Hey guys… I know this is a really dumb question for most of you but here it goes…
What is the difference among limiting and compresing ?
Thanks in advance !
Someone else may explain it better, but this is the way I think of it.
Limiting sets a max. that cannot be exceeded - so, you can keep the sound from clipping, but it does not “level out” the sound. It is usually applied last, after compression.
Compression is a way of turning down the sound as it exceeds a threshold - the sound is turned down by a ration by how much it is over that threshold. With a ratio of 1:2 every 1db about the threshold the sound is turned down by 2 db.
Loud enough and it will still clip, but the sound is “leveled out.”
The dynamics of the sound are effected adversly if too much compression is used.
Bax
Yeah–
Limiting will protect equipment and prevent clipping.
Compression, to oversimplify, is to make it sound more even.
Fuggle
Fuggie, you are going to have to learn to spand far more time on extra words. Your answeres are very short and make too much sense!
Thanks guys !
I agree with Bax, Fuggle put it really simple.
So far so good… now, Why does it happen that when I use limiting, my signal seems to be so much louder ? (or is it me just imagining it)
Hi, Gamma–You’re probably not imagining it.
Maybe your program gets louder because of a low threshold setting? If you don’t want it jumping up in volume, try decreasing the threshold.
(I can’t edit the message, for some reason)…
Or, if you like the way it sounds and it’s too loud, try reducing the output gain?
This stuff can make my head hurt…I don’t understand why a signal gets louder with a low threshold and a ratio of greater than 1:1 when the output gain is unity.
I like to try find a level where the sound file looks like a “wave” file. Compression, especally ones that can deal in the different sound pitches (Low, Mid range, and highs) can really bring the sound to the front, but it does not “Cut off” the sound, it squishes it, the sound is compressed. You could think of it a denser sound, more sound energy. Loud recordings have wave forms that are almost squared off, they hsve very little dynamic range, and the wave form looks like it. The amount of sound energy (the volume level has beeen turned way up past where the sound would be completely distorted/clipped exce-pt the the compressor/limitor Hold them back like a dam holds large amounts of water - there is plenty of energy to move well past the dam, ut it holds it in place. The peaks and valleys are small - the sound stays loud because the softer sounds have been made louder and the louder sound squished into a space/sound range where they are as loud as they can get without clipping “Too” much. Some clipping/distortion is done on purpose.
I await the shorted, clear version . . .
Hi, I like graphs to explain limiting and compression see below.
http://www.mediacollege.com/audio/processing/compression/
http://www.mediacollege.com/audio/processing/limiter/
Nick
Limitting - is relative to the loudest parts of a source.
So if you have a source with a -6db peak, and
you set the ‘limitter’ to -3db, then only those
-6db peaks are going to reach -3db, and the rest
of the source material will be raised in amplitude
accordingly. (proportionally)
Compression on the other hand (which can be used as a limiter also)
can raise all of the source to the set limit by amplifying
the lower source material disproportionally.
i.e lower levels recieve more amplification.
But there are many other things you can do with compression.
Thanks to all guys… I think I got it now…
I guess it wont be long before the next question but for now, back to my music !
Cheers !
Quote: (sevenOfeleven @ Sep. 22 2008, 7:56 AM)
Limitting - is relative to the loudest parts of a source.
So if you have a source with a -6db peak, and
you set the 'limitter' to -3db, then only those
-6db peaks are going to reach -3db, and the rest
of the source material will be raised in amplitude
accordingly. (proportionally)
Compression on the other hand (which can be used as a limiter also)
can raise all of the source to the set limit by amplifying
the lower source material disproportionally.
i.e lower levels recieve more amplification.
But there are many other things you can do with compression.
Err, mostly with parts missing.
Downward Compression (as opposed to upward compression): this reduces everything over the set threshold by the ratio specified. If you have a ratio of 2:1, a threshold of -20db, an input signal of -10 db, your output will be -15db. If the ratio were 3:1, your output would be -16.67db.
Limiting: Limiting in the traditional sense is a compressor set to a ratio of 10:1 or higher.
PEAK limiter or brickwall limiter (also called a clipper): This is a limiter where nothing can be pushed passed the threshold and it clips the peaks of the signal rather than compress them as in a traditional limiter. Then, usually make up gain is applied to make things louder.
So when to use which? A peak limiter is generally used in the mastering stage simply to make things louder. An old school regular limiter is what you use to tame the errant snare hit and protect speakers. You would not want to use a peak limiter like a Weiss or Waves L2 for this purpose.
Peak limiters clip transients where as a traditional limiter will not. A traditional limiter simply turns down the gain (depending on the attack/release settings a traditional limiter and really smash a transient, but used properly it won't). A peak limiter chops off transients (clips them) so that make up gain can be applied and does not clip the output.
A traditional limiter does not necessarily apply make up gain. Don't forget the effect of make up gain on a signal as it is not always desirable.
Thanks, Bubba. Good post.
Anyway–So, I’ve been dwelling on the issue of why Fasoft Compressor 1.1 raises the actual volume when the threshold is lowered, even when the output gain’s at unity. I couldn’t think up an answer, but it really does seem to, especially at around -45dB (not necessarily that I’d DO that in the real-world) with a ratio of 3:1.
So, not understanding WHY it’d do this, I pluged a record player up to my RNC1773 compressor, which is an offboard unit. The RNC does what I’d expect it to do: lowering the threshold with a ratio above 1:1 lowers the volume of the program material, full-stop.
Can anyone please help me understand why Fasoft Compressor 1.1 behaves counter-intuitively?
Fuggle
Fuggle:
I think it may have to do with the fact that the FASoft compressor defaults with it’s “compensate gain” turned on - I’m guessing it’s a post-gain stage. Adding nTrack comps to tracks will make them louder by default. Turn off the compensate gain switch and the compressor behaves like you would expect.
Woodmeister, star & flag for you…thanks!
That weirdness was really shaking me up.
F
Quote: (Bubbagump @ Sep. 22 2008, 2:06 PM)
Don't forget the effect of make up gain on a signal as it is not always desirable.
What he said...
Bubbagump
You are correct. When I said limiting, I was actually referring to
normalizing (and a particular type of normalizing). And of course
you are correct about compression.
There is only one kind of normalizing… and that is adding or subtracting a set gain to all samples in a file. Don’t let Sound Forge’s normalize to RMS mumbo jumbo fool you. That is just a crummy implementation of a peak limiter.
Again thanks 1000 for all this info, I’m getting a pretty good grasp of the concepts…
Any way, thinking of this I should have guessed before asking that well… a compressor compresses and a limiter errrr limits lol