Hello everyone
I recently bought Zebra2.8 and unfortunately I encounter a strange problem which simply makes it impossible to use it professionally. I work with a iMac, OSX 10.13 and with Ableton Lite. The problem is reproducible here with the Initialize patch and I have verified that no velocity-dependency is present.
The perceived volume is not constant when I play notes on my keyboard (either my LaunchPad or the computer keyboard). It is constant when I quickly play notes, but I can reproduce a volume drift by pressing one note in intervals of about a half second (note hold duration doesn't matter). Then the notes gradually become louder. When I make a pause of several seconds, the volume reverts to the initial level and the volume drop is really significant. Sometimes the volume changes without these reproduction steps as well. If I record the test and play it, the same volume drift can be heard. I don't see anything in the recorded data that would explain the volume drift. Changing the velocity in the recorded data doesn't change anything.
I have excluded several possible causes: it is not the DAW (the problem is reproducible with both Garageband and Ableton Lite). Playing samples in audacity is 100% constant, thus the operating system is most likely also not the cause. I have never encountered similar volume drifts outside of the Ableton environment. Ear adaption is also excluded: the effect can be easily verified using the microphone feedback in the OS X preferences.
The volume feedback/green bars in Ableton report the same level during the test described above, so there is a mismatch between acoustic volume and volume displayed in Ableton. It is almost as if the sound is played with increased compression, but I don't have a compression effect enabled.
Restarting Zebra2 or Ableton doesn't change anything.
I simply would like to ask whether anyone has encountered such a problem before and may have an idea what could be causing this.
Thank you and best regards
Zebra2: unwanted volume drift
-
- KVRer
- Topic Starter
- 10 posts since 21 Apr, 2016
-
- KVRAF
- 4130 posts since 11 Aug, 2006 from Texas
Hi starbird, it could be several things and I can't really tell without actually loading the preset you've modified. Because Z2 is so complex and flexible its probably better to post your patch for analysis, if that's possible.
Feel free to call me Brian.
-
- u-he
- 28065 posts since 8 Aug, 2002 from Berlin
-
- KVRist
- 83 posts since 19 Jul, 2008
A couple of ideas but without hearing an example its hard to say...
1. On the init patch, or any patch with high sustain, it's normal for the overall volume to increase if you hold the sustain pedal while playing multiple notes - it's the effect of sonic material added to other sonic material. You may also hear this on patches with a long release time, and it's especially prevalent when repeating the same note- repeated notes don't replace their predecessors but instead add a new instance of the note.
On some synths - especially models of older poly synths - this is less noticeable because the EGs and sustain pedal respond differently to multiple notes, although you'll usually still notice it in a series of notes that are physically held down through a phrase.
2. Is phase reset on or off for the oscillators? If off, theoretically this can cause differences between notes that could be perceived as volume changes, though I'd think they'd be minor.
I just went through a bunch of synth plugins testing this. It's noticeable on Zebra's init patch, but I also expect it there. It's less noticeable on some of Arturia's synths, more on others - it all depends on the original synth's behavior. I don't notice it in Pianoteq, but I wouldn't, because the notes don't sustain at full volume - by the time a subsequent note is played the note before it has gone down in volume - and for repeated same notes, the model is calculating the effect of the same sound source being restruck.
Similarly, my old Korg PS3100 behaved the way Zebra does, but some other polysynths I played that belonged to others did not.
You may be experiencing something totally different - again, can't tell without hearing it - but at least in my DAWs/hosts (Reaper, Bidule) Zebra behaves completely as expected in terms of volume consistency.
1. On the init patch, or any patch with high sustain, it's normal for the overall volume to increase if you hold the sustain pedal while playing multiple notes - it's the effect of sonic material added to other sonic material. You may also hear this on patches with a long release time, and it's especially prevalent when repeating the same note- repeated notes don't replace their predecessors but instead add a new instance of the note.
On some synths - especially models of older poly synths - this is less noticeable because the EGs and sustain pedal respond differently to multiple notes, although you'll usually still notice it in a series of notes that are physically held down through a phrase.
2. Is phase reset on or off for the oscillators? If off, theoretically this can cause differences between notes that could be perceived as volume changes, though I'd think they'd be minor.
I just went through a bunch of synth plugins testing this. It's noticeable on Zebra's init patch, but I also expect it there. It's less noticeable on some of Arturia's synths, more on others - it all depends on the original synth's behavior. I don't notice it in Pianoteq, but I wouldn't, because the notes don't sustain at full volume - by the time a subsequent note is played the note before it has gone down in volume - and for repeated same notes, the model is calculating the effect of the same sound source being restruck.
Similarly, my old Korg PS3100 behaved the way Zebra does, but some other polysynths I played that belonged to others did not.
You may be experiencing something totally different - again, can't tell without hearing it - but at least in my DAWs/hosts (Reaper, Bidule) Zebra behaves completely as expected in terms of volume consistency.
filmmaker/composer - http://www.brookhinton.com
-
- KVRer
- Topic Starter
- 10 posts since 21 Apr, 2016
Hello everyone
Thank you for all your contributions I appreciate this very much.
