Envelope VS Automation (Now with CV, MIDI ...and drama)

This forum is for discussing Reason. Questions, answers, ideas, and opinions... all apply.
RobC
Posts: 1832
Joined: 10 Mar 2018

14 Mar 2018

Okay, so I'm pretty sure that if I use an envelope to modulate a filter for example, it will be much more detailed than when automating the filter frequency in the sequencer. Cause technically, most go from 0 to 127, while cv/internal routing seems to be far more than that. However, the reason why I would prefer automating in the sequencer is that... well, you aren't limited to just ADSRetc. ~ though it certainly could use some vector curving.

I hope the Easter Bunny considers bringing us a RE with CV generator knobs (and buttons), that have maximum detail, not just 0-127. (Or is there already one?) I mean, if the mixer can go to 0-1000, then the CV signal and internal routing must be super detailed, or no?

What does it matter you may ask: if you create very dynamic sounds and want to draw different curves for each note, then no envelope is sufficient; and 0-127 is not enough for snappy transients.
Last edited by RobC on 19 Mar 2018, edited 1 time in total.

User avatar
Carly(Poohbear)
Competition Winner
Posts: 2871
Joined: 25 Jan 2015
Location: UK

14 Mar 2018

Totally agree, I tend to start with modulation then I like to move it to automation as you can have more control over when it's applied however as you stated you then lose the detail (in a lot of my cases that's not so much an issue).

Some devices do have a greater range than the knobs say, prime example would be the C8x4, on the front the values go from 0 to 127 however if you monitor what is coming out of the back it's a different case, nothing close to 0-1000 more like 0-512 range.

Now if you want real fine control then make use the pitch bend wheel it has a range of -8191 to 8191, e.g. Use Thor, draw your automation in the pitch bend, route that to a CV out, control the device you want with it...



PoohBear

-------------------------------------------------------------------------------------------------------
Recent Music made with Reason
(Electronic) Poly Madness
(Electronic) Solar Dance
(Electronic) Into The Night
(Upbeat) The Players Took Me There Made in Reason 9 with heavy use of the players
(Electronic) 5 Steps (Video added) Updated
------------------------------------------------------------------------------------------
My Soundcloud Page ....... Nektar Mappings
------------------------------------------------------------------------------------------

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

14 Mar 2018

Automation is moving knobs, so it won't matter how much resolution you add if the knob has 127 steps. Some controls go well beyond 128 steps, btw, so in those cases you're hearing up to 32 bit resolution.

What examples have you heard that sound 'stepped'?
Selig Audio, LLC

User avatar
O1B
Posts: 2037
Joined: 26 Jan 2015

14 Mar 2018

0 to 127 or -8191 to 8191 makes no difference if the knob your using doesn't have the depth
and using shift with pitch wheel moves it 70 clicks... without, 240... not 1 by 1.

Envelope vs Automation?

Sync? Free?
CV?
and, then there's 'what Controller do I' use? Encoder, Fader, XY, what resolution?....
I think it boils down to how that auto/envel. moved. Use both.

How about a power ring...
Image

and CV..?
Image
Carly(Poohbear) wrote:
14 Mar 2018
Totally agree, I tend to start with modulation then I like to move it to automation as you can have more control over when it's applied however as you stated you then lose the detail (in a lot of my cases that's not so much an issue).

Some devices do have a greater range than the knobs say, prime example would be the C8x4, on the front the values go from 0 to 127 however if you monitor what is coming out of the back it's a different case, nothing close to 0-1000 more like 0-512 range.

Now if you want real fine control then make use the pitch bend wheel it has a range of -8191 to 8191, e.g. Use Thor, draw your automation in the pitch bend, route that to a CV out, control the device you want with it...

PoohBear

RobC
Posts: 1832
Joined: 10 Mar 2018

14 Mar 2018

I'm honestly confused how some devices' values work - while testing in the sequencer some show values up to 1000, yet 1 step starts at 4, so not like 1,2,3,4 etc.
Now with the pitch bend, I'm surprised - I'm really getting confused by when even moving it with the mouse, it usually jumps values, yet I didn't remember, that in the sequencer it in fact does go up and down with 1 step at a time, so that's detailed true to the numbers. This will come in handy in the meantime, thank you very much!

That is true, but when I use a CV generator knob, like the one mentioned, I would route its cv out to the desired device's cv input. Then the range would matter.
Sadly, not every knob has super fine resolution.

