Dynamic Compression – benign or evil?

I’m not talking about the ‘loudness wars’ and the way modern popular music is compressed to within an inch of its life, but merely what might pass as an ‘audiophile recording’, yet which will most likely have had a modicum of compression applied. This will have happened because it is conventional audio wisdom that the dynamic range of the real world has to be reduced in order to reproduce it on a hi fi system. Even classical recordings will generally have had some compression applied; replay systems that can reproduce a symphony concert at “realistic volume” are supposedly a myth.

I don’t agree that compression should be mandatory. For a start, I think that dynamic range is often confused with maximum volume level. Dynamic range per se is a fairly academic measure: effectively the ratio of the loudest and quietest levels in the recording. If it is a live recording, then the ‘noise floor’ is defined by wheezing, snuffling, shuffling, coughing, sweet wrapper rustling, chattering audience members. If not, it may just be the air conditioning and the involuntary vocalisations of the musicians (which seems to be quite a common problem). The dynamic range, as such, certainly isn’t going to trouble 16 bit audio. The maximum achievable level is a different issue.

From experience, sitting halfway back in a concert hall, a full symphony orchestra is not all that loud; I am confident that large speakers can reproduce the maximum level and the dynamic range in someone’s living room. (Clearly some people think that in order to do this, the speakers have to be the 50 piece orchestra – which of course would not be possible – rather than to merely reproduce the sound level as heard at a sensible distance.)

Small speakers, however, cannot reproduce a symphony orchestra or rock concert at a realistic ‘audience volume’ (- or at least not the full frequency spectrum) and people don’t always want to listen at high volume, anyway. The common technique of close-miking instruments tends to hardens the sound. Compression is usually assumed to have a benign effect that ‘takes the edge off’, allowing people to listen at moderate volumes without ‘the quietest bits’ dropping into inaudibility below the level of the microwave oven and washing machine. Dynamic compression is mandatory with vinyl because… well, for many reasons. Using it, it is still possible to perceive excitement, and the musicians playing loudly, even though the absolute increase of volume is small.

It never used to bother me, but it sometimes does, now. Thinking about it, why should something so drastic be completely benign, anyway? I can think of several weird things that it is doing:

  • The obvious thing is that it is continuously modulating the gain (duh!) and this has surely got to increase the work that the brain has to do in tracking what is going on. Sometimes individual elements in the mix will have their own compression applied, and sometimes it will be more like blanket compression that modulates several sources at the same time – just depending on who applied it and at what stage in the production.
  • The system pulls its punches on dynamics, denying the listener the sheer thrill of a great crescendo or the contrast of an unexpected ‘quiet bit’.
  • The strangest thing, I think, could be what it does to our perception of the frequency content of what we’re hearing. It partially neutralises the Fletcher Munson effect, meaning that we hear more of the bass and treble extremes in the quiet bits than we should, and are denied the full spectrum in the loud bits. Acoustic sources naturally change their frequency content as they are played harder, but with compression, we don’t get the full corresponding volume increase. It probably sounds like nonlinear distortion – another thing for our brains to cope with. In addition, the frequency content may be being explicitly modulated using a technique called multi-band compression.

Some classical labels pledge not to use dynamic compression, and they usually seem to include more ‘ambience’ than other recordings i.e. are not miked so closely. In my experience they can sound amazing. However, they must be played back at ‘realistic’ volume – not a great drawback in my opinion!


One thought on “Dynamic Compression – benign or evil?

  1. I agree with you. I think that an orchestra can manage a maximum dynamic range of about 80 decibels but the usual maximum is about 50. Where I live the back ground noise floor is about 30 to 35 decibels and, therefore, I can play a concert without having to adjust the volume because the orchestra is not playing too loudly when going “full steam” or about 80 decibels at my ears. I usually play pop records at around 70 to 75 decibels which is about 40 decibels above the noise floor.

    When I was a young lad living in the country in Wales the background noise floor would have been much lower. When I ran to school I rarely saw a car and I could hear one coming from half a mile away. A concert hall in those times was very quiet indeed you could easily hear the softest of sounds of an orchestra.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s