low frequency instruments mostly but guitars get a lot of compression, especially in metal.
It's fine with distortion - pre distortion, but then they will generally recompress the guitars at the mixing stage and squash the shit out of it so it loses it's dynamic and life.
Ocasionally it's used as an effect ie: Spiderbait, Smashing Pumpkins, but those guitars are hyper-effected, so what are you going to lose anyway, they are achieving a synth like sound.

Man, check those before and after remastering shots... it's fucked.

What really grinds my gears is the over use of compressors y'know, it's lazy.

Look at fine classical releases and jazz releases. shit all compression there, just perfection at the miking stage and balance in the mixing stage, if you mike correctly you get less spikes, so the overall sound can be brought up in the mix safely

quiet passages are meant to be quiet for a reason!

Att: Engineers
Use you ears people, that's what they are there for!

//Oh OK, so apart from the intentional use of compression you mention above (i.e., Smashing Pumpkins et al), you mean the over-use of compression is a sly way to cover up for a band that can't get its sound right, or an engineer that didn't get the accoustics right in the studio?

No not quite.
Generally guitars are close miked so room acoustics don't play a big part in the sound, only if you add distance mikes for natural reverb.
A little bit of compression on guitars in the channel before the amp will bring up lower (Frequency) and quieter Frequencies.
Where adding more compression during the mixing stage will flatten the sound and remove the dynamics(ie: If you play quieter it will be at the same volume as harder played parts.) and it sounds squashed.

I just got a pre-release of the feelers new album. (And yes, it's horrible).

A classic example of the loudness wars moving to New Zealand...

[ http://nzmusicreviews.blogspot.com/ ]

Good one.

I was actually trying my best to paraphrase everything I'd read on this and the production thread that's going on over the NZ Music forum. I'm just hoping I got it (vaguely) technically right.

// then the resulting flat sound has been cranked up (literally) to 11 at the final mastering phase.

that is a great summary. I think you also notice it really intensely with ads. There are some which are just like an aural onslaught. It would be interesting to know if those are the most compressed. I'd pick out warehouse and flight centre ads, but it probably applies to lots of them. Those are just noticeable because they're annoying as well.
I also wonder if it's responsible for the interesting dynamic changes when you make your own mix CDs, because it often seems that some tracks are louder than others.
Also, following on from Jet's comment, I wonder if the compression is why I really, really can't abide RHCP's last few albums. I've really never got what is good about them.

//In that case, what (how much?) constitutes "good" use of compression?

It's all relative really. but you can hear it if you really know what to listen for, such as at the end of a note you can hear the noise floor (The sound of the room etc, when nothing is playing) rise up, it's real bad if they use a slow noise gate (which reduces the volume on an instrument or track when volume drops below a specified level) then you get a sort of surge sound where the sound rises then falls away but at the same density (or pressure)

It's weird: I'm no producer or engineer or whatever, but this whole compression has piqued my interest. I'm doing some remedial reading right now, so I'm slowly picking up what those technical terms you use mean.

One thing really makes me wonder, tho'. Why bother trying to make a recording louder overall... can't we listeners just turn up the volume and still enjoy the contrast between instruments?

Sorry, to attempt to be a little more helpful, the two rationales I know of for increasing the overall volume are:

If your song is louder than others it will stand out more on the radio

and

if the overall volume of a song is louder it's better to listen to in a busy environment, whether that's travelling or in an office cubicle or cafe.

First one becomes quickly moot, of course...

Second one's not shit, but obviously it's easy to go too far, and as an idea it's thinking of music firstly as a functional thing rather than as a composition, isn't it? I guess if the artists want the music to work in such an environment then that's no problem.

I have had plenty of times when I've wanted to listen to something delicate and intricate when on a train or wherever, cranked up the volume a bit and suddenly been munged by a loud part hiding around the corner...

//Has anyone else noticed the distortion/clipping on the voice of the flight centre ad guy.

I suspect that may have something to do with the fact that his voice isn't actually his - they use the same guy in the ads in Australia and he speaks in a totally different voice with an aussie accent.

We presume tv is all about programmes interrupted by ads. Its really ads with programmes in the middle to draw people in.

It seems to me that russ has read this thread.

CD quality has diminished hugely and a lot of new releases sound so compressed that they are wearying to listen to, he says. In fact, I've gone back to LPs the new Lucinda Williams is the latest to hit the platter.

I think part of the "vinyl is better myth" needs to be dispelled, I think a majority of vinyl releases are actually lower quality than cd because I doubt that most major releases are mastered twice (once for CD and once for vinyl). I guarantee they send the same master to the vinyl press as the CD manufacturer. and then it will be re-mastered at the Vinyl press.

Once it is recorded to CD the damage is already done, frequencies are stripped out by the CD as is already mentioned in this thread.
Some labels will send high-resolution DATs, but not everyone.

I know that vinyl has a greater frequency range but vinyl is incapable of recreating frequencies that aren't there.

In regards to what I'm saying about CD's

http://en.wikipedia.org/wiki/Sample_rate

The NyquistShannon sampling theorem states that perfect reconstruction of a signal is possible when the sampling frequency is greater than twice the bandwidth of the signal being sampled, or equivalently, that the Nyquist frequency (half the sample rate) exceeds the bandwidth of the signal being sampled. If lower sampling rates are used, the original signal's information may not be completely recoverable from the sampled signal. For example, if a signal has a bandwidth of 100 Hz, to avoid aliasing the sampling frequency should be greater than 200 Hz

Read through this thread.
http://www.nzmusic.com/topic.cfm?show=all&i=16546

Vinyl has a different frequency response to CD (Ie: different frequency bands need attenuation or boosting) So an album gets Re-Mastered at the vinyl press, Regardless of how well it was mastered. They have to it's not an option, it's the same as needing to slightly eq your stereo to your room.

[ http://murl.se/22464 ]