Re: Confusion about Fading Amplification

On Fri, Aug 5, 2011 at 1:34 AM, Joel Roth <[email protected]> wrote:
> On Thu, Aug 04, 2011 at 09:30:32PM -0400, Mark Armendariz wrote:
>> * Note: I assume attachments are not kosher on this list, so I posted
>> this question on StackOverflow as so you can see my resulting
>> waveforms:
>> http://stackoverflow.com/questions/6950482/how-do-multiple-amplitude-fades-work-on-ecasound
>>
>> -----
>>
>> I want to fade a track in and out as specific time codes.  For
>> example, I would like to take an audio file, and:
>>
>> Start it at 100% Volume,
>> Fade it to 20%  at 2 seconds
>> Fade it to 100% at 4 seconds
>> Fade it to 20%  at 6 seconds
>> Fade it to 100% at 8 seconds
>> Fade it to 20%  at 10 seconds
>> Fade it to 100% at 12 seconds
>> Fade it to 0 at 14 seconds.
>
>
> I see you've been using multiple -kl2 and -ea operators.
>
> While that may be possible, you may find it simpler
> to create a single envelope using -klg. For the above example (not
> tested) you could try this:
>
> ecasound -a:1 -i:tone.wav -ea:100 \
> -klg:1,0,100,7,1,2,0.20,4,1,6,0.20,8,1,10,0.20,12,1,14,0 \
> -o:test.mp3
>
> The arguments for -klg are:
>
> fx-parameter, low value, high value, count (no. of pairs),
> pos1, value1, pos2, value2,...
>
> The values are specified between 0 and 1.
>
> Hope this helps...
>

Joe,

Thanks so much for your response.  I had JUST figured out the issue
with kl2 when your email came in, so I tried out your approach as
well.

Your example didn't quite work, but it got me started.   Here's the
actual command line that did what I described in my previous email:

ecasound -a:1 -i:tone.wav -ea:100 \
-klg:1,0,100,14,2,1,3,0.20,4,0.20,5,1,6,1,7,0.20,8,0.20,9,1,10,1,11,0.20,12,0.20,13,1,14,1,15,0
\
-o:test.mp3

The main difference is that it specifies the "from" points - the end
of the spaces between the fades (better illustrated on SO).

As for using kl2 (for the record), I realized that the "from"
parameter on kl2 should be 100 every time, and if increasing the
amplitude, the "to" parameter needed to be (100 / from * to).  So if
going from 20% to 100%: 100 / 20 * 100 = 500.  If going from 45% to
90%: 100 / 45 * 90.  Of course, now if you drop to 20% of the current
amplitude, it's actually 18% (.20 * 90), so going back to 100 would be
100 / 18 * 100 = 555.56

I explained the whole solution (with the final correct screenshot) as
an answer on StackOverflow

http://stackoverflow.com/questions/6950482/how-do-multiple-amplitude-fades-work-on-ecasound/6952211#6952211

Joel, I think I may end up using klg in my scripts as that seems to be
easier to dynamically generate without having to think about the
relative amplitude for each effect.  I really appreciate your help.

