Backwards Compression

Here's something different, for me anyhow.

http://forum.recordingreview.com/f122/backwards-compression-vocals-32741/

I've got to try this.

His explanation seems overly confusing to me. Isn't he just suggesting that you compress a vocal track after it's reversed to get a smoother result?

Terry D.

Yes, exactly.

Good but old trick that has been done for decades. Can be quite effective. I believe Steve Albini does this sometimes.

Think I heard someone doing this on drums one time and all it ended up sounding like to me was the drum sound on Queen's "We Will Rock You". I remember thinking there were other ways to get that effect.

If there's any background noise just before each phrase, that would get amplified. Of course, that could easily be limited or muted before applying the compression.

The point is simply that when running backwards, the attack time of the compressor doesn't matter, because tails tend to be slow. (It wouldn't work as well with very staccato vocals, like on Mack the Knife.)

Techniques like this were more important back in the day. Today, with software that can essentially look ahead, you can have zero or even negative attack times. In theory, anyway -- I don't think I happen to have any compressors that do. (Such compressors wouldn't have low latency when used that way.)

His explanation seems overly confusing to me. Isn't he just suggesting that you compress a vocal track after it's reversed to get a smoother result? Terry D.

Not too sure myself Terry, still got to have a go at it myself, but a further explanation has been posted later on.

I think it is to compress, play/export the reversed track all at the same time, then flip the track back again, without compression. Not sure what the compression ratio should be, but I think up to 20:1 is mentioned somewhere and can be used on a variety of instruments.

He gives credit where he got it from.

I am looking to see if the book exists, yep comes up at the top of Google, yes, Mike Stavrou is his name, whether he was the first to do, that is another question.

when you additionally tune down the ambience (room mics or the artificial reverberation) a few cents, the vocals will sound a tiny little more sad

I stopped reading here:

hee hee, i stopped reading at the words compressor and vocal

Today nobody who makes music in high class sonics uses a compressor on vocals. When part of vocal phrases, or words are to soft, you simply add a few dB in the editor on that spot until the vocals don't disappear in the music.

Boy... that's like... not true.

If you follow the poster's instructions literally, he's telling you to flip it backwards, then flip it forwards again and expect it to magically different. Definitely a step missing there.

just think of using such a process on a recording of V

If you use the process on Michael J. Anderson, it just comes out normal.

I believe he actually does say to apply compression when it is backwards:

It's a confusing, unnecessarily wordy description of how to do it, but he does say to apply compression when the track is reversed.

To me music is about experimentation and having fun, not be a hero, if you don't like don't do.

Yes, I found that line quite interesting but I read the article anyway.

I think its an interesting technique and I guess it could work but I`m looking for that affect on a vocal I use my own vocal technique and if I`m recording someone who does not have the same control, I`ll compress with multiple compressors.

EB

Yeah, I figured it was about flipping the clip, compressing, then flipping back. That's kind of a yawn to me. I'll admit I never did a lot of in depth experimentation with it but it never seemed all that promising.

And I'm a lot more inclined to flip something backwards and leave it that way.

To me music is about experimentation and having fun, not be a hero, if you don't like don't do.

Good you don't work as structural engineer, or stress analyst. Then crossing a bridge would be no fun.

More complex works I compose from the beginning backward, a good final in my mind and I know it is worth working to the opening and then add the ouverture. Intermezzi however follow a complete other schedule, and commisions of pop songs I give to the local bedlam in order to be crazy enough to break into todays market successfully.

No offence intended, but I am not really into classical music myself, after hearing a few pieces it all sounds much the same to me, and today's (the past 30 years) pop market is a real yawn.

I like to listen to and create new and interesting music/sonic/studio/experimentation, and move from creation to creation, life is far too short to stick to what the big wigs in the music scene think is good for the masses.

Anyway I posted this for people who still enjoy experimenting with music/audio.

I find that anyone who dismisses almost any genre as all sounding the same isn't really listening. Especially one as rich and varied as Western classical music. That even applies to stuff I could live without, like electronica and trance, or minimalist, or just about anything. It's a bit like saying "all Orientals look alike". It's diagnostic of ignorance or inexperience. Deciding after listening to "a few pieces" confirms the diagnosis.

If there's a genre for tunes like Cage's 4'33", I'd admit they'd all sound pretty much the same. (lol)

BTW, I must be "nobody".

No no, your respond was excellent.

And I like to read what people say, no matter how wrong they are.

The term backwards compression can be mistaken with a backwards compressor which do exist.

Waves multiband and others have that feature. What the article describes is compressing a wave file played in reverse

then rendering the comp to those reversed transients. Then they flip it back for listening. Any artifacts will be displaced

from where you'd normally hear them.

Basically you're swapping around the attack and reverse gate times by doing that. Dont know how well it works but I may give it a try.

What Harbal does something even better with their new release. Since the levels are already scanned and not reliant on a look ahead

detection, the entire file gets dynamically smoothed out in a single step. The result is like simular to true volume riding. The results

I been getting really knock my socks off because you arent dicking around trying to get attacks and releases to match the transient tempos.

Heres a description on it.

http://www.har-bal.com/