Totally Messed Up Envelope Behavior

[Elhardt]

I can live with quite a few shortcomings in a software synth, but when a basic module like the envelope generator is screwed up, it affects many patches in a negative way.  I'm getting the silent treatment from Arturia tech support which in itself pisses me off.  I explained the problem and he just concentrated on the times not matching the real stage times (they're off by a factor of seven times) and then said they modelled it after the Mg they have.  While the time issue is annoying, it's not the main problem.

The problem is the attack stage spends about half its time stuck at its maximum level.  I've been told by a synth technician that this is NOT the way a Mg EG behaves.  I've never encountered any EG that acts like this.  This leads to everything from sluggish envelopes at shorter attack settings to long periods of the attack stage just sitting at max level until it finally moves to the decay stage.  Here is an example of the EG patched into osc pitch with a 7 second attack.  As you can hear it reaches about max level half way through then just sits there for the other half before reaching the decay stage and falling off.

http://home.att.net/~synth6/MMV_EG.mp3

There has got to be some future fix for this.  Either a switch on the module to select a normal EG type, or another EG module altogether, or a global setting for the entire patch.  Here I thought the MMV would be my final synth of choice for most things and then this damn problem rears its ugly head.

I also find it a bit ridiculous when Arturia tries to claim they're modelling Mg behavior as an excuse and yet their fixed filter bank is really an EQ, not a bank of parallel bandpass filters as the real Mg has.  That's another thorn in my side.  Though an EQ can be equally useful, they should allow the EQ to be replaced with an actual fixed filter bank.  Even the formant filter is an EQ.  Again, that's not right either.

-Elhardt

[Sweep]

Is this the modulation envelope generators or the two main VCA envelope shapers?

[Elhardt]

Is this the modulation envelope generators or the two main VCA envelope shapers?

That particular mp3 shows one of the modulation envelopes.  However even the two VCA envelopes act the same way with the same type of attack curve.  They're all screwed.

I'm also getting more annoyed by the times being off by a factor of 7.  It seems ridiculous to intentionally write something like that into a synth.  And yesterday I just noticed that the decay and release times are different per knob position.  I was wondering why the release times on my timpani patch seemed to be longer than the decay times.  Then I noticed the decay and release have completely different time settings dispite that the actual values around the knobs on the graphics are the same as each other.  Boy it gets confusing.

-Elhardt

[Sweep]

Do you want to put up the patch you set up for the MP3 as well, so I can see if I get exactly the same result on mine and have a look at it. That seems more vaulable than me just replicating your sound. If I can download the actual patch as a file and then run it on mine and compare with the MP3, that may throw some more light on this.

[Elhardt]

Do you want to put up the patch you set up for the MP3 as well, so I can see if I get exactly the same result on mine and have a look at it. That seems more vaulable than me just replicating your sound. If I can download the actual patch as a file and then run it on mine and compare with the MP3, that may throw some more light on this.

Okay, I put it up at the following link:

http://home.att.net/~elhardt/EG_Problem.amc

It's certainly nothing special.  Just patching an EG into an osc and triggering it from the keyboard.  Since it's a lot easier for the human ear to hear changes in pitch than it is for hearing changes in loudness or brightness, that's why I attached the EG to an osc to show the problem.

It seems that the attack curve shape is just an inverted decay or release curve, and that's the problem.  It's necessary for the decay or release to level out into a flat horizontal level to smoothly fade into the sustain stage or fade into silence.  But the attack isn't supposed to fade into the decay or sustain stage.  It doesn't on any other synth in history that I know of.  The attack is always supposed to only keep rising up in level for the attack time setting, then stop when it hits the decay stage.

I've tried altering the attack shape by running the EG through an exponential mixer and also the modulation input into that same mixer.  But I end up with more of an "S" shaped attack.  It didn't help much.

-Elhardt

[Sweep]

I tried the link, but there's no Modular V file accessible. What comes up is the symbol for Quicktime Player with a question mark in it.

[Elhardt]

I tried the link, but there's no Modular V file accessible. What comes up is the symbol for Quicktime Player with a question mark in it.

