how do you know if your compressor is compressing?

[foxfire]

i made myself one of tonepad's orange squeezer2's, http://www.tonepad.com/project.asp?id=5. anyway i don't really know what a compressor does? i don't have a switch wired up on it yet so i can't really a/b it just yet. as far as i can tell it is working correctly in that the volume pot works as it should and i get a faint distortion like they said it would but, since i don't know what it's supposed to do i can't say for sure?

[Processaurus]

Hi, lots of guitarists don't know how compressors work, but they can help a lot with people being able to hear what you are playing clearly, and smoothing out volume anomolies in your playing. 

Compressors are like a monkey with exceptionally good reflexes turning down a volume knob when things get too loud.  You can tell if they are working by playing soft, and playing really hard, and seeing if the difference in volume between the two are different from when the thing is bypassed.  The compressor should turn down the really loud stuff.

The OS is on the subtler end of the pedal compressor spectrum.  It doesn't have the crazy sustain of the dynacomp/ross and others like that, but it sounds nice, and is handy for things like evening out the volumes when you go back and forth from chords and single note lines.  I like how mine sounds with clean guitar, if I turn it on when I'm playing with a loud band, it's interesting, all of a sudden you can hear what is being played better, without it necessarily seeming louder.  That's from notes being evened out in volume, and the decay of the note being a bit louder in volume, because the attack was turned down by the monkey.

[ulysses]

i dont know why compressors are called compressors - they should be called "maximisers"

in every day life a compressor makes something smaller - like a rubbish truck that compresses your trash.. or a zip file that compresses your data..

if you have ever used audio editing software you may have used a function called "normalize" - which makes quiets section of an audio file as loud as the loud parts of the track - giving you a track of constant volume..

to test this theory - play a string really softly and then play it hard - the note should come out of your amp at the same volume..

it is amazing how many people own a compressor and dont know what it actually does - you get a lot of responses like "makes it tighter mate" - sounds like a hella-lot of mojo to me.

cheers
ulysses

[micro]

i dont know why compressors are called compressors - they should be called "maximisers"

They can be more specificlly known as "dynamic range compressors" because they compress the dynamic range, or, the range from the softest sound to loudest.

A compressor for a guitar often differs from one used for mixing or tracking vocals and well, all other signals, for the simple fact that
it has less controls. This is because it is designed to use common settings specific for guitar, rather than being able to adjust for multiple signal sources.
Setings you might find on a compressor are:

Threshold- Set the minimum level for which the signal has to reach for your compressor to engage
Ratio- The ratio of total signal to compression/ example: 2:1 says for every 2dB of audio in it will compress 1 dB, 4:1 says for every 4dB in it will compress 1dB.
Attack- How fast the compressor engages once signal has reached the threshold.
Release- How fast the compressor "lets go" of the signal once it has began processing
Makeup Gain- Since your dynamic range has been squashed a bit, what was your loudest sound is now softer, so your makeup gain allows you to bring the overall
volume up to "make up" for lost volume.

http://www.youtube.com/watch?v=RWmrTgGBXuY
http://en.wikipedia.org/wiki/Audio_level_compression

Hope this helps foxfire


Wow I really geeked out on this one. 

[dirk]

i don't know why compressors are called compressors - they should be called "maximisers"

in every day life a compressor makes something smaller - like a rubbish truck that compresses your trash.. or a zip file that compresses your data..

if you have ever used audio editing software you may have used a function called "normalize" - which makes quiets section of an audio file as loud as the loud parts of the track - giving you a track of constant volume..

to test this theory - play a string really softly and then play it hard - the note should come out of your amp at the same volume..

it is amazing how many people own a compressor and don't know what it actually does - you get a lot of responses like "makes it tighter mate" - sounds like a hella-lot of mojo to me.

cheers
ulysses

A compressor does the same to audio as a zip does to data. After compression, the amount of information in the audio signal is diminished. (Information theory states that if a signal changes less, the amount of information in it is also less.)
Normalization is just gain, not compression. The amount of information stays the same.

