Re: [ARSCLIST] "Difference Track"

[Andrew Hamilton <ahamilton@xxxxxxxx>]

Hi Garr,

     I'll give this a try.
     If a full track head was not available for the transfer and one of the
2-track head's channels has noticeably more hiss or surface noise than the
other, throughout the program, your best bet might be simply to mute the
noisier track and make up any desired gain on the remaining quieter channel
(this can in some ways be better than using a full track head and getting
all the signal but also some bad noise that exists on only one longitudinal
half of the tape, or floor of the groove), or try summing both channels
together at 12:00 panorama.  That will increase the gain of the Mid
component of the program by 6.02 dB (roughly ;).   But the makeup gain
applied to the Side component of the program will be less than that.  That
is because the Side (aka Difference) component is only full of things that
happen on one channel or another, but not on both at the same time with the
same intensity so there's always less than a doubling of the noise when they
are summed (and this works better with program on which the louder noise
moves from channel to channel).

     But, since the Side component noises will still be in the mix of the
two summed channels, if you really want to get rid of the noise that exists
in the Difference component to the L+R mix, you can do the following:

     Send channel 1 (normal polarity) and channel 2 (normal polarity) to a
pair of channels.  Put both of the copies at 12:00 panorama and invert the
polarity of one of the copied channels.  Upon changing the polarity of one
of the copies you will have changed these new channels from being a Sum
(L+R) bus to being a Difference (L-R) bus.  Now invert the polarity of the
Difference bus.  Sum together the playback of the inverted Difference bus
with the original (normal polarity) program.  Voila.  No more Side
contributions at all, as the inverted Difference bus will null the playback
of the original Difference component of the original hard L + hard R mix.

  

________________________________________________________________________

  

     As a side bar, I'd like to revisit what Glenn Meadows taught me on his
Web Board some years ago.  Acoustic summing uses the same math as electrical
summing.  So consoles that have a pan rule which is less than -6.02 dB at
12:00 (for ex.) are designed to work in rooms that have imperfect acoustical
summing - due to standing waves that cancel wanted compressions and
rarefactions.  He added that SSL boards used to be calibrated to -4 dB at
12:00.  This was in contradistinction to the plethora of mixers which employ
the typical -3 dB pan rule, because rooms specified for SSL consoles would
normally be well-enough designed to sum L+R at at least +4 dB.  The
mastering room built by Tom Hidley and Sam Kinoshita that he used to work at
in Nashville (can't remember the name ;)  summed around +5.9 dB, according
to their tests.  SSL have long since adopted the -3 dB pan rule.  Yamaha
digital boards, incidentally, use a -0 dB at 12:00/+3 dB at L or R pan rule.
Coping with this is most counterintuitive when playing back a stereo mix on
adjacent mono channels...  unless you like the extra DSP.



Andrew





On 11/22/08 12:10 AM, "Garr Norick" <big_garr_2002@xxxxxxxxx> wrote:

> Hello,
>  
> I have read that when mono recordings (such as Records and Cassettes) are
> Digitized in stereo, using stereo playback equipment, there is a "Difference
> Track"... i.e., on a digitized recording that originated from a cassette being
> played on a stereo deck, that one of the resulting channels will usually have
> a great deal more tape hiss, and lower fidelity, than the other... I have also
> read that certain digital recording/editing/restoration software (namely Adobe
> Audition, and Diamond Cut Pro) can find and eliminate this track, presumably
> by finding and eliminating any frequencies that are not common to the two
> channels... I have Diamond Cut Pro... is there anyone here who is familiar
> with Diamond Cut Pro, and can tell me how to use this function in the
> software? Thanks in advance.
> 
> 
> 

--