Voices Outputting At Different Levels

[MaxwellsDaemons]

I have a P12 module thats exhibiting some odd behavior. I noticed the other day while playing it that some voices sounded noticeably lower in volume than others. So after investigating this further I decided to do some tests.

I had the unit on for several hours so I reset globals, cleared the filter calibration, and ran it again fresh. Then I initiated a basic patch, which is just a sine wave with no other processing going on, full gain, with wave rest so the sines all start at the same phase. Then I just recorded a sequence of 24 quarter notes at C3 (about 130.8 Hz), that way all the voices would cycle through twice. What I saw was the amplitudes of each voice fluctuate in a range of about +/- 3db. The pattern was the same for the first 12 notes as the last 12. Also I was sequencing this from my DAW, no velocity differences, no plugins, just the P12 straight into a line in on my MOTU interface.

I then read in the manual in the troubleshooting section that difference in the level of the voices could be due to the voice spread panning the voices in the stereo field, but I verified that this was set to zero. Furthermore, I recorded the left and right channels separately, and they all shared the same pattern of amplitude fluctuations. I also did the test again using C7 (about 2093 Hz) and while the amplitudes were still not all the same, this time they were different by about +/- 2db from each other.

So my thoughts are that in the oscillator section the amplitudes of the voices should all be pretty darn close to the same, unless of course you've got some processing going on that makes them different on purpose. But thats not so with the basic patch. One sine wave full amplitude, filters open, gain all the way up, no effects, no distortion, etc... So what I am thinking is that somewhere in the analog section, namely the filters and amplifier stages, these level differences arise.

Is this just normal analog synth behavior, or is this a sign my P12 is due for servicing? Its hard for me to tell. I wouldn't expect them to be bang on the same of course, but I would think they would be similar enough to not be noticeable.

Either way I am interested to hear what other P12 users experience. Thanks!

[timboréale]

My P12 has fairly consistent output level across all voices. I've run into some situations where I had a similar effect to what you describe, though I don't remember offhand what I did to fix it (it may have been a few runs through the calibration routine at different operating temperature points). I ended up sending my board in for service at one point, but they found nothing wrong with it. IMO this thing has so many quirks its hard to say what's "correct" behaviour from it, but it sounds so good it's worth working through.

Some off-the-cuff thoughts:

Are the _same_ voices low or high, not just from rep to rep but between start and stop or different notes? Do which ones are high/low change after you re-cal the filters (e.g. down to a specific voice /chip circuit on the board, comparing to the voice # light when it sounds)?

One thing to note: there are definitely small differences in how fast each filter operates. If you're using very short notes that are mostly in the "attack" phase length (even with a wide-open filter) you're likely to see more variances between voices than if you use longer ones - so make sure you're testing with fairly long (e.g. eighth or longer at 120bpm, ideally quarter, for reference) notes to be sure you're seeing things consistently.

If you don't figure it out, I'd contact support directly and ask them. There might be a modulation you forgot to reset when making your init patch, or perhaps there's some stray voltages being read from expression or volume pedal inputs that are affecting things, or who knows, maybe your unit does have an analogue output issue of some kind. If it does, sending in just the voice card for service is possible - that's what I did - and they are very reasonable about cost and turnaround as long as you have realistic expectations given they're still a small firm.

[Soundquest]

Just tried my P12 (keyboard version) no issues.  Identical output on all voices.

[jdt9517]

Check the voltage of the WallWort transformer.  S/B 13.5 volts.  If it isn't replace the transformer.  Common problem which causes lots of gremlins.