RMS Limiter/Compressor - does that even exist?

[Compyfox]

Hi!


Sorry to sound totally noobish, but I do this on a purpose now. Let's say I have the following case example:

I have tracks, that have a loudness (RMS) of about -12dB, but some of those tracks have "RMS spikes" that go up to -10dB. But I don't want to let them go there. The problem is... Limiters can't do that. Limiters react to the Peak Level, but not to the RMS level.


My question is fairly simple (but mabye stupid too):
I'm looking for a brickwall limiter/compressor that doesn't react on peak, but RMS. (example: I setup my peak limit in the limiter to -0.3dB, but in this case I want to set the maximum RMS level to -12dB).

Is there anything out there VST/DX (no matter if commercial or not) that can do this? Best if you could answer as soon as possible.


Thanks. And if I was talking "nonsense", feel free to laugh about me.

[RapRogue]

Psp mastercomp
Neodynium

Prolly more...

[Compyfox]

And... do they really react like "okay... don't go higher than an RMS level of about -10dB"?

[RapRogue]

You can do that with high ratios, and fast attacks. Then it would be kind of like rms limiting.

If you are thinkin about rms layering Neo with the
unique zone mode would be your best shot.


http://www.kvraudio.com/get/1252.html

http://www.kvraudio.com/get/1859.html

[Compyfox]

Did someone actually "try" to do this? I mean... RMS limiting rather than peak limiting?

EDIT:
Just tried it with a simple compressor:
Threshold -8dB
Ratio: 3.0:1, 5.0:1, 7.0:1 and 8.0:1 (max of this one)
Attack: 0.1ms to 5ms

I hear noticable pumping. *sigh* Maybe it's just a stupid idea anyway.

[RapRogue]

Depends on what you want to do. Imo it works well with vox, in particular layering many voices.

Not everything benefits from rms compression though.
Which I assume you noticed.

If you want to raise overall volume a limiter maximizer is better. If you want to control the constant level of something then rms.

[Compyfox]

Well I want to use this as "master limiter" for spikes in the RMS level only. But it does more damadge than wanted. But I want to release this on DVD (AC3), and AC3 has a limit of -10dB to -8dB RMS. But some tracks go over. And I don't want to sacrifice some bass.

[RapRogue]

Ah, doubt you will achive that as it just kills cold any
dynamics. If you don't want to fiddle with EQ or key
filtering.

Voxengo elephant is a good limiter with peak rms
readouts. You might want to use very slight rms
compression and a limiter.

[Compyfox]

should look into that. But I guess there's no choice just to turn down the gain of the limiter. Yay for dynamic tracks... wasn't any different in the 90ies anyway.

[RapRogue]

You could always make a softer mix, try and put voxengo span on masterbus, and then aim at the intended level.

[Compyfox]

Well I'm using InspectorXL and I already did what I could without sacrificing too much of the sound. I'll try the one or other thing. But I guess I'll get nothing but pumping. So I leave it "as is".

If it pumps in the AC3 file, I simply turn down the gain of 1dB. Simple and clean. Thanks for the suggestions.

[P.T.]

Did you try a high threshold like -30db or thereabouts?

If you are trying to keep things below -10db then you need to have the threshold below -10db.

[C00kie]

AC3 has a limit of -10dB to -8dB RMS. But some tracks go over.

RMS is an average level, while I hear you talking about peaks. Peaks should do not much to the average. Do realise that -10dB average level is f**king LOUD! There will be peaks already touching 0dB (if you look close enough) which is the actual limitation of AC3 (as is in 16bits audio for compact disks)

[Compyfox]

HIgh thresholds like -30dB produce even more pumping while listening to the RMS (the compressor, not me).

Speaking of whcih... I'm not talking about "peaks" as in PMM. I'm talking about "spikes" as in RMS. Big difference. I'm trying to master something in K-12, and those who're familiar with that system know that K-12 is aimed at Broadcast.

The thing on K-12 however:
YOu have an RMS level of -12dB while you still have a so called "amber zone" of +4dB (which means -8dB RMS max!) for mezzoforte passages. AC3 can easily handle that, I extracted the "bonus tracks" from the "The Matrix Ultimate Edition" and analysed them. All tracks were around -10dB to -8dB depending on the pace of the music.

For me K-12 (with full amber zone) it's fairly hot (to others it isn't), however for very dynamic orchestra material (even if it's a digital one) a bit hard to handle and compared to recordings like "Lord of the Rings" (which is K-12 btw). I can't get rid of the spikes going into the warning zone (-7,9dB to -6dB, which is a NO-NO to get in there). And due to the pace of the tracks (I didn't write them, I'm only engineerig them) they're fairly "quiet" most of the time.


It's a bit tricky here, but I guess I have to live with it.

[RapRogue]

Big RMS adjustments are harsh, even more so on dynamic
material. You could always chain several compressors,
that way you should get a smoother response. The thing
here is to not stress the compressor.

first compressor on lowest RMS value with slow attack,
and last compressor on highest rms value and fastest
attack (limiter like). Guess that would be kinda like
'soft knee RMS compression' Parallel
compression is another option. I prefer doing such in
SFP or maybe XT?

Try soniformer in expander mode for soft passages.

Or just decrease the overall volume like you said