Ableton Forum

Seeing these terms more and more as I read about mixing and mastering practices.

Can someone please exlpain the two?

google? wikipedia? in short ..peak is quicker and should show you the absolte maximum level.. so its better for tracking

while rms reacts slower and therfore rather shows the general loudness of a signal.. but ignores very fast overshots...

As far as I know, these both words are terms for the performance of Speakers they can last.

RMS is the sound intensity a speaker can keep playing without taking any damage.

Peak is the highest sound intensity a speaker can put out, but only for very short moments before it would take damage from it.

For example:

A Speaker with:

120W RMS >> permanent performance without damaging the membran.

300W Peak >> only for short moments it can last @ this output energy, for example the noise of a car crash. On longer time it would take damage.



I tried to paraphrase it because my english not very experienced considering this expert-knowledge, but i hope its enough to understand.^^

Greetz

thank you. I did google and wikipedia. I enjoy reading this forum and the different styles of explanations people have. Much appreciated.

RMS is used for signals that oscillate and is often used for power measurements. measuring power from a peak measurement tells you how much power or energy or force a system produces but since many systems can take powerful energy for brief moments an RMS value gives a value that better reflects the level of energy the system can take or produce for a sustained amount of time..

take a speaker, it can handle peaks over its maximum rated value for short periods of time. you'd use an RMS value to see what kind of CONSTANT energy it can handle.

another way to look at it, look at a sine wave. it spends very little time at the peak but most of the energy is in the lower 2/3 - 3/4 of the wave. an RMS value for a sine wave is 1/[sqrt(2)] or 0.7 or about 3/4, whatever's easiest for you to calculate in your head. (when considering a sine wave you have to think of it as a rectified sine wave since negative values drive the signal the same way positive values do.) non-sinusoidal signals have their RMS calculated differently, it's math I used to be able to do, not anymore, integrals and shit. generally speaking, people use off the cuff 2/3 to 3/4 measurements.

DIjaKeR wrote:As far as I know, these both words are terms for the performance of Speakers they can last.

The terms can be applied to speakers but that's not the basic meaning of them.

Just think of it as Peak is the instantaneous level that a signal hits at a given time, and RMS is a sort of a statistical average of the signal.

For metering, a peak meter will move fast and show you every transient, however an RMS meter will generally move more slowly and not hit as high a level as the peak meter. Makes sense you understand the definitions.

For a constant sine wave the square root of 2 factor holds but music is a constantly changing signal (technically a big load of sine waves ) so the RMS will change over time relative to the changing peaks.

Hope this makes sense, not the most techincally rigorous explanation!

+1 - maybe not the most technical terms, but I think of RMS as average level and Peak as the peak (highest) level.

in terms of audio levels for DAWS and Mastering.

For example, most CDs have peak levels approaching 0dbFS but an RMS of between -8 and -13db depending on genre (lower for classical and jazz etc).

Dynamic range is the difference between the peak and the RMS level.
http://www.pleasurizemusic.com/ watch the TT Dynamic Range meter graphic for a visual demo.

The Dynamic Range plugin can be downloaded free for win or mac from here
http://www.brainworx-music.de/en/download

In the context of audio metering, RMS doesnt make alot of sense on its own without some standard around the effective period over which RMS averaging occurs (or more likely a moving average) and some standard around how this is visualized in a meter (both combining as the meter balistics, or responsiveness and decay etc), otherwise what you see on one meter isnt going to be the same as the same signal through a different meter that uses different metrics.

The most basic RMS calculation if one assumes no DC offset is the measuring over a single cycle of a fundamental, in pratice its measured over a much longer period and probably using a moving average weighting system to save recalculating an entire sliding window (even if there are other easy shortcuts to calculating a slding window for this).

So for the purposes of our needs in mix metring, I suggest reading up on Bob Katz K-System metering. (http://www.aes.org/technical/documentDo ... m?docID=65)
The proposal covers alot more than just the metering, and is a very worthwhile read for anyone who works with audio and want to attempt to get some consitency from session to session.

VU meters are in theory supposed to be roughly calibrated to common balistics as well, but the VU meter concept predates any standards (AFAIK) as well as originating for monitoring speech levels in a way that looked meaningful for the kind of rate of signal level change encountered in normal speech on the moving coil meters of the time.

on a analog setup absolute peakmetering is not possible.. but there is headroom.. and with this headrom its possible to rely on rms metering.. its even better becaust it shows how mucvh energy a signal has... so the soundengineer can better judge by the metering where to place the fader..

without headroom you cant use rms at all.. you will catch overshots