Well, the biggest problem was with dynamics for instance. I just couldn't get the same transients like with an envelope. It sounded a bit weak. Things get really audible when you turn the audio level up and start off with setting the decay from zero (silence).
But if you try modulating pitch with a simple CV generator knob, that's audibly gonna jump (side example).
Sometimes there are clicks that an envelope generator can create, but the 127 range can't.

One of the Rack Extensions show there are (or could be) values like +/- 0.00001 - talk about detail!

So yeah, the pitch wheel on Thor seems like a temporary solution, but a simple RE would be more efficient.

User avatar
Carly(Poohbear)
Competition Winner
Posts: 2871
Joined: 25 Jan 2015
Location: UK

14 Mar 2018

O1B wrote:
14 Mar 2018
0 to 127 or -8191 to 8191 makes no difference if the knob your using doesn't have the depth
and using shift with pitch wheel moves it 70 clicks... without, 240... not 1 by 1.
But with automation you can move it one by one and that is what I'm saying and then you use it like any CV signal.

Also when moving the pitch wheel with shift you can go down to 24 (using the very precise setting)..

RobC
Posts: 1832
Joined: 10 Mar 2018

14 Mar 2018

I enjoy taking time fine tuning things by the way and doing a lot of post work. Envelopes sure work in live situations better, but at the sacrifice of that, automating a super fine CV generator knob that modulates things via CV, opens much more possibilities at the same resolution. More of a sound design, or electronic music thing.

Edit: haha, lame pun: "post work. Envelopes"
Last edited by RobC on 14 Mar 2018, edited 1 time in total.

EdGrip
Posts: 2343
Joined: 03 Jun 2016

14 Mar 2018

I'd like to understand this subject better too.
Though I can't think of examples, I'm sure some devices have 128-click knobs which can be audibly steppy when you move them with the mouse, but if you modulate that same parameter internally, or via CV, the resolution is greater. So why not make that resolution apparent at the knob?
Steppy knobs on soft synths, man. Nobody needs that. Filter cutoff at high resonance.

So if more resolution is available internally, can automation bypass the knob to get to it? I've never tested this.

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

14 Mar 2018

RobC wrote:
14 Mar 2018
I'm honestly confused how some devices' values work - while testing in the sequencer some show values up to 1000, yet 1 step starts at 4, so not like 1,2,3,4 etc.
Now with the pitch bend, I'm surprised - I'm really getting confused by when even moving it with the mouse, it usually jumps values, yet I didn't remember, that in the sequencer it in fact does go up and down with 1 step at a time, so that's detailed true to the numbers. This will come in handy in the meantime, thank you very much!

That is true, but when I use a CV generator knob, like the one mentioned, I would route its cv out to the desired device's cv input. Then the range would matter.
Sadly, not every knob has super fine resolution.

Well, the biggest problem was with dynamics for instance. I just couldn't get the same transients like with an envelope. It sounded a bit weak. Things get really audible when you turn the audio level up and start off with setting the decay from zero (silence).
But if you try modulating pitch with a simple CV generator knob, that's audibly gonna jump (side example).
Sometimes there are clicks that an envelope generator can create, but the 127 range can't.

One of the Rack Extensions show there are (or could be) values like +/- 0.00001 - talk about detail!

So yeah, the pitch wheel on Thor seems like a temporary solution, but a simple RE would be more efficient.
What you are describing, e.g. transients, have nothing to do with the value resolution and everything to do with time resolution, while modulating a filter or pitch are involving the value resolution - totally different things!

But let's compare using automation to control gain vs an envelope via CV. I'm using the Global envelope in Thor, controlling Selig Gain, vs drawing automation, both with an "instant" (0 ms) attack time. Both show a little smoothing on the attack, both taking about 2ms to reach full level. This example shows no difference in attack times using automation vs CV.
The attached screenshot shows about 15 ms total time:
Screen Shot 2018-03-14 at 2.51.17 PM.png
Screen Shot 2018-03-14 at 2.51.17 PM.png (43.27 KiB) Viewed 2233 times
Selig Audio, LLC

User avatar
O1B
Posts: 2037
Joined: 26 Jan 2015

14 Mar 2018

Check!

I did catch that. But 'moving' it ('drawing' automation) vs 'playing' it will probably lead to drastically different results. But, I'm probably on a tangent now...

Easy day/eve out there.
Carly(Poohbear) wrote:
14 Mar 2018
O1B wrote:
14 Mar 2018
0 to 127 or -8191 to 8191 makes no difference if the knob your using doesn't have the depth
and using shift with pitch wheel moves it 70 clicks... without, 240... not 1 by 1.
But with automation you can move it one by one and that is what I'm saying and then you use it like any CV signal.

