There
are an enormous number of variables involved in
the recording of audio and in the reproduction.
Perhaps that could be the attraction of the
ongoing endeavour to capture a stereo soundstage
and retaining as much of the reality as possible
in the playback, knowing that it will never be
perfect!
The
major areas of variability are in the choices of
microphones, the variations in the way they are
used to capture the stereo soundstage, how many
are in use, their placement in the acoustic, the
characteristics of that acoustic, and the
judgements made in what ever ‘monitoring’ room
is available, apart from the loudspeakers in use.
Whatever is finally available to the home
listener, is then subject to another plethora of
variables, and it is with this aspect I want to
start.
THE
PLAYBACK
A
number of years ago I recorded, in dozens of
different listening rooms and different playback
systems, the sound of one particular CD. In each
case, as near to a similar sweet spot, symmetrical
to the speakers, at like volume levels, the
recordings were made via binaural mics in my ears.
Playback comparisons were then made via quality
open back headphones.
These
playbacks were all entirely different, let alone
the one where the speakers were out of phase and
another where the tweeters were defunct! Apart
from differences in loudspeakers, there was their
placement, symmetrical to the room or otherwise,
the room sizes, their dimensional proportions and
different reflections, low frequency resonances,
furniture layout and floor covering.
There
is obviously never going to be a ‘playback
standard’, but there are perhaps some ways of
reducing the multifarious number of effects
mentioned above. Inherent in this is the
construction of your property - solid brick
interior walled rooms are different to timber
framed panel constructed interiors, as I noticed
when I moved to the latter type, which I straight
away preferred.
Traffic
resonance was obviously noticeable in the earlier
home, but not so now. However this would also be a
factor in the LF reproduction of the loudspeakers,
as also would be the size of a room, apart from
the loudspeaker’s design characteristics and
positioning – you can’t win!
Loudspeaker
manufacturers ‘favour’ a number of choices of
design types. For electromagnetic drivers there
are sealed boxes or various ported types, an
extension of this being the transmission line.
Electrostatic loudspeakers justifiably have their
enthusiasts, however you do have to consider their
positioning, due to their bi-directional output.
There
was a characteristic of a well known electrostatic
loudspeaker, that I used to have. This was the ‘point
source’ of the full frequency range reproduced.
This is also a noticeable facet of some dual
concentric coned mid and high frequency drivers I
have used. With two or three driver designs, it is
likely that they will be vertically in line, or at
least the mid and high should be.
These
aspects are considered to be an important factor
in obtaining clear imaging. So, if your
loudspeakers are as symmetrical to the room as
possible, away from corners and walls, preferably
angled in to the ‘sweet spot’ listening
position, it is worth checking out that you get a
clear centre image with a mono recording. With
passive loudspeakers you could temporally wire
both from the same amplifier output to directly
test the speakers themselves. All decent playback
systems should have a mono push button to easily
check all is well, give or take any waxing up of
our ear’s listening abilities!
The
timbre tonality differences, or colourations,
between the original sound and what is reproduced
are likely to be more from the loudspeaker
playback than due to what the microphones added,
and of a different nature also. An aspect to
consider is that boxed loudspeakers can radiate a
sound from their side and rear panels, and
electrostatic loudspeakers are inherently free of
this box effect possibility, apart from needing to
take into account their ‘out of phase’, rear
output.
THE
RECORDING
I
don’t know if it is definable what we should
expect from the reproduced stereo soundstage? My
view is that the loudspeakers, as individual sound
sources should not be directly noticeable. The
stereo should start in the middle and spread out
towards the loudspeakers, with a good imaging and
balance of instrumental sources between the
loudspeakers and a ‘detachment’ of the
recording venues ambience behind and wider than
the performers.
Good
location imaging of the performers in the
reproduced stereo soundstage, can allow the
listener to choose to focus on a particular
performer or group of performers. This asset is
less likely if the musical balance is poorly done
with an unwanted dominance of some other sources,
give or take what the composer indicates in the
score, and how that is interpreted by the players,
or by the conductors in the cases where they are
involved.
It
was interesting that in the early days in the move
to stereo, there was a ‘ping pong’ approach on
some recordings, where you had ‘mono’
instruments coming direct from each of the
loudspeakers! Mind you, I have recently heard a
5.1 Surround Sound DVD of a string quartet, where
they have placed the violins fully in the front L
and R loudspeakers, with the viola and cello
coming from the two rear loudspeakers – I don’t
know of any concerts where you can get the
audience to sit in the middle of a string quartet!
However,
microphones are definitely used closer to the
performers than almost any of a typical audience.
