How's my mastering chain?

[jeffgowins3]

Stereo seperator (do I even need this? I swear I can't hear one of this on most pro recordings)
EQ
Compressor/Limiter

[Adam!]

Seems fine. I like:

1. EQ (Usually a steep highpass @ 20 hz, a subtle lowpass at 18khz, and a bass boost around 60-120hz)
2. Tape Saturation Plugin (I use Stienberg's Magneto--old, but good--and aim for 3db reduction on only the loudest drum hits) to catch any really crazy transients
3. Compressor (4:1 ratio, soft knee, 10ms attack, longest release possible, and never more than 1-2 db attenuation)
4. Brick-wall Limiter (fastest release possible, 3-4 db peak attenuation max)
5. Clipper (1-3 db of gain. Most VST clipping plugins will let you choose the "hardness" of the clipping; I like to use the hardest setting that I can get away with)
6. Master Fader set to -0.3 db, to keep the peaks from hitting fullscale and making your CD player flip-out.
7. 16-bit dither

You'll notice that instead of using one compressor or one limiter to do 6-9 db of gain reduction, I split it up between multiple dynamics processors. I find doing it this way... um... awesomer.

[Billy's Little Trip]

Puce, you are a plethora of studio info. I'm happy knowing that I'm in the ballpark of db settings you described.

Let me ask you this. I just started setting my lows, like 100hz and under, to mono and widening my stereo highs a bit more. It sounds good to me like that. What do think about this practice? I did it on my "Let Me In" song. I feel I'm getting so close to what I like, but just not quite there.

By the way, I do know that if I ever do a song that has serious left and right panning of the kick and low tones, I can't do the mono lows thing.

[Adam!]

I just started setting my lows, like 100hz and under, to mono and widening my stereo highs a bit more. What do think about this practice?

I think that is a solid business plan.

[Kill Me Sarah]

Seems fine. I like:

1. EQ (Usually a steep highpass @ 20 hz, a subtle lowpass at 18khz, and a bass boost around 60-120hz)
2. Tape Saturation Plugin (I use Stienberg's Magneto--old, but good--and aim for 3db reduction on only the loudest drum hits) to catch any really crazy transients
3. Compressor (4:1 ratio, soft knee, 10ms attack, longest release possible, and never more than 1-2 db attenuation)
4. Brick-wall Limiter (fastest release possible, 3-4 db peak attenuation max)
5. Clipper (1-3 db of gain. Most VST clipping plugins will let you choose the "hardness" of the clipping; I like to use the hardest setting that I can get away with)
6. Master Fader set to -0.3 db, to keep the peaks from hitting fullscale and making your CD player flip-out.
7. 16-bit dither

You'll notice that instead of using one compressor or one limiter to do 6-9 db of gain reduction, I split it up between multiple dynamics processors. I find doing it this way... um... awesomer.

These are all on the master, yes? Do you also EQ, Compress, Limit, etc. each individual instrument/track? Or only the master?

[Hoblit]

I like to use hardware compression on my guitar on the way in. I get better sustain on my guitar and I can keep the general volume lower and still hear all of the guitar.

I also add a little, LITTLE compression to the drum tracks to 'beef' them up just a little so most of the tones come through and don't get lost in the frequencies of the other instruments.

PUCE is a WAY better producer than I am and I don't have enough time to put into my work to get that much better. Thats why I'm here, learning from Puce myself. I'm mixing and mastering down six new Ghost Town Gridlock songs and I'm trying to pick up important info on the way.

[Adam!]

Do you also EQ, Compress, Limit, etc. each individual instrument/track? Or only the master?

I will often compress and EQ individual tracks, unless they don't need it. A couple weeks ago Jolly Roger posted the source files to a song he had recorded, and Deshead, BLT, and I mixed it and posted what we did to each mix. I posted a long essay about what I had done, which I think is pretty representative of how I might normally treat individual tracks in a mix. I think JR has taken down his original mix of the song, but the other ones are still up.

Anyway, your gateway to all this can be found right here.

