Noviss's (the producer) guide to compression.

  • Thread starter Thread starter ethos
  • Start date Start date
ethos

ethos

Swag Type Heavy
this has come in pretty handy for me, maybe other people will find it helpful.

---------------------


Well, Compression is something most people don't really notice, but it can turn your mix from ill but sounding like crap, to ill and sounding nice...

Most peeps use compression, all the producers and beatmakers use Compression, doesn't matter if your Dre or anyone else, thery use compression... It makes your beats sound just a bit nicer...

Terms to know:

Attack: How fast a compressor will react once the threshold is breached. 0ms will result in immediate action.

Decibles (db): Measure of sound pressure.

Gain: Used to increase or decrease compressed sound. (measured in DB)

Knee: A compressor characteristic that affects the way a compressor behaves.

Milliseconds (ms): Attack and Release times measured by milliseconds.

Ratio: How much a signal is compressed. With a compression ratio of 3:1, a signal which is 9db over the threshold level would be reduced to 3db. A signal of 3db over the threshold would be reduced to 1db.

Release: How fast the compressor will return to its normal state after the signal has moved below the threshold. 0ms will result in immediate return.

Threshold: Threshold level determines which signals are subject to compression. With a threshold of -5db, all signals above this level (-4db < ) would be compressed by the set ratio.

----------------

Compressing instrument tracks...

Drums: Perhaps the most important element in a hip-hop track. DJ Premier, Pete Rock, Alchemist, Havoc, RZA, Marley Marl, Jay Dee, and Timbaland. What do all these producers have in common? Their thumping drums. Now imagine if all those beatmakers had used weak drums. Premier's "Come Clean" probably wouldn't be considered a classic, nor would Pete Rock's "T.R.O.Y". Compression is very much needed on drums, especially in the hip-hop world. What exactly does compression do to help? Fatten, thicken, louden, and sharpen. Deep, rumbly kick drums and sharp, snappy snares. Ah, the wonders of compression.

Threshold: -10db to -15db
Ratio: 6:1 to 8:1
Attack: 3ms
Release: 10ms
Knee: Hard
Gain: +5db to +7db
---

Percussion: Although not all hip-hop tracks contain, or need percussion, a lot of the newage pioneer beatmakers are using bongos, congas, triangles, steel drums, as well as other percussion instruments. Percussion doesn't require a lot of compression because usually, the percussion track rests behind the drum track. Bongos, congas, and the likes usually have an immediate popping sound that doesn't need compressing, so the attack should be set slower than drums.

Threshold: -3db to -7db
Ratio: 3:1 to 6:1
Attack: 5ms to 7ms
Release: 15ms
Knee: Hard
Gain: +2db to +4db
---

Bass: A common problem with bass is that the low notes seem to disappear into the mix while the higher notes stick out like a sore thumb. With many instruments, reverb could solve this problem. However, using reverb on the bass track usually gives it an undesirable effect. By using compression, you can bring up the lows, and submerge the higher notes into the mix. Often times, there is an initial "pluck" to the bass sound, and it can be more beneficial to let this sound slide through uncompressed.

Threshold: -4db to -9db
Ratio: 4:1 to 8:1
Attack: 3ms (if there is a plucking sound, use an attack closer to 7ms)
Release: 100ms on short bass sounds / 300ms on long bass tones
Knee: Hard
Gain: +2db to +4db
---

Brass / Wind instruments: Brass and wind instruments require a "transparent" type compression. Any obvious processing can noticably ruin the sound. Brass and wind instruments have a lot of variety in playing styles. Trumpets can be played expressivly loud, and a smooth, mellow flute will need much different processing.

Threshold: -2db to -4db
Ratio: 6:1 (lighter instruments) to 15:1 (deep brassy instruments)
Attack: 3ms (If a transient sound needs through uncompressed, use 6ms)
Release: 300ms
Knee: Hard
Gain: Varies

Guitars: When working with acoustic guitars, compressors tend to reveal themselves more so it's a good idea to use a very "transparent" compression. If working with electric guitars, make small increases to the ratio and threshold.

Threshold: -2db to -3db
Ratio: 3:1 to 4:1
Attack: 3ms (If there is an initial pluck, use 5ms)
Release: 30ms to 60ms
Knee: Soft
Gain: 0db to +1db
---
Samples: If you're a sampled based producer (specifically, phrase sampler), chances are you don't get to compress several instruments in different ways. Using the following numbers, you'll be able to smooth out the entire sample without too much limiting.

