Thoughts on Compression!!! A ha.....

[51m0n]

On 28/03/2019 at 09:57, Jus Lukin said:

Another way to look at it could be that they both reduce dynamic range. In that regard they are the same.

However, clipping, intentional or otherwise, in the simplest terms will be finding the limit of a particular component or circuit, and putting in more signal than it can deal with, hence anything at or past the limit is just not produced with any gain. Looking at the wave form it appears the sound changes shape, and therefore takes on a different timbre. This why fuzz is often used to replicate synth sounds- the wave has been turned quite 'square'.

Compression however, only ever turns the volume of the whole signal up or down. To understand them in their most basic sense, they were designed to stop mix engineers having to ride the fader to keep a very dynamic track from bobbing in and out of the mix or to avoid unwanted clipping where a loud bit would hit physical limits and get a bit square. With compression, the wave stays the same shape, but is at different volumes at different times. We can see this in the need for a release control- at the threshold, the volume is reduced so that the signal is not so loud, but if it is not turned back up, the signal below threshold will stay quiet. The comp needs to turn that volume back up, unlike a fuzz, which simply gets less dirty as it becomes less saturated with signal.

Valve amps blur the lines as they behave differently to a fuzz, but the fact that they can sound cleanish in that hinterland at the point of saturation belies the fact that they are still clipping- just really nice and smoothly.

As I said, the graphs don't help as they can look very similar, too. But while you could draw the dynamic action of a fuzz and limiter in the exact same way, the difference is that the fuzz can't get any louder at the threshold, so changes the shape of the wave as more signal is shoved in. The limiter turns the volume down so that it doesn't cross the threshold, then turns it back up as it falls away.

Equate the behaviour regarding dynamic range of a very fast hard limiter (ratio >20:1) with a lowish threshold to a fuzz, and a soft knee compressor with a very low ratio, a medium fast attack, and a very low threshold to a tube amp.

Clearly compressors/limiters are not producing the same levels of saturation (an awful lot do produce some levels of saturation at certain settings though, certainly both an La-2a and an 1176 add saturation artifacts, that's why they are 'magic boxes'), and understanding the effect on the dynamic range of  these different types of effect, and it is absolutely true that tube amps change dynamic range in much the same ways as compressors can do, and in a more extreme way fuzzes change dynamic range in much the same ways as limiters do. The saturation effects are the differences, and their psycho-acoustic effect makes what tube amps do much more obvious to us. But they both 'compress' the dynamic range.

The term is an over simplification, but the result (with respect to dynamic range) is very much the same. Otherwise mastering limiters couldn't get you to within 3dB Crest for an entire track. Crest is the measure of the difference between the peak loudness of a track and the average loudness of a track, the mastering wars were all about getting Crest as low as possible. Overdo this and you rob tracks of punch, dynamics and ultimately emotional content, plus it sounds stinky poo. Didn't stop marketing boys always opting for the louder master for years though, louder almost always equates to better unless you are trained to know the truth of what you are hearing, or have metering to help you. For instance when you apply  serious compression to a track you have to change your reverb/delay levels to avoid the track disappearing down a well, that is because quieter parts of the signal are more loudly perceived as a result of the compression, make up gain make them louder compared to the rest of the sound that goes over the threshold of the compressor, the ration of reverb to dry (typically louder) signal changes, a lot. Same with a simple bass signal. Compression will bring out artifacts in your playing that you otherwise wont hear, poor muting, string squeaks etc, whilst 'containing' other issues (thuds on muted strings as you play) dependant on which contain the most energy in the signal as a whole and what the particular compressor side chain is best at 'hearing', or set up to hear, hence the use of low pass filters on compressor side chains to retain dynamics.

Compression and limiting give you vastly more control over the effect in terms of envelope and transient manipulation, for better or worse, and this is where they shine. But in terms of dynamic range there are many similarities in simple terms.

[Jus Lukin]

-

Edited March 3, 2022 by Jus Lukin

[51m0n]

59 minutes ago, Jus Lukin said:

@51m0n, as with the other thread, it looks like conceptually we are in complete agreement.

Sure, there's the semantic matter that a reduction of dynamic range is a compression of the range, but I think that's a bit of a red herring, and on a forum in which full or part featured compressors are so easily dismissed, I'm keen to highlight the specific benefits of them, particularly apart from dirt boxes in this case.

Understood, but I consider less full featured compressors as an evil to be fought at all times.

