"Mastering" limiters are forcing me to change the way I mix.

[NashBackslash]

It was usually about making a good mix that's tonally pleasing, has plenty of punch (especially on the bass, kickdrum and snare), without comparing your mix to other commercial CDs (especially loudness-wise) or anything like that. As long as it translated well in anything else other than my studio's monitors, it's good.

Then the stereo tracks would be brought elsewhere for a real mastering engineer to finalize, or if the client absolutely wants a little more loudness out of the mix, I can probably slap a generous amount of limiting on it and see how it goes.

But when the metal guys come in, it has to be LOUD LOUD LOUD. This was very challenging for me because no matter how well were the tracks recorded, no matter how good the mixing was - it would all be ruined when the "mastering" limiter is pushed to the edge.

Hihat jumps out like a motherfucker, the "click" on the kick drum disappears, guitars end up masking every other instrument, any ambience and reverb - especially on the drums - makes the mix sounds like it came from the 80's. But what I'm really concerned about is the SNARE. I spend a great deal of effort and time recording a great sounding snare. I make sure the compression on it really makes it smack the listener. But with the limiter at 12 dB of gain reduction, the snare disappears. It sounds gay and mushy and crappy and not even the world's best drum samples loaded into Drumagog can save it.

Maybe there's some secret lost art to using the limiter that no one wants to pass around. I've tried all kinds of release settings but none of them can save my mix's punch when the loudness is 12 dB (or more) louder.

"So just turn down the limiter" yes good idea! Problem is, their mix won't sound like Trivium or Lamb Of God. No offence to those bands, because I am especially impressed with Jason Suecof's work on the latest two Trivium records. Very punchy snares.

I'm not going to argue with them about the loudness war. I think talking about that is like pretty much beating a dead horse by now. They are paying me money and putting their trust on me to deliver them what they want. On top of that, these guys are great; they can play. It's just that they want a really LOUD CD. So I'll do whatever it takes to make sure they get it.

So what do I do? I mix with the limiter on. Yes, before anything happens at all, I insert the limiter plugin into the Master bus. I used to use some "go to" freeware limiters like TLS Pocket Limiter, BuziMax 3 and Kjaerhus' Classic Limiter, but a few months ago I purchased Kjaerhus' MPL1. Unfortunately, it didn't seem to me that it would fix my lack of punch problem but I did notice that the MPL1 produces less artifacts on my mix than the freeware ones.

Sometimes when I feel like it, I will even TRACK (and monitor the signals through the DAW; after all my computer can handle it) with the limiter at the Master bus just to get a better picture of how will the final track sound like.

This has forced me to do things differently than what I am used to do as far as balancing the instruments go. The drums, especially the snare will be waaay in front. Then the bass tucked under the drums. The guitars will be pushed back A LOT. I also mix a lot LESS reverb than I am used to (which has always been just subtle, now it has to be even more subtle!).

As for equalization, I find that when mixing with the limiter on, I have to cut down a little more low end than I am used to otherwise the limited mix will just sound smudgey.

It'll sound really funny when the limiter is disabled but I really can't think of any other way to retain the drum punchiness when the limiting is extreme.

So does anyone have any similar experiences they'd like to share? I'm particularly interested in hearing from the real mastering engineers here; do your expensive hardware limiters really just have that magic bullet to make a mix loud and punchy at the same time?

[chessrock]

When people tell you they want it louder, just turn up the volume and say "Is that better?"

.

[LogicDeLuxe]

Originally Posted by NashBackslash

Maybe there's some secret lost art to using the limiter that no one wants to pass around.

Yes, there is. It says: never ever let it reduce about 12 dB of every beat. If applied to a degree that it only limits some of the hits, it usually will work rather transparent. If someone wants it LOUD LOUD LOUD, show him the video in my signature and let him rethink.

I'm not going to argue with them about the loudness war. I think talking about that is like pretty much beating a dead horse by now.

Yeah, that's where such a demonstration video comes in handy.

It's just that they want a really LOUD CD. So I'll do whatever it takes to make sure they get it.

If quality doesn't matter, smash it through a multiband compressor, than apply a tape saturation effect (which reduces some dB), and if it's still not loud enough, just let it clip a few dB after the master limiter is applied. Red Hot Chili Peppers "By The Way" looks and sounds pretty much like mastered in such a way.
You really should convince him that sound quallity DOES MATTER, though.

So does anyone have any similar experiences they'd like to share?

