why do limiters destroy low end punch?

[saemskin]

perhaps my ears are developing along nicely. Lately I have begun to notice any limiter plugin makes the low end come out gurgling. Some more than others, but its there.
For example, Buz maxi is completely unusable on anything with a frequency below 700 Hz. Our beloved Precision Limiter isnt nearly that harsh, but its still messing things up, even with no gain reduction.

What is going on here guys? Is there a truly transparent limiter on the market?

 

[dt]

You are hearing correctly. There are very few transparent limiters, and Buzimax isn't one of them. You can't push a limiter hard in most cases and still maintain a smooth response, but many affect the sound regardless.

Sony's Wave Hammer (DX plugin included with Vegas and Sound Forge) is probably the most transparent brickwall limiter I've found, and I've tried most every plugin limiter out there (PC at least). Just disable the compressor and think very conservatively - a low threshold (-1dB down, going negatively) adversely affects transparency with most limiters, so keep it to catching the highest peaks. I also don't use a plugin's gain reduction, but make it up in other ways.

 

[csl]

what? it's messing up the sound without any gain reduction? never noticed that myself.

 

[Trebor Flow 2]

Csl

I've only one thing to say ....

TC Powercore MD3, Brickwall Limiter. Stunning. If you can afford it the Brickwall on the MD3 package will blow you away, it leaves the transients totally untouched it's an amazing piece of software and for $1000 it should be.

I have never heard a better digital limiter and I've heard \"m\" all. The P Limiter is very good but the BL in MD3 really is something special.

Trebor

 

[csl]

yeah, i've heard good things about that bundle, but after shelling out on the uad-1 precision stuff i can't justify buying more mastering processors, especially when i'm just a bedroom producer. look really good though.

 

[Eurocide]

The high energy of low end frequencies are often a problem for limiters.
With this following litte analytic test you can check out, how limiters behave.

1. load a 0dbFS pink noise in your host application
2. PL (or whatever) as insert.
3. at \"unity\" gain setting (i.e. 0db) there should be no effect at all.
4. increment the gain and you will hear a fluttering in the low end.

The character of this \"fluttering\" determines the character of the limiter.
Maybe you like it or not - matter of taste.

If you want to push your limiter to run in the loudness race put a Cambridge before the PL an set the low cut filter to type \"Butterworth 6\" or \"ecliptical\".
Then increment SLOWLY the low cut frequency until the fluttering stops.
Usually it stops at around 32Hz in the PL with +3 to +6db gain.

But if you need the sub basses (e.g. D'n'B, techno, whatever) you should not use this cutting trick.

I'v already heard some good-selling electro productions where the bass was cut up to 45-60Hz just to win a pole position in the mad loudness race :lol:
I don't recommend this extreme setting at all =;

 

[Staccato]

Are any of you willing to clue me in as to why a discussion such as this does not mention (at least so far) attack and release times with respect to the effect they have on the low end?

 

[Dan Duskin]

limiters removes both lows and highs.

The attack of the limiter is rolling off the high-end on peak transients, removing highs... and the slower release of the limiter is moving up and down in the oposite direction of low-end waves, reducing the lows.

The combination of the two might sound like \"less low-end punch\"... the punch part being the high-end transient and the low part being the lows.

Two methods to eliminate this is to use a limiter with \"look-ahead\" and/or to use a clip-based limiter. Both have some nasty side-effects when pushed hard, but they can help remidy the above problem.

 

[Staccato]

The high energy of low end frequencies are often a problem for limiters.
With this following litte analytic test you can check out, how limiters behave.

1. load a 0dbFS pink noise in your host application
2. PL (or whatever) as insert.
3. at "unity" gain setting (i.e. 0db) there should be no effect at all.
4. increment the gain and you will hear a fluttering in the low end.

The character of this "fluttering" determines the character of the limiter.
Maybe you like it or not - matter of taste.

If you want to push your limiter to run in the loudness race put a Cambridge before the PL an set the low cut filter to type "Butterworth 6" or "ecliptical".
Then increment SLOWLY the low cut frequency until the fluttering stops.
Usually it stops at around 32Hz in the PL with +3 to +6db gain.

But if you need the sub basses (e.g. D'n'B, techno, whatever) you should not use this cutting trick.

I'v already heard some good-selling electro productions where the bass was cut up to 45-60Hz just to win a pole position in the mad loudness race :lol:
I don't recommend this extreme setting at all =;

Is the fluttering also related to attack and release times? I think yes. If you set this up to try it, see if the fluttering also goes away when lengthening attack and release times, and thus lowers the frequency of where the rolloff eliminates the fluttering. In this pink noise experiment, you'll find there's not a transient wavefront associated with any particular frequency that follows, making proper adjustment of the limiter impossible. The settings are particularly important when limiting program material with more or less consistent transients, (music) followed by associated notes. (low frequencies such as kick and bass) Lengthening the release time, however, will not endear you to the loudness mongers.

 

[Eurocide]

Is the fluttering also related to attack and release times? I think yes. If you set this up to try it, see if the fluttering also goes away when lengthening attack and release times, and thus lowers the frequency of where the rolloff eliminates the fluttering. In this pink noise experiment, you'll find there's not a transient wavefront associated with any particular frequency that follows, making proper adjustment of the limiter impossible. The settings are particularly important when limiting program material with more or less consistent transients, (music) followed by associated notes. (low frequencies such as kick and bass) Lengthening the release time, however, will not endear you to the loudness mongers.

