Just a question about Compression Ratio.

[bouldersoundguy]

Yea. Thinking about it I'm finding that distorted guitars generally sound pretty equal, I guess that this is because you generally play quite hard and steady due with distortion because you want to deliver impact?

Distortion is dirty clipping. It chops off the high parts of the wave and fills in with harmonics. The natural result is a less dynamic signal.

I've often got clean intros and stuff that have more dynamics and they need a little bit of compression definitely.

Compression or level automation can work, depending on various things.

Although I guess sometimes people would want to add compression to add the effect that it can give too maybe?

People use it as an effect all the time. The line between controlling levels and compressing for effect is rather blurry. Change it from the way it is to the way you want it to be.

 

[PhilLondon]

So what about a slightly driven tone, like edge of breakup? That still has a fair amount of dynamic range right?

 

[ashcat_lt]

It chops off the high parts of the wave and fills in with harmonics.

I really don't like this statement. It brings to mind a picture of a wave form with the high parts clipped off and the areas around the zero crossings "filled in" with something that wasn't there before, and that is completely not what happens in distortion. The "added harmonics" are exactly a result of the clipping and happen while the signal is clipping. The portion of the wave which does not attempt to pass the clipping ceiling is not affected at all - it's exactly the same as what went in. The reduction in dynamic range comes completely from the fact that the peaks aren't allowed to go as high as they otherwise might. Yes, it's kind of OT, and kind of nit-picky, but...

 

[bouldersoundguy]

I really don't like this statement. It brings to mind a picture of a wave form with the high parts clipped off and the areas around the zero crossings "filled in" with something that wasn't there before, and that is completely not what happens in distortion. The "added harmonics" are exactly a result of the clipping and happen while the signal is clipping. The portion of the wave which does not attempt to pass the clipping ceiling is not affected at all - it's exactly the same as what went in. The reduction in dynamic range comes completely from the fact that the peaks aren't allowed to go as high as they otherwise might. Yes, it's kind of OT, and kind of nit-picky, but...

I agree and debated wording it like that. But it seemed a reasonable way to get the message across. Distortion doesn't add anything where there is no clipping. I was thinking more in spectral than temporal terms.

How about this: The clipping causes harmonics that redistribute some of the energy from lower frequencies to a range of higher frequencies. By spreading the energy across the spectrum it sounds less loud and so the difference between the loudest parts and the quieter parts is reduced.

 

[PhilLondon]

I think this was missed.

If you have double or triple tracked guitars would you compress each track individually or do it via a send to a compression bus? I tried this with an EQ bus and it didn't really do anything, I'm not sure if you can even do it.

Do you add compression to the mix bus when before you master it, or are you talking about when you're mastering? Or could compression get added to both the mix bus and when it gets mastered, or either?

 

[ashcat_lt]

How about this:...

IDK. I've never really thought about it in those terms, and I'm not sure how technically accurate it is. I now what you're saying, but I think it's just making things more complicated than necessary. (Yes, RAMI, I know...pot...kettle...) It really is as simple as the fact that distortion doesn't let the loud parts get as much louder than the quiet parts as it otherwise would, so the dynamic range is reduced. The harmonic content is a byproduct of that action. Essentially, hard clipping (real serious distortion) is exactly the same thing as a compressor with 0 attack, 0 release, and 0 RMS time. In fact, one might say that attack, release, and RMS integration are ways to make a compressor work without sounding like distortion.

 

[bouldersoundguy]

I think this was missed.

If you have double or triple tracked guitars would you compress each track individually or do it via a send to a compression bus? I tried this with an EQ bus and it didn't really do anything, I'm not sure if you can even do it.

Whatever works. Over the last few years I've found myself using less bus processing whether it's a submix bus or the main bus.

Do you add compression to the mix bus when before you master it, or are you talking about when you're mastering? Or could compression get added to both the mix bus and when it gets mastered, or either?

I save overall compression for the mastering stage. Some people like to mix into a compressor. It's personal preference. Either way some form of dynamics processing will probably be used in mastering, maybe even several.

 

[RAMI]

Distorted guitars almost never need compression. They're already compressed. If you're talking clean guitars, then Boulder has the answer above: "Whatever works".

 

[ashcat_lt]

I think this was missed.

If you have double or triple tracked guitars would you compress each track individually or do it via a send to a compression bus? I tried this with an EQ bus and it didn't really do anything, I'm not sure if you can even do it.

Usually, when we say "send", we're talking about parallel processing, like we might do with a send to a reverb or something. You end up having the original signal from it's original mixer channel plus the signal from wherever the send goes. This is almost never appropriate for EQ, and while it is used for certain specialty applications of compression, doesn't really work the same as series compression. IDK for sure that you actually meant "send" in that context though. If by send, you mean something more like "assign", so that the only way the guitars are heard is through the bus itself, then you're closer to the right track.

I personally feel like it's best to control dynamics at several levels in more complex arrangements. Control the individual guitars (distortion/overdrive tends to do it for you, cleaner guitars might need automation and/or compression), then maybe bus them and squish them together a bit, then mix that with the rest of the band (the other instruments have been treated similarly as necessary) and squish the whole thing together a bit, take care of any "aberrant peaks" that come from multiple instruments accidentally pushing the same direction at exactly the same time, and general "finishing".

Do you add compression to the mix bus when before you master it, or are you talking about when you're mastering? Or could compression get added to both the mix bus and when it gets mastered, or either?

Yes, any or all of those, as necessary/desired/preferred.

 

[Pinky]

Do you add compression to the mix bus when before you master it, or are you talking about when you're mastering? Or could compression get added to both the mix bus and when it gets mastered, or either?