Threshold: -2db to -4db (If the sampl#e is recorded bad, and there's lots of peaks, use a higher threshold around -8db)
Ratio: 2:1 to 3:1
Attack: 2ms
Release: 400ms
Knee: Hard
Gain: +1db to +3db
---

Full Mix: The final mix doesn't require much compression, although some hip-hop songs have been compressed with up to a 4:1 ratio, most aren't needed that much. A final compression should act as a limiter, keeping the signal close to the 0db mark.

Threshold: -4db to -7db
Ratio: 1.5:1 to 2.5:1
Attack: 5ms
Release: 200ms to 500ms
Knee: Hard
Gain : Varies
__________________

Note that these arn't set in stone, theyre just starter points.

.peace.
 
nice post man, compression is the ultimate use efx in the studio beside eq.
 
mmmm great post. Really helped me. Thanks man

A quick question with the drums and percussion. Do you take your snare, hi-hat etc and compress them all together or do you compress each seperately?
 
i compress them all together.
sometimes my cymbal is quiet, and i'll maximise that seperately.
definitly experiment though.
 
Anyone care to post a starters guide for vocal compression...?
 
i'm bumping this up. good info, i gave you a rep point.
 
Maybe it's just me... but when I am mixing drums I do kick, snare, bottom snare, overheads, toms, etc. individually. I try not to overcompress things esp. snare and kick. I go more for a ratio of 4:1 to 6:1 with the threshhold more around -5 or so with the attack time between 5 and 10 ms and the release around 125 ms. If you have good trained ears, I feel these are good starting points for kick and snare especially. Maybe it's just me... If I am completely off please be sure to let me know though
 
vcgeetar said:
Maybe it's just me... but when I am mixing drums I do kick, snare, bottom snare, overheads, toms, etc. individually. I try not to overcompress things esp. snare and kick. I go more for a ratio of 4:1 to 6:1 with the threshhold more around -5 or so with the attack time between 5 and 10 ms and the release around 125 ms. If you have good trained ears, I feel these are good starting points for kick and snare especially. Maybe it's just me... If I am completely off please be sure to let me know though


i think it depends on your preference and the sound you are trying to get, i personally don't think that there isn any real blueprint to processing when it comes to creativity, but there is a common place that i notice a lot of engineers tend to share.
 
Well I am more inclined to doing alot of rock work esp. when it comes to my own bands and projects with friends. So I guess I am kinda trying not to get into the newer "mainstream" style mixing which you hear on rock radio stations where everything is overcompressed to all hell and there is no dynamics whatsoever.
 
vcgeetar said:
Well I am more inclined to doing alot of rock work esp. when it comes to my own bands and projects with friends. So I guess I am kinda trying not to get into the newer "mainstream" style mixing which you hear on rock radio stations where everything is overcompressed to all hell and there is no dynamics whatsoever.


yeah, i feel you there. it's just that the masses or moving toward the over compressed sound. from what i've learned it's all a competition on making songs POP. if the song isn't abused by compression yes, a music connesiuer ( i hope i spelled that right) will enjoy it, but these days louder is better. i guess no one likes to just turn the volume on thier stereo up anymore.
 
jugalo180 said:
i guess no one likes to just turn the volume on thier stereo up anymore.

Very true. If only I had an old 2" tape machine I could get mixes like one of my favorite engineers... Jay Baumgartner(check out the new helmet cd size matters... its REALLY loud yet not overcompressed) i think its just protools. since they cant get that loud badass sound of tape the overcompress and quantize everything so its no longer natural. its ashame dont ya think?
 
vcgeetar said:
Very true. If only I had an old 2" tape machine I could get mixes like one of my favorite engineers... Jay Baumgartner(check out the new helmet cd size matters... its REALLY loud yet not overcompressed) i think its just protools. since they cant get that loud badass sound of tape the overcompress and quantize everything so its no longer natural. its ashame dont ya think?


i agree, i am hoping to record a few songs for my next cd to 2". even though they will be mastered with the other pro tool tracks and quite possibly compromised by over compression, i hope to retain most of the benifits of recording to tape.
 
Now, I actually compress everything individualy and differently.
I have a totally seperate mixer for my drums, and apply compression to each drum as needed.
 
thanx yo...i kno i was asking for something like this when i asked what everyone was using lol....good to see ppl are lookin out for each other on here
 
The general ideas are pretty good, but a couple of things:

Threshold settings, it's hard to say, because they depend on the input level. If your threshold is at -5, a stronger signal will get too compressed, a weaker one not enough.

These seem like decent places to start, if instead of a certain threshold setting, you set the threshold high and dial it down until you are happy, and then use the make-up knob to bring your levels back up.
 
Back
Top