In addition to that there are often difficulties
in using the layout, that some groups of
performers always want follow, to get a decent
recorded musical and stereo soundstage balance.
Consider Brass Bands and smaller ensembles like
String Quartets.
The
former have two major groups of performers
directly facing each other, with a third group, at
the back of the performing area, with the
percussion behind them. If only those two groups
to the conductor’s left and right could be ‘opened
out’, their instrumental sounds would come over
much more directly and ‘naturally’ on the main
pair above and behind the conductor!
With
String Quartets etc, they are obviously so used to
rehearsing in much smaller places, than a concert
platform, and are so used to directly hearing each
others contributions. In the String Quartet’s
case the balance from the second violin and the
viola is often compromised, unless the quartet can
be persuaded to open out a bit, to suit a single
stereo pair mic approach.
There
is one potential pit-fall with location
recordings, that can be so obvious when you get
back to base. You hear the live ambience in the
venue while you are setting up, and when you get
to hearing it in the ‘control room’, you can
easily misjudge the amount of ambience you are
getting, because it always seems a lot less than
you were used to in the hall.
Back
at your usual playback room you can often find you
have captured more than you want! One good way
around this is to play back, in the location
monitoring situation, a known recording of the
same genre you are about to record, and have it
playing back while you are getting to the stage of
hearing what the session microphones are about to
provide you with!
Classical
music recording sessions are often made in Concert
Halls or Churches. Temporary studios, in fact.
Monitoring rooms are often almost cupboards!
Loudspeaker placement and the room’s acoustics
add to the judgements made on the musical balance
and also with the amount of ambience being
captured, as outlined above.
So,
on to variety of approaches which are possible at
the recording stage and in the subsequent Post
Production, as it is known. There are many
different microphone techniques in use. Different
recording engineers tend to specialise with a
particular approach. It must be said, that their
experiences with a given way of doing it, can mean
they have really honed their technique over time.
Some
believe in a minimalist approach and often get
away with a simple pair of mics, suitably placed
in a given acoustic, with the layout of the
performers often suiting the results, in terms of
balance and imaging. This single pair approach can
be with directional mics (usually Cardioid)
human ear spaced apart and angled, say 90 to 110
degrees. Not many use two vertically coincident
directional mics angled at 90 degrees, one above
the other – little stereo width is a noticeable
result
Others
prefer to use non-directional mics (Omni)
spaced apart more than the ear spacing, mentioned
earlier. There is in fact a basic reason for the
use of non-directional mics – they have a more
extended low frequency range than directional mics
usually have. For some, the down side is poor
imaging and a tendency for the stereo to collapse
to one or other loudspeaker, if the listener moves
even a little from the sweet spot. However, in
this ‘horses for courses’ world, the use of
Omnis with piano recordings, is fine!
Regarding
the use of two microphones, there is a system
called Mid & Side and used by some. Here you
have a Mid mic facing to the centre soundstage,
directional or non directional, and vertically
coincident with it, a Side bi-directional mic (Figure
of Eight) picking up the sound from either
side of the Mid mic. The signals are mixed with
the Mid mic in phase to both the left and right of
the stereo with the Side mic mixed out of phase to
the stereo.
The
technique uniquely covers the centre of the sound
stage, the stereo starts in the middle and spreads
out symmetrically towards the loudspeakers, with a
variable stereo width instantly available, with
the amount of the Side mic signal fed in the mix!
There
is a famous three mic rig known as the Decca tree.
Three large diaphragm Omnis are used, two spaced
around a metre apart and a centre one out in
front, around half a metre forward. Omni
directional mics are known to have varying amounts
of directionality at frequencies above 5kHz and
this could improve the soundstage imaging being
obtained.
There
is now a sort of renaissance in the use of Ribbon
microphones, from a number of microphone
manufacturers, many with on-board amplification
which does reduce their noise floor. Regarding the
mostly commonly used capacitor microphones, with
various diameters of circular diaphragms, there is
a manufacturer in Sweden who uses rectangular
diaphragms which have been judged to have ‘ribbon
mic’ qualities.
Regarding
the use of ‘special’ cables, I am not aware
that these are often involved in Studio or
Location recording situations. The crucial feature
regarding microphone cables in the professional
recording field, is that the feeds are all
balanced, in early days with transformers and
nowadays, more likely with IC circuitry. Mine you,
we don’t deliberately run them close to mains
cables, especially if lighting feeds are involved!
Some
recording engineers do not like extra spot mics,
but there are occasions where they do need to be
used to get a musical balance. The skill is that
they should not really be noticed in the mix!
