Compression Revisited (long) Beating a Dead Horse?

[LawrenceF]

Ok, I hate to beat a dead horse until he wakes up and enters the Kentucky Derby but this no compression thing is becoming really interesting. Sit back and join me on my continuing journey to become a really good audio engineer.

 

1. One of the things I've done as I matured as an engineer was try to get my final mixes to sound as close to a good commercial recording as possible. One of the first steps in that process was aquiring the skills to capture signals that were high quality enough to stand up to that kind of examination. That took (still taking) a while to learn to do consistently with the multitude of signal types I record at my studio. Fine.

 

2. One of the next stages in my growth plan was to learn how to mix those signals in a way that would be arguably comparable to those types of "world class" recordings. This can take years but I continue the quest. Different songs require different decisions. There are no set rules. Compression, EQ, reverb, delay, at what balance, etc, etc, etc. Fine. I've made some definite progress there.

 

3. One of the "final" logical steps seemed to be getting that mix to sound "tight" as commercial recordings do. This led me to study the use of compressors to "tighten" up a mix for people who won't be using professional mastering. It worked well. I pat myself on the back. I AM SPARTACUS! Fine.

 

4. Now along comes Bruce with his theory of "little or no compression". It confused me at first because I liked the way my recordings sounded and so did my clients. I couldn't figure out why I should perhaps experiment recording and mixing without comps. Fine? Not.

 

5. After fighting every urge in my body to revert to what I've been doing for years and remixing some songs without any compression, only automation, it suddenly hit me like a brick. When I play those mixes back in the car, on the stereo, on the walkman or mp3 player I hear something I hadn't heard before even with the mixes that I was very happy with. What was it? Realism. Especially on vocal tracks.

 

Uh... Realism? Whaddya mean?

 

It's hard to put into words but I'll try. A good "tight" mix in the car can be a thing of beauty, especially on a song with a really good beat and percussion. I strove for that knowing that even when my lead vocal was squashed 4:1 to sit better in the mix it would still sound good. Now when I hear the same song without compression it's not the "tight" mix that I notice but the realism. It's much closer to feeling like you are in the music than just enjoying a good mix/song. An entirely different perspective... one that I am starting to prefer over the typical "tight" mix which has been my benchmark. I think I finally get it.

 

I was taught to "control that signal so it sits better in the mix" and here comes Bruce saying (paraphrasing here... "uh..., you're screwing with the essence of the signal. Leave it alone!" . Even worse (I hate to admit), I've heard a couple of people say it before but didn't give them the proper due because they don't have his credentials. Shame on me for being close minded.

 

I'm totally dumbfounded by what I'm hearing. I still reach for comps on my digital console for certain signals purely from habit and stop. I'd sure like to hear from anyone else who has taken a serious look at eliminating compression to make sure I'm not crazy. With the headroom we've got available today why not?

 

Any input appreciated. Thanks for reading this long post.

 

Lawrence

[Thomas K]

What about how you recorded the material in the first place. Wouldn't it start with mic placement to determine the signal? Is that not how Bruce came up? How many times have you really nailed a song because of mic placement and not had to do that much to it in the mix because it sounded so damn good.

 

I hadn't thought about riding faders in the mix instead of compression. I'm mixing a song right now. I will try it.

[where02190]

With the ease an ability of DAW mixing and editing, particularly detailed volume automation editing, your mixes can be even tighter and more professional sounding if you forgo the one size fits this entire track compressor for detailed volume editing. customize each dynamic change, rather than have a global setting for each track via a compressor.

 

Whether a track is to be professionally mastered or not, theat final polish needs to take place on the mixed track, not during mixdown. Yes get it close, but leave the mix dynamics to a mastering stage, even if you're doing it. You'd be amazed at the differnt approach you'll take after a few days away from a mixed track that has'nt been buss compressed.

[John Sayers]

I really like the way you can add 3db gain to a word or a phrase or even a "T" in digital.

 

cheers

john

[sign]

Lawrence, much depends on the quality of the player/vocalist.

 

There's such a thing like multing. I always use multing on vocals that are nice, sometimes I use light compression like 1.2:1 and another compression like 3:1 and mult them to another two tracks.