Also when moving the pitch wheel with shift you can go down to 24 (using the very precise setting)..

RobC
Posts: 1832
Joined: 10 Mar 2018

14 Mar 2018

I didn't find any possibility to bypass a knob's resolution during automation. Though who knows, gonna (try) reading all the 1300 pages of Reason 10's Operation Manual soonish.

Alright, now it makes more sense. I haven't perfectly experimented automating dynamics though! Hence the example was kind of half-assed. (Not to mention that I'm very sleepy and barely with it, so bare with me.)

But yeah, better resolution has its use in the right places. 0-127 is rather good for setting things up on board. Not saying that it doesn't work for some things, but numbers like 0.00001 and 0-127 did make me worry.

That's what forums are for though, to discuss things, help each other, and clear up misunderstandings, and throw in ideas.

User avatar
Carly(Poohbear)
Competition Winner
Posts: 2871
Joined: 25 Jan 2015
Location: UK

14 Mar 2018

O1B wrote:
14 Mar 2018
Check!

I did catch that. But 'moving' it ('drawing' automation) vs 'playing' it will probably lead to drastically different results. But, I'm probably on a tangent now...

Easy day/eve out there.
Carly(Poohbear) wrote:
14 Mar 2018


But with automation you can move it one by one and that is what I'm saying and then you use it like any CV signal.

Also when moving the pitch wheel with shift you can go down to 24 (using the very precise setting)..
arh, are you saying a knob on your controller ?

Well a lot of that comes down the to the mappings\codec as well, I know the Nektar does send out delta's so making the scale larger should work for controlling something like the bend wheel and down to 1 to 1.

User avatar
Carly(Poohbear)
Competition Winner
Posts: 2871
Joined: 25 Jan 2015
Location: UK

14 Mar 2018

RobC wrote:
14 Mar 2018
I didn't find any possibility to bypass a knob's resolution during automation. Though who knows, gonna (try) reading all the 1300 pages of Reason 10's Operation Manual soonish.

Alright, now it makes more sense. I haven't perfectly experimented automating dynamics though! Hence the example was kind of half-assed. (Not to mention that I'm very sleepy and barely with it, so bare with me.)

But yeah, better resolution has its use in the right places. 0-127 is rather good for setting things up on board. Not saying that it doesn't work for some things, but numbers like 0.00001 and 0-127 did make me worry.

That's what forums are for though, to discuss things, help each other, and clear up misunderstandings, and throw in ideas.
I don't think you can bypass a knobs resolution however some have more precision than people realise.

Don't forget to play about with the option: Preferences (General Tab) Mouse Knob range.. i.e. very precise.

In the case of the C8x4 you have finer control with a mouse than you do with automation.

With the bend wheel with very precise you can go down to 24 (clicks\step values) where as in automation it's 1 to 1..

RobC
Posts: 1832
Joined: 10 Mar 2018

15 Mar 2018

Hold on, I was thinking about building a MIDI controller. Regarding the deltas, would that mean that Reason could handle very fine MIDI messages? Then again, it would be a catastrophe with an encoder, I think. I mean, if it would send out 0.00001 with one click - figures what time it would take to jump just one value on a 0-127 knob.

I think I have OCD, haha - so very precise is always on for me, and I even press shift if I use a mouse when I want even more precision. ...and I want more.

Didn't realize that, it should be the same with automation on the 8x4...

RobC
Posts: 1832
Joined: 10 Mar 2018

15 Mar 2018

I must add though, that no matter what a better resolution never hurts, be it pitch, audio level, tone, waveform, etc.

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

15 Mar 2018

RobC wrote:I must add though, that no matter what a better resolution never hurts, be it pitch, audio level, tone, waveform, etc.
If there were such a thing as infinite bandwidth, this would be true.


Sent from some crappy device using Tapatalk
Selig Audio, LLC

RobC
Posts: 1832
Joined: 10 Mar 2018

15 Mar 2018

Just the resolution of Reason's internal values would be enough.

Say, as you said earlier, that some audio weirdness couldn't happen just because of 127 range only:
why did it happen that my kick drum (where I separated the 20-40 Hz range to a wav file with a FIR filter, and square window, very high resolution - which sounded normal ~ ignore the FIR ringing now) made some crackling sounds even with an envelope I think as the audio levels went down to silence. I remember I just tested simple decaying dynamics on it - though the same happened with Reason's IIR Mclass EQ when I just gently swept the sub bass region.