[Kill Me Sarah]

Do you also EQ, Compress, Limit, etc. each individual instrument/track? Or only the master?

I will often compress and EQ individual tracks, unless they don't need it. A couple weeks ago Jolly Roger posted the source files to a song he had recorded, and Deshead, BLT, and I mixed it and posted what we did to each mix. I posted a long essay about what I had done, which I think is pretty representative of how I might normally treat individual tracks in a mix. I think JR has taken down his original mix of the song, but the other ones are still up.

Anyway, your gateway to all this can be found right here.

Awesome, I'll refer to this as much as possible when I mix our next Nur Ein song... you better not be feeding me misinformation!

[Renwick]

Let me ask you this. I just started setting my lows, like 100hz and under, to mono and widening my stereo highs a bit more.

I'd like to try this mono lows thingy but i'm not really sure how to do it. Does it just have to do with panning? Any tips would be appreciated. Oh yeah, I'm using Cubase LE. Thanks!

[Adam!]

I just started setting my lows, like 100hz and under, to mono and widening my stereo highs a bit more.

I'd like to try this mono lows thingy but i'm not really sure how to do it. Does it just have to do with panning?

The way you'd do it at the mastering stage involves Mid-Side processing, which is a big ol' can o' worms. It's probably easier to get the same result in the mix by panning the kick, snare and the bass to the center (which I'm assuming you do already) and then highpassing every other track (excluding the toms) at 150hz or higher.

[Kill Me Sarah]

I don't know if this is the right thread for the question, but it's recent and all the right eyes seem to be watching it

I'm feeling a bit overwhelmed as I delve into some of this mixing/mastering stuff. It's like I start to peel away layers and the layers keep coming. Since my intent is to gradually improve, I was wondering if someone might make a recommendation as far as how I should go about learning step-by-step. Sort of a self-followed curriculum as it were. I think this question is longer than it needs to be but what I'm asking is. What is the first, easiest thing I could learn to do during mixing that will have the most obvious impact. Once I've mastered that, then what? And so on? Just a kick in the right direction would be great.

[ken]

I showed Martin how to use highpass filters the other day and it excited him to no end. Here's what you do:

Apply an EQ that eliminates low frequencies on the majority of your tracks, leaving only the kick drum in the low end. For example: Guitars - everything below 100 gone. Vocals - everything below 250 gone. Bass - everything below 60 gone. See how much you can get away with cutting out before your track doesn't sound good anymore. Suddenly, you can hear the kick drum and bass.

Ken

[roymond]

Apply an EQ that eliminates low frequencies on the majority of your tracks, leaving only the kick drum in the low end. For example: Guitars - everything below 100 gone. Vocals - everything below 250 gone. Bass - everything below 60 gone. See how much you can get away with cutting out before your track doesn't sound good anymore. Suddenly, you can hear the kick drum and bass.

This post alone makes me want to get back into recording asap. I always lose my bass and kick.

[Reist]

I think JR has taken down his original mix of the song, but the other ones are still up.

I don't think I took those down ... I should check.

EDIT: Looks like I moved it. It's fixed now.

[Kill Me Sarah]

Okay, that all sounds straightforward enough, except I don't know how those numbers relate to what I'm looking at

This is what my EQ plugin looks like

[Billy's Little Trip]

I showed Martin how to use highpass filters the other day and it excited him to no end. Here's what you do:

Apply an EQ that eliminates low frequencies on the majority of your tracks, leaving only the kick drum in the low end. For example: Guitars - everything below 100 gone. Vocals - everything below 250 gone. Bass - everything below 60 gone. See how much you can get away with cutting out before your track doesn't sound good anymore. Suddenly, you can hear the kick drum and bass.

Ken

I never really thought about this, but it makes perfect sense. I often find myself cutting lows out of my guitars because it kills the bass. So I guess I was on the right track, I just need to be more aggressive about it.
Great info Ken, thanks.

[Billy's Little Trip]

KMS, doesn't your recording program have some kind of on board EQ?

Here is what my Cubase LE looks like. The EQ is in the lower right of this picture.

