question about compressing a final mix/or the master channel

[a27thletter]

ok, when i get my mix sounding the way i want it and all and i feel that it needs a little taming with some compression where is a good place to start? i know its a very broad topic, and most of it makes sense to me but the thing thats confusing me is the threshhold. on a final mix, how do you know where you want the threshhold at? do you set it to hit the peaks? set it to start somewhere lower? sorry for the ignorance.

 

[legionserial]

What I usually do, if it's a case of just a bit of taming, is pinpoint where the major peaks are, decided roughly where I want to bring them down to, and play around with a compressor and a peak meter, until those peaks aren't peaking any higher than I want. Of course the factor of screwing up the overall sound comes in, which is where the tinkering around comes in. To try and keep it sounding good (ie not over compressing), whilst keeping the peaks down.

That said, I'm not exactly an accomplished user of compressors. My motto is usually "twiddle the knobs until you are clear on how each will affect the sound, then twiddle some more until you get the desired result". I'm not a big fan of assigning numbers to things, I just play it by ear...

In your case I would just suggest playing with the threshold until you know how it's going to affect the sound. I know it sounds lame, but I personally see more value in knowing your gear blindfolded, than I see in reading the numbers on the dial. Perhaps that makes me a 'tard, I dunno. I guess I'm more of a practice guy than a theory guy.

End of waffle....

 

[drbyatch]

threshold is where it will first start effecting it.

say your threshold is -5 dB, anything that goes over that -3 db or whatever will be compressed according too the ratio.

and yeh i agree with what legion.. it is also what i do. fidel.. sound cool. awsome... NEXT SONG...(said like jack black)

 

[SouthSIDE Glen]

ok, when i get my mix sounding the way i want it and all and i feel that it needs a little taming with some compression where is a good place to start? i know its a very broad topic, and most of it makes sense to me but the thing thats confusing me is the threshhold. on a final mix, how do you know where you want the threshhold at? do you set it to hit the peaks? set it to start somewhere lower? sorry for the ignorance.

I'll leave it to the real ME's (sending out the bat sigmal to Tom, John and Brad ) to give definitive answers - as well as correct mistakes in this post. But I'll try to get the conversation going beyond just "do what sounds good" .

1. Get the mix right
Leave as little to the premastering stage as possible. If individual tracks are peaking too hard or the spectral balance is uneven, try to corrct those before or at mixdown, not after. Try not to fall into the habit of doing in the mastering what can be done in the mixing. Intentionally fixing the mix in mastering is as mis-targeted as fixing tracking in the mix; it should be done only as a last option or when the earlier option is not economically or temporaily feasible.

2. Taming wild peaks
When excess peaks in the 2mix cannot be avoided, then try to tame those first before trying to tighten up and gain push the whole mix. If you are using a visual digital editor, try to tame the wildest peaks via manual edits first rather than blanket compression or limiting; at least try to get the overall crest factor fairly even through the song before applying blanket compression or limiting.

3. Tightening the mix
Once the worst peaks are brought in line with the rest, try applying light compression to what's left in stages or multiple coats instead of one big thick leayer of heavy compression or limiting. Try setting the threshold to somewhere just above the RMS level first, and apply some light compression somehwere in the 1.5:1 - 2:1 range. Then go back, drop the threshold a couple of dB, and apply a second coat of light compression below 2:1. Repeat one more time, if necessary(use your ears to figure how many coats you want), another couple of dB down on the threshold and apply a third coat in the 1.2:1 - 1.4:1 range (give or take).

4. Pushing the gain
Once you have a fairly tight, good sounding package, you can adjust the overall level.

Here it depends upon whether you're creating a single or assembling an album. In the case of an album, you'll want to adjust the levels so that the *apparent volume* - i.e. how it actually sounds to the human ear and not what any meters or measurements or visual cues show you - matches the final apparent volume of the quietest (least dense) track on the album. In the case of a single, you also what to use your ears and not your eyes, but this time you might want to get a level that balances loudness with quality regardless of surrounding songs.

If steps 1-3 have been followed properly, you may only need to use standard peak normalization at this point. Before anyone lights their flamethrowers, remember the song has already had one manual and 2-3 programmed coats of dynamics control, further flattening is often not actually necessary.

However, as most folks here just won't accept that fact, then maybe a final push pass through a hard limiter just to squeeze an extre couple of dB of RMS out f the song wuld placate them. But keep it easy and keep your ears tweaked; rarely should more than 2 dB of extra gain reduction via limiting be necessary before the sound starts getting grainy and starts "falling apart".

5. Use guidelines, not recipes
While the order of steps above and thei general principles behind them are pretty good guides IMHO, they are not hard-and-fast rules or recipies that apply to all occasions. Use your ears, apply only the number of coats that are necessary, and don't push any harder than you have to. Vary the settings as necessary as determined by your own ears, and ignore what any numbers are telling your eyes.

Mixes are like pieces of shaped wood, the key is to pull the natural grain and beauty out of them, and each piece will have it's own grain and other characteristics requiring it's own degree of sanding and varnishing. Just planing them down and slathering on a thick coat of latex paint so that all pieces look the same may be easiest, but it gets real boring to look at real quick.

IME, YMMV, IMHO, 5 cent depost in MI, void where prohibited by law, etc.

G.

 

[Massive Master]

Follwing up with SSG's post - If the whole mix is "nice" but there are freakish peaks going on, you can either (A) leave it to the mastering guy OR even better, (B) find out WHAT is going on and fix it at the track level. There's nothing wrong with taming dynamics as early as possible in the chain (track, group, etc.) before trying to compress the main buss.

 

[djclueveli]

no one metioned anything about attack and release setetings too. isn't that a big factor when compressing?

