Free way of measuring compressors

[wlinart]

Found this online: and CMT (Compressor Measurement Toolbox) Manual
It's meant for measuring the characteristics of compressor plugins, but also works with hardware. And it's free

 

[Bobby Baird]

Interesting.

 

[user 37518]

This is pretty standard stuff, its called measuring the step response of a system and it applies to anything, from compressors, eq's, etc... and you could do the same thing with a digital scope thou... square waves have extremely fast rise and decay times (what you are actually measuring is its step response to a positive and negative pulse), so even though you might be accurately measuring the compressor response to a step input, when it comes to music, signals usually don't have such fast transition times. If you use a band limited square wave, this will soften the transition edges and give a more "musical" representation of how compressors work on music signals. You could also use other type of signals, like a sine wave with a gaussian envelope (a sinewave multiplied by something like a Hanning window) also know as a wave packet, etc...

 

[wlinart]

This is pretty standard stuff, its called measuring the step response of a system and it applies to anything, from compressors, eq's, etc... and you could do the same thing with a digital scope thou... square waves have extremely fast rise and decay times (what you are actually measuring is its step response to a positive and negative pulse), so even though you might be accurately measuring the compressor response to a step input, when it comes to music, signals usually don't have such fast transition times. If you use a band limited square wave, this will soften the transition edges and give a more "musical" representation of how compressors work on music signals. You could also use other type of signals, like a sine wave with a gaussian envelope (a sinewave multiplied by something like a Hanning window) also know as a wave packet, etc...

I know there are other ways to do these kind of measurements, this is just an easy way

 

[JohnRoberts]

Back several decades ago when I was actively designing dynamics processors I made a custom tone burst generator to excite and review response to signals steps.

[TMI] for today's too much info about tone bursts, my DIY burst processor didn't have its own internal sine wave oscillator so accepted an external audio input. I incorporated multiple adjustments. Controls varied burst on time, burst off time, and on/off ratio. The variable on/off ratio was created by feeding a dry signal feed around the gated signal. That burst off was performed by a JFET mute that was synchronized to only switch on/off at zero crossings (reducing clicks), further the gating was coordinated to include only even numbers of zero crossings, so the gated sample did not inadvertently have DC content from an odd extra half cycle (reducing thumps). I spent many hours on my design bench coaxing out misbehaviors from marginal gain elements of the time. We didn't have modern VCAs back then. If I couldn't make a dynamic processor misbehave with my burst box, it could pretty much handle all difficult music signals. For another final test of my designs I was able to feed music into my burst box, and punch up the crest factor (peak to average) of typical music. Not what I would consider a useful musical effect, but very effective for bench testing dynamics processing. [/TMI]

edit [TMI-2] in light of modern technology we should be able to generate a smart tone burst using digital computer sound cards or the equivalent [/TMI-2]

JR