Actually not. Everyone here I ever did a master seems quite happy with a master at -14 dBFS RMS sine. Sometimes, it takes some explaining what's going on, but it sure is worth the time.

Originally Posted by chessrock

When people tell you they want it louder, just turn up the volume and say "Is that better?"

I like that idea.
Too bad some customers seem to be too boneheaded to turn up the volume on their own amplifiers.

[SonicAlbert]

Originally Posted by chessrock

When people tell you they want it louder, just turn up the volume and say "Is that better?"

Seriously, I've seen people do that. The other time to do that is when the client wants the mix "warmer". Louder sounds warmer too.

The problem comes when you put the track on the radio and it's softer than the tracks around it. When that happens, the soft track really does seem smaller than the others.

that happened to me on my first album, which got a lot of play on independent stations and internet radio. I happened to hear a stream of it on the internet, and my tracks (which were not limited at all and sounded great) were softer than the surrounding tracks. That's the last time I'll ever let that happen. You simply *have* to match whatever the current levels are, just try to do it better than the hacks that are flattening tracks to oblivion.

I personally would suggest mixing the tracks without the limiter, get the mix you want and then print it. The put the limiter on and do the volume thing. If you have to change the mix, then so be it. But you already have archived a non-squashed mix that you can go back to, and that you can use as comparison.

The other trick is to put the mix through two or more limiters in series, limiting a little with each one. Rather than slamming it all with one limiter.

[NashBackslash]

Originally Posted by SonicAlbert

The other trick is to put the mix through two or more limiters in series, limiting a little with each one. Rather than slamming it all with one limiter.

Interesting. I'll have to give this a try. Can you explain to me how is this different than using one limiter?

[LIMiT]

another way is to use a couple of compressors in a row - compressing 1-2dbs - then the limiter...

I prefer to mix with a comp on bypass, and turn it on once and awhile to see whats getting lost (or gained..ie bass).
My final mix will have compression on it...it gets printed.
Then I import it (or sometimes submixes of sections) into a mastering session where *if the client insists on having it uberloud* - I'll do what I mentioned above.

If it's going to a mastering engineer...I ask him what he/she wants, which is usually an unmastered version, and a compressed mix of how i intended it to sound.

-LIMiT

[masteringhouse]

But with the limiter at 12 dB of gain reduction ...

Something is seriously wrong if you need to do this. Most MEs I speak to (including myself) only use about 2-3 db worth of reduction through a limiter, though this may be after a bit of crunching the converter.

Volume comes in many ways, it's more than just removing dynamic range and raising the level. Yes, most CDs (particularly Metal) are over compressed, but the mix has to be able to withstand that abuse too.

One thing to try particularly with drums (if you aren't already) is parallel compression. You can make drums sound huge (and louder) while still retaining some of the transients. Don't over compress as in slapping one on a track and leaving it on stun, it takes the life of a track and when you try to limit more later makes things sound like mush. Metal is very dense music, you need to be able to leave as much space as possible both dynamically and in the frequency ranges of the instruments to allow them to breath. Use HPF and LPF to remove unnecessary freqs above and below tracks where it doesn't contribute anything, etc.

One area where I hear issues in Metal mixes is having way too much bottom in the kick. Yeah it sounds cool by itself, but this low end eats up a lot of headroom if you want it to come out in the mix, that headroom might be needed elsewhere. Rap mixes can have this issue as well. Another thing that I hear in a lot of Metal mixes is a really grating high mid on the guitars. You really can't make things good and loud if it's already irritating at a lower volume.

I'm not saying that your mixes are suffering from the above (I haven't heard them) just some hopefully helpful tips.

[NL5]

Originally Posted by LIMiT

another way is to use a couple of compressors in a row - compressing 1-2dbs - then the limiter...

Can you explain your "multiple compressor theory" a little bit?

[Farview]

Originally Posted by LogicDeLuxe

You really should convince him that sound quallity DOES MATTER, though.Actually not. Everyone here I ever did a master seems quite happy with a master at -14 dBFS RMS sine. Sometimes, it takes some explaining what's going on, but it sure is worth the time.I like that idea.

I can normally get metal mixes to sit at -13 or -14 rms just by normalizing them. I need the limiting to get them to -10 or -12 rms.

You need to not put the drums so far out in front of the rest of the mix, the farther you put them out, the farther you have to beat them back. That's what makes it sound bad.

