Translate Europa/Grain envelope to standard ADSR values?

[RandomSkratch]

Had a look through the manual and did not see anything written regarding this. Is there any way to create an envelope the same way as a standard ADSR envelope is made? I'm not talking shape but rather values. If I want a 50ms attack, is that even possible? The total envelope length knob isn't helpful (or I don't understand it).

Or is this a strictly "by your ears" setup? Which is only good if your ears are good...

[selig]

Had a look through the manual and did not see anything written regarding this. Is there any way to create an envelope the same way as a standard ADSR envelope is made? I'm not talking shape but rather values. If I want a 50ms attack, is that even possible? The total envelope length knob isn't helpful (or I don't understand it).

Or is this a strictly "by your ears" setup? Which is only good if your ears are good...

The envelope’s total time is represented across the display, and by dividing the display you can approximate the time of each stage.
For example, if the overall envelope is 1 second, and the attack is 1/4 of the length of the display, it’s 250ms long. If it’s 1/8 it’s 125ms, etc.

But the envelope curves change the way we perceive the stage times and that’s difficult to reproduce in other envelopes. Once the envelope curve moves away from linear, it can speed up or slow down the perceived length of each stage making the actual time less relevant.




Sent from some crappy device using Tapatalk

[RandomSkratch]

The envelope’s total time is represented across the display, and by dividing the display you can approximate the time of each stage.
For example, if the overall envelope is 1 second, and the attack is 1/4 of the length of the display, it’s 250ms long. If it’s 1/8 it’s 125ms, etc.

But the envelope curves change the way we perceive the stage times and that’s difficult to reproduce in other envelopes. Once the envelope curve moves away from linear, it can speed up or slow down the perceived length of each stage making the actual time less relevant.

Sent from some crappy device using Tapatalk

Thanks Selig, that's what I suspected. Approximating each time can be a little annoying - not sure why they didn't include snap values (looking at you Synchronous!). At least that would not require a bunch of math.

I get it how the curve can adjust the perceived time but there is still going to be an absolute calculable time which should be indicated in a tool tip somewhere. EG. The first node is X MS from the start, the next node is Y MS from the previous node. If the entire time is set, I don't see why these can't be calculated.

[selig]

I get it how the curve can adjust the perceived time but there is still going to be an absolute calculable time which should be indicated in a tool tip somewhere. EG. The first node is X MS from the start, the next node is Y MS from the previous node. If the entire time is set, I don't see why these can't be calculated.

Curious how you would use these times if calculated for you?


Sent from some crappy device using Tapatalk

[RandomSkratch]

Curious how you would use these times if calculated for you?


Sent from some crappy device using Tapatalk

Then I could see that my attack time is 50ms or release is 1 sec, or the time between these two spots is 1/4 bar. If you want to create rhythmic modulation, why would you omit snapping?

Absolute values are very helpful when learning (for me they are). So if I'm following along with a tutorial or if I'm trying to recreate a patch from another synth and I know the ADSR values, I would be able to set them in Europa to see if the synth could duplicate the sound. Without having the experience of patch design or understanding everything there is about synthesis, doing things blindly can be challenging. At least when you compare numbers you can see if you're on the right path.

[selig]

Curious how you would use these times if calculated for you?


Sent from some crappy device using Tapatalk

Then I could see that my attack time is 50ms or release is 1 sec, or the time between these two spots is 1/4 bar. If you want to create rhythmic modulation, why would you omit snapping?

Absolute values are very helpful when learning (for me they are). So if I'm following along with a tutorial or if I'm trying to recreate a patch from another synth and I know the ADSR values, I would be able to set them in Europa to see if the synth could duplicate the sound. Without having the experience of patch design or understanding everything there is about synthesis, doing things blindly can be challenging. At least when you compare numbers you can see if you're on the right path.

I see two things (separate but related) you are saying here: the need to see absolute values to translate from other synths (which won’t work for the curves, btw, which can be an essential part of a patch recreation), and the need to create rhythmic “snap to grid” envelopes, which is a great and common suggestion (and I’m really not sure why they left them off).

For the second, we’ll just have to wait. But for the first, understanding the curve shape is essential.

For example, you’ll never match Subtractor and Thor’s envelops because Subtractor is exponential and Thor is linear. Once I discovered that, I found you can make Thor’s envelopes match Subtractor by multiplying them by themselves, but you can’t make Subtractor’s envelops linear.

How I checked this, and something you may want to try, was to use a simple waveform in both synths (like sine), and record the envelopes to an audio track playing a single note in the middle of the frequency spectrum. You can see the curve SHAPE is different, which affects our perception of curve TIME. If it’s not clear enough, take a screen shot and import into a simple drawing app and zoom the vertical axis to see the difference.

As a rule, more exponential envelopes will sound more percussive or “snappy”, which is why some synth’s envelopes can sound very different even when you set them to the same exact values.


Sent from some crappy device using Tapatalk

[RandomSkratch]

I see two things (separate but related) you are saying here: the need to see absolute values to translate from other synths (which won’t work for the curves, btw, which can be an essential part of a patch recreation), and the need to create rhythmic “snap to grid” envelopes, which is a great and common suggestion (and I’m really not sure why they left them off).

For the second, we’ll just have to wait. But for the first, understanding the curve shape is essential.

For example, you’ll never match Subtractor and Thor’s envelops because Subtractor is exponential and Thor is linear. Once I discovered that, I found you can make Thor’s envelopes match Subtractor by multiplying them by themselves, but you can’t make Subtractor’s envelops linear.

How I checked this, and something you may want to try, was to use a simple waveform in both synths (like sine), and record the envelopes to an audio track playing a single note in the middle of the frequency spectrum. You can see the curve SHAPE is different, which affects our perception of curve TIME. If it’s not clear enough, take a screen shot and import into a simple drawing app and zoom the vertical axis to see the difference.

As a rule, more exponential envelopes will sound more percussive or “snappy”, which is why some synth’s envelopes can sound very different even when you set them to the same exact values.


Sent from some crappy device using Tapatalk

Yes sorry it ended up being two separate things. I know once you start implementing curves the numbers fall apart but for an initial setup it's beneficial (getting into the ballpark).

Something you said about Subtractor and Thor made me realize (and with the recent SAW-1 RE). Sub and SAW have 0-127 values and Thor shows actual MS. I always wondered why Sub didn't provide time and now it makes sense - because of the exponential aspect of the slider. With SAW it has a curve feature which modifies the value. But what I still don't know is why can't they also show a MS value - it must have one, no? Or is it incalculable? Or would it be useless because of the perception of time being shortened or lengthened? I believed you referenced this but 1 sec linear sounds different than 1 sec with a steep curve.

So I guess the answer to my question is essentially no, you can't translate the values.