Compressing audio files?

Hey there,

Is it important to compress a mix down? We are having a problem at the moment with our on air audio files. Over the years several different people have done the production duties, using several different programs and techniques etc, which sometimes has a noticeable difference in volume. Someone is saying if we compressed them all that would solve this problem is that true?

I've done a search on this subject on this forum here and most of the links are dead now.

Hope someone can help.

Cheers.

like everything else with audio editing and processing there is no thorough & simple answer.  Current broadcast signals are heavily compressed.  Majority of commercial stations will, among other things use some form of Optimod to guarantee consistency

If you are trying to duplicate a commercial radio experience the easiest thought not necessarily the cheapest approach is to use the gear they use

that's the simple answer, though not the thorough one

'compression' is general label used to reference a variety of dynamics controlling techniques.  (expansion, limiting, gating are other labels for other dynamics processes).  When you compress an audio signal you reduce its dynamic range, using a half dozen parameters you make the loudest portions quieter.  Then by using make up gain you increase the volume of quieter portions.  There is of course no 'free lunch' and whether compression alone is what you need to achieve your goals depends on a lot of things: genre, musical content, quality of recording, quality of mix (it is not particularly difficult for phase issues to mean that all extreme compression accomplishes is making a confused mess of sound louder (with make up gain)), complexity of material.

It's not unusual for a snare drum's loudest transient to be 9-12 dB louder then anything else happening during those 10-30 ms.  Since, for the sake of argument, that 'loudest' bit controls the upper limit of all the rest on a recording and no one I know is fast enough to ride the faders to 'duck' the snare transient.  What you can do is use a hardware or software processor to quash that loud transient by a specific ratio based on both amplitude and temporal (attack and release) parameters.  Snare drum gets quieter, make up gain makes everything else get louder.  And the snare being what it is you don't lose (necessarily) it's accent emphasis.  You smooth out the dynamics to provide room for everything to be heard.  Or you quash hell out of everything and end up with a brick of noise . . . depends on what your goal is.

You can also use compression and expansion in conjunction to remove energy from sustained notes.  A bass player with primarily live experience might not truncate the sustain of his note sufficiently for particular production goals.  You might still want his timing and 'sound' .  So you might leave his transients in place and quash the sustain (or you might 'gate' the signal)

There are times when rather then using EQ it is appropriate to band limit compression on some voices to reduce clutter in cluttered sections of a piece.  The voice you want to emphasize is centered around 400 Hz, so you might define a band centered @ 400 Hz in the other voices and quiet them with compression, without make up gain

main point here is that compression is not just one thing deployed in one way & by itself  a preset fix  for everything in sight tool (that's the optimods job)

to be any more specific it would probably help to here samples of material that both failed and managed to achieve the goal for which your striving

good luck

I've done a search on this subject on this forum here and most of the links are dead now.

Hope someone can help.

When I find dead links I either remove them if they are external failures, or get them working again. But as you can appreciate, there are rather a lot of them. This situation is not helped by an alteration to the forum software that means that internal links made before the change no longer work. If you come across one of those, then copy just the topic number from the link (topic=nnnn.n where the last n is often 0) and alter a working page URL so that it has the copied link number instead of its own at the end - then you will get to it.

So the end of the URL will say topic,{nnnn}.0.html and that will display the whole thread.

I appreciate above posts that are really beneficial for the stared thread.

I did try to compress so many audio files before but I was not success to get 100% quality but via reading above posts I can do it easily.

Thanks!