How to apply an exponential fade-out in Cubase or Reaper?
-
TheAdventuringPanda TheAdventuringPanda https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=353719
- Banned
- 46 posts since 15 Mar, 2015
Is it possible to apply an exponential fade-out in Cubase or Reaper? Searching both manuals for the word "exponential" turns up nothing.
Reaper has options for fade-out curve-shapes, and a couple of the options look concave like an exponential curve would look, but not all concave curves are exponential, and the manual doesn't clarify whether any of these options are exponential. I tried asking on Reaper forum, but I couldn't pass the are-you-a-human test to register on their forum.
Cubase lets me design fade-out shapes with splines, but it's not mathematically possible to construct an exponential curve out of spline-curve-segments. In another thread, someone reports being able to apply an exponential fade-out in Cubase, but he hasn't provided a description of his method that enables me to reproduce it.
Reaper has options for fade-out curve-shapes, and a couple of the options look concave like an exponential curve would look, but not all concave curves are exponential, and the manual doesn't clarify whether any of these options are exponential. I tried asking on Reaper forum, but I couldn't pass the are-you-a-human test to register on their forum.
Cubase lets me design fade-out shapes with splines, but it's not mathematically possible to construct an exponential curve out of spline-curve-segments. In another thread, someone reports being able to apply an exponential fade-out in Cubase, but he hasn't provided a description of his method that enables me to reproduce it.
-
TheAdventuringPanda TheAdventuringPanda https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=353719
- Banned
- Topic Starter
- 46 posts since 15 Mar, 2015
The Cubase fade-editor (Cubase manual p. 250) provides no way to apply an exponential fade. It provides ways to apply fades, but not fades which are exponential. The manual says:SJ_Digriz wrote:select the event, drag the fade handle, double click on the fade to bring up the fade editor.
"Curve Kind. These buttons determine whether the fade curve consists of spline curve segments (left button), damped spline segments (middle button), or linear segments (right button)."
It's not mathematically possible to construct an exponential curve out of spline-curve-segments or linear-segments. The word "exponential" has a definite mathematical definition, and it is key here.
-->Kullervo wrote:I replied in the other thread already
Thanks for this info. The options in Cubase, though, are just spline-curve-segments and linear-segments -- according to the manual and everything I find while operating the program -- and it is mathematically impossible to construct an exponential fade using spline-curve-segments and linear-segments.Kullervo wrote:'sample editor', that's what I meant, sorry.
You can choose the curves, which are not named in Cubase, but are in WL :
(here they are in WL, the menu is different in Cubase, but I'm on my laptop now)
-
- KVRist
- 279 posts since 13 May, 2003
but you do get to see these curves, right ?TheAdventuringPanda wrote: Thanks for this info. The options in Cubase, though, are just spline-curve-segments and linear-segments -- according to the manual and everything I find while operating the program -- and it is mathematically impossible to construct an exponential fade using spline-curve-segments and linear-segments.

I just took them to be the same as in WL (exp+,exp, log, ...) even if they are approximations..?
I think you'll have to take it to the Steinberg forum if you want more details on their math
- KVRAF
- 6097 posts since 5 Jul, 2001 from Just about .... there
Typically when people ask about exponential curves for audio editing they just mean a curve that is weighted early or log weighted late. So pedantically, no they don't have "exponential" curves. But, neither does any other audio program, because you can't choose which exponents you are using. You simply get a weighted curve (usually not sample accurate so its pointless anyways).
And yes, spline curves can fully represent an exponential curve to sample accuracy and far beyond that. Just not the ones in cubase because you can't get enough resolution points. Of course that will be absolutely crucial to absolutely no one.
So if we are done with our pointless pedantic douchebaggery .. have a nice day.
And yes, spline curves can fully represent an exponential curve to sample accuracy and far beyond that. Just not the ones in cubase because you can't get enough resolution points. Of course that will be absolutely crucial to absolutely no one.
So if we are done with our pointless pedantic douchebaggery .. have a nice day.
If you have to ask, you can't afford the answer
-
TheAdventuringPanda TheAdventuringPanda https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=353719
- Banned
- Topic Starter
- 46 posts since 15 Mar, 2015
Yes, it looks similar in Cubase:Kullervo wrote:but you do get to see these curves, right ?TheAdventuringPanda wrote: Thanks for this info. The options in Cubase, though, are just spline-curve-segments and linear-segments -- according to the manual and everything I find while operating the program -- and it is mathematically impossible to construct an exponential fade using spline-curve-segments and linear-segments.
Seeing your new image, showing 4 segments connected by three points, it does look like WL and Cubase are using the same math -- both are using splines (finite-order polynomials) to approximate exponential curves (infinite-order polynomials).Kullervo wrote:I just took them to be the same as in WL (exp+,exp, log, ...) even if they are approximations..?