All a compressor does is top of the loud sound levels.
Compressors do also have a gain knob (normalization). This is what makes the signal higher in level. Thus making it louder. And because the compressor reduced the peaks of the audio signal, low level signals will be closer to the higher levels. Making the overall impression of sound level much higher than with just gain.

[ulysses]

All a compressor does is top of the loud sound levels.

i think you've got your compression theory round the wrong way. loud signals stay the same - low signals boosted up.

if you want the theory check out the page i wrote on the john hollis flatline compressor - its in the gallery - there is also a thread here somewhere inwhich mark hammer backs up its findings - low signals boosted up.

cheers
ulysses

[axg20202]

if a compressor only raises low levels, how do you explain the most common parameters of compressors, namely "gain reduction" and "knee" or "threshold"? The threshold of a compressor is set by the knee, beyond which gain reduction is applied. The more aggressive the settings, the more the peak gain is reduced. A compressor does squash, so it is named appropriately. The whole point is to control spikes in level, not to bring up the lower levels - that's normalization, which is not the same thing.

[MartyMart]

Ulysses, you're not wrong but you've just confused "compression" with "expansion" , an expander
will bring up percieved level of the quiet stuff and compression has the opposite function.
The two ARE very similar in nature though.
MM.

[Paul Perry (Frostwave)]

What micro said.
You are reducing the distance between the loud and soft parts of the music.
Which enables one - if you want - to crank it up more without overloading the speakers. Which is why compressed music CAN sound "louder". But it doesn't HAVE to.

And what leads to some confusion, is that because compressing the dynamic range moves the noise floor upwards toward the centre, compressors are often combined with a kind of noise gate (specifically, reducing the gain when the sound is below a certain threshold).

[foxfire]

well alright! ask and ye shall receive. so i guess i should just slap on a switch and go from there. thanks everyone for your explanations. Rylan

[dirk]

I see that there's still some confusion about the matter.

All a compressor does is top of the loud sound levels.

i think you've got your compression theory round the wrong way. loud signals stay the same - low signals boosted up.

if you want the theory check out the page i wrote on the john hollis flatline compressor - its in the gallery - there is also a thread here somewhere in which mark hammer backs up its findings - low signals boosted up.

cheers
ulysses

Ulysses, you're not wrong but you've just confused "compression" with "expansion" , an expander
will bring up perceived level of the quiet stuff and compression has the opposite function.
The two ARE very similar in nature though.
MM.

As pointed out (by a.o. Frostwave), compression reduces the distance between the loud parts and soft parts of the music. It does this by reducing the louder levels. Above a certain threshold level, the loud parts get less loud. This frees up headroom, and now you can adjust the gain to make the sound louder. It is the gain adjustment, after compression, that makes the softer parts louder.

An expander is the opposite of a compressor. Hence it makes the distance between soft and loud parts bigger. Below a certain threshold softer sounds get even softer. The noisegate mentioned by frostwave is the most extreme of expanders. Soft sounds (containing only noise) get softer (turned down completely) and you only hear the loud parts.

[axg20202]

This topic has been done to death now, but a good example of how compression reduces the peak level is in the situation known as 'ducking', which is a common use for compressors for radio stations and club DJs. In this situation, the threshold of the compressor is under the control of a second signal - the DJ's voice (this technique requires a compressor with a 'side-chain'). When the DJ talks, it reduces the threshold of the compressor and 'ducks' (compresses) the programme material (the music) so you can hear what he's saying. When he stops talking, the gain is returned to normal.