 

So you leave the original vocal with the nice transients alone and add light and heavy compression to taste, let the song tell you what it needs.

 

A couple of years back I've worked two years on the debut album of a very nice female singer. We've put very much of ourselves into that album and everytime a song was finished, I've put it on DAT as a rough mix.

 

The record company decided to mix it in New York and I was very unhappy with that mix, everything was compressed and squashed, even the grand piano was compressed and had lost it's beautiful tone.

 

Yesterday I got a Luxman A371 amp and connected it to my IMF RSPM speakers (17hz to 40khz) and grabbed the CD with the rough mixes. The sound was absolutely gorgeous. No compression at all and mixed in ten minutes.

 

Bottom line: leave it like it is, like Harvey says: 'if you polish too much, you may lose the edges'.

 

Peace, Han

[KenElevenShadows]

As I mentioned in another thread, I am veering away from using compression when tracking vocals, and use automation to "compress" the vocal (gain-riding). But that's not to say that compression is necessarily bad. Just like any other tool, there's good uses and bad uses.

 

You must listen to the song. The song tells you what to do. Yes, it does. Ignore those little voices in your head. Listen to the song.

[Thomas K]

There's such a thing like multing. I always use multing on vocals that are nice, sometimes I use light compression like 1.2:1 and another compression like 3:1 and mult them to another two tracks.

My patchbay is normaled. How would I go about multing my lead vocal? Do I have to have three pre's to do it. Any further explanation of how to do this would be appreciated.

 

Thomas

[kylen]

Originally posted by sign:

[QB]There's such a thing like multing. I always use multing on vocals that are nice, sometimes I use light compression like 1.2:1 and another compression like 3:1 and mult them to another two tracks.

 

So you leave the original vocal with the nice transients alone and add light and heavy compression to taste, let the song tell you what it needs.

For the DAW crowd this sort of multing has a one-plugin solution (of all the available DAW solutions) using a real trasparent VST plugin called Voxengo Polysquasher - it has a mix knob so you just dial in the amount of dry (uncompressed) or wet (compressed) mix you like. There are other plugs in that price range also that have mix - instant multing. I think the host needs plugin delay compensation (PDC or ADC) for Polysquasher to work in a mixing scenario - like Sonar4 and it's VST Adapter have.

http://www.voxengo.com/psquasher/

 

Want to use Volume automation because compression hits too hard (hehe), but don't like Volume automation because it takes to long? Here's one solution:

 

http://www.kvraudio.com/get/1288.html

 

Blue Cat Digital Peak Meter will write an automation curve that follow the peak envelope of your audio. The Pro version ($6 I think!) will write an inverse curve. You can then assign this envelope to Volume or any parameter of an inserted effect (that has automation capabilities) like the mix knob of Polysquasher - back the compressor off when you don't need it - dial it in when you do.Offset the envelope in time - set up controls before events happen, etc. Science Fiction.

 

Just some stuff to debate...

[Tedly Nightshade]

Me, I've only used compression in a desperate attempt to salvage something badly tracked. I've never enjoyed using it, and always found it to trash the sound quite promptly. Nonetheless, there are recordings out there that use compression artfully that please me.

 

I have had an interesting look at compression by having some stuff mastered recently. What I found- compression changes the groove! It can fuck it up badly, or it can shape it into something unrealistic and unnatural but cool.

 

It's been very strange to get something back from mastering, moderately compressed (no squashing) and the groove is history. Back and forth with mastering, describing how the swing ought to feel, and it comes back grooving again (less compressed).

 

Many grooves, apparently, are artifacts of compression and aren't anything live people can do! Although of course we all may try to cop grooves achieved through compression when we play unrecorded.

 

I really think a big reason we're so nuts about metronomic (inhuman) timing in recordings these days is that metronomic timing responds better to compression than human timing. So often the hook points of a live groove between humans amount to what I call a "transient dance". The percussive peaks are what are hooking together to make the groove. And these peaks happen in a different place in time than the meat of the notes. Compress it, the transients are less prominent, and the meat is more prominent- and the timing sounds completely different.

 

Curious stuff!