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

15 Mar 2018

RobC wrote:Just the resolution of Reason's internal values would be enough.

Say, as you said earlier, that some audio weirdness couldn't happen just because of 127 range only:
why did it happen that my kick drum (where I separated the 20-40 Hz range to a wav file with a FIR filter, and square window, very high resolution - which sounded normal ~ ignore the FIR ringing now) made some crackling sounds even with an envelope I think as the audio levels went down to silence. I remember I just tested simple decaying dynamics on it - though the same happened with Reason's IIR Mclass EQ when I just gently swept the sub bass region.
Internal values are all 32 bit float in Reason, unless the developer internally limits them (as in the case of a semitone or octave control of pitch).




Sent from some crappy device using Tapatalk
Selig Audio, LLC

RobC
Posts: 1832
Joined: 10 Mar 2018

16 Mar 2018

Now I have no clue why 32bf knob resolution would be that much of a drain, but I'll leave that to developers now.
I mean, I'm interested in what's under the hood, close to the core, but I prefer to focus on the creative part.

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

16 Mar 2018

RobC wrote:Now I have no clue why 32bf knob resolution would be that much of a drain, but I'll leave that to developers now.
I mean, I'm interested in what's under the hood, close to the core, but I prefer to focus on the creative part.
Not sure what you’re referring to - how much of a drain are you talking about, and how does this release to the topic of resolution?

I assumed you were implying the internal resolution was not enough for your work, and I was pointing out it’s 32 bit float - which is pretty extreme.

I wonder if what you are experiencing has nothing to do with resolution?

Maybe post specific examples of what you’re talking about so we can be on the same page - could turn out there’s a simple fix for what you’re experiencing!

Sent from some crappy device using Tapatalk
Selig Audio, LLC

RobC
Posts: 1832
Joined: 10 Mar 2018

16 Mar 2018

The internal resolution is perfectly enough.
The problem is when we don't just tweak on-board knobs, but start modulating them. The knobs themselves. Be it with automation, or via the combinator's CV in. As you know, some jump huge steps with rotation, compared to the internal values. It's nowhere as smooth as when for example an LFO modulates something internally or through cv that isn't an on-board knob.

The best example to smoothness is if you connect a CV8X4 CV Generator to a synthesizer's pitch cv, like Subtractor, play a static note, and if you hold the shift button and start rotating the knob on the CV Generator, the pitch audibly jumps. So it doesn't change smoothly.

User avatar
Carly(Poohbear)
Competition Winner
Posts: 2871
Joined: 25 Jan 2015
Location: UK

18 Mar 2018

RobC wrote:
16 Mar 2018
The internal resolution is perfectly enough.
The problem is when we don't just tweak on-board knobs, but start modulating them. The knobs themselves. Be it with automation, or via the combinator's CV in. As you know, some jump huge steps with rotation, compared to the internal values. It's nowhere as smooth as when for example an LFO modulates something internally or through cv that isn't an on-board knob.

The best example to smoothness is if you connect a CV8X4 CV Generator to a synthesizer's pitch cv, like Subtractor, play a static note, and if you hold the shift button and start rotating the knob on the CV Generator, the pitch audibly jumps. So it doesn't change smoothly.
I thought I would do a quick test....

I was able to change the midicodec to the range from 0 to 99999 which means I can now turn a knob on my control and move the C8x4 knob up by 0.00001 at a time, got excited by it but then.............................. it stopped when I hit 0.00127 DOH!!!! back to the drawing board.....

------------------------------------------------------------------------
NOTE this thread would have been great in a CV\IO section.....

EdGrip
Posts: 2343
Joined: 03 Jun 2016

18 Mar 2018

If we're involving physical MIDI controllers, their encoders probably only have 128 steps. (I know the Alesis ION is one exception, I don't know if there are others)

User avatar
Carly(Poohbear)
Competition Winner
Posts: 2871
Joined: 25 Jan 2015
Location: UK

18 Mar 2018

EdGrip wrote:
18 Mar 2018
If we're involving physical MIDI controllers, their encoders probably only have 128 steps. (I know the Alesis ION is one exception, I don't know if there are others)
it was the Nektar P4, it's my understanding they can send out delta's, I will have to play about with it a bit more :)

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

19 Mar 2018

Carly(Poohbear) wrote:------------------
NOTE this thread would have been great in a CV\IO section.....
Was just thinking would’ve been perfect for an Automation section…
;)


Sent from some crappy device using Tapatalk
Selig Audio, LLC

Post Reply
  • Information
  • Who is online

    Users browsing this forum: xbitz and 28 guests