and you can get it to need 4, 5, 6 and 7 note ons before firing the 4, 5, 6 or 7 samples in rapid succession on the 4th, 5th, 6th or 7th note on... 8) There appears to be some method in the madness. (...) Or not... now I can no longer induce the error - playing with the ENV4 times (on hat 1 and hat 2) seems to have "Macsorcised" the pesky little demon... yet another workaround (or a miracle cure, ) Quote Link to comment Share on other sites More sharing options...
Eric Cardenas Posted June 24, 2012 Share Posted June 24, 2012 It's the Adaptive Limiters summed latency on the perc 2 AUX that is giving you this behavior. You can delete the unused one if you want. The behavior you are experiencing is related to Latency inducing plug-ins with a Multi-Output Instrument and in this case plug-in latency exceeding the sample length in UB. Deleting one of the Adaptive Limiters will reduce the Plug-in latency from 100 ms to 50 ms. Quote Link to comment Share on other sites More sharing options...
Eriksimon Posted June 24, 2012 Share Posted June 24, 2012 It's the Adaptive Limiters on the perc 2 AUX that is giving you this behavior.You can delete the unused one if you want. The behavior you are experiencing is related to Latency inducing plug-ins with a Multi-Output Instrument and in this case plug-in latency exceeding the sample length in UB. That seems like a reasonable explanation, but imho it is not. There's just no good reason why the latency of the AdLimiter on another channel strip (not even in the signal path of the gluggy notes) would make it necessary to click or play a midi note twice in order for it to play two shots on the second note. Also, there's no time factor involved. You click a note once, no sound. Wait a second, minute, hour or day, click a second time, voila, two rapid double hihat hits happen(say that quickly five times ). Something appears buffered somewhere somehow. To put it cautiously. And it also happens if the lookahead is set at 20, in fact, the lookahead parameter seems to have no effect on the glug (glitchbug). That the length of the produced (the ENV4 decay setting) note is a factor, is clear. I'll try see if the bug still happens with all PLC turned off then... (testing...) This seems to support your diagnosis, dr Eric "House" Bradley. Bug only happens with PLC set to All. I must admit though that removing one of the AdLimiters indeed resolves the issue, reintroducing it also reintroduces the bug. And since there cannot really be any good reason for using two Adlimiters on one channel strip anyway, it is the best workaround or even solution for this issue. Remove the second AdLimiter. Quote Link to comment Share on other sites More sharing options...
Eric Cardenas Posted June 24, 2012 Share Posted June 24, 2012 Try setting the Lookahead of a AD-limiter > 80-something and you will encounter the flam. Quote Link to comment Share on other sites More sharing options...
Eriksimon Posted June 24, 2012 Share Posted June 24, 2012 Try setting the Lookahead of a AD-limiter > 80-something and you will encounter the flam. Any lookahead setting (on either AdLim) gives the flam here. Have you already tried the neat 2, 3, 4, 5, 6, and 7 flam variations (sustain off, shorten decay time on ENV 4)? Bug-dressage! In fact, I'll try to recreate this from a new empty project, to see if it is indeed project corruption or an application buffering bug. But since David said the issue imported with the tracks, I'd suspect app bug rather than corrupt project - which in a way is good news for boafzilla. And as bugs go, this is tolerable, especially since it needs all those specific settings for it to happen AND it is easily cured IF it happens. Quote Link to comment Share on other sites More sharing options...
boafzilla Posted June 24, 2012 Author Share Posted June 24, 2012 Wow! I didn't expect a full-blow discussion to come through on this one. Thanks for all the troubleshooting, guys Bug only happens with PLC set to All. I just tested this theory, and you are totally correct! I had seen a mention of this setting in another thread (can't find it now) with a similar problem (delay when routing UB through multiple channels), but I never thought to try it out So, this seems to be at least part of the fix... Quick Workaround: insert a Channel EQ on the Aux receiving UB 5-6, leave the EQ on but flat. For some reason, it solves the hit and miss double hat issue. As does this... It's the Adaptive Limiters summed latency on the perc 2 AUX that is giving you this behavior. And this... Meters showing signals while the project is stopped, no input monitoring armed, no record armed, UB return channels showing signal when UB is not playing [...] I can't say that I was seeing these issues, but if you were I believe you! This whole project is messed up... [...] some UB voices showing signal on the Aux, but the signal not being routed to the Stereo Output, etc etc. You know, now that you mention this one, I can recall early on in the early phases of creating the template that this project was based on, I was having really strange issues with the Aux channels created for my multi-output UB. Something like the channels were only audible if routed to a very specific Aux channel (like Aux 12-16 or something). So, maybe this all stemmed from the original template? In case you folks have some time, it's attached... Scary thing is, I have a good number of other projects based on this exact template. :( :( I'd suspect app bug rather than corrupt project - which in a way is good news for boafzilla I guess! If this leads to a patch in 9.1.8, sweet! boafzilla v0.1.logic.zip Quote Link to comment Share on other sites More sharing options...
