( free ) Limiter, once again !

EDIT :  freeware W1 Limiter is one of the better freeware limiter that is somewhat neutral.

To buy:  Fabfilter Pro-L is the " almost neutral " sounding limiter out there... !

And this measurement tool / loudness meter : 
http://www.klangfreund.com/lufsmeter/ 

Rock / Pop etc ...
If i keep the short term RMS level above -9.5db as max, and long therm RMS at -11.5db as max,
( personally i wont push it more than that ) , then there should be no issues for bad sound.

Classic / Acoustics etc
For Acoustic / Singer / songwriter / ballads etc, i would also ease off 2db more to catch the subtle
nuance's of the music, going at  short therm RMS -11.5db , and long therm RMS -13.5db
to really catch the " movement of the spirit " 

Metal Music etc
Just push it no one cares  , hey what did you say .... hey !!  you....  "&//&% !

To many push music to hard so they loose transients, clarity and definition in today's music and its too bad

Ease down, breathe, Relax, sit back and feel the energy and life itself
don't forget to join the :  transient & dynamic movement ( TDM ) 

Jan S

Thanks for the tip, I was just about to research freeware limiters, I guess now I can save the time!

What is more neutral than Fabfilter Pro-L
maybe another like the waves L3 ?

The issue is that more or less " all Limiters "  have some EQ " coloring " to the sound
it add some quite often from 400-1.2khz, not much, but i can hear it.
(guess that go for all kinda plugins anyway ....) , but the goal is to get a " neutral "
sounding Limiter, without altering frequency / stereo image or presence of the instruments in the mid's.

Some limiters also tend to take " away " some from the mid's
and others add/boost a little. Every Limiter is different.........

Loudmax, is one of the limiters that take away some from the mid's
and also bring some frequencies down also in the 500-1k range.

So a good test between limiters is, between loudmax and W1 Limiter
those are quite different, and easy to tell the difference.

the best way to compare, is to use music that's not to busy
with the arrangement, but open and detailed with nice tone
where you can hear / feel the vibe.... !

The W1 Limiter, is more mid ranged than Loudmax.

So they'r having their different classes, i would do Loudmax at, Dance / pop music
and W1 Limiter for classical / instrumental type of sound, if i should choose.


when i testing by ear i do it like this :

first i boost my stereo amp about + 6db and listen to the sound....., then turn down the amp again
and activate the limiter... comparing the stereo width / tone (harmonics) and the change
in the frequency range, compared to the boosted volume of my stereo amp.

If i boost with my stereo amp, it will be a way nicer boost, than with a limiter
a limiter tend to add more " harsness " to the sound.

Maybe that is why Mastering engineer's do use expensive outboard gear to avoid those problems
and get the best sound.

Cause there's no doubt that hardware is better than software....

Yes the best way is like you say, not to use more than  1-3 db gain
the more gain , the harsher the sound gets, can also hear that on a
Amplifier when it reaches above 50% of output power, the frequency range start to change
and it slowly distort more and more, as volume is cranked up.

But Hardware do sound better than software. It is more expensive and
require more work to use outboard gear, than plugins.

Best thing is probably at the final stage use some outboard to hit the 
music with a few DB more, with a hardware valve Limiter.

Just boosting the volume will not, by itself, alter the nature of sound. It's only when you distort the audio that new harmonic and inharmonic components are added that can range from pleasant enrichment to harsh annoyance.
 
A limiter by nature is a distortion device. Clever designs can mitigate or disguise the distortion, but they all distort and they all add something new to the sound. This is true for both software and analog limiters. The only way to avoid that completely is to not use a limiter at all.
 
The difference between digital and analog devices is that digital processes can only mimic the way analog amplifiers naturally saturate. You have to really drive a vacuum tube hard before the distortion becomes unpleasant, and as it transitions into unpleasantness it does so in a gradual way. With digital devices, the onset of unpleasantness is abrupt and profound.
 
Another difference is that analog devices are simpler, with fewer options. They're easier to understand and use, and more difficult to abuse. Digital equivalents tend to offer many options, some of which are dangerous. A tube device will be designed to not react fast enough to do the worst kind of damage. But software can react to individual samples if you tell it to. At 44.1 KHz, that's about 50 times faster than the threshold at which you're likely to start causing audible distortion.
 
So when I hear somebody say they get much better results with outboard gear, my first thought is that they were probably mis-using their digital gear in the first place.

Yeh, Pro-L is a little obtuse. The first time I looked at it, my first thought was "why on earth does a brickwall limiter have an Attack knob?"
 
Then I figured out that the core of Pro-L's secret sauce is a transient detector, with separate algorithms for transients versus slow-moving values. The Attack control is really adjusting the point at which one algorithm hands off to the other. Make it short and you only process the fastest transients separately. Make it very long and you essentially bypass the transient processor, turning Pro-L into a generic limiter.
 
In Dynamic mode, Pro-L kicks in a transient enhancer in front of the limiter. Kind of like Dolby pre-emphasis but for dynamics. It's my favorite mode.
 
And yes, lookahead is a kind of a reverse attack setting. The shorter the lookahead value, the longer Pro-L waits before mitigating a peak. Longer lookahead values make the process more transparent. But I'm using "transparent" to mean you don't hear the limiter working, not in the way most of the kids over on KVR use the term to mean "still sounds OK when I turn it up REALLY loud".

bitflipper
Then I figured out that the core of Pro-L's secret sauce is a transient detector, with separate algorithms for transients versus slow-moving values.

 
IIRC, Elephant has 8, maybe more of these algorithm choices.

I went and looked at the details again. Elephant 3.1 has 10 styles of limiting that you can manually select.