Masterclass by Tony Andrews of Funktion One. Presented at PLASA 2009.

Sound Quality

 

 

 

In today"s world general sound quality is far lower than it was
25 years ago not in spite our digital technology but because
of it; in fact we have actually gone backwards. Early digital
was dreadful because of the limits of the technology and the
expense of doing it properly. Add that to the narrow band
width of the internet and we now have the ubiquitous MP3
and generations of people who have been deprived the
experience of quality sound.
Knowledge and understanding is not an ever growing bubble
of progress, it is in fact like money. It can be gained and lost.
Where are the people who built the pyramids and the classic
Greek Temples today! My point is that during the
development of digital we have either forgotten, or never
heard, high quality audio. Despite the fact that we now
possess the technology to easily transcend 25 year old
analogue technology we don"t, vinyl is still the most human!
I have spent 40 years working with loudspeakers because
when I started in audio they were the limiting factor on the
final result. Funktion One now has loudspeakers that in
some respects outclass studio monitors even at huge power
levels. The limiting factor today is the source material, signal
processing chain and the medium it is stored to and
transmitted with. MP3s are the final insult in a series of
events leading to bad quality source material. Good quality
sound is clean, transparent, spectrally (tonally) balanced,
free of distortion and has sufficient resolution (quantity of
clean information) to produce a stereo sound stage of depth
and dimension. To my mind the stability and vividness of the
stereo sound stage is the best indicator of good quality
program material and reproduction equipment.
Sound quality can be compromised right from the start with
bad choice of samples, instrument or vocal. For example, a
web download, an 8 bit sample, badly recorded vocal and

overloaded or clipped sample resulting in distorted or edgy
sounds. Some people actually want it to sound this way – I
can offer no help with that situation apart from suggesting
therapy.
The next place where audio can go wrong is in the mixer and
the mixing process. Assuming that any analogue to digital
converters being used are of high enough quality, the mixer
itself must have full frequency response; low distortion levels
and free of any artefacts or spurious audio nonsense of its
own. Professional equipment at unity gain (same level
output as input) ideally should not degrade the original signal
any more than a length of wire would.
Digital audio quality is often described by sample rate and bit
depth (44.1kHz 16bit for example) with the assumption that
the bigger the number the higher the quality. If only life were
that straightforward, there are many other factors
determining digital audio quality. However, all things being
equal a higher sampling rate will give greater resolution,
particularly noticeable at higher frequencies and a greater bit
depth will retain more of the low amplitude information such
as harmonics and reverb tails which give sounds their
richness, depth and uniqueness.

 

Now we come to the all important digital gain structure.

Input gains should be such that as much of the bit depth as

possible is utilised without compromising the dynamic

headroom.

 

 

 

In digital audio the bits are distributed across the input level

range of the channel or device (minimum to maximum level).

In the case of human hearing it goes from 0dB to the

threshold of pain (130-140dB) so the dynamic range of

human hearing is 140dB.

If the input signal is too quiet (rare) the effective bit depth of

the signal is reduced so lowering quality, conversely if the

input signal is too loud (more common) dynamic range is

squashed (no room to breathe) and even worse if it goes

over – known as clipping – it produces a most noxious

square wave form (poison for the mind). After MP3s this is

probably the most common cause of bad sound in today"s

world, particularly when it comes to the output stage of club

mixers most of which do not sound good in the first place.

To sum up if you do not use all the level you can without

clipping you will lose low level audio information.

Although this is a departure from the recording process I

have been describing it is a good moment to mention that

myself and thousands of other sound engineers in the world

have spent years begging djs to keep their output levels out

of the red because once the line drive amps are overloaded

or clipped the signal is trashed and there is no retrieving it.

In fact the idea of being as loud as possible is primitive and

pointless as the sound guy will have limiters and/or

compressors to protect his system from unprofessional

behaviour anyway.

Returning to the recording process bear in mind that every

effect and plug-in used has a cumulative, degenerative

effect. Do not go in and out of the analogue and digital

domains more than is necessary because every time it

happens the signal is going through A to D or D to A

converters which are notoriously difficult and expensive to

get good. Most of them will degrade sound quality. Less is

nearly always more in audio. Preservation of the integrity of

the original signal is what should be striven for. So we finally

get to the stereo end result which may be nice and clean and

of sufficient resolution. Two horrible events may now take

place. Firstly, it will be imagined that we are in some kind of,

"how much volume can we cram onto this medium"

competition.

 

Top waveform nice level, bottom waveform silly level

 

Same as above but expanded

It will have the life compressed out of it leaving no dynamic

headroom. The final insult will be downgrading to MP3 for

the web so people can fill their heads with sonic mush from

their IPods.

To put some perspective on the MP3 format, a CD wav file

at 16 bit 44.1 KHz is not enough information for true audio

and an MP3 is only 10% of that! 24 bit 96 KHz is more like it

and could exceed analogue in transmission of detail and

information. When I listen to an MP3 version of a CD wav

file I hear bass with missing low frequency extension and

also the notes are shortened so the tails are missing which

altogether makes it sound weak and insipid. More annoying

are the high frequencies which are the opposite of

translucent and crystalline, being fizzy and crunchy. I can"t

stand it. The worst thing of all however, is the loss of space

and dimension depriving the listener of involvement, feeling

and spiritual motivation. High quality stereo audio generates

a 3 dimensional left, centre, right sound stage. I think most

people appreciate this. When one is deeply, emotionally

involved with quality stereo sound a space of light, colour

and energy opens up within and the listener becomes the

sound. It is a place where much can be experienced and

learned which is why I am banging on about sound quality.

MP3s are just about ok being used with the internet and

IPods but they have no business being used in professional

situations such as clubs and shows. Just imagine taking a

low resolution internet picture and then enlarging it to the

size of a house wall. Up close it just looks like a pixelated

mess. The same thing happens with audio.