Audio Schizophrenia
Am I alone in thinking that the
audio industry has become somewhat schizophrenic these days? On the one hand,
we're striving to improve performance by raising both sampling and bit rates, 24
bit/96kHz being one example. On the other, we're into data reduced formats like
the popular MP3, Minidisc and Dolby Digital, which work by throwing away up to
90% of all the audio information in places. Don't you find that strange?
It's funny how the 16 bit / 44.1kHz format of CD has been widely criticised as
"inadequate", yet it appears to me that not only do the public appear
to be quite happy with CD, they also appear to be quite happy with all the data
reduced formats, which are of lesser quality!
CD's 16 bit dynamic range is considered insufficient these days, yet that equates to 96dB. If you were to record an orchestra with 16 bits and not compress it, you'd immediately get complaints from some that the dynamic range was too great and "could you bring up the quiet bits"? And that's just from 16 bits! CD's are routinely compressed to make them more "acceptable". When I worked for Decca in London I re-mastered Karajan's 1959 recording of Richard Strauss's Also Sprach Zarathustra with the Vienna Philharmonic Orchestra. As usual, I transferred the original AME analogue tapes in as pure as manner as possible and the dynamic range was stunning, considering the age of the recording. And, when it was issued on CD, there were still complaints about the huge dynamic range!
Now look at radio. The trend in recent years is for
everything to sound LOUDER. To this end, ever more compression is used.
In the UK, Classic FM has virtually no dynamic range at all and I have heard
Haydn's "Surprise" Symphony where Classic FM's surprise was that the pp passage
just before the famous loud chord was louder than that chord! Even BBC Radio 3,
once the benchmark of high quality sound, where you could almost believe that
you were "there", has in recent years both gone down the road of fakery (you
wouldn't recognise London's Royal Festival Hall from a Radio 3 broadcast as they
have added so much artificial reverberation) and compressed its output,
particularly during the day. Worse, BBC Television, which now broadcasts
more Proms than ever, has joined its radio cousin in the bad sound dept. TV has
always sounded worse than radio, mainly because of the need to cater for viewers
listening through the TV's own speaker, which invariably is of poor quality. But
as the quality of those speakers has improved over the years, the BBC's sound
seems to have got worse rather than better. Listening to the 2006 Proms on BBC4
and BBC2 via high quality B&W speakers revealed that the fiddling of the dynamic
range in particular had plummeted to new lows. So low, in fact, that it amounts
to musical vandalism akin to that of Classic FM, who at least don't pretend to
give quality sound. Complaning gets you nowhere, as a retired BBC engineer
friend found out. His comment was that the BBC has become supremely arrogant as
far as technical quality is concerned, yet that quality is as bad as it has ever
been. Isn't that strange in an era where we have the capability to have superb
quality sound yet seemingly choose not to?
What's more, all this compression is done by a machine called the Optimod. The engineers only have to decided on the level of compression (maximum on Classic FM?) and the box does the rest. One presumes that the box had a good musical education. The pop stations are even worse. It sounds to me like a dynamic range of about 10dB. Ironic when so many new recordings are made on 24 bit equipment with a dynamic range of around 124 dB and then it's all squashed down to about 10. They could have saved an awful lot of money and made the original recording on Midi Disk for the same result! The move to Digital radio, which might promise better sound quality, especially as the home listener can control the dynamic range of the material he/she is listening to, has proved to be a great disappointment. As always, money got in the way. Why take up space broadcasting a few radio stations in high quality sound when you can broadcast many in crap sound? Many people, myself included, have kept our analogue tuners (I bought my Pioneer in 1975 and it's still going strong) because they give better sound quality than new digital ones.
Higher sampling rates are also becoming common. The advantages of sampling at 96, 192 or even higher rates are much touted these days. But who can hear it and do the public, who seem to be perfectly happy with their 44.1kHz data reduced formats, care? How many microphones, amplifiers and loudspeakers can actually capture and reproduce up to 44kHz or more? Most tail off above 20kHz. More fundamentally, can we humans hear it? Only children can hear up to about 18kHz and the average adult can't hear more than about 15kHz. The only advantage of high sampling rates that I can see, is that it pushes the low level ultrasonic artifacts created by anti-aliasing filters present in all digital recorders up by an octave, but this will be very subtle.
I had an e-mail from someone the other day worried about microphone self-noise. This is another of those oddities, as the noise floor of most venues is louder than the noise floor of any decent professional mic. In fact, finding a quiet venue has become rather difficult these days, due to traffic, aircraft and all manner of other noises.
I find it ironic that today, when the quality of both professional and domestic equipment has never been better, that we still feel the need to fiddle with the sound. In the '50's & 60's there was a real need to both limit (or compress) and EQ a program due to the constraints of the medium (LP) and the relatively poor quality of domestic replay equipment. Today, that's no longer the case. All the domestic replay mediums can handle a huge dynamic range and today's loudspeakers can reproduce sound with great accuracy. So why do we still EQ and compress our CD's and DVD's? It's that 'boys and their toys' thing again. The desire to play with equipment is as great as its ever been, even when it now does more harm than good.
On a similar tack, with 5.1 SACD recordings, we have the potential to reproduce stunningly natural recordings in the home, yet most engineers simply see 5.1 as more channels to play with. Makes me wonder - why bother?