All about producing and mastering audio for disc, the web and beyond

Saturday, September 4, 2010

What is Loud? (redux)

“When I was a kid, I was filling balloons with oxy/accetaline mixtures and tying a firecracker fuse in the bottom of them and lighting them. I had one go off at point blank range as soon as I lit the fuse. That seemed pretty loud at the time. Of course, I’ve never been around anything really loud, just things that seemed loud at the time.” – anon internet

“Music is the space between the notes” – Claude Debussy

“We think the record sounds pretty good but really, we just want to be sure it’s loud.” – a renowned American String Quartet that shall remain nameless

What is Loud?

This question is at the heart of so much of our work. How to measure volume, either in an absolute sense or a relative one. How loud should you listen, how loud should the vocal be in a mix, how loud should you record a track or a mix?

The related question, What is TOO loud may in fact be more germane. If something is TOO loud than something else gets lost. In a live sound reinforcement context too loud might be described as a point when the level becomes painful….you lose your hearing.

When recording too loud is when unintended distortion takes place…you lose fidelity.

In the context of mixing, too loud might be when one sound obscures others…you lose the balance.

In mastering too loud might mean that you lose the space between the notes.

In the days of analog recording and the early days of digital there was a relationship between the nominal level of a recording (‘0 VU’ or +4dbm or –20 dbfs) and the peak level (‘+20 VU’ or +24 dbm or 0 dbfs). In fact when mixes were heavily compressed you might see that 20db dynamic range shrink to 16 or even 14 db, thought that was the extreme. So the loud part of a song would peak at a level at least 14 db hotter than the nominal level of the music. And the quiet sections of the music would reside lower than that. There were technical and aesthetic considerations that helped form this ‘equation’.

Now, with our all powerful dsp chips we can actually make recordings that have a nominal level of –8 or 9 dbfs or even LOUDER! Now, often the quiet sections of the music are living in the area that was previously reserved for the med-loud sections.
Have you ever compared two level matched version of a mix, one at a ‘mildly compressed volume and one at a highly compressed volume to see which is louder?

The reasons for this? It could be that technology has improved and we simply can make hotter records. It could be that we live in an culture riddled with attention deficit syndrome and the only way to get people’s attention is to yell all the time and hope to be heard. It could be that the typical MP3 playback system is poor and dynamics don’t translate well on the poor equipment playing over the sound of a subway train.

While it’s not likely we can change the louder world, whatever the reasons for this phenomenon, we need to understand what the implications are.

The answers to these questions emerge in the context of every project and they bear asking. I leave you to draw your own conclusions.

What does it mean to music when the difference between loud and soft is disappearing?
What does it mean when the average signal level presented to a D/A converter is so hot?
What does it mean that the excursion of a peak of a drum hit is now only 3-6db hotter than the rms level?
What does it mean when clipping distortion is an acceptable practice in recording?
What does it mean when listeners are routinely subjected to high listening volumes and levels of distortion in the recording?
What does it mean when engineers, producers and fans accept a hotter level as ‘standard’?
What does it mean when the level of a recording is sustained at a high enough level that you never get to hear the low level detail/ambience/etc.

What does it mean when you never get to hear the space between the notes…………………………

No comments:

Post a Comment