Mark

>
>> I've been testing this with a constant tone generated by ecasound so
>> that I can open the resulting file in Audacity and see the results
>> visually.  As far as I can tell, increasing the amplitude is relative,
>> while decreasing it is not.  It seems that if I fade the amplitude up
>> that it affects the volume of the whole track and not just at the
>> specific time I set the fade, which is where I'm getting lost.
>>
>> Note: I seem to have figured out the "why" as I was coming up with
>> examples, but I'm still hoping you can explain if this is correct or
>> if there is a "better" way.
>>
>> An example of what has me scratching my head
>> -----------------------------
>>
>> # generate the tone
>> ecasound -i tone,sine,880,20 -o:tone.wav
>>
>> # Just the test to see that i can fade start it at 100 and fade it to 20.
>> ecasound -a:1 -i tone.wav -ea:100 -kl2:1,100,20,2,1   -a:all -o:test_1.mp3
>>
>> # Fade it out and in
>> ecasound -a:1 -i tone.wav -ea:100 -kl2:1,100,20,2,1   -ea:100
>> -kl2:1,20,100,4,1   -a:all -o:test_2.mp3
>>
>> # Fade it out and in with a peak of 500
>> ecasound -a:1 -i tone.wav -ea:100 -kl2:1,100,20,2,1   -ea:100
>> -kl2:1,20,500,4,1   -a:all -o:test_3.mp3
>>
>> # Fade it out from 500, out, and then back to 500
>> ecasound -a:1 -i tone.wav -ea:100 -kl2:1,500,20,2,1   -ea:100
>> -kl2:1,20,500,4,1   -a:all -o:test_4.mp3
>>
>> # Fade it out from 500, out to a low of 10, and then back to 500
>> ecasound -a:1 -i tone.wav -ea:100 -kl2:1,500,10,2,1   -ea:100
>> -kl2:1,10,500,4,1   -a:all -o:test_5.mp3
>>
>> # Fade it out from 1000, out to a low of 10, and then back to 1000
>> ecasound -a:1 -i tone.wav -ea:100 -kl2:1,1000,10,2,1 -ea:100
>> -kl2:1,10,1000,4,1 -a:all -o:test_6.mp3
>>
>> # The eventual result I'm looking for
>> ecasound -a:1 -i tone.wav -ea:100 -kl2:1,500,20,2,1 -ea:100
>> -kl2:1,20,500,4,1 -ea:100 -kl2:1,500,20,6,1 -ea:100 -kl2:1,20,500,8,1
>> -ea:100 -kl2:1,500,20,10,1 -ea:100 -kl2:1,20,500,12,1 -ea:100
>> -kl2:1,500,0,14,4 -a:all -o:test_7.mp3
>
> Hi,
>
>
>
>> -------------------------------
>>
>> The best I can tell from these results is that the amplitude of the
>> whole track is relative to the difference between the low and the peak
>> of all the fading effects.  I'm not sure if this result is expected,
>> but it's very confusing.
>>
>> Also, in the last result, the fades are no longer taking a full second
>> each.  In order to figure out why that may be, I took the final
>> fade-to-zero off and the durations were back to normal.  This does not
>> seem like expected behavior.
>>
>> -------------------------------
>>
>> # "Fixing" the fade durations
>> ecasound -a:1 -i tone.wav -ea:100 -kl2:1,500,20,2,1 -ea:100
>> -kl2:1,20,500,4,1 -ea:100 -kl2:1,500,20,6,1 -ea:100 -kl2:1,20,500,8,1
>> -ea:100 -kl2:1,500,20,10,1 -ea:100 -kl2:1,20,500,12,1 -a:all
>> -o:test_8.mp3
>>
>> -------------------------------
>>
>>
>> As a side note, I've also tried changing the -ea values to the
>> "current" amplitude with every line. It didn't make any difference (no
>> matter what I set -ea to)
>>
>> I have the very latest installed from git (2.8.1+dev). I had these
>> same issues with 2.7.0, which is why I upgraded and eventually found
>> myself here.
>>
>> Am I doing this wrong?
>>
>> Thank you for your time.
>>
>> ------------------------------------------------------------------------------
>> BlackBerry&reg; DevCon Americas, Oct. 18-20, San Francisco, CA
>> The must-attend event for mobile developers. Connect with experts.
>> Get tools for creating Super Apps. See the latest technologies.
>> Sessions, hands-on labs, demos & much more. Register early & save!
>> http://p.sf.net/sfu/rim-blackberry-1
>> _______________________________________________
>> Ecasound-list mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/ecasound-list
>
> --
> Joel Roth
>

------------------------------------------------------------------------------
BlackBerry&reg; DevCon Americas, Oct. 18-20, San Francisco, CA
The must-attend event for mobile developers. Connect with experts. 
Get tools for creating Super Apps. See the latest technologies.
Sessions, hands-on labs, demos & much more. Register early & save!
http://p.sf.net/sfu/rim-blackberry-1
_______________________________________________
Ecasound-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ecasound-list