boafzilla Posted June 24, 2012 Author Share Posted June 24, 2012 Deleting one of the Adaptive Limiters will reduce the Plug-in latency from 100 ms to 50 ms. You know... maybe if I just watched my levels better and fixed my compressor settings I wouldn't have to rely on Adaptive Limiters in the first place... Quote Link to comment Share on other sites More sharing options...
boafzilla Posted June 25, 2012 Author Share Posted June 25, 2012 Just a quick bump... Can anyone else take a look at that template I uploaded and see if maybe some of the issues are stemming from there? I don't have a very deep understanding of the intricacies of Logic's quirks/features, so a pro set of eyes and hands would be super appreciated. Quote Link to comment Share on other sites More sharing options...
Eric Cardenas Posted June 25, 2012 Share Posted June 25, 2012 Just a quick bump... Can anyone else take a look at that template I uploaded and see if maybe some of the issues are stemming from there? I don't have a very deep understanding of the intricacies of Logic's quirks/features, so a pro set of eyes and hands would be super appreciated. The problem is not your template, though it is too complicated for it's own good. Too many busses routed to busses routed to busses... As I stated earlier the problem is the delay that the Adapltive Limiter's Lookahead is adding to an AUX in a Multi-Output instrument. It miss-triggers when that delay is exceeding the length of the sample wile PDC tries to compensate for that latency added. Try this: Start with an empty template (44.1 kHz) and initiate a Multi-Output version of UltraBeat. Create two Auxes by clicking on the little plus sign on the UB channel strip. Chose the preset Drag and Drop Samples. Use the Hi-Hat Sample from your previous example and load it on an empty slot. Route it to Output 3-4. Insert an Adaptive Limiter on the 5-6 Aux. Open the disclosure triangle at the bottom of the plug-in window and enter a Lookahead of 100 ms. Create a backbeat Hi-Hat pattern on your original UB track. Press play... you will hear the same miss-triggering. Press stop. Now decrease the Lookahead to 70 ms or less and press play again. Hope that makes any sense. Quote Link to comment Share on other sites More sharing options...
boafzilla Posted June 26, 2012 Author Share Posted June 26, 2012 Hope that makes any sense. It does, thank you. So are you saying any plugin (Logic-native or AU) that has a lookahead function of some kind can introduce this behavior? Thanks again for your detailed explanation. I think this thread can be marked solved. Quote Link to comment Share on other sites More sharing options...
Eric Cardenas Posted June 26, 2012 Share Posted June 26, 2012 Hope that makes any sense. It does, thank you. So are you saying any plugin (Logic-native or AU) that has a lookahead function of some kind can introduce this behavior? Thanks again for your detailed explanation. I think this thread can be marked solved. You're welcome man! What I am saying is that a plug-in or plug-ins that induces more latency than the sample length is culprit. A nice thing to try is to take your Hi-Hat sample and drag it in to the arrange. Now time stretch it to about half its size (Option - Drag the right region border. ). Import that into UB and you will have to decrease the Lookahead to roughly 30 ms or less for it to work. Quote Link to comment Share on other sites More sharing options...
boafzilla Posted June 26, 2012 Author Share Posted June 26, 2012 Oh I see what you're saying now. So, the issue is not plugin-specific... I just tested this in an EXS24 and it definitely behaves the same. I guess my last question is (and my apologies if this was already answered), what's causing the issue to only show up when the sample is routed to a non-Main channel (i.e 3-4, 5-6). For example, here's what I did with the EXS24: Quote Link to comment Share on other sites More sharing options...
Miguel972 Posted July 3, 2012 Share Posted July 3, 2012 Hi guys, I had exactly the same problem and and a quick Google search pointed me to this thread. So thank you guys for your troubleshooting because you've made me save a lot of time Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.