[Csnd] Control signal for audio fader

Date	2013-05-06 14:14
From	Jeff Taylor
Subject	[Csnd] Control signal for audio fader
	Hi, I am attempting to set up a standard mixer audio fader value (to be multiplied against another signal) that is controlled with a signal of range 0-1, but I am having some difficulty finding the curve to use. I've attempted to use a base 10 log curve, scaling and offsetting the control signal such that the desired min and max values are achieved at the extremes of the control signal range, but the curve is so severe that almost the entirety of the control signal range is taken up by a very small slice of the possible resulting range. I want it to work like a mixer where there is more resolution near dB(0) that at the bottom of the scale, but what I currently have is far too much. What kind of curve is typically used for these applications? Also, what is a good minimum dB level to use to effectively get "silence?" Since the log curve has an asymptote at the y axis it is impossible to get exactly 0. Is there a handy way to calculate what it should be for a given bit resolution? -- Electronically, Jeff Taylor

Date	2013-05-06 14:28
From	Justin Smith
Subject	Re: [Csnd] Control signal for audio fader
	if the input kdb is 0-1, try something like this: kamp = ampdb(-60 + (kdb*60)) On Mon, May 6, 2013 at 6:14 AM, Jeff Taylor <taylor.jeff.d@gmail.com> wrote: Hi, I am attempting to set up a standard mixer audio fader value (to be multiplied against another signal) that is controlled with a signal of range 0-1, but I am having some difficulty finding the curve to use. I've attempted to use a base 10 log curve, scaling and offsetting the control signal such that the desired min and max values are achieved at the extremes of the control signal range, but the curve is so severe that almost the entirety of the control signal range is taken up by a very small slice of the possible resulting range. I want it to work like a mixer where there is more resolution near dB(0) that at the bottom of the scale, but what I currently have is far too much. What kind of curve is typically used for these applications? Also, what is a good minimum dB level to use to effectively get "silence?" Since the log curve has an asymptote at the y axis it is impossible to get exactly 0. Is there a handy way to calculate what it should be for a given bit resolution? -- Electronically, Jeff Taylor

Date	2013-05-06 14:28
From	peiman khosravi
Subject	Re: [Csnd] Control signal for audio fader
Attachments	faderCurve.txt
	Hi Jeff, I have attached a file that I use in maxmsp. It's a nice curve with an analogue feel to it. It shows the midi -> linear amplitude mapping. In Csound, you can just read the numbers into a table (with 128 points) and then use tab to read through the table values. Best, Peiman www.peimankhosravi.co.uk \|\| Subscribe to RSS Feed To receive emails about upcoming concerts visit this page & enter your email address in the 'follow' section. On 6 May 2013 14:14, Jeff Taylor <taylor.jeff.d@gmail.com> wrote: Hi, I am attempting to set up a standard mixer audio fader value (to be multiplied against another signal) that is controlled with a signal of range 0-1, but I am having some difficulty finding the curve to use. I've attempted to use a base 10 log curve, scaling and offsetting the control signal such that the desired min and max values are achieved at the extremes of the control signal range, but the curve is so severe that almost the entirety of the control signal range is taken up by a very small slice of the possible resulting range. I want it to work like a mixer where there is more resolution near dB(0) that at the bottom of the scale, but what I currently have is far too much. What kind of curve is typically used for these applications? Also, what is a good minimum dB level to use to effectively get "silence?" Since the log curve has an asymptote at the y axis it is impossible to get exactly 0. Is there a handy way to calculate what it should be for a given bit resolution? -- Electronically, Jeff Taylor