I did some more analysis of my own and found a simple yet surprising explanation, although I still don't know in detail where the problem is located. I hope that I can help other people with similar problems to solve them without going to the frustration I had to endure.
The first thing I did today was to make a recording with my test procedure (with the Initialize preset) and to export it as a WAV file, then load it into Audacity and play it there. And there came the first surprise: the volume drift was clearly audible there as well. And I immediately noticed that the amplitudes were absolutely tiny, even though the perceived volume wasn't that low. But the amplitudes did not increase visually even though I could definitely hear the acoustic volume increasing. So I exported the data as 8 bit PCM and looked at the sample values directly and found the amplitude to be around 3/128. I didn't see the amplitude increasing in the acoustically louder parts.
Then I used the 'Amplify' effect to increase the amplitude, and the volume drift was gone! Now that was definitely surprising. I don't have a explanation for this effect, but I wouldn't be surprised if there was some hidden volume adaptation logic buried somewhere in the operating system which manipulates the volume if certain conditions are met, such as the amplitudes being tiny.
So why was I affected by this problem in the end? During the last couple of days I was working late and wanted to reduce the total volume to not bother any neighbors (I wasn't on head phones), and Ableton's master volume is pretty convenient to control, especially with the Launchpad. For me the overall volume wasn't that small, but ear adaptation in a quiet environment and the apparently pretty powerful sound output of my Mac have probably conspired here as well.
In the end a quick test showed that the volume drift is gone when setting the master volume to 100%.
I am fully aware that reducing volume is a no-go when doing production work, but I was testing/jamming and so didn't care too much about small reductions in quality and I definitely didn't expect this weird volume drift effect as a consequence of the volume reduction.
Another lesson learned. Thanks for listening, best regards
Thank you for all your contributions I appreciate this very much.
I did some more analysis of my own and found a simple yet surprising explanation, although I still don't know in detail where the problem is located. I hope that I can help other people with similar problems to solve them without going to the frustration I had to endure.
The first thing I did today was to make a recording with my test procedure (with the Initialize preset) and to export it as a WAV file, then load it into Audacity and play it there. And there came the first surprise: the volume drift was clearly audible there as well. And I immediately noticed that the amplitudes were absolutely tiny, even though the perceived volume wasn't that low. But the amplitudes did not increase visually even though I could definitely hear the acoustic volume increasing. So I exported the data as 8 bit PCM and looked at the sample values directly and found the amplitude to be around 3/128. I didn't see the amplitude increasing in the acoustically louder parts.
Then I used the 'Amplify' effect to increase the amplitude, and the volume drift was gone! Now that was definitely surprising. I don't have a explanation for this effect, but I wouldn't be surprised if there was some hidden volume adaptation logic buried somewhere in the operating system which manipulates the volume if certain conditions are met, such as the amplitudes being tiny.
So why was I affected by this problem in the end? During the last couple of days I was working late and wanted to reduce the total volume to not bother any neighbors (I wasn't on head phones), and Ableton's master volume is pretty convenient to control, especially with the Launchpad. For me the overall volume wasn't that small, but ear adaptation in a quiet environment and the apparently pretty powerful sound output of my Mac have probably conspired here as well.
In the end a quick test showed that the volume drift is gone when setting the master volume to 100%.
I am fully aware that reducing volume is a no-go when doing production work, but I was testing/jamming and so didn't care too much about small reductions in quality and I definitely didn't expect this weird volume drift effect as a consequence of the volume reduction.
Another lesson learned. Thanks for listening, best regards
-
- KVRer
- Topic Starter
- 10 posts since 21 Apr, 2016
I was a bit too early to declare success. It looks like the volume adaptation already kicks in with moderate volumes. Setting Ableton's master to 100% is not sufficient for some of the sounds I tested (unless I turn up the oscillator and mixer volume in Zebra2 as well). I am pretty surprised that no one else seems to have this problem (?)
-
- u-he
- 28065 posts since 8 Aug, 2002 from Berlin
-
- KVRAF
- 23102 posts since 7 Jan, 2009 from Croatia
Perhaps a rogue plugin on a master track somewhere (a limiter or compressor)?
-
- KVRer
- Topic Starter
- 10 posts since 21 Apr, 2016
I have found a thread which seems to discuss the same problem as the one I have:
https://forums.macrumors.com/threads/ne ... n.1555906/
There is a link to another thread, where post #9 describes exactly what I am experiencing:
https://forums.macrumors.com/threads/an ... t-18971933
If what they suggest is true then it's an issue limited to the internal speakers. That explains why so few people have noticed the problem since no one does serious sound work with the internal speakers. My boxes have died and my headphones are temporarily unavailable so I had to resort to the internal speakers until I get new boxes/headphones.
https://forums.macrumors.com/threads/ne ... n.1555906/
There is a link to another thread, where post #9 describes exactly what I am experiencing:
https://forums.macrumors.com/threads/an ... t-18971933
If what they suggest is true then it's an issue limited to the internal speakers. That explains why so few people have noticed the problem since no one does serious sound work with the internal speakers. My boxes have died and my headphones are temporarily unavailable so I had to resort to the internal speakers until I get new boxes/headphones.
