--- In "John Lundsten"
<> wrote:
> This is just plain wrong, you would have to create a very odd
recording with
> the forest ambience in anti-phase for this to happen.
This is not "plain wrong" at all; certainly not in the context of
the narrator/forest ambience example I was giving earlier...
When properly decoded from MS to LR, everything captured by the S
microphone becomes 'anti-phase' [your term, see * below] and
therefore cancels out entirely when summed to mono, unless the
decoding or mono summing has not been done correctly. That's
precisely how MS creates a stereo signal - the S signal in the right
channel is an inverted polarity version of the S signal in the left
channel. The L and R versions of the S signal are of equal amplitude
but inverted polarity, and therefore should cancel out entirely when
summed to mono. Here's the decoding maths:
L =3D M + S
R =3D M - S
Therefore:
Mono =3D L + R =3D (M + S) + (M - S) =3D 2M
Simple and straightforward. When summed to mono, the level of the M
signal doubles (+6dB) while the S signal is removed altogether.
If the S signal doesn't cancel out entirely when summed to mono,
you've got either an MS decoding problem or a mono-summing problem.
[*By the way, 'anti-phase' is not the right term here; the two
versions of the S signal are perfectly in phase insofar as there is
no time difference or phase difference between them, they both
begin, end, and cross zero at exactly the same times; the only
difference between them is that one has its polarity inverted
relative to the other. Phase and polarity are not the same thing...]
To clarify the narrator/forest example I gave earlier, let's say the
narrator(or whatever subject of interest - a close-up frog at the
edge of a pond full of croaking frogs in a field of chirring
crickets?) is 12 inches in front of the M mic, and the S mic is
capturing the forest sounds to provide the ambience to go with it.
The M mic signal will have very little of the forest ambience
relative to the narrator's voice, the majority of the forest
ambience will be in the S mic signal. When summed to mono, the S
signal disappears and takes most of the forest ambience with it,
while the M mic signal doubles in amplitude. This creates a
significant difference between the stereo and mono versions of the
same recording, so in that respect it is reasonable to question the
validity of claims that MS is mono compatible because it depends on
what you mean by 'mono compatible'.
If your definition of 'mono compatible' means it sums to stereo with
no comb filtering problems, then MS does this very well, as does XY
(discounting minor phase differences between channels due to the
inability to put both capsules in exactly the same physical point in
space, a problem suffered equally by *all* coincident stereo
techniques *including* MS but with the possible exception of the
Soundfield).
But if your definition of, or requirement for, 'mono compatible'
means the mono version of the signal retains all the information
contained in the stereo signal, MS fails by its very design while XY
succeeds very well.
You make a good point about the centre image. An XY pair means
sounds in the centre are captured off-axis by two microphones,
rather than directly on-axis to one microphone. Unless the two mics
in XY have very good off-axis response then the centre image won't
be as solid as it will with MS. So if you need a strong centre
image, use MS. If the side information is more important, use XY.
For the record, I am a huge fan of MS with a cardioid M capsule; it
is my primary microphone rig, I use it all the time and enjoy it
very much.
I have a Schoeps MS pair, which is a wonderful implementation of the
MS technique and produces a fantastic stereo image when decoded
properly. But from time to time the problem described above hits me.
Most recently, I used my MS pair to record a direct-to-stereo album
for an acoustic folk ensemble. For one of the more complex pieces
with lots of performers and instruments, I had to place the
musicians in a reasonably wide arc around the microphone, resulting
in a stereo image that extended from hard left to hard right. I knew
the balance and overall sound would change when summed to mono, but
I thought "who's going to play this in mono these days?"
Well... a short time after the album was released, one of the
musicians called me complaining that it sounded very wrong when
played on Radio National (an Australian broadcaster who apparently
transmits in mono). After telephoning the station to enquire why the
album sounded so wrong, he was told that the recording had poor mono
compatibility because, when collapsed to mono, the instruments on
the far sides of the recording dropped considerably in level (6dB or
so relative to sounds in the centre) as did the reverberation,
changing the entire balance and sound.
Fortunately for me, the same station awarded it 'Album of the Week',
which probably saved my butt from the financially-compensating wrath
of the musicians!!!
|