You are totally right, but PL doesn't give us the possibility to adjust the attack time - other brickwall limiters like the L1 & L2 also. My experiment is not for the "punch" but for the sensibility of brickwall limiters to low frequencies and why a lot of people kill (important) sub frequencies just for the loudness race.
sorry if I should have hi-jacked this thread.

 

[Staccato]

Eurocide, thanks, I did not know that about those limiters.

 

[saemskin]

thanks for the info thus far everyone, great stuff really.

So here is the problem then. I cannot use this 200$ limiter on the master buss, simple as that. It ruins my tracks, and I didnt notice this effect until recently because its in my default template and I set it at -2.5 or -3 and just watch the gain reduction to keep it under 1 or 1.5. That's been how I mix up to this point. I have over half an album that I've already recorded and sent to the mastering company that's been mixed this way.

I understand that the MD3 is the shit, and it's not out of the question. But right now I am really trying not to buy anything new so I can use all of my available funds for promotion of my album when the time comes in August.

I still put the PLim on the master buss, but now I've turned the mix level all the way down and just use the meters to watch the output level. I prefer the numerical display over watching a meter rise and fall. I now have to attempt to mix my tracks correctly to keep the master level even, and its obviously harder than I anticipated because I have not been successful. I do not want to compress the piss out the individual elements, but I'm struggling here.....

Suggestions? Anyone, anything? I'm all ears.

one thing I considered was to create 4 or 5 sub-busses inside my DAW. Grouping similar sounding channels together. One for kicks, one for basslines, one for percussive elements, etc....
Then maybe I can focus on those 4 or 5 busses and making them \"fit\" rather than the jumble of up to 50 audio/sample/instrument channels that comprise the song. I started to set this up in my most recent piece, but I'm wondering if I'm backing myself into the same corner of limiting reliance, only just spread out more?

Have I simply made a eureka discovery that will ultimately make my mixing skills better, or am I fucked?

 

[cAPSLOCK]

That is certainly a good technique. Regularly used in just about any standard mixing scenario. It's what your busses were designed for. Then you have control to do kind of what a multiband compressor tries to do, but yours will be a better way in most circumstances..

On some mastering limiters you have a selectable low roll off. This is an important part of getting 'loudest' since most of the audio energy CAN live down in the sub frequencies. So with any limiter across busses or the stereo pair you want to lose all the lows you don't need but only the ones you don't need.

cAPS

 

[Staccato]

Yes, the multiple stems will work well, if you're sending to a Mastering House, do you really need to worry about the limiting? If you send off with -6dB peaks and no limiter, any Mastering House would be happy. Once something is previously limited, it may not be easy to undo it if at all.

 

[svs95]

limiters removes both lows and highs. <snip> The combination of the two might sound like "less low-end punch"... the punch part being the high-end transient and the low part being the lows.

Almost right, Dan! Actually, you are correct in what you say, it's just that there is something nobody's saying yet.

The "punch" in music, at any frequency, has a name. It's called "dynamics." Music cannot have dynamics if it's squeezed into the top two bars of a peak meter all the time!

You want the attack part of a kick drum to tower over the rest of the mix? Then you simply have to leave the rest of the mix lower than the attack of the kick drum. In order for something to be big, something else has to be smaller! IT REALLY IS AS SIMPLE AS THAT! It's surprising that some people are only just now realizing this!

 

[Dan Duskin]

limiters removes both lows and highs. <snip> The combination of the two might sound like "less low-end punch"... the punch part being the high-end transient and the low part being the lows.

Almost right, Dan! Actually, you are correct in what you say, it's just that there is something nobody's saying yet.

The "punch" in music, at any frequency, has a name. It's called "dynamics." Music cannot have dynamics if it's squeezed into the top two bars of a peak meter all the time!

You want the attack part of a kick drum to tower over the rest of the mix? Then you simply have to leave the rest of the mix lower than the attack of the kick drum. In order for something to be big, something else has to be smaller! IT REALLY IS AS SIMPLE AS THAT! It's surprising that some people are only just now realizing this!

True.

But let me mention another technical point...
You can retain a lot (even most) of the punch in an overly squashed mix/master by doing something very simple... let it pump! i.e., if the attack of the master-buss/mastering compressor is very slow, and the release is fairly slow, the compressor will pull the volume down and let the punchy parts through. This is a different kind of dyanmics... it's dynamics created via a compressor (as opposed to a natrually dynamic recording, via less compression and limiting). You can also acheive some of this (but not all) with a limiter by making the release time very slow.

PS: this is technical info, not an opinion of which is better.

 

[svs95]

You can retain a lot (even most) of the punch in an overly squashed mix/master by doing something very simple... let it pump! PS: this is technical info, not an opinion of which is better.

Yep, you're talking about reconstructive surgery. I've had to use it too. I can guess which one you think is better, but I'll keep it to myself. :lol:

 

[bob humid]

You can retain a lot (even most) of the punch in an overly squashed mix/master by doing something very simple... let it pump! PS: this is technical info, not an opinion of which is better.

Yep, you're talking about reconstructive surgery. I've had to use it too. I can guess which one you think is better, but I'll keep it to myself. :lol:

PSP MasterComp "Some Fresh Touch"

0.7 : 1 decompressor...

sorry. pollen allergy hit me. I am not going to explain this..

robert