You have to be more careful about the frequency response of each instrument. Everything needs its own place. Also, the second and third harmonics of an instrument won't have as much dynamic power as the fundamental. You really don't need a bunch of 40hz in your bass guitar, the 150hz and 800hz will get the richness of the bass without taking up the dynamic headroom. Same for kick, guitar, vocals, etc...

[mshilarious]

Originally Posted by LogicDeLuxe

Everyone here I ever did a master seems quite happy with a master at -14 dBFS RMS sine.

I envy you for your clients. I did a demo on a jazz fusion track; said they wanted commercial volume levels. The mix was -20; somehow -13 wasn't enough

Once my MIDI pipe organ is done though, I am done with those clients. It's all pipe organ, all the time! It's the new dance craze! The, uh, rondeau . . . everybody's doing it!

[LIMiT]

Originally Posted by NL5

Can you explain your "multiple compressor theory" a little bit?

Depending on the mix (of coarse) -

On a rock mix, I would compress it about 4db ish making it kinda pump (without hearing the artifacts too much) but just to clean it up a bit.
Then you hit 1 more (sometimes even 2) compressor to tighten it back again.

What it does is that it doesn't tax too much the one compressor, so the effect is more transparent as it flattens a bit more with each compressor
(-without squashing it too too much-)

The one plug that is absolutely remarkable to master with is the sony inflator. It's REALLY clean .

I tend to compress more on the first one to make it pump a bit (good with big beats) just so it 'makes your head bob with the music' (so to speak), then close it off a bit by reducing the "peaks and valley" by hitting a few more compressors with a small reduction gain on each, and the limit.

For faster more aggressive mixes, I use faster attack and release settings...with not so much pumping on the first comp.

-LIMiT

[LogicDeLuxe]

Originally Posted by Farview

I can normally get metal mixes to sit at -13 or -14 rms just by normalizing them. I need the limiting to get them to -10 or -12 rms.

While I see no reason to do this, it probably wouldn't hurt either. It's nothing like limiting 12 dB this thread was about.

Originally Posted by mshilarious

I did a demo on a jazz fusion track; said they wanted commercial volume levels.

That hurts, I can imagine.
In my experience, some kinds of jazz are the 2nd most dynamic music right after classic and are in deed best mastered slightly lower then the usually prefered -14 dBFS rms sine.
I've seen jazz CD's with excessive mid range turned up which makes them apprear loud. At least, such a thing can be cured, in contrast to excessive compressor usage.

[NashBackslash]

So I take it that nobody is doing the thing I am doing (mixing with the limiters turned on).

The main reason why I do that is because everytime I put the limiter and try to get the loudness to match commercial CDs (we're talking about metal here), it'll end up sounding nothing like how I mixed it. Among the effects of how drastically my mix changes have been mentioned in the original post.

I have this thinking where "if limiting it is going to screw up my mix, might as well mix with the thing on". If I mix first and limit later things become so unpredictable. So with the limiter on, I'll try to make it sound good.

Also, just a note that I monitor my mixes at conversation level.

So how exactly is it done? I don't think my mixes themselves are bad because when I turn down the commercial CD to match my unlimited mix, it sounds pretty much close. Equal tonal balance in all frequency spectrums, very punchy, instruments are all clear, no bass mudiness etc.

I keep the highest peaks for my unlimited mix at around -12 to -6 dBFS because that seems to be the advice spread around the internet. The problem with that is the the average volume for the song will be really low. As a client put it "your mix was so soft I had to turn the knob a lot to hear it, and when the next CD came up, it was so loud I fell off my seat".

I'm really confused here. On one hand, people are saying do not make your mixes have peaks close to 0 dbFS, leave some headroom, bla bla bla. On the other hand, there's just no way the mix can match the loudness of other CDs AND RETAIN THE PUNCHINESS with such a level.

I appreciate the people who say over compressed/limited mixes just sound like crap but this isn't the subject here. It's too late to discuss about that. I came here seeking advice on how to make my mixes just as loud as the other CDs but still have the punch (especially on the snares, I'm really fussy about my snare sounds) because that's what my clients want and they're paying me money to do it. I can't really just tell them off and nag or lecture them about why limiting isn't good and all that.

Plus, I just know it's possible to have loud but punchy masters. I'm hearing them on some of these CDs.

EDIT: For clarification, I am not having problems with my mix. I'm not having problems with sitting my instruments, no clarity problems and all that. I love my mixes. As I've said, if I turn down the commercial CD and play it next to my mix, they'll sound close. What I'm really having problems with is to get my mix to match the perceived loudness as the commercial CD. Whenever I try that, the whole thing becomes mushy and my punchiness is gone. Which was one reason that led me to create this thread; do I have to mix really, REALLY differently (as in, unconventional or unusual) than what I am used to just to make that limiter be able to boost my mix's loudness?