On a PC computer you normally right click (that means right mouse button) on the link and then you can download it to disc.  It won't try to actually run it with some application like it will if you left click it.  It sounds like you might be on a Mac and I'm not sure what the procedure is there.  I haven't used my Mac online in years.

However, it's such a simple patch.  If you're trying to hear and see the problem, all it takes is patching an envelope into a modulation input in an oscillator and turn the nut to the right a bit until you can clearly hear the affect.  Basically forget I even posted an mp3 and a patch and just patch something up to hear it.

BTW, are you from Arturia?  I don't even know who I'm talking to.

-Elhardt

[Sweep]

I asked for your file because I can't see that making a patch that replicates your sound is going to help much. In fact it hasn't. It isn't replicating the sound that's needed, but replicating the problem. I'd like to run exactly your settings (i.e. your actual file of your patch) on my synth to look at the settings and try to see what's wrong.

I've programmed absolutely masses of very useable sounds on this synth and I've had absolutely no problem getting the envelopes to do what I need them to do. It may simply be that the problem you're referring to hasn't affected my patches, but at the moment I'm not clearly seeing what your problem is and why you find the envelopes so unuseable.

I tried right-clicking to get the file but without success. The options I get that way still don't get me the file. If you want to attach it to an email and send it to me at topographic.oceans@virgin.net I'll load it up that way.

To reply to your question about me - no, I don't work for Arturia.

[Elhardt]

I emailed it to you as you requested.  Note that I always try my own links and make sure I can download my own file, so I don't know why you couldn't.

>> asked for your file because I can't see that making a patch that replicates your sound is going to help much. In fact it hasn't. It isn't replicating the sound that's needed, but replicating the problem.<<

The problem is what my file is all about.  It's not about my sound.  I'm showing the screwed up shape of the attack stage.  So any patch that can replicate the problem is all that's needed if you want to see for yourself.

>>I've programmed absolutely masses of very useable sounds on this synth and I've had absolutely no problem getting the envelopes to do what I need them to do. It may simply be that the problem you're referring to hasn't affected my patches, but at the moment I'm not clearly seeing what your problem is and why you find the envelopes so unuseable.<<

Well I explained some of the problems it causes.  I first noticed it when I was trying to do a Vangelis type brass sound.  I wanted a slow brass-like attack and have it fall away with the decay.  The sound got brighter on the attack but then just sat there stuck for a while before finally falling off with the decay.  So right away it was causing problems.  I wanted to do a lot of W. Carlos type sounds where I might want a soft attack on a percussive like sound, but it gets sluggish because of this problem.  And then there is the issue of when you want to use the EG for wide modulations of pitch or whatever because the LFO's can't be started on a key press, but with this EG problem, it doesn't come out right.

Point is, this is not how envelope attacks are shaped in the real world.  I've programmed a lot of sounds on the MMV too in the few weeks I've owned it, but that's because I've programmed sounds that don't rely on the above conditions.  They're all mostly percussive or organ type envelopes.

Also, I was asking if you were with Arturia because I couldn't believe they were actually going to look into this problem.  Guess they're not.  Nor will they look at the other problems I've posted.

-Elhardt

[Sweep]

I received your file, but I'm unable to load it on my Modular V. I suspect we have different versions. All my files have an .amb suffix, and I notice yours is .amc, which might explain why my synth crashes when I try to load it.

Anyway, as I mentioned, the reason I asked for the file was to see exactly what settings you used to get the sound you got. I was hoping that might throw some light on why you say you have a problem. I could obviously replicate your sound, but as I can also produce plenty of other envelope shapes I can't see how that would help clarify what's so wrong in your view.

At first I wondered if the problem might be the slope time and slope level controls, which is why I asked about which envelope shapers were involved. But if that was the case you'd presumably have either checked the manual about those or else noticed you didn't have the problem when you checked how other people have created patches for this synth.

I'm still not clear why you have a problem. If all possible envelopes stuck for several seconds between attack and decay, fair enough. But they don't.

[Elhardt]

Since the slope knobs only have to do with an additional decay stage, they don't come into play.

