This may sound like a really dumb question but it goes more into the inner understandings of said instruments instead of their practical use. I was kind of studying up on the physical workings of distortion and how they are basically just waveshaping. But as far as I am concerned, compressors and limiters meet a very similar goal with input and output levels. To me it seems that the biggest difference (and probably the one that gives it the effective difference as well) are the attack and release time controls.

So what I'm wondering about is how do compressors and limiters regulate the transients and overall volume of the material I put into it without (ideally) clipping the original signal in any way? It's not like those instruments actively pull down my mixer faders at times the signal comes past the threshold. For that I would have to use volume automation (which is a very good alternative to sidechain compression) and is something very different. I kind of know that the avoidance of clipping within compressors has to do with the threshold and attack and release times but that's not really specific enough. Because unfortunately that doesn't tell me exactly why I can avoid clipping when setting these controls the right way and when I could achieve certain amounts of clipping within my compressor if I wanted to. So how exactly do those instruments avoid clipping while regulating volume and how could I purposefully induce clipping from said instruments if I wanted to?

I hope this entire topic is not too vague or theoretical because it probably gets into the physics of the entire ordeal quite a bit. But I sometimes have a very hard time just taking things for granted without wanting to really understand their inner workings. Little side question, since compressors and limiters adjust the transient/volume levels of your material but with possibly clipping the material when dialed in the wrong way. Would I possibly get any clipping if I would pull down volume faders on my mixer fast enough?

 

If you want to experiment a little, open an audio channel in your DAW and put a sine generator, say 1 kHz, the compressor of your choice and a fft analyser like izotope Insight. Try with various input level. I know that is a "static" situation but you can see a lot of things. Vary the usual parameters... attack threshold release etc. Some compressors have only odd harmonic dist, others have a sort of artificial noise floor in order to seem more analogue, others have the super rare 2nd harmonic dist., others a lot of aliasing... a bit funny and interesting. But in the end the final judgment.. your ears

 

Isn't that quite a similar concept though? The compressor is automating signal level based upon signal input strength, and the way it attenuates the signal is configured by ratio/attack/release settings. It's like telling an invisible hand to turn down the input by x amount (reduction amount being expressed in a ratio to original signal strength) when signal exceeds the threshold, and the nuance of how sharply/quickly to do so is controlled by attack/release. Extra control parameters to fine-tune the metaphorical invisible hand would include a soft/hard knee setting.

The difference with fader automation is that it will (generally) affect the signal at the end of the processing chain, rather than the point of insertion like a compressor/limiter does.

Edit: Clipping is often the result of either makeup gain or setting a reduced threshold. Think about Waves L2 as an oft-abused example. If you set the ratio to infinite (total reduction when signal exceeds threshold) and attack to instant (meaning ANY transients are instantly acted upon so none sneak above the threshold pre-attack stage) then anything exceeding the threshold gets its peaks lopped off very abruptly. The signal above threshold becomes a straight-up flat plateau. This isn't such a problem if it's a very short blip, but when significant amounts of program become clipped like that then the distortion is very audible. That's why setting the threshold below isolated transients and into a significant portion of signal is bad form. Waves L2 (being a limiter) is designed to work exactly like that, except that it lets you control how little/much damage is being done by altering the threshold. Clipped peaks are the most effective way to provide more headroom for makeup gain, but ANY reduction of threshold below ceiling is guaranteed to damage your signal. Clearly a lot of people think it's worth the trade-off, though (Death Magnetic, *ahem*).

Excessive makeup gain (gain that exceeds the headroom provided by the signal reduction during compression) works the same way because digital 0 becomes the threshold rather than the adjustable arbitrary threshold. Instead of bringing the threshold down into a large chunk of signal, you're just pushing the signal above the ceiling which is essentially a fixed threshold. The net result is the same: nasty, flattened-off, clipped peaks.

 

Last edited: Jun 27, 2016

Well most peak-limiters have a look ahead function built-in (Some allow you to switch in off and other allow you to enter how much you want them to be able to look ahead into the future of the in-coming signal and prepare for processing it). As has been mentioned your best bet is to play around with a variety of them until your heart is content. The clipping applied is normally of the soft variety but indeed can turn into square waving your mix as has been brought up, That is why even with some of the best DAC which money can buy there is only a rather small window/sweet spot where "shredding" the mix through them to obtain that really slick, crisp yet transparent result is available. As clippers do not have and should not the ability to look ahead to see/prepare for what's coming, That's the major difference as well as they do not raise the RMS level either (Hence why you'll find them after compression and peak-limiting or used in such a manner)

FabFilter's Pro-L is one I'd recommend for fiddling around with all the various parameters which work in conjunction with one another to offer one of the best peak-limiters available, Elephant 4 by Voxengo is also a good one for that...Compare against say Waves L-Series and the flexibility on offer/to hand becomes so immediately apparent/obvious, However I do also like L3 as it has it's own charm/thing going for it.