The Cubase manual clarifies the math they're using: splines. Steinberg confuses things by labeling one of their curves "Exponential", but now that I see your WL image showing 4 segments, it's clear Steinberg was speaking falsely. Like a car salesman selling you a car with "Porsche" written on the side, but the car is really a Hyundai.Kullervo wrote:I think you'll have to take it to the Steinberg forum if you want more details on their math
Oh really. Good to know what "people" really mean when they say exponential, and that they don't really mean exponential. Thanks for your insight into "people".SJ_Digriz wrote:Typically when people ask about exponential curves for audio editing they just mean a curve that is weighted early or log weighted late.
But really I'm using no fewer than 3 audio programs which do let me choose which exponents I'm using.SJ_Digriz wrote:So pedantically, no they don't have "exponential" curves. But, neither does any other audio program, because you can't choose which exponents you are using.
Really though, I'm using 3 audio programs which do give me sample-accurate exponential fades.SJ_Digriz wrote:You simply get a weighted curve (usually not sample accurate so its pointless anyways).
In other words, it isn't absolutely crucial to you, for whatever you're doing. You being one person, despite your attempts to speak for everyone. You being totally wrong about what I'm doing with my own copy of Cubase, despite your pretending to have the slightest idea.SJ_Digriz wrote:... will be absolutely crucial to absolutely no one.
So butt-hurt. Over a question about a computer program. Get a life, so your DAW can be just a computer program to you, and not whatever crutch you're currently using it as?SJ_Digriz wrote:So if we are done with our pointless pedantic douchebaggery .. have a nice day.
- KVRAF
- 6097 posts since 5 Jul, 2001 from Just about .... there
[quote="TheAdventuringPanda"]
But really I'm using no fewer than 3 audio programs which do let me choose which exponents I'm using.
and so when you apply them they do nothing more than what cubase offers. But have fun anyhow.
But really I'm using no fewer than 3 audio programs which do let me choose which exponents I'm using.
and so when you apply them they do nothing more than what cubase offers. But have fun anyhow.
If you have to ask, you can't afford the answer
-
- KVRist
- 279 posts since 13 May, 2003
just to clarify: my second image is from cubase, not WL.TheAdventuringPanda wrote: Seeing your new image, showing 4 segments connected by three points, it does look like WL and Cubase are using the same math -- both are using splines (finite-order polynomials) to approximate exponential curves (infinite-order polynomials).
-
TheAdventuringPanda TheAdventuringPanda https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=353719
- Banned
- Topic Starter
- 46 posts since 15 Mar, 2015
whatever you say, butt-hurt Cubase fanboy in Alabama. (Readers outside the U.S., can read up on Alabama for context into this individual's dysfunction.)SJ_Digriz wrote:But have fun anyhow.
-
TheAdventuringPanda TheAdventuringPanda https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=353719
- Banned
- Topic Starter
- 46 posts since 15 Mar, 2015
oh, so maybe WL is doing real exponentials after all. Thanks for your help on this.Kullervo wrote:just to clarify: my second image is from cubase, not WL.TheAdventuringPanda wrote: Seeing your new image, showing 4 segments connected by three points, it does look like WL and Cubase are using the same math -- both are using splines (finite-order polynomials) to approximate exponential curves (infinite-order polynomials).
- KVRAF
- 6097 posts since 5 Jul, 2001 from Just about .... there
nope sorry, WL interpolates that curve to the automation resolution. That picture you see is just a picture. Let me ask you what ratio on the taper means exponential to you? There are an infinite number of curves that can be defined as exponential mathematically.TheAdventuringPanda wrote:oh, so maybe WL is doing real exponentials after all. Thanks for your help on this.Kullervo wrote:just to clarify: my second image is from cubase, not WL.TheAdventuringPanda wrote: Seeing your new image, showing 4 segments connected by three points, it does look like WL and Cubase are using the same math -- both are using splines (finite-order polynomials) to approximate exponential curves (infinite-order polynomials).
If you have to ask, you can't afford the answer
-
- Banned
- 12367 posts since 30 Apr, 2002 from i might peeramid
ftr, sj is right, but his criteria are above your scope of interest. don't bash him, five years from now you'll know what he was talking about, and realise he is trying to inform you 
btw you're all all of you buggered up
your slider is already logarithmic because decibels, so a linear automation contour would produce an "exponential" amplitude decay. hehehehe. you can hate me, i'm right.

developers are the only people who will ever hear what an actual linear decay sounds like
"whump - whar'd my sound go??" (oh edit that's not true.. several wav editors include a linear fade.. bloody useless, amplitude is log)
btw you're all all of you buggered up
developers are the only people who will ever hear what an actual linear decay sounds like
"whump - whar'd my sound go??" (oh edit that's not true.. several wav editors include a linear fade.. bloody useless, amplitude is log)
you come and go, you come and go. amitabha neither a follower nor a leader be tagore "where roads are made i lose my way" where there is certainty, consideration is absent.
- KVRAF
- 6097 posts since 5 Jul, 2001 from Just about .... there
lol, if they didn't understand my statement ... this one is going to soar over heads at 30k ft.xoxos wrote:btw you're all all of you buggered upyour slider is already logarithmic because decibels
If you have to ask, you can't afford the answer