I wouldn't be so averse to a 2 or 3 control compressor if they were properly and clearly marked controls (not things like 'glimmer' or 'compression' that are at best unclear and at worst marketing bollocks) and if they had comprehensive metering (so at least 6 LEDs switchable between input/output/gain reduction in dBs, and even better if you could 'zoom' the dB scale, never seen that in a hardware meter though).

Anything less and they are unfit for purpose, they are a 'magick' box that is only perceived to be doing something when they are doing so much as to actually damage the player's experience.

I have lost count of the number of times I've heard "compressors are rubbish, they crush all my dynamics", which is utter nonsense, what that actually means is "compressors are rubbish, because I can't hear them working until I am doing bad stuff to my playing experience, and I don't understand why". Harsh but true.

[51m0n]

1 hour ago, Jus Lukin said:

Sure, there's the semantic matter that a reduction of dynamic range is a compression of the range, but I think that's a bit of a red herring

Why is that a red herring?

In a mix I have found a change in level of less than 0.5dB can completely change the balance between 2 instruments.

So by that merit if I reduce the dynamic range of one of those instruments by just 0.5dB, then use make up gain to get it as loud as before, it will dominate the other instrument, whereas before it may have been masked by it.

At serious gig volumes everyone's ears are compressing the crap out of everything, so getting louder than something else requires a far more heavy handed approach with a static volume knob than even a very minor change in dynamic range over the time of each note. Whether that dynamic range change is the result of a tube amp beginning to saturate, or a compressor just lifting the sustain level of the note envelope a dB is irrelevant.

[Jus Lukin]

-

Edited March 3, 2022 by Jus Lukin

[51m0n]

19 minutes ago, Jus Lukin said:

Only in the context of a public discussion in which at any moment someone might say that compressors don't do anything, or won't do anything a dirty amp won't! ☺️ 

You say that it doesn't make any difference where the change in dynamics comes from, but my whole point was that to those who poo-poo compressors and don't know how they differ to a dirty (clipped) amp, the distinction is worth making.

I said I'd regret bringing it up...

Don't regret it, its a valid point, I wanted clarification of where you were coming from. Its something that has bothered me too.

Too often people (and myself included, I confess) say things like, "Nah mate you don't really need a compressor, you are driving a tube style saturating preamp/fx pedal, so you're golden".

Of course this is a massive massive over simplification. What is your goal with the compressor for one thing, what are you trying to achieve/fix/improve?

Say you are Mr R God, and you love your grinding pick playing, you enjoy driving your all tube behemoth lead sled 400w super amp until its sweaty and grindy and all kinds of grrrr, but you are losing that crisp pick attack somehow. You hear compression might help, but you dont really get it, and the all analogue optical tube pedal with one knob doesnt seem to do anything but crush your dynamics. You come on basschat and some utter twunt (ie me...) says, you need a compressor, you go on a bunch of threads and get told no you dont mate, you're golden. In fact that over simplification isnt true. A fully featured compressor set up right would allow you to do things like allow a bunch of attack through, hold the release for 300ms at a healthy ratio (say 4:1) and use the makeup gain to make that appear to be unity, the effect on the just breaking up amp would be to give the pick transient a massive boost into the front of the amp, but hold the sustain of every note at just the 'right' level to maintain the grind a tiny bit longer, fattening everything up.

But we cant be bothered to fight the good fight, because compression, you either study them and play with them for ages to really understand them in the context of a mix, or you misunderstand them and get grumpy or say something daft like, no professional bassists ever use a compressor in their rig live, leave that to the engineer...

 

Edited April 4, 2019 by 51m0n

[Cuzzie]

I must say, so far as a compression thread goes, this has been pretty appropriate and @Jus Lukin @51m0n bringing clarity to the matter contemporaneously is always appreciated.

With quality drip feed like this more people may begin to understand.

[51m0n]

1 hour ago, Cuzzie said:

I must say, so far as a compression thread goes, this has been pretty appropriate and @Jus Lukin @51m0n bringing clarity to the matter contemporaneously is always appreciated.

With quality drip feed like this more people may begin to understand.

Nah, we're about to start kicking off dude, ever since he brought my mom up in the conversation, I mean, that stinky poo is unforgiveable

[51m0n]

A really common thing to do in mixing land (yeah I know, but bear with me) is to eq signal hitting the compressor and then eq it again after that. George Massenburg (massive 'hero' of mine) explains this sort of thing so well in this video:-

If you are still interested he does something interesting with eq and compression in this one too:-

If you are wondering who the hell George Massenburg is, he literally invented the parametric eq, legend or what!

[chris_b]

And he produced, amongst others, some Little Feat records.

[Ricky 4000]

Great thread. 👍