???Stereo Enhancement???
I've used one in FL Studio, found a free one on the net somewhere, but how do they work? I've been trying to find some documentation but with no avail. Has anyone built a stereo enhancement patch? Any help would be much appreciated. Thanks!
According to 'Enhanced Sound - 22 Electronics Projects for the Audiophile',
by Richard Kaufman, a 'shuffler' was once a popular feature of many British
recording studios, and it's use was for enhancing the stereo image.
"The device adds the right and left channels, to give an M, or middle
signal, and subtracts one channel from the other to give an S, or side,
signal. A balance control allows one to attenuate either the M or S signal
without affecting the overall loudness level."
The same technique is used in M-S recording using a cardioid and
figure-eight mic; parallel the figure-eight's output into two channels,
phase-reverse one channel, and pan left and right. Add to the omni, which is
panned central.
cheers
Roger
On 11/12/2008 19:40, "Nathan" wrote:
>
> I've used one in FL Studio, found a free one on the net somewhere, but how do
> they work? I've been trying to find some documentation but with no avail. Has
> anyone built a stereo enhancement patch? Any help would be much appreciated.
> Thanks!
Speaking of M-S encoding, I recently built this little example patch for doing L-R to M-S conversions. Lot's of stuff you can do to tweak things in M-S format before converting to L-R again.
Andrew B.
There's one that I found in Logic or something that basically works on
the spectrum. You can take 4-8kHz and pan it to the left, with a
curve, or something like that. I can't really think of a non-FFT way
to do it, but it might be cool as an exaggerated effect, especially on
human speech--so that you isolate certain vowels in certain speakers,
for instance. Someone run with this?
Andrew
Logic 8 also has a nice binaural panner, which can do a lot of fun things.
Also, though I the previous explanations were adequate, I once did a bit of planning to do something similar in msp. After a little bit of search, I found this thread in a dsp forum which might be useful-
the one i've seen at some mastering studio is:
- take a stereo signal, split it into left and right and flip the phase and add back to the opposite sides (left to right, right to left) and play with the level (very quietly for the most part). the louder the flipped signals, the wider the stereo image. (also the louder the signals, they might mess up the sound in the center though) you can play with multiple eqs as well, i.e. adding multiple eqed flipped signals, so that you can widen the image in the desired frequency ranges.
flipping the phase would always result into some kind of mono incompatibility, depending on the degree of teh flip. not that this matters today very much, but actually technically it would alter your signal in a way, that it comb filters itself, when played back mono. you should watch the phase in the stereo field, in any case.
aside this the logic way of creating stereo signal splitting signal into 8-12 bands while panning alternate band to the extremes makes sense and is mono compatible as well.
jrp
Andrew, this looks awesome. Do you happen to have a 4.63 version of your MS patch? Or maybe you could supply some screenshots, so i coudl rebuild it? Would love to.
jrp
Another thing you can do is to EQ the file slightly differently. Then
split the original and EQed files into left and right channels. Now
mix the EQed version with the original with a little delay (from 40
to 200 ms) while the left-right balance of the original and the EQed
version are different (e.g. turn the volume of the left channel of
the EQed version down and turn the volume of the right channel of the
original copy down - by 1 or 2 dbs). This creates a kind of spectral
movement in the stereo field and gives the impression of a deeper and
wider stereo image.
P
On 12 Dec 2008, at 09:16, uhito kiyosue wrote:
>
> the one i've seen at some mastering studio is:
> - take a stereo signal, split it into left and right and flip the
> phase and add back to the opposite sides (left to right, right to
> left) and play with the level (very quietly for the most part). the
> louder the flipped signals, the wider the stereo image. (also the
> louder the signals, they might mess up the sound in the center
> though) you can play with multiple eqs as well, i.e. adding
> multiple eqed flipped signals, so that you can widen the image in
> the desired frequency ranges.
> --
> mesa.elech/tele
Strictly, shuffling is a frequency dependent technique that boosts lows
in the side signal to give the impression of wider bass (below ~700 Hz
we rely more heavily on inter-aural level differences for location,
above that on phase difference).
It's effectiveness is dependent on the way the sound was originally
recorded, as different stereo miking techniques rely to different
extents on level and phase differences.
Here's a paper by Gerzon:
http://www.audiosignal.co.uk/Resources/Stereo_shuffling_A4.pdf
The trick he describes at the end with a
channel-swapping-phase-inverting feedback chain is pretty effective if
used in moderation (and AFAIK is similar in basis to Bob Katz's
mystifyingly-patented K-stereo process).
Another similar trick involves decorrelating left and right signals so
that they are optimally out of phase at any given moment, which enhances
separation. See
Kendall, G. "The Decorrelation of Audio Signals and Its Impact on
Spatial Imagery", Computer Music Journal, Vol. 19, No. 4. (Winter,
1995), pp. 71-87.
I've not made an attempt at seriously implementing this (with modulating
filters & all), but it seems to be prone to introducing pronounced
phasiness on certain types of material. YMMV.
--
O
Roger Carruthers wrote:
> According to 'Enhanced Sound - 22 Electronics Projects for the Audiophile',
> by Richard Kaufman, a 'shuffler' was once a popular feature of many British
> recording studios, and it's use was for enhancing the stereo image.
> "The device adds the right and left channels, to give an M, or middle
> signal, and subtracts one channel from the other to give an S, or side,
> signal. A balance control allows one to attenuate either the M or S signal
> without affecting the overall loudness level."
> The same technique is used in M-S recording using a cardioid and
> figure-eight mic; parallel the figure-eight's output into two channels,
> phase-reverse one channel, and pan left and right. Add to the omni, which is
> panned central.
> cheers
> Roger
>
>
> On 11/12/2008 19:40, "Nathan" wrote:
>
>> I've used one in FL Studio, found a free one on the net somewhere, but how do
>> they work? I've been trying to find some documentation but with no avail. Has
>> anyone built a stereo enhancement patch? Any help would be much appreciated.
>> Thanks!
>
>
>
Andrew and Roger, thanks for the enlightenment! I had never heard M-S processing. Very cool, opens up a lot of possibilities!
These are great tips, wish we had more threads like this.
It makes me realize how mastering is an art in it of itself.
