compression .....

  • Thread starter Thread starter yiordanaki
  • Start date Start date
Y

yiordanaki

New member
Hey

I was just wondering if anybody agrees with me in saying that it is possible to overcompress.

Also if you like, I'd like to hear peoples opinions on what compressing does to a waveform? (because ive heard clashing stories already) and what are the technical reasons audio engineers use it in music?

thanks
yiordanaki
 
Overcompression is rearing it's ugly head all over the place these days in a misguided attempt to be louder. A good compressor, properly applied should have only one effect on a waveform - it should decrease the dynamic range, that is, reduce the difference in level between the softest and loudest parts. A poorly designed compressor, or using settings at the extremes of their ranges can produce all kinds of waveform distortion and unwanted artifacts.
It has become popular to use compression as an 'effect', rather than use it for it's original intended purpose as a dynamics control. This has led to a lot of what I call "compressor abuse" - while it might enhance the occasional instrument or voice for a particular style of music, it usually just sounds like crap.

My suggestions for using compression:
DON'T use it as a replacement for fader moves; riding the fader sounds better 99% of the time.
Use conservative settings, and learn how to use the attack and release controls - if you can hear the compressor working, you did it wrong.
Leave the final "loudness" of the overall mix to an experienced mastering engineer. If you just can't resist compressing the mix, use the best compressor you can buy, rent or borrow - and burn 2 copies of the mix, one with compression & one without.

Lecture ends... :D

Scott
 
I am not an audio engineer, so, someone might have to correct me on this.

Over-compression can be used when you intend it to be an effect(Punchy/up-front) rather than just making the dynamic more subtle(compressing)

A very noticable change in a wav form after a hard compression is that you have a wav that has very little spikes sticking out, which can alomost look like a straight bar than a sin wav.

People use compression during mixing for better control. If the first 2 notes of your guitar playing is alot louder than the rest of the guitar playing, you are gonna have a make a decision:

1) turn down that entire guitar track, so the first 2 notes don't stick out in your mix. But then you are gonna have problem hear the rest of the playing if you turn the overall volume of that track down.
2) You compress that track, so the first 2 notes don't jump out, and still being able to hear the rest of the track. But, you might lose that dynamic you wanted.

So, it's a compromise you have to make.

AL
 
Thanks for the replies,

Very informative.
I'm studying Electrical engineering at the moment and something ive learned so far in my signal processing studies is that compressing is the act of reducing the no. of samples used to approximate the wave. And the way i see it, this gets rid of all the spikes or imperfections in the wave. However, the human voice (or a particular instrument) isnt perfect and so this really gives an unhuman feel if it is excessively done.

What are anyones views on this definition? maybe i'm talking about something else.......

yiordanaki
 
yiordanaki said:
However, the human voice (or a particular instrument) isnt perfect and so this really gives an unhuman feel if it is excessively done.
You may be thinking of another effect.... compression plays with the dynamics of the levels, as well as shaping the punch/sustain -- too much of it and the dynamics get squashed, but you won't get a voice to sound "inhuman" (well... except maybe if you use an Alesis 3630........!) ;)
 
yiordanaki said:
Thanks for the replies,

Very informative.
I'm studying Electrical engineering at the moment and something ive learned so far in my signal processing studies is that compressing is the act of reducing the no. of samples used to approximate the wave. And the way i see it, this gets rid of all the spikes or imperfections in the wave. However, the human voice (or a particular instrument) isnt perfect and so this really gives an unhuman feel if it is excessively done.

What are anyones views on this definition? maybe i'm talking about something else.......

yiordanaki

Actually you are confusing Data Compression with Audio Compression. Data Compression (MP3 and others) does remove some data from the original sample and should be avoided at all costs until you are finished and have to convert a final master to MP3 for the net.

Audio Compression reduces the dynamic range but it does not remove any data from the file.
 
thanks Tex and BBS.

Thats what i needed to hear. clears it up for me.

Thanks
 
A1A2 said:


People use compression during mixing for better control. If the first 2 notes of your guitar playing is alot louder than the rest of the guitar playing, you are gonna have a make a decision:

1) turn down that entire guitar track, so the first 2 notes don't stick out in your mix. But then you are gonna have problem hear the rest of the playing if you turn the overall volume of that track down.
2) You compress that track, so the first 2 notes don't jump out, and still being able to hear the rest of the track. But, you might lose that dynamic you wanted.

So, it's a compromise you have to make.

AL

Sorry AL, this is what I was talking about with my first comment. Just use the fader to bring those first 2 notes down, then bring it back up for the rest of the song. No compromise involved. If you have a problem instrument (like a bass with a 'dead' string) where you would have to be moving the fader up and down several time each measure for the whole song, that track might be a candidate for compression, but if you had fader automation available, I'd try that first.

Scott
 
DigitMus:
I think we both posted at the same time, I would have not posted if I saw yours first :) But, yeah, I would say the only thing you have to give in for using automation is the time/skills, which I lack :(

AL
 
A good compressor, properly applied should have only one effect on a waveform - it should decrease the dynamic range, that is, reduce the difference in level between the softest and loudest parts.


See, thats what I don't understand about compression. I mean I understand, what it does, it's the application of compression I don't understand.

Let's say you're playing/recording a classical piano piece. Dynamics play a BIG role in the performance. Many piano players, myself included, work very hard to achieve the proper dynamics called for in a piece, then you want to compress that, and in effect, destroy those dynamics?
I don't get it?
:confused:
 
mike, well, maybe don't! Compression isn't manditory to be used for everything! Its a handy tool to have when needed. Vocals usually get it, so the whispers are as loud as the screams, otherwise you'd either not beable to hear the person, or they'd be so loud they'd be clipping. (mic technique is more important, but this is just an example) It can be used on a lot of things very well, where you need to level out the dynamic range, but as you just stated, a piano performance is not one of those things.
 
Yeah!
Laughing at myself now. I now it's not manditory.
Seems like though, if you hd a vocalist, who whispered sometimes, and screamed at others, you could ride your faders at mixdown? Rather than compress the whole track.

Obviously compression, and it's proper use is not one of my stronger points.
 
DigitMus said:
A good compressor, properly applied should have only one effect on a waveform - it should decrease the dynamic range, that is, reduce the difference in level between the softest and loudest parts.

Michael, for the reasons you have stated, compression is used only very sparingly by most classical engineers.

But the quote you excerpted from Scott's post (above), while often true, is not completely accurate if taken literally. As I'm sure Scott knows, (and I appreciate it's hard to remember to include everything when posting off the top of your head...), a good compressor can be used to have other effects on a waveform besides reducing dynamic range.

For instance, if you use a slow enough attack time, the dynamic range of the attack transients (which are often the loudest spikes) won't be reduced at all, although the average program level may be. By playing with attack and release times, a good compressor can also be used, for instance, to increase the sustain time on electric bass notes. Such common uses of compression do not necesarily reduce the total dynamic range of the track.
 
Man, this is exellent......

Thanks for replying with this info everyone.


yiordanaki
 
Back
Top