First of all, I never use compression or limiting in my mastering. But this is just my choice, I like dynamic range to be preserved as it is on the actual mix. But analog compression is a good thing, used in mixing, it's indispensable. Here is why: Compression is a tool. In the hands of a good operator, it is crucial. Too much dynamic range on recordings (the softest sound to the loudest sound) is bad. Not enough, worse. There is a art/science/technique to making a good recording. The compressor; overused, bad, used correctly, perfection! When music is mixed, human hands work the faders but the compressor is there to help. I'd venture to say that 99% of all recorded music of the last 90 years was created with a compressor/limiter. When electric recording came in around 1925, the Western Electric microphone went to the cutter with a fixed-groove revolving beeswax disk, 80 RPM usually (standardized at 78 RPM around 1934.) The music, of course, was recorded live, the cutting amplifier was only about 1 or 2 watts and the saturation of the tubes caused a pleasing compression that made it possible for a giant dance band to be recorded with only one microphone and mastered on the disk at full volume without overload distortion. You could hear the tuba in the back, the horns, drums, guitar, bass, reeds, vocal, everything with just one microphone. Why? Tubes really don't overload, they just compress when stressed. I've heard thousands of 78's from 1925 to the start of tape in 1949 and I've never heard any distortion on any of them. Some sound better than others, but the saturation of the tubes prevented any ugly odd order harmonics to screw anything up, even on a piano recording from 1925. (Compare that to the 25% piano distortion on the peaks during 1961's Bill Evans At The Village Vanguard!) When cutting amps got bigger, more accurate in the 1930's, the Victor and Columbia engineers discovered that it was harder to control the sound of the music because it was becoming too dynamic. The saxes would be lost because the horns were three times as loud. More microphones were needed, causing the orchestras to sound unnatural, isolated. The pleasing saturation of a low power tube stage was needed to make it all gel. Thus, they found a way to make the tubes work against their nature, instead of amplifying, they de-amplified the sound. Compression via outboard compressor/limiters! The best were the RCA Opto-limiters, using optical sound like in a movie studio and making it work in reverse. Later units could choose (by operator switching) between compression and limiting. In most old recordings (and most new ones) compression is always in the circuit, reducing dynamic range during the music and the limiter part only comes into play after a certain level is reached. The RCA engineers decided that around 15 db of dynamic range (a hell of a lot, actually) was just about right for late 1930's big band recording. Both the compressor and limiter were used on an Artie Shaw or Benny Goodman recording, the limiting only kicking in during the loud brass passages. "Why do we need this at all?" You ask me this all the time. Too much dynamic range is very unlifelike. In the concert hall, the sound bounces around and compresses nicely before it reaches your seat. Try listening to an orchestra OUTSIDE. Sounds like dead crap, right? You need the concert hall just like recordings need compression. Make sense? Let me think of another way to talk about it.. Well, think of a recording as a good Chili sauce that you're making for a party. You have the ingredients on the counter, chopped onions, tomato, seasonings, etc. In a multi-track recording, each instrument on a track can be considered in these terms. If you do, it will be easier to understand. The drums are onions, the bass, spice, the guitar, something else, and so on. Now, you know that in order to make a good sauce, the ingredients need to be cooked and simmered to perfection for everything to blend together just right. Over done, bad, under done, you can taste each ingredient. Urggh. The same goes for mixing music, but we cannot mix music without the fire to simmer the various instruments together. This is what the compressor does. It is the crucial step to get something to sound not like a bunch of separate things, but as a whole. I've remixed a lot of stuff without compression and I cringe when I hear these today on the radio or wherever. They sound like band demos and not real recordings. Two hands can't control faders and dip something lightning fast, impossible but the compressor fuses it all together in such a good way (if used correctly) that you don't even realize it's working magic, you just know you like the sound. BTW I typed all of this on my iPad, I'll try and correct typos when I have a chance. Questions, comments welcome.