[SOLVED] MSF ADSR - part 2

[mevla]

I still have some problems with the ADSR that does not follow what other synths are doing. On other synths, like ZebraHZ shown here, having a sustain of 0 means that the release will pickup where the decay ends. However, with MSF, with a sustain of 0, if the note being played exceeds the decay time, the note event will find itself at 0 (circled in the screen shot below) and there will be nothing for the release to pick up. So no release for such notes.

It can be a feature. It's possible to play notes accordingly and have them with long release as long the player is playing accordingly shorter notes that the decay value and then plays notes with a longer decay value to have them chopping abruptly with no release at all.

It can be a feature, but as a default behaviour ?

bitwigADSR.jpg

MSFADSR.jpg

Env1 is for oscillator, Env2 for filter. The oscillator settings are about the same as for MSF above.

zebraHZADSR.jpg

[jmg8]

What would the release do, if the the value is already at 0?

[mevla]

There will be nothing to pick up to release, as I mentioned. Yet, with ZebraHZ with a sustain of 0, there's a bridge being made between the decay and the release. The release picks up at the decay and forgets about any sustain, since the sustain is at 0. And really, when the sustain is at 0, it means that the user does not want any sustain, no ?

EDIT: And really, when the sustain is at 0, it means that the user does not want any sustain, and does not want it replaced by silence, no ?

I've erased the one used for this demo, but no sweat, it was very basic anyways. Anything should do however, single OSC, some waveform and an ADSR shaped like the one above, with notes of various lengths being played. I'll rig a patch anyways.

[jmg8]

I just don't understand. Please could you make a basic preset in MSF. Make it so that something obvious should happen, but it doesn't, due to the issue you have. I will take a look for you.

[mevla]

OK, I have a patch. How do I do to produce one of those ASCII strings that I can copy here ?

[jmg8]

Click copy (In MSF)
Then here in the forum, use full editor and at the top there is a code insert feature. Paste it in there.
If you cant work it you can email me the settings at jmg8@live.co.uk, for this option use CTRL and copy to create a settings file and attach it to the email.

[David]

OK, I have a patch. How do I do to produce one of those ASCII strings that I can copy here ?

Here's a visual for how to use the copy/paste code function in MSF and in this forum for sharing stuff:

You use the icons highlighted in the screen grab below, in the Toolbar, which copy whatever you've got loaded, or replace that when you paste. Here you click the icon in the full reply editor, and paste into the space between the code brackets, in the usual way.

[mevla]

I could get the copied preset into MNotepad but then a MNotepad export only results in a file with:

<?xml version="1.0" encoding="utf-8"?>
<MNotepadpresets/>

And nothing else.

I'm running Linux. I can paste presets from here to MSF, but it looks like that copying does not make it outside of the windows bridge process space. I was hoping that a MNotepad export would result in a text file of the preset, but it does not. Only the XML header is saved as shown above. Good to know as I started to take notes using MNotedpad on the master track of some projects.

MNotepad.jpg

[mevla]

And so, this I can do.

1 oscillator, no FX no nothing else. A bit of a jagged edge on a sine to make it more distinct.

MSFADSR2.jpg

With this envelope.
Attack 550ms
Decay 15ms
Sustain all the way down.
Release 1026ms

MSFADSR3.jpg

And most importantly these notes. BPM 110, 1/4 note 545ms

http://www.sengpielaudio.com/calculator ... potime.htm

MSFADSR4.jpg

[mevla]

So it seems that the sustain must be not at the minimum, but at the maximum.

What is the plateau seen at the top with the decay and hold being at 0 ?

MSFADSR5.jpg

[jmg8]

So you worked it out?
I have just tested and the behaviour I was getting was completely expected. If you hold a note long enough for the env to decay to 0, then the release will have no effect. Right.
But I guess your last post means you figured it out already.

[jmg8]

So it seems that the sustain must be not at the minimum, but at the maximum.

What is the plateau seen at the top with the decay and hold being at 0 ?




MSFADSR5.jpg

Shouldn't be there logically, but its not really a big deal.

[mevla]

So you worked it out?
I have just tested and the behaviour I was getting was completely expected. If you hold a note long enough for the env to decay to 0, then the release will have no effect. Right.
But I guess your last post means you figured it out already.

At least it's possible to say that what you see is what you get. It drops at 0 then there's nothing for the the release cycle to pick up.

The point though is that in absence of any sustain, a bridge is made between the decay cycle and the release cycle, which for instance ZebraHZ makes. Put the sustain at 0 and you'll get for those same notes what you'd think you'd be getting: the release cycle run for each and every note. Because there's no sustain anymore. And it's not replaced by silence.

With MSF (and most probably MPS) it's the contrary. One has to put the sustain not all the way down but rather all the way up at 0 db to achieve the same result.

I reckon though that having an extra option of a silence zone in lieu of sustain could be an additional creative feature. So it's all good - one just have to know to inverse the principle most commonly found elsewhere.

Cheers and thanks for your feedback.

[mevla]

So it seems that the sustain must be not at the minimum, but at the maximum.

What is the plateau seen at the top with the decay and hold being at 0 ?

MSFADSR5.jpg

Shouldn't be there logically, but its not really a big deal.

Well it's actually something. Since to achieve envelopes that other synths are producing the sustain in MSF has to be at top blast (0 db) instead of not there at all (and not replaced by silence) this plateau appears and is actually in the path of the sound.

The envelope will stop at the beginning of this plateau for as long as the MIDI keyboard key is pressed, and upon release will continue on with the release cycle.

So it does not feel natural regarding what's expected, eg. a non-existent sustain, with the decay cycle immediately flowing into the release cycle.

Imagine a vibraphone. Or an acoustic piano. One presses a key and the decay goes into release all by itself, while the key is still pressed. This is natural. And is replicated by many synths out there.

Hive2 will just behave like that. A key is hit, the sustain is at zero, so it will naturally flow from the decay cycle to the release cycle all by itself, as a user would expect from physical experience (acoustic piano, acoustic guitar, marimba, drums, etc). The screen shot was taken while still holding the MIDI keyboard key and the sound went from decay straight to release.

hiveadsr.jpg

However with MSF the envelope, with an equivalent sustain of "0" (1) will squarely stop until the MIDI keyboard key is released. The screen shot was taken while still holding the MIDI keyboard key and the sound is stuck there at that place shown by the line, waiting for the MIDI key release to go on:

msfadsr6.jpg

(1) What is the alternative to get rid of the sustain cycle ? All the way down to silence is not a solution.

[Gamma-UT]

Hive2 will just behave like that. A key is hit, the sustain is at zero, so it will naturally flow from the decay cycle to the release cycle all by itself, as a user would expect from physical experience (acoustic piano, acoustic guitar, marimba, drums, etc). The screen shot was taken while still holding the MIDI keyboard key and the sound went from decay straight to release.


hiveadsr.jpg



...
All the way down to silence is not a solution.

Surely if you hold a key down and wait for the decay to zero, the envelope never really moves to the release cycle (except to play further silence)? There is only decay in your example, which fades to zero as you'd expect. There is no release that can be heard.

Kontakt has the same behaviour as Melda here and it's as I would expect. I chose Kontakt because it shows you when voices are being processed. It continues to process the sound (albeit silent) while the key is held down even when sustain is zero and the note has gone through the full decay.

I imagine Hive works the same way: it's just not showing you on the GUI, although it's possible that the synth stops processing as there's nothing left to do. But that would mean the same thing: there's no audible processing of the release phase.