I'm running version 2.0, so perhaps that's the .amb vs .amc problem.

And as far as the problem, I keep explaining why it's causing problems.  It's as if there is a hold stage inbetween the attack and decay stage because of the way the attack is shaped.  All envelopes do get stuck at that hold stage.  But if the attack is only a few milliseconds you won't hear it.  The longer the attack stage is, the equally longer that stuck part of the envelope lasts.  500 ms of attack also gets you 500 ms of stuck time.  10 seconds of attack gets you an additional 10 seconds of stuck time.  It's always there.  How much you can have before it causes problems in your sound is different based on what sound you're trying to patch.

That attack problem just got in my way on yet another type of sound where I was trying to synthesize a fast bowed string sound.  The attack stage is shaped like one quarter of a circle.  That means the first part is very vertical and percussive, so to to give it a soft attack requires increasing the attack more than on a typical EG.  The problem then is that stuck time increases and the sound doesn't quickly end.

An attack stage that's shaped like a quadrant of a circle is going to cause all kinds of problems.

Now I'm off to list other bugs I've found.  I'm now getting patches that are bombing out when played or patched.

-Elhardt

[Sweep]

Oh well, nothing's perfect I suppose. But Klaus Schulze sold his original hardware Mg Modular and then endorsed the Modular V, so it can't be so bad. And another synth pioneer who I won't name because we have a personal connection said they're very impressed with it. Quite a few respected synth musicians seem to feel the same.

I'm sure there's room for improvement, but only Kenneth Elhardt seems to think it's `totally screwed.' Funny, someone called Kenneth Elhardt keeps turing up on different synth discussion groups, almost always to moan about something that's apparently `screwed' or `useless.'

Personally, I'm delighted to have a software Mg modular that cost £170, doesn't fill my entire studio, has polyphony and patch memories and doesn't need regular maintenance. I can live with it being `totally screwed.' Then again, none of these `bombing out' problems happen with my synth, or any of the other bugs that seem to be happening on one person's computer. They didn't happen to Tangerine Dream last year when they risked running three (I think it was) Modular Vs on stage at the same time. Lucky, that.

[bg]

Sweep, are you disputing the EG behavior described by Elhardt, or just trying to impune him in general?

[Elhardt]

I'm wondering myself what Sweep is getting at.

I believe every bug I've posted probably shows up on other people's computers except the memory leak, because that one is where MMV interfaces with the computer and depending on OS, audio card and drivers, or computer (Mac vs PC), it will be different.

An envelope where the attack is shaped like part of a quarter of a circle is just messed up no matter how you look at it.  It's interfering with my sounds and it's not how real envelopes work.  So there is no defending it.  Even if it doesn't affect 95% of my sounds, the fact is that it soundn't interfere with any of them, and I shouldn't have to think about using another synth for those sounds it does affect.

And without trying to sound demeaning to others, it's their rather simple or basic use of some of these synths which is seems why they don't notice these problems.  I'm always amazed by this.  The first day I used my MOTM modular synth I noticed my sounds getting brighter the faster I played.  Turned out the motm EG climbs up in voltage the faster it's triggered.  This should be a problem for everybody who owns one of those EG's, yet nobody noticed back then.  In all the years since, only 3 other people I know of have been caused such trouble because of it, they went on to replace them with other EGs.  This really shows the sad state of affairs when problems with one of the most used modules isn't even noticed.  I can't believe I'm the first to try to hardsync a sinewave osc, or sync more than one osc to a source, or use the 2nd PWM input.  Yet it appears I'm the only one finding these problems.

I'm not disputing the fact that the MMV is a good synth for the price when it works properly.  But I'm just coming across bug after bug on almost a daily basis.  If I were using the MMV on stage like TD, I wouldn't use patches that bombed out when played either.  Most of the bugs I'm reporting aren't ones that bomb out the computer anyway.  They're things that don't work properly and should be fixed.  It's the same sloppiness I see in almost all software I buy and I'm getting really tired of it.  MMV has been out for years now.  The bugs I'm finding should have been found long ago and already been fixed.  Don't know why it takes me to find them.

-Elhardt