[ulysses]

marty im not talking about an expander - an expander raises gain above a threshold and increases dynamics

im talking about upwards compression vs. downward compression having the same net results - low signals boosted up.

downwards compresson (in simple terms) = take a signal, half its volume, then boost the low parts up based on how loud the input signal is then pass to another boost stage (gain knob) becasue the signal is still a maximum of half the original volume.
upwards compression (insimple terms) = take a signal, boost it right up and lower boost if the input signal is high.

therefor, with an upward compressor if a signal is inputed at full volume, no boost is applied and the clean signal passes - with a downwards compressor, the signal is always reduced and then boosted back up.

paul said, floor noise is raised up - he did not prove your theory - he proved mine.

two different methods of achieving the same result - low signals raised up.

cheers
ulysses

[slacker]

two different methods of achieving the same result - low signals raised up.

In a "downward" compressor the quiet parts aren't boosted. The high volume parts are reduced in volume and the quiet parts are left unaffected, so you end up with a signal who's dynamic range is reduced and it's also quieter than the original. You can then boost the entire signal back up the the point where the loud parts are the same volume as they were originally and that gives the impression that the quiet parts have been boosted.
Like you said it's the same end result, but thats where the confusion is coming from.

[dirk]

I'm talking about upwards compression vs. downward compression having the same net results - low signals boosted up.

downwards compresson (in simple terms) = take a signal, half its volume, then boost the low parts up based on how loud the input signal is then pass to another boost stage (gain knob) because the signal is still a maximum of half the original volume.
upwards compression (in simple terms) = take a signal, boost it right up and lower boost if the input signal is high.

therefor, with an upward compressor if a signal is inputed at full volume, no boost is applied and the clean signal passes - with a downwards compressor, the signal is always reduced and then boosted back up.

I don't understand your explanation of downward compression.
But:
Upward compression is parallel compression.
So a compressor as I described earlier, parallel to the normal signal.

[Pushtone]

One way you can test your build is to put a booster in front of it.
Feed the Orange Squeezer a super hot signal and you should hear some serious compression.




I have found a compressor after an Envelope filter is a good combo.
My MXR EF filter can produce huge dynamic range swells that can be painfull in the ear.
Put a comp behind the filter and you can even out those peaks and bring the level up.
try it with an auto-wah.

[ulysses]

One way you can test your build is to put a booster in front of it.
Feed the Orange Squeezer a super hot signal and you should hear some serious compression.

i believe you will hear some serious volume reduction + the regular ammount of compression.

parallel compression is not the same as upwards compression.

also, my explanation of normalization was a very poor one - i frequently normalise select quiet sections of an audio track where i dont want all of the track compressed - and that is what i was trying to describe - so that is where the confusion of my explanation on normalization was coming from..

i will draw a diagram to explain how downward vs upwards compression works. give me 15 mins or so.

cheers
ulysses

[ulysses]

here is a diagram of how compression works - in simple terms.

the original sig in all cases is a sine wave with say a 5db boost in the second half of the waveform. it will be used to demonstrate compression.

edit: the dotted lines represent where the first half of the original signal used to be before it was boosted.

DOWNWARDS COMPRESSION
i think you are getting confused with the progression from A to C  - C looks like A has had the 5db boost squished - what has actually happened is the whole signal has had its volume reduced and the lows in the original signal have been boosted up. the booster D brings the reduced overall volume back up to the original level - as you can see the net result is  - lows are boosted

PARALLEL COMPRESSION
a dry signal and a compressed version of the dry signal are mixed to give 1. the original dynamics of the signal 2. the fattness of raising lower signals up.

take care
ulysses

[Pushtone]

One way you can test your build is to put a booster in front of it.
Feed the Orange Squeezer a super hot signal and you should hear some serious compression.

i believe you will hear some serious volume reduction + the regular ammount of compression.

Of course there will be a reduction in volume but that is what the pot is for on the OS.

Make-up gain.

Set it to the point where the bypass signal is at the same level as the OS and you should hear the squish.


This thread has gone away from the posters question and deep into the theory of compression.
I am suggesting a test (not a real world one either) to ensure the build is working. Thats all.

[slacker]

Sorry but the downwards compression example is wrong, the high volume signals are reduced not the low level signals boosted. So in your example you'd go from A straight to C where the highs have been turned down to the same volume as the lows. Then you boost the whole thing to get to D.