I tend to get the track/instrument mixed to the point that additional pre-mastering compression on the mixed bus would end up just interfering with the mastering. Yes, it can be done as others have said, but it's stealing dynamics before the often overly aggressive mastering compression has its chance to steal them. lol

 

[PhilLondon]

Thanks for the info guys. It's very valuable.

I feel that I'm getting better, as people are telling me that most of the things they have issue with are just preferential things like the sound of the kick or whatever. In terms of levelling and EQing and (guessing) compression settings I feel like I'm getting somewhere.

Ive been leaving my mixes for a week or so and coming back and now I'm not finding any issue with how they sound in terms of level and unwanted frequencies, so in a month or so I might start thinking about getting a mastering house to master the better ones.

Any advice on what I should do before I send them to them? I'm fully aware that I'm going to come across very new at it to them, but any advice to soften the blow is dry much appreciated. I'm also aware that my mix could result in a poor master, so I know that I'm taking a risk.

Thanks.

 

[PhilLondon]

I guess that question is for another thread!

Anyway, thanks a lot, it's been very helpful. Watch out for my upcoming reverb thread!

 

[RAMI]

Sorry Phil. I'm sure I'm not the only one that missed your last post.

To answer your question, when you send your final mix to a master-bater, just make sure your mix is as good as you think it can be. Don't have that thought in the back of your mind that any flaws will be fixed by mastering. Put nothing on your master bus like compression or anything else. Leave headroom volume-wise. You shouldn't be going near clipping at all. Peak at about -6 or even -10. Put it this way, it can be too loud, but it can almost never be too low.

 

[Bobbsy]

I just "get a feeling" so I'm going in sideways here --

IF you're trying to keep the VOLUME of certain sources within a certain area, the first line of defense in that case is the fader / fader automation.

Just putting that out there... Just in case...

Fader automation is just "intelligent" compression with continuously variable threshold and ratio!

I must admit I tend to apply light compression to most tracks to get "in the ballpark" then use automation for my final tweak to get the balance just right (or at least my version of just right).

Anyhow, another good resource to learn about compression are the Rane notes on the topic available HERE.

Finally, yes, use your ears to make the decisions on compression...but understanding the numbers can really speed things up by letting you get close for the first try rather than randomly tweaking three or four inter-related controls until you like the sound.

 

[rayc]

The PDF is easier to work with if you can't spend ages readinga a computer screen...
http://www.rane.com/pdf/ranenotes/Dynamics_Processors.pdf

 

[PhilLondon]

Thanks a lot guys. I've got a little more to do then I'll ask for some recommendations on a mastering house. Thanks.

 

[PhilLondon]

Sorry to bring this up again. I've had some time to look at some of my mixes. I understand it in theory, but I don't get it in practical terms at all.

I want to compress the vocal. The level of the vocal is around 20-15. I set the threshold to 20, the ratio to 5:1, the attack to something quick like 10ms, and the release to 500ms. It doesn't do anything, in fact the signal is now louder without any gain added! it will only kick in if I put the threshold up to 10 and increase the ratio drastically. None of the signal goes to 10db, so how come it works?

When I want to affect the drums it doesn't do anything unless the threshold is lower.

If I just fiddle around with it I can get the compression working, and I can hear when it starts to mess it all up and then I back off, but I don't know if I'm adding way too much compression or not.

What am I doing wrong here?

 

[mixsit]

Firstr off let's clear these up..

... The level of the vocal is around 20-15.

You mean -20 to -15(dbFS just to be clear.

I set the threshold to 20, the ratio to 5:1, the attack to something quick like 10ms, and the release to 500ms. It doesn't do anything, in fact the signal is now louder without any gain added!

'Without any gain added' as in you didn't turn it up somewhere else-- But how about 'Auto Make-up Gain on the comp? If so, turn it off!

... it will only kick in if I put the threshold up to 10 and increase the ratio drastically. None of the signal goes to 10db, so how come it works?

You mean put it up to -10? That's higher threshold and equals wanting less compression. Leave the ratio alone (5:1 whatever) and lower the threshold (HIGHER NEG NUMBERS) untill you see and hear some reduction. Regardless of the numbers that is where that comp is kicking in.

 

[mixsit]

"Leave the ratio alone", get rid of 'auto functions'. In other words get one function (threshold in this case) under your belt before confusing things by changing too many things at once.
Also.. check to make Shure the comp doesn't load with an output boost by default- A Stupid Pet Trick they'll put in to make you think the plug sounds Louder = Better!

 

[PhilLondon]

Firstr off let's clear these up..

You mean -20 to -15(dbFS just to be clear.

Yes, sorry.

'Without any gain added' as in you didn't turn it up somewhere else-- But how about 'Auto Make-up Gain on the comp? If so, turn it off!

That's off. Yes.

You mean put it up to -10? That's higher threshold and equals wanting less compression. Leave the ratio alone (5:1 whatever) and lower the threshold (HIGHER NEG NUMBERS) untill you see and hear some reduction. Regardless of the numbers that is where that comp is kicking in.

Yea, so I'll have the threshold set to -20 and nothing, then to -10 and it does, then I was trying to add compression to a snare and it only became compressed when I lowered the threshold, back to -30, -40. I don't get it. Maybe the volume was too low?

Are you kinda saying just set the ratio to not too extreme a number and then play with the threshold, and the compression will kick in and I'll hear it and can go by ear, regardless of understanding it?

"Leave the ratio alone", get rid of 'auto functions'. In other words get one function (threshold in this case) under your belt before confusing things by changing too many things at once.
Also.. check to make Shure the comp doesn't load with an output boost by default- A Stupid Pet Trick they'll put in to make you think the plug sounds Louder = Better!

Nah all that's off. Thanks.