Going too close can certainly be evident, and an
often consequence is that more and more are then
put out! If the aim is missing to achieve the
important ‘depth perspective’ in the stereo
soundstage, then it is likely that there are many
close-up mics out there. I don’t know if there
is a world record for the number of mics used on
an orchestra, but sixty are known to have been
used……..!
An
interesting development, in the current use of
spot mics, is for their feed into the mix, to be
delayed in milli-seconds related to their distance
from the main mics. More easily done nowadays than
in the past, and now practised by many recording
engineers in getting a proper musical balance and
a depth perspective in the stereo soundstage.
Something
that is not fully pursued are the ‘subtle’
mechanical resonances which may ‘colour’ the
sound being captured by microphones. Certainly it
is obvious with mic stands on wooden floors. Best
overcome by using blocks of thick foam under the
legs of the mic stand, the weight of the stand
being really effective in stopping those
vibrations reaching the microphones. However,
perhaps we don’t consider the less obvious
effects of vibration transmission up the cables to
the mics, or do some stands resonate to loud
impacting sounds?
POST
PRODUCTION
Razor
blades, with reel to reel tape, was the skilled
way of editing in the past. Nowadays computer
program editing, and the use of the visual
waveform, is an amazing advance. The amount of
editing ‘needed to be done’, varies a lot. It
is recognised that too many ‘cuts between takes’
do not always maintain the ‘feel’ of a
performance.
There
is a particular aspect in post production that is
also important to the listener’s experience.
Wide excursions of the dynamic range, which are
perfectly fine at a live performance, do need to
be sensibly dealt with for any domestic playback.
I
am certainly not advocating the mad situation,
current with most ‘pop’ music, where the
levels through out the songs are pushed up to the
digital limit, with the whole thing ‘bashing
zero’. This is in pursuit of ‘loudness’, as
the selling point! The old Beatles stuff and other
CDs of that era are not like that!
With
classical music it is perfectly acceptable to trim
any wide excursions in level, done progressively
to keep the sense of dynamic change, and in that
way allowing the overall average level of the
performance to be raised and more interesting to
listen to.
I
know of two CDs where the dynamics, as recorded at
the sessions, and with the headroom that has to be
there to avoid a sudden peak going ‘over the top’,
were not allowed to be trimmed. A purchaser of one
complained that the levels were ‘all over the
place’, and he had get up to vary the levels
himself, to enjoy what he had paid for.
Another,
which was at a particularly low level for the
opening three minutes, had the reviewer’s
complaint that he made five attempts to get the
level right to suit the 30dB leap at the very
pronounced tutti. It is believed that the
composer, on receiving his issued CD copy, turned
up the initial volume to suit the music at the
start, and at the three minute point, his tweeters
ceased to ‘exist’!
WHAT
ABOUT ‘SURROUND’ AUDIO REPRODUCTION?
Are
we missing out on improving our whole audio
listening experience by not progressing to having
the ‘stereo’ ambience lifted out from just
being between our stereo speakers, to it being
available all around us, not just horizontally,
but including the height aspect that exists in our
real world around us?
Future
posting on this ‘Saturn Sound’ site will cover
the many ways the writer has ‘researched’ this
extension to our listening experience, albeit with
the very unlikely event that it will ever ‘take
off’, as it did with stereo over mono.
Many
differing microphone rigs have been used along
with the need to have a ‘mastering standard’
and a ‘properly’ set up playback system. You
can imagine the greatly increased number of ‘variables’
that this endeavour has revealed!
FINAL
THOUGHTS
Here
is a sort of ‘Unanswered Question’, to use the
title of a work by composer Charles Ives. Is there
likely to be an important phase relationship, over
the audio spectrum, heard in the live recording
situation, that is not maintained at any replay
situation, particularly due to microphone
imperfections in that area, along with what phase
corruption every loudspeaker playback, must
contribute to it - just a thought?
To
conclude, it is definitely a fact that the number
of variables, at the recording process and then
even more of them at the listener’s playback,
are the main stumbling blocks in always achieving
a good musical listening experience. It makes me
wonder if the 24bit 96k and beyond will ever be
that important – especially when we have DAB
Radio at around a seventh or below, the bit rate
of the CD Red Book standard. That apart from what
the iPod and MP3 deliveries do and what happens
with music downloads.
Some
years ago I was approached by a well known
ensemble, who asked if I could lend them or hire
some mics to record their next concert, as they
had just acquired a 96kHz sampling DAT recorder. I
asked them what kind of mics they would like? They
said ‘It didn’t matter’. It was a very short
phone call ………!
Up-dated
13/03/11