setting up ratio on compressor

[pascual]

this is something i've been guessing at for far to long....so i ask.

what should i look at when setting up the ratio to compress a particular signal?

[MrM]

The perfect ratio for success
Just about any music you hear has gone through a compressor several times. Each individual instrumental line, each vocal, the entire mix, and again at the mastering stage.

Obviously compression is important, but it's tricky to get the hang of the controls at first.

Compressors really should not come with preset settings, although some do. Each situation will be different and the engineer needs to use his or her ears.

The best place to start is with the 'ratio' control. This is the main control governing the 'strength' of the compression effect.

A ratio of 1:1 means no compression at all is happening. Not much point in that then. At the other end of the scale, a ratio of 20:1 means that no matter how much you keep on increasing the input level, the output doesn't get any louder, at least hardly so.

One of the most frequent uses of the compressor is for vocals, simply because compressed vocals sound nicer. Here, you want the compression effect to be noticeable, otherwise there would be no point.

A setting of around 3:1 or 4:1 is a good place to start, and probably you will be able to leave the setting there.

A lower setting of around 1.5:1 or 2:1 is suitable if you have a recording that is too quiet in some parts, too loud in others. A low ratio should make the compression effect hardly audible but will balance out the levels nicely.

A full-on ratio of 20:1 or more creates 'limiting'. This is used so that you don't go into the red and spoil your recording through distortion. However, the limiting effect is clearly audible, so you would just use it as a standby and aim to keep your levels below the limiting threshold.

A ratio of around 3:1 to 4:1 is a good starting point for most types of recording.

Source: http://www.record-producer.com/learn.cf ... type=title

[MrM]

I am being very nice, and even add a second reply with much more info, take a look at this PDF: http://www.tcelectronic.com/media/drone ... ompres.pdf

It contains info on ratio's for Vocals, Guitar, Bass, Drums, Piano and Organ!

That's hot!

[Lagrange]

hmm so let me understand this. There is no actual way to measure or plot the amount of ratio based on the audio?? I mean this from a technical standpoint.. I guess then you have to go by the sound you WANT rather than the sound you HAVE?? Interesting..

[jens]

this is something i've been guessing at for far to long....so i ask.

what should i look at when setting up the ratio to compress a particular signal?

actually you usually first set the threshold and then adjust the ratio while listening carefully and looking at the compression-meter...


a threshold of around -20db tends to be audible while something like -12db tends to be inaudible...

[MrM]

hmm so let me understand this. There is no actual way to measure or plot the amount of ratio based on the audio?? I mean this from a technical standpoint.. I guess then you have to go by the sound you WANT rather than the sound you HAVE?? Interesting..

There is a way, take a look in a wave-editor and search for the highest peak (usually this is a feauture in the editor). Then look at what level the average sound is; then you will know the distance in dB, now you can decide for audible or non-audible compression.

Say the loudest peak is at -2 dB and the average level is at -25 dB, then the distance is 25 - 2 = 23 dB.
For non-audible compression, a ratio of 4:1 will make those 23 dB's into 23/4 = 5.75 dB (I'll round this up to 6 for easy readinng and typing), so the peak will still be at - 2 dB and the average level will be at - 8 dB.

Now you can also play with attack settings, for letting the attack of the signal thru or lessen the attack of the signal.
You can also play with release time, to make sure it follows the input; setting the compression so that the compression meter is always way at the extreme is no good; the meter has to move.
Unless you want it to sound that way

[kylen]

hmm so let me understand this. There is no actual way to measure or plot the amount of ratio based on the audio?? I mean this from a technical standpoint.. I guess then you have to go by the sound you WANT rather than the sound you HAVE?? Interesting..

You're funnin' I know. But seriously I've taken a look at a bunch of comparative rms/peak values from different parts of the spectrums of reference material as well as my own. I think it's kinda hard to come up with a fixed formula - in a world where milliseconds and db/octave are either very obvious on todays reproductive equipment ( ) or the pshchoacoustic affects thereof are...I just kinda treat ratio as a "size" knob - do I want a sound (or freq range) relatively larger or smaller. Usually ratio ends up anywhere from 2-4:1 or thereabouts. But that's when using a compressor to kinda ride the envelope (either peak or rms). I guess you're right - if you use deep thresholds (beginning 30db below relatively) then there is a lot lower ratio like 1.03:1 or something you might start at. I guess it matters if you want smooth or 'compression with an attitude'. Where' bmanic and Kingston? They know all this stuff...

[Lagrange]

hmm so let me understand this. There is no actual way to measure or plot the amount of ratio based on the audio?? I mean this from a technical standpoint.. I guess then you have to go by the sound you WANT rather than the sound you HAVE?? Interesting..

There is a way, take a look in a wave-editor and search for the highest peak (usually this is a feauture in the editor). Then look at what level the average sound is; then you will know the distance in dB, now you can decide for audible or non-audible compression.

Say the loudest peak is at -2 dB and the average level is at -25 dB, then the distance is 25 - 2 = 23 dB.
For non-audible compression, a ratio of 4:1 will make those 23 dB's into 23/4 = 5.75 dB (I'll round this up to 6 for easy readinng and typing), so the peak will still be at - 2 dB and the average level will be at - 8 dB.

Now you can also play with attack settings, for letting the attack of the signal thru or lessen the attack of the signal.
You can also play with release time, to make sure it follows the input; setting the compression so that the compression meter is always way at the extreme is no good; the meter has to move.
Unless you want it to sound that way

This is great cause I have a plugin that measures the average! COOL!!

[MrM]

No you don't want to MEASURE the average, cause that will be based on the peak signal as well; you will have to SEE 'how much louder' the peak is, relative to the peak of the average level.
Or just do it by ear...
But personally I rather look then listen

http://www.petethomas.co.uk/recording-compression.html



Take a look at the drumtrack, to level out the levels, take a close look (with different zoom level) and look at the peaks and the high level of the body...