[Lectric_Lab]

I'm no Pro ME for sure, but I always do that stuff outta the box using analog gear running to a masterlink.

[LogicDeLuxe]

Originally Posted by NashBackslash

So I take it that nobody is doing the thing I am doing (mixing with the limiters turned on).

As you talked about "mastering", this wasn't that obvious. Why not limit the individual tracks during mixing instead? If done right, you certainly won't have to limit 12 dB during mastering.

So how exactly is it done? I don't think my mixes themselves are bad because when I turn down the commercial CD to match my unlimited mix, it sounds pretty much close.

CD can be mastered quite differently. It depends much on which CD you do the comparison. You have to analyze the CD in question. I noticed many heavy metal CDs are overdriven.

I keep the highest peaks for my unlimited mix at around -12 to -6 dBFS because that seems to be the advice spread around the internet.

As long as your mix doesn't clip, it is pretty much irrelevant where it peaks, as the final volume is adjusted during mastering anyways. If you're mixing in float format, even clipping can be recovered with the master limiter. I really don't care that much during mixing.

I appreciate the people who say over compressed/limited mixes just sound like crap but this isn't the subject here.
...
Plus, I just know it's possible to have loud but punchy masters. I'm hearing them on some of these CDs.

If you could upload an excerpt of your mix, I'd like to take a look into it and decide what can be done to it during mastering.

[masteringhouse]

Originally Posted by LogicDeLuxe

As long as your mix doesn't clip, it is pretty much irrelevant where it peaks, as the final volume is adjusted during mastering anyways.

It's a good idea though to have at least 6 db of headroom before peaking in order to let the ME work. For example if you are using a digital EQ as the first process and boosting any frequencies you could get overs. As a result you would need to perform a gain reduction before the EQ. A process that can create quantization distortion and could have been be avoided if there were a little room.

Regarding limiting the mix, there's nothing wrong with this it's done all of the time. The thing is not to overdo it when creating the final mix so again there is enough headroom to work with while mastering. The type of limiting that's used during mastering will very likely be different than smashing it by 12 db with a limiter though, so I wouldn't gauge things against a reference like this. I would try slight limiting on the mix and lowering the level of a commercial CD to match as a reference. If things sound good there job well done. Leave the rest to a capable ME.

[mshilarious]

Originally Posted by LogicDeLuxe

While I see no reason to do this, it probably wouldn't hurt either. It's nothing like limiting 12 dB this thread was about.That hurts, I can imagine.
In my experience, some kinds of jazz are the 2nd most dynamic music right after classic and are in deed best mastered slightly lower then the usually prefered -14 dBFS rms sine.
I've seen jazz CD's with excessive mid range turned up which makes them apprear loud. At least, such a thing can be cured, in contrast to excessive compressor usage.

Well actually when I was done, I was pretty happy with it. It was maybe 1-2dB louder than I would have otherwise made it, but it didn't suffer from it. It was fairly slamming stuff, not crunchy guitars or anything, but it was busy, no sparse quiet sections or anything, so it was a reasonable volume. But how loud do people want this stuff?

[LogicDeLuxe]

Originally Posted by masteringhouse

It's a good idea though to have at least 6 db of headroom before peaking in order to let the ME work. For example if you are using a digital EQ as the first process and boosting any frequencies you could get overs. As a result you would need to perform a gain reduction before the EQ. A process that can create quantization distortion and could have been be avoided if there were a little room.

In that case, I would question the skills of the ME, as he apparently either has dated software/hardware or he doesn't know how to use it properly.

- Reducing by 6 dB doesn't introduce more quantisation noise than a mix done 6 dB quiter has to begin with.
- In 24 bit recording, the noise floor is significantly higher then the quantization noise anyway.
- Importing the mix as float doesn't introduce quantization noise at any kind of editing in the first place.

I always master in float, and the very last thing to do is the 16 bit quantization with noise shape dither. And if it unexpectedly clips after EQ'ing, I can just lower the volume afterwards, since nothing gets lost in float. If the source peaks at 0 or at -30 dB, it really doesn't make any difference to me.