 

[SouthSIDE Glen]

no one metioned anything about attack and release setetings too. isn't that a big factor when compressing?

You're right, they are very omportant variables in the use of compression. But it's difficult to describe any kind of guidelines here with regard to the topic of general mastering because A & R are more specific task-dependent controls than anything else. Each "coat" of premastering compression will be affecting content with different kinds dynamic characteristsics, as will ones applied to individual and differing tracks before the mixdown, so while general guidelines regarding gain reduction can be given, guidelines regarding the A&R settings will change dependant upon the specific task at hand.

Just follow the general rules of A&R usage depending upon whether you're trying to mostly tame peaks or thicken the gumbo at any stage in the game, and adjust to keep pumping and breathing to a minimum. If you're not sure of the basics of A&R usage in that regard, check out the compression tutorial in my signature line.

G.

 

[a27thletter]

very helpful information, thanks to all of you!

 

[mixsit]

no one metioned anything about attack and release setetings too. isn't that a big factor when compressing?

Yes a huge one. IMO perhaps more impact on the compression effect than ratio. So where to begin. How about handles.
Get 'thine ears wrapped around the difference between 5, 10, 25, and 50-100 ms. Start with a modest ratio -say 3:1, and a fairly fast release (How about 100ms) on a kit or a raw mix. Notice what gets through. Think about what is poking up in the mix, it's shape, and how it compares to the time setting. And how the gain reduction is connected to the time vs the shape. Peaky shapes have larger swings, grab hard set fast' (relative).. Expand the idea out from there.

Move on to release. Again, the shape, or in this case often tempo', is relative. For a given amount of gain reduction a faster release can also tend to allow more forwardnesses. So in some ways release is also a presence control.

 

[masteringhouse]

All of the above is good and valid info, including twiddling with the knobs until it sounds right.

For setting threshold I often start with a high ratio (10:1 or more) and a fast attack. Then bring the threshold down until you start to see a reduction. This is where your peaks are (top threshold). From there it depends on what I indented on using the compressor for in the first place. If it is to tame peaks, adjust the attack and release so that you aren't adversely affecting things like the kick, snare, or other transients. Usually somewhere around 50-75 ms gets out of the way of these, but in some cases you may want to tame hits a bit more, so a faster attack is required. Then listen to how long you want the comp to "grab" or "fatten" and adjust the release so that it doesn't pump, distort, or make the track sound "lifeless".

If going for an overall comp, say to add some density to a track, reduce the ratio to somewhere around 2:1 or 4:1 and lower the threshold (after you find the peak) to somwhere around 3 db of reduction. From there adjust attack so that the comp allows the track to "breath" and gets out the way when it isn't needed. Release time are often in the 200-300 ms range for this. From there "rinse lather and repeat".

There are no "silver bullet" settings, just keep honing your compression skills much like you would practicing an instrument.

 

[TerraMortim]

ok, when i get my mix sounding the way i want it and all and i feel that it needs a little taming with some compression where is a good place to start? i know its a very broad topic, and most of it makes sense to me but the thing thats confusing me is the threshhold. on a final mix, how do you know where you want the threshhold at? do you set it to hit the peaks? set it to start somewhere lower? sorry for the ignorance.

not exactly your question but just another word of advice to do with master bus compression. I always find it's easier to get a better result if you put the compression and all of that on before you start the mix..and mix into it the whole time, instead of getting mix right and throwing compression on it it after. It just seems to gel things in easier, and there will be no suprises by throwing a compressor on an allready perfect mix.

 

[Middleman]

not exactly your question but just another word of advice to do with master bus compression. I always find it's easier to get a better result if you put the compression and all of that on before you start the mix..and mix into it the whole time, instead of getting mix right and throwing compression on it it after. It just seems to gel things in easier, and there will be no suprises by throwing a compressor on an already perfect mix.

I agree about mixing into a buss compressor. The challenge with this approach is getting the compressor to breath in time with the music. Until you establish the pulse of the track, you don't have an accurate bead on attack and release. Not for all types of music however, but I generally wait until the basic framework is up i.e. drums, bass and at least one chordal track, perhaps a vocal rough. Then I go after the mix buss compressor and match the pulse of the song. This helps later when implementing drum compressors for kick, snare and overall drum buss. With a little mix buss compression action I tend to go lighter on the individual drum compressors. A little more clarity is the ultimate result. This really helps if you are ITB because all plug ins smear.

Across the board, because compressors can color things, the mix buss compressor at the right time keeps you from chasing a dead end on EQ and individual instrument compression. These can build up in a bad way if you wait until the end to add the buss compressor because the buss compressor can take the EQ in a direction which is not desirable. I'm talking mainstream music here not for classical or jazz unless you have a really great transparent compressor or limiter and highly dependent on what you are trying to convey.

In addition to the above, tracking the bass with full compression but always have a clean DI backup is a safety valve, the same for vocals). You generally should have a vision of the way the song will go at mix time but keeping duplicate clean versions of things is kind of a tracking backup mentality gives you some options. Possibly a little anal from an engineering perspective but musicians change their minds and it's good to have some options come mix time.

The approach I take, loosely based around the Michael Bruaer approach, is to move backward into the mix from the two buss down to individual tracks. In most cases this creates a build up of compression which is lighter at each stage and more controlled than starting with track compression and moving towards the main buss. I just found that trying to apply track compression first, especially ITB, you end up process loading your plug ins and chasing the ugly, filtered mix.

Another approach is to have a "mixers out" around the buss compressor to A/B things. If I need more clarity outside the mix for a vocal or solo, I keep an extra non-compressed buss for experimentation. Compressors are so addicting your mix may need medical help. This approach keeps a path back to reality.