That's because there is no one "technique", but rather many, and many tools. I've already listed some of them in this thread. It's not really pertinent to describe everything in specific because it wouldn't be accurate for every track. Some of the tools are (not intended to be a comprehensive list): • Compression on a per-track basis • Limiting on a per-track basis • Peak limiting on a per-track basis • Overall compression • Overall limiting • Overall peak limiting • Overall clipping Compression, limiting and peak limiting can be applied to the entire spectrum or to portions of the spectrum with individual control over characteristics via a crossover. Clipping is a valid processing tool, and not always a distortion generator, as it takes time for distortion to become audible, and if a very short peak is clipped off the results can be completely inaudible. Generally speaking, compression increases long-term average loudness, limiting increases short-term average loudness, peak limiting also increases short-term loudness by controlling maximum peak amplitude, and clipping permits the others to work with less audible side-effect by acting only on short inaudible peaks. Split-band processing can also act as a form of dynamic EQ while permitting higher average levels with less audible side-effect. However, side-effect of gain control, usually a type of "pumping" effect, is not always undesirable. There are differences between how levels are detected, which affect how a processor response. They can be peak-based (peak limiting and limiting), peak, RMS or average based (compressors). And there can be differences in attack and release times for each, as well as differences in the thresholds for each, as well as release-gate thresholds that alter release times based on instantaneous levels. It's a Swiss Army Knife of tools. All of this can be used, some of it, or very little. The mix and balance of any of these can be adjusted relative to the others, and even the order in the chain can be changed, though typically track processing comes first, the overall compression, then limiting, peak limiting and possibly clipping. All tools could be used so gently as to be imperceptible, and so aggressively as to be the loudest track ever, and everywhere inbetween. The point is, there's no one magic tool or formula, and there's also no knowing what was done without first-hand information. Looking at a waveform in a DAW doesn't tell you what was done in anything other than very general terms. You can't, for example, assume there was clipping used because the waveform looks flat-topped. That can be done with peak limiting too. Does that answer your question?