[Reggie]

Here's a few stoopid things to try on your master bus:
1. A decent shaping/somewhat surgical EQ. Even though EQing the individual tracks is the best, sometimes you can still end up with a buildup/deficiency in parts of the spectrum on the final mix that are hard to attribute to one instrument or the other. See what lows and low-mids you can live without, and see what parts of the midrange you can stand more of. Might be good to keep a popular metal CD handy.

2. A transparent compressor. For metal, perhaps try a mid-fast attack and release. Play around with the range under 30ms attack, and under 500ms release. Whatever amount of gain reduction sounds good. I have an outboard compressor I like to use for this role - a sort of clone of an S*S*L mixbuss compressor.

3. A sweetening EQ. I like the Pultec plug for this. Just doing broad adjustments to the high and low end balance usually. Maybe a bit of a high end shelf boost.

4. A look-ahead style limiter. Increase input to where you are getting maybe 3db of steady gain reduction at most, with occasional spikes of 6-8db of gain reduction. Some of these have an Auto setting that automatically adjusts the release as it sees fit. I tend to turn this off when I am really wanting something loud and crunchy, turning the release down as fast as I can before it starts to crackle and crap out, then back up a hair.

5. If you are really wanting to get crazy, next feed the mix out of your D/A, into some sort of box where you can add gain as transparently as possible, and back into your A/D ignoring what goes into the red and just listening to the incoming signal. Turn up the gain to where you can just start to hear crappiness, then back it down a hair. Should add a little more level to the final master for you, and metal dudes are usually not concerned with having prestine, unclipped masters.


If this still isn't loud enough for them, then just tell them to go away because there really isn't much more you can do to mangle it further anyway!

[Farview]

Originally Posted by LogicDeLuxe

While I see no reason to do this, it probably wouldn't hurt either. It's nothing like limiting 12 dB this thread was about.

That's my point. I'm getting -12 to -10 dbfs rms sine but I only have to get 3 or 4 db of reduction. I do it with a compressor and a limiter chained together. So neither of them are doing more than 2 or 3db of reduction by themselves.

[fishkarma]

Originally Posted by LogicDeLuxe

If the source peaks at 0 or at -30 dB, it really doesn't make any difference to me.

apart from bring up the noise floor if its at -30db when you make it a ''regular'' volume right?

[Farview]

Originally Posted by LogicDeLuxe

I always master in float, and the very last thing to do is the 16 bit quantization with noise shape dither. And if it unexpectedly clips after EQ'ing, I can just lower the volume afterwards, since nothing gets lost in float. If the source peaks at 0 or at -30 dB, it really doesn't make any difference to me.

This assumes that you are mastering in the box. If you actually have hardware, the gain staging becomes more important. Even if you are at 32 bit float, it will get bounced down to 24 bit when it hits the converters.

[NashBackslash]

Okay, let's try this; can anyone here post a before mastering limiting and after mastering limiting audio example of a metal song?

You don't have to post the full song, even just a section will do. I'm particularly interested in hearing how some of you can retain the snare's punchiness on really dense sections.

Then if you're generous enough, a walkthrough on what you did, the sequence, in the boux, out the box, whatever. That'd be really nice.

I just want to make sure I'm understanding everyone here.

METAL SONGS ONLY!

[LogicDeLuxe]

Originally Posted by NashBackslash

Okay, let's try this; can anyone here post a before mastering limiting and after mastering limiting audio example of a metal song?

Here you are. It is from timboZ mixing competition at this forum some time ago.

Mastered at -14 dBFS rms sine (as I usually do), the master limiter had almost nothing to do, it just could have been normalized as well: http://www.abmischung.de/snippet_14.ogg

Same volume, but peaks are limited to -6 dB: http://www.abmischung.de/snippet_8-6.ogg

And finally boosted by 6 dB, reaching -8 dBFS rms sine, which is a typical level for excessive compressed CD: http://www.abmischung.de/snippet_8.ogg

Then if you're generous enough, a walkthrough on what you did, the sequence, in the boux, out the box, whatever. That'd be really nice.

Pretty much the same as I suggested above, except for the hard clipping: multi-band-compressor, emulated tape saturation and finally the master limiter.
If that's still not enough, just turn it up and let it clip slightly. Many metal CD's do clip.

[Farview]

Here is an example. The unmastered version has been normalized to peak at 0dbfs, so you can see the volume difference from the mastering compression and not the difference plus the headroom that I ended up with while mixing.

http://www.farviewrecording.com/audi...unmastered.mp3


http://www.farviewrecording.com/audi...or_nothing.mp3