Saturday, February 28, 2009

FFW03: Level headed

It's late 1998... popular music is getting increasingly louder and I'm feeling increasingly exasperated. I gave up listening to commercial radio years earlier due to the excessive amount of processing used to make it loud, and now commercial CDs are sounding just like commercial radio - continually loud and in-your-face, with no contrast, no room to breathe, and no space to appreciate the little things that make music great.

The saddest part about this time in history was that recording technology had advanced to a point where we could have all the dynamic range imaginable - more than could be reproduced through an analogue playback system, in fact - and yet sound engineering had 'advanced' to a point where it wanted virtually none of it. The Loudness War, as it came to be known, was in full swing. Since then, loudness has triumphed and sound quality appears to be lost forever - at least in commercial music. But like a beacon of hope, the third and final part of my interview with Rupert Neve, published in this issue, closed on an uplifting and reassuring note. As long as designers like Mr Neve are around, there's still hope for quality audio!

Here's First Word #03...

Level headed
“Not more Chesky!” cried Philip Spencer, AudioTechnology’s sales director, as I tossed a new set of discs in front of him. “Chesky! Chesky! Chesky! It’s all you ever listen to. You’d think they were the only record label in the world!”. “Well,” I said defensively, “they are one of the only labels releasing 24-bit 96k recordings on DVD…” I could see I’d gotten his attention. “You, er, want to hear some?” “Yeah, right, with what?” He cautiously surveyed the room, then fixed a triumphant stare on me. “In case you haven’t noticed, we don’t have a DVD player…” Moments later, Sony Australia’s Peter Norman waltzed through the door with his usual boyish grin. “Hi Greg, here’s that DVD player you wanted to borrow.” While Peter and I got in each other’s way setting up the DVD player, Philip reluctantly tore the shrink wrap off the discs while muttering something about doing some real work. “But Philip, this is work”, I taunted.

Peter had brought in a Sony DVP-S715 DVD player, which plays DVDs, CDs and Video CDs. It’s a fine sounding machine, with a good solid drawer, a well-appointed remote control and, most importantly, 24-bit 96k converters. It’s one of a select number of DVD players capable of playing 24-bit 96k audio, and we were about to hear this emerging format ourselves for the first time.

The source material was Chesky Record’s excellent Super Audio Collection & Professional Test Disc. Along with some very useful test tones, Chesky have compiled a selection of their 24-bit 96k recordings onto the DVD Movie format disc, which supports two channels of 24-bit 96k audio. These are master recordings of tracks that have previously been released on Chesky CDs, so it’s easy to do a direct comparison between 16-bit 44.1k and 24-bit 96k versions of the same recording. Which is exactly what we did.

So what does this new 24-bit 96k format sound like? Personally, I think it’s stunning. It’s not a radical difference from 16-bit 44.1k, but it’s the right difference. If human perception has a quality threshold – a minimum level of sonic quality required for things to sound natural and real – then 16-bit 44.1k audio falls below that threshold, while 24-bit 96k clearly exceeds it.

One of the standout tracks on the Chesky DVD is an acoustic cover of Stevie Wonder’s ‘Isn’t She Lovely’, performed by Livingston Taylor. This beautiful recording takes full advantage of the 24-bit 96k technology’s wider dynamic range, higher resolution and extended frequency response. The CD version is noticeably lack-lustre when compared to the DVD, yet, as far as CD quality is concerned, it’s still a standout recording. Which brings me to the point of this editorial: dynamic range. It’s the wonderfully subtle dynamic range of this recording, on both CD and DVD, that really got me thinking.

One of the theoretical rules for digital audio states that you get 6dB of dynamic range per bit (actually, it’s 6.0206dB, but who’s counting?). So a 16-bit recording has a dynamic range of 16 x 6dB = 96dB, while a 24-bit recording has 144dB (in practice, problems such as thermal noise restrict this to around 120dB). How much dynamic range do we need? Let’s consider the dynamic range of human hearing. If the softest sound we can hear is 0dB SPL (Sound Pressure Level), then the threshold of pain is somewhere around 130dB SPL. That gives human hearing a dynamic range of 130dB. Yet many of our latest commercial recordings have a paltry 6dB or less!

Dynamic range – the ability to play softer or louder – is one of the most important forms of musical expression. So why are we so hell-bent on reducing it? There is a commonly held belief that for a recording to be commercially successful, it must be as loud as every other commercial recording, or louder! Because we can only get so much level onto a CD, this perceived loudness is created by applying powerful compression during the mastering stage, thereby sacrificing dynamic range.

Many musicians compare their mixes against songs heard on commercial radio. Invariably, the radio songs have more ‘punch’, so musicians want more punch (i.e. compression) applied to their mixes to make them competitive. What they fail to understand is that commercial radio stations apply considerable processing, including multiband compression, to their audio before transmitting it – and that’s what produces the extra ‘punch’. Broadcasters have their reasons for doing this; some good, some bad. Nonetheless, what we hear on commercial radio is a heavily processed version of the artist’s mastered mix. Using commercial radio as a mixing or mastering reference is a big mistake.

Once upon a time, all mastering was done by talented specialists who understood the concept of dynamic range and the processes used in broadcasting, and strived to produce a finished result that translated well into any environment. These people truly deserved the title of ‘mastering engineers’. Many are still in business today, but staying in business means doing what the client wants – including heavy compression. It compromises the sound quality, and it therefore compromises their professional reputation.

Worse yet, the true professionals now face a particularly ugly competitor: backyard wannabes with PCs, soundcards, six months experience and the nerve to call themselves mastering engineers! Ask these people to define ‘mastering’ and they’ll tell you it’s the process of making each mix as loud as possible. As Charlie Brown says, “Good grief…”.

Maintaining your dynamic range is an important part of maintaining your musical expression. Remember, it’s the soft parts of your mix that make the loud parts ‘loud’. I’m tired of listening to music that’s compressed to the max and continuously in-your-face. I’m tired of listening to commercial CDs that are designed to make my high quality monitoring system sound like commercial radio. It sounds bad, it’s fatiguing to listen to, and it implies that I, as a listener, have a limited attention span. It’s just as insulting as the excessively loud canned laughter you hear on TV sitcoms. AND IT’S JUST AS ANNOYING AS PEOPLE WHO WRITE EVERYTHING IN CAPITAL LETTERS.

So, when it comes to mastering your mixes, keep a level head and say ‘no’ to heavy compression. (Your mixes will still sound punchy on the radio, because commercial broadcasters will still be doing what they’ve always done.) As for me, I’m saving to buy a DVP-S715 DVD player and as many well-recorded 24-bit 96k DVDs as I can get my hands on. Philip, on the other hand, is waiting for the Spice Girls to be remastered at 24-bit 96k. In the meantime, he’s borrowing my ear plugs…

No comments:

Post a Comment