So yeah meddle around/get your hand's as dirty as you can comparing is all I can recommend, As the above posts by the fellow gentleman and ladies, Always good folk around here that cover everything/highlight every important aspect, It's why I really enjoy the forums here, Non of the uber-salty, bitter types that are found on most of the others, Also just plain snobbish sorts that equally rub me the wrong way/Just too aggressive in making their opinion the right one and everyone else ought to bow down sort of attitude. Thank you AudioSEX for not being that way, It means a great deal and muchos <3

All the best to all as always

Dean

 

Last edited: Jun 27, 2016

Thanks for the input guys, I did some fiddling around with a compressor, oscilloscope and spectrum analyzer already but still some stuff eluded me. The thing is that I just used to take compressors and limiters for granted and never really used clippers in any of my songs nor did I think about either of their inner workings. I guess I didn't really acknowledge them and their uses because I did whatever I wanted to do in separate steps by adding a deliberate compressor/limiter and distortion or saturation plug-in. But now I started reading up on the physics and none of what it did actually made sense to me anymore.

What really helped me to come to a certain understanding was reading that I misinterpreted the compressor or limiter time controls. To be honest, I think most people do but don't really pay attention to it. I always thought that a compressor attack or release time meant that it would let through the peak surpassing the threshold for that amount of time before acting and it would then act instantly after set time. But from what I've just read is that it's more like a lag time you put in for your compressor or limiter to "buffer" their volume adjustment. So where something like a clipper or waveshaper wouldn't have any time controls and act instantly giving you distortion or saturation, the compressor or limiter takes the given time to make the adjustment and by doing it gradually like this. So as long as the adjustment time isn't shorter than a single run of your waveform, you shouldn't impose any unwanted distortion/saturation on the actual signal.

 

This is a "beginner's guide", but I found useful information here, and I got a better understanding about different types of compressors. http://music.tutsplus.com/tutorials/the-beginners-guide-to-compression--audio-953

 

Most the stuff in here is obviously old news, but the little part about the type of compressors was nice so thanks for that. I may search a bit more about the inner workings of these different type of compressors. Most of the other educational tools and explanations I used to learn how to use the compressor don't really deal with that part as far as I know. They just go "here, this is a compressor and it impacts the sound likes this" but never really highlight the difference between compressors and usually even downplay it by stating they all accomplish the same thing mostly so why consider any differences at all. Having never worked on analog gear (too young mostly) knowing about the different hardware compressors really helps out.

 

Also, I have read there is no standard for attack. It is not the time for how long it takes for the compressor to reach the it's full ratio, but about two thirds of the ratio is more close to common. But it can varyhttps://www.attackmagazine.com/features/columns/gregory-scott-demolishing-the-myths-of-compression/.

 

( http://www.dancemusicproduction.com...ons-tutorials/sessions-04-dynamic-compression is quite good)

 

Last edited: Jun 27, 2016

In my mind, and sometimes if it right for the source that I am recording, judicious use of overdriving a preamp thereby creating distortion is just what is needed to put a track a little deeper in the mix or even more upfront (think rock and roll vocal), to help define it's whereabouts in the sound field . This would be instead of or in concert, parallel with a compressor to achieve certain desired affects. Depending on where, how and with what instrument you are recording, you can accomplish different objectives. One can reign in, for example, a harsh high hat by lobbing off the transient of each hit to perhaps avoid using eq later on down the line and save yourself and your listeners ears, and in addition bring it further away from the front of your mix. Or dial in just enough gain to help a slightly dynamically erratic singer not to jump around and too high in level (smacking the ceiling and pulling in digital growl from the converters) at the same time as letting the fizz in this instance take center stage in the center of your mix. Using the input transformer of your preamp as a soft compressor is a nice way to add color to your tracks. You have to use your ears to find the sweet spot for each source. It takes some practice, and every preamp has it's own mysterious ways, so experimentation is vital to making good choices in picking where to implement this technique , understanding, mastering and enjoying the soft clipping of your audio signals. Output voltage is your friend.

 

http://www.soundonsound.com/techniques/compression-made-easy
This one is probably a safe bet for one that covers all the areas, Although not as in-depth as some other articles on specific areas but don't mention the fundamentals. I started before working fully ITB wasn't feasible so working with hardware definitely helped but I tell you what, When working ITB was do-able and available, It took me a while to get used to setting up properly for the input side of each, Especially when having to switch...Oh I'd start of often with almost no signal but a lot of tape noise and then a clipping monstrosity on the other side to start off, I was just glad when it went down to one setup to stop messing with/frying my brain, I'm sure you can imagine the look of "excitement" on band's faces when I'd smoked too much and setup/gain staged for the wrong bleeding format

Many a

All the best, Dean

 

They do actually "actively pull down the volume fader"! They're using a circuit to track the current input level over time and then reacting accordingly. That's the key to understanding the idea behind sidechaining too. With sidechaining, you're just feeding this detection circuit a completely different track, but letting it act on this one.

There's no special magic in them from that POV, it's literally just automatic gain reduction. The magic is that they can do it inhumanly fast for you, which can give its own desirable sound to the result ("ooh that's phat").

They can still clip at certain settings if your signal has peaks that are too unruly. The analog models sometimes improve the peaks they clip because they "round them off" with their analog clip behavior which is a little different and softer compared to hard-chopped off digital peaks.