[signboy]

I was shown a great trick for the bass & kick specifically:

find the frequency that the kick is centered in by moving an eq spike around while listening to just the kick.
when you've found it, leave just enough of a spike to accentuate the kick without making it sound funny.
then put a notch in the bass eq at the same frequency.

this keeps the bass out of the kick's way, and keeps it from sounding muddy.

[bz£]

Forgive me if this seems a little basic. KMS' picture is a graphic EQ. These are popular in consumer grade equipment, partly because they are relatively simple to use, but they are not particularly useful in a recording studio environment. (*)

What you really want is a parametric equalizer. They are harder to use effectively but you need to learn what you can do with them. It is well worth a little effort. A parametric EQ has, typically, four controls: type, centre frequency, Q and dB.

There is probably a better name for "type," but what I mean is a choice between (typically):
- high pass
- high shelf
- bandpass
- band cut
- low pass
- low shelf
This lets you choose the basic "picture" of the EQ curve. A high "pass" eq lets everything above the centre frequency go through unchanged, but stops low frequencies, and a low pass is just the opposite. A high "shelf" applies a cut to everything belowe the centre frequency, but that cut is the same for all the lows. Bandpass is a combination of high pass and low pass: it cuts everything far away from the centre frequency (in both directions).

"Q" is a measure of how sharp the EQ curve is. It's hard to visualize, but a Q of about 0.7 is a reasonable number for a baseline. Lower Q means a very sharp curve: a band cut at, say, a centre of 60 Hz and a Q of 0.1 is a good way to remove hum from your power supplies (in America) without affecting nearby frequencies too much.

The "dB" measurement is just how much you cut or boost. As a rule of thumb, don't boost more than about 3 dB unless you really can't help it; stronger cuts are usually safer. High (and low) pass filters don't have this; the drop-off is a function of the filter (you'll often see something like 12 dB per octave).


Finally, most parametric EQs are either three- or four-band. Each band is a separate equalizer, but often you won't have the full range of choices for each band. Typically you will have a low-pass, a high-pass and two band-pass/cuts.


I'm sure if you look around online you can find more details about how these things work. A good place to start is to find any post by deshead, read it carefully and then follow whatever links are in his signature. Hell, if I was sober I would just direct you to him without wasting my time typing any of this crap.


So what's that all mean?

When Ken says "cut everything below 100 Hz" he means you want a parametric EQ with a high-pass filter centered around 100 Hz. The end result of this is that sounds at 100 Hz will get about a 3dB cut, and everything below 100Hz is going to be cut significantly more.

And, why is this okay (and even good)? A typical electric guitar, in standard tuning, has a fundamental frequency of around 81 Hz. That's the open low-E. (Almost) everything below that is just noise. Sound energy is also a function of frequency; lower frequencies have much more energy. So it's important to get rid of all this junk that you don't even hear because it is silently dominating your mix. The results may surprise you; you'll be able to turn up the volume on your tracks without overloading anything and your song will sound both louder and better.


You can simulate this effect on your graphic eq by setting the lowest two bands all the way down, and maybe lowering the third band a little, but it's not really ideal.


Finally: signboy: another way to accomplish this is to sidechain the kick into a compressor on the bass track, so the bass is compressed more when the kick is active. This is (prehaps surprisingly) easier to do in hardware, because a lot of software compressors don't have sidechain inputs.



(*) They're okay; I've used graphic EQs for recordings before. I'm not saying they are useless, just that a parametric is more likely what you need.

[Kill Me Sarah]

Forgive me if this seems a little basic.

You could not possibly get too basic for me. This is very helpful. So what Ken would be referring to is something like this?

[bz£]

Exactly.

What I called "Q" is "Bandwidth" in your picture. Same thing.

[Kill Me Sarah]

Exactly.

What I called "Q" is "Bandwidth" in your picture. Same thing.

Awesome, thank you. Now I'll work my way through and see if I can make sense of the rest of it.

I actually dreamed about EQ last night. I coudln't get it shaped right in my dreams either but in my dream I was moving flocks of birds in the shape of the frequencies I wanted, so it at least looked pretty.