Dudes, I really respect your time and energy. But I do not understand how the compressor works. The only thing that i understood was the dB part.
I am not sure if this explanation helps. If you think of it in two dimensions, one dimension of peak signal level (y-axis) and you have the compressor in 'peak' mode and one dimension of time (x-axis) and then you imagine that the audio interface can output a max amount of signal, say in your case +18 dBu (which equals to 0 dBFS). Furthermore, let's you now set the threshold level at -4 dB, then that means the threshold is a horizontal line at +14 dBu. Visualize that when the signal exceeds past that line from below that line to above the line, then if the ratio is set to 4:1, it means that if the signal level goes above the line by +1 dB so that it peaks at +15 dBu, then the signal level is lowered by the compressor by -4 dB from +15 dBu down to +11 dBu. If the attack is 0 ms, the signal level drop kicks in instantly the signal passes the threshold. Then if the hold is 0 ms, it does not stay at the signal level drop for any time, which means the release time start kicks in instantly. Let's say the release time is set to 40 ms, then it means it takes now 40 ms for the signal level drop held by the compressor to go from the state "applied 100% amounting to -4 dB signal level reduction" to "applied 0% amounting to 0 dB signal level reduction".
This is what the compressor does, you can imagine it as a volume fader that constantly moves down and up extremely quickly whenever the signal peak level exceeds the threshold signal level. But keep in mind that the compressor also has a make-up gain, so let's say the make up gain on the above example has been set to +4 dB, then it means that the output of the compressor, whether it is in a state of compression or in a state of non-compression at any given moment, it will always gain/increase the signal by + 4dB. Now this then means that all of the signal that did not exceed the threshold level, is gained in signal level. So you have some of the
signal going into signal level reduction and some of the signal going into signal level boost, in the above case it means that the signal was reduced in signal level by -4 dB, but the makeup gain made it +4 dB, so the net gain was +0 dB, no gain in this case for the signal that passed the threshold, all other signal was boosted by +4 dB due to the make up gain. The result is that the peak-to-rms of this signal has now dropped, meaning the dynamic range has dropped, and the rms/overall/average signal level has increased, all due to the compressor compressing the signal. With a normal gain knob the rms signal level increases by boosting the gain knob, the difference being that when you boost the signal past the output signal capacity of the audio interface at let's say +18 dBu (amounting to 0 dBFS inside of the DAW), then the signal will start to distort at the peaks as soon as the peak signal level exceeds +18 dBu, because the audio interface cannot output that much signal, its design/construction causes nasty clipping distortion whenever that happens. So in order to allow more overall volume to occur without clipping the signal, you basically use a brickwall peak limiter, which is basically a compressor with very high ratio and very low attack time. So now instead of gaining the signal and quickly getting clipping distortion, you can now keep boosting the volume more before the clipping occurs, because the compressor ensures the peaks are kept below the ceiling (below 0 dBFS, below +18 dBu).
Now, what your max input and output signal capacity of your audio interface is, that is a separate discussion, but a very important one! You can bet that (at least the high end) pros are not using weak gear.
High end masters with or without a brickwall peak limiter applied, can typically for a CD quality AAC encoded signal have the ceiling/True Peak at -0.6 dBFS, which means no peaks are let beyond that level. A typical example of this would be on Carrie Underwood's hit mix "Heartbeat" on her latest CD album mixed by Chris Lord-Alge. I highly recommend that you follow this practice for a better sounding CD master.
Now, why you use compressors is due to a number of different reasons. But it is important to know that compressors control signature/timbre of the sound, because compressors have different attenuation and de-attenuation curves, different gain characteristics (dynamic/non-linear or linear) and different features in what you can control and how they can behave, some even have a dynamic threshold. All of that means you now can shape the sound in different ways depending on the dynamic characteristics of the compressor. Tons of more around this, but it's important to understand that because various compressors have various nature due to how they were designed to act on the signal, they themselves have a signature on the sound. And it's basically about finding sweet combos between sound source(s) and comps. Sometimes the best is to combine a set of sound sources and various comps. An example of that would maybe be a bass and a kick drum, into a particular series of compression that when the signal is "signed"/"characterized" that way, it just sounds awesome. So the art is really to become good at understanding various combinations between playing, sound source(s) and compressor(s). You might have heard about the 1176 compressor, I would say that is maybe one of the most loved compressors out there because of its overall soft/warm signature on the sound.
The creator of a compressor effect can create the compressor to have dynamic behavior on all of the compressor settings mentioned. It can be totally content dependent, partially content dependent, it can have look ahead, oversampling, expand capability and so on. So the compressor, although it appears simple, it's not so simple, because it's really up to the creator of the compressor to decide what kind of impact the compressor should have on the signal. This is why the compressor is such a great effect for instance in sound shaping. In combination with compressor techniques like side-chaining and two stage compression, all of a sudden you can now start to control the dynamic nature of not only individual sounds, but in the interaction between sounds by applying it on groups of sound sources, between groups and on the mix itself. How good that sounds is basically a matter of how great you are at working with dynamics and how nice the compressors are to the signal. It's also an artistic choice, you might for instance want some more hard characteristic on some sounds in the mix, to enhance the soft characteristics of some other sounds and stuff like that. It becomes very deep what you can do and what the effects become. You might want your mix to pass a series of dynamic profiles in vary advanced ways to attach the listener to the mix in very specific ways. Only your creativity, quality of technique, quality of application and access to various compressors are the limits...
Your best win-reward is not in understanding the compressor to the extreme detailed level, your win-reward is in the perspective of compression, your access to compressors and in your level of creativity with the compressor in order to create attractive vibe.
