compression, RMS?

[fn86]

can some1 explain how and when one would use rms compression? and is rms compression actually an function that a compressor may have or may not have, or is it a compression technique that can be applied to any compressor?

i used to think about compressors as a tool for squeezing the dynamics to a more flat waveform, but this rms thing has gotten me all unsecured and sdfsdfsdfew.

and rms, is it the AVERAGE MUSIC LEVEL ? of is that something else.

Often when i use compression for the think i thought it was for, squeezing the dynamics, i use the waves paz meter VST, and get the rms (average levels) from the meter, and set the treshold of my compresosn a litle bit over that treshold, isnt this what a compressor in general are used for when mixing? and should i call this a dynamic squeezing or peak squeezing type of compression

when its time for the mixdown, i usually apply a compression on the master bus, with like a -40 treshold, medium attack and release, and a low ratio of 1.1:1 or something like that. what is this type of compression technique called? is this the actual technique that i should call dynamic squeezing, and the first one i stated is actually called peak compression?

i need some basic info, the difference between rms, peak and dynamic squeezing compression!

 

[moses]

yup, RMS (Root Mean Square) is some kind of average level, it is the power of a signal.

there is a hot discussion about what compressor type is better for what purpose. the only important difference:

Peak comps acts much faster and sounds much less 'stable' on complex material than RMS comps, because they use a very short-term averaging. they are (trying to) react on the max-values, the peaks. they are good for controlling the maximum level of a signal. limiters are always peak-compressors.

RMS comps acts on the long(er)-term dynamics. RMS sounds very clean & transparent, but they are in most cases ineffective compressors (it's hard to gain more than 1-2dB average level with them), they are more sound-'shaping' devices to bring an album "together" in level.

but in fact, every peak-comp can be tweaked to act like an RMS. simply increase the attack (attack and release are in fact averaging settings). but no RMS comp can be tweaked to a peak-comp. 'RMS' is marketing if you ask me, making a bad feature to something good. an analogue 'RMS' comp is simply a very slow (technically bad) compressor. sounds nice in most cases, why? because it doesn't really compress the dynamic range.

you are on the right way. your PAZ technics shows you already understood most things.

there are two standart comp setups for threshold and ratio: low ratio and low threshold (averaging), or high ratio and high threshold (limiting).

 

[fn86]

man, that was a bomb ass post you made! you pretty much answered all my doubts i had and made me feel a lot less insecure! some questions though:

so there actually are two different type of COMPRESSORS? im just curious then, cause i guess the most of the vst computers i have used are peak compressors, and i guess then that they are very very more common then rms compressors?

how do a rms based compressor work? does it have the same kind of settings? i would guess that it is the treshold part of a rms compressor that works in a different way technicaly?

about rms:

"they are more sound-'shaping' devices to bring an album "together" in level"

you mean they are used to gel sub-mixes or mixes together to sound cohensive?


and about making a peak compressor into a rms type:

"simply increase the attack (attack and release are in fact averaging settings)."

with increase, do you mean set an attack that reacts a little slower ?(from 5 ms to 60 ms for example)

and what do you mean with atack and release are in fact average settings?


if you could answer those questions in the same simple way you answered the oringial quesiton, i would apreciate that man!

 

[moses]

so there actually are two different type of COMPRESSORS? im just curious then, cause i guess the most of the vst computers i have used are peak compressors, and i guess then that they are very very more common then rms compressors?

yes there are different types of 'averaging': peak & RMS, but more in the analogue domain. in digital-audio, it is easier and much cheaper to build peak-compressor.

(there are very slow comps in use, they are called 'leveler')

how do a rms based compressor work? does it have the same kind of settings? i would guess that it is the treshold part of a rms compressor that works in a different way technicaly?

no, only the timing. the averager is just 'slower'. the averager is a filter with independent attack and release parameters. the precise definition of RMS averaging is: "square the input, filter it, and take the squareroot of the filtered signal". use more attack (around 30ms), a low threshold and a low ratio to get an aproximation of RMS with a peak-comp.

"they are more sound-'shaping' devices to bring an album "together" in level"
you mean they are used to gel sub-mixes or mixes together to sound cohensive?

yes, this helps. you can equal the average levels without affecting the peaks to much. RMS is a very soft processing.

"simply increase the attack (attack and release are in fact averaging settings)."

with increase, do you mean set an attack that reacts a little slower ?(from 5 ms to 60 ms for example)

yes, slower.

and what do you mean with atack and release are in fact average settings?

attack and release are parameters of the averaging-function (envelope-follower).

check this for more (very nice) in-depth information on compressors:

http://www.tangible-technology.com/dynamics/comp_lim_ec_dh_pw2.html


http://66.249.93.104/search?q=cache...amics+sidechain+hidden&hl=de&client=firefox-a

 

[fn86]

in your posts, when you write "the averager", you mean an rms-compressor right?

and

attack and release are parameters of the averaging-function (envelope-follower).

with this you mean that attack and release are two of the controls on an rms-compressor, just as they are on a peak-compressor?


basicly, i do a mixdown and left everything else to be mastered from a professional. Would you say that i should only use peak compression in this stage of the production, or should i put some rms-compression on the master buss? and are there any scenarios where rms-compression on different elements in a mix would be great? and are there any special release times that are used for rms-compression in general? short or long?

 

[moses]

the "averager" is the envelope follower. the envelope follower is basically a filter 'damping' the (internal) rectified sidechain signal controling the gain reduction. it has 2 parameters, attack and release. they define the 'window' of the averager. attack + release = windowsize.

yes, attack and release are both paramaters of peak-comps and RMS comps. but RMS-comps have a greater 'window size'.

RMS compression is used when you don't want to compress "short-term"-dynamics. your sound will still sound fresh after processing. but the gain in loudness is low compared with a peak-compression.

i don't want to confuse you. technicaly, RMS is something else than peak. but it is very easy to emulate this with a peak compressor, because of the wider range of control.

no matter what stays on the front panel,
here a the internal settings of those two comp families:

peak: attack ~1ms to ~500ms, release ~25ms to ~1000ms

RMS: attack ~30ms to ~1000ms, release ~30ms to ~3000ms

 

[fn86]

aight man, i guess i just dont understand this:

the "averager" is the envelope follower. the envelope follower is basically a filter 'damping' the (internal) rectified sidechain signal controling the gain reduction. it has 2 parameters, attack and release. they define the 'window' of the averager. attack + release = windowsize


and maybe it's for the best, i mean ,maybe that is to advanced for me, cause i dont understand it



and one last question, besided the rms compression has a slower attack and release, does it function in another way technicaly than a peak compressor?

 

[fn86]

or maybe i do get it, the envelope is basicly the attack-release function of the compressor in an other word?