Orban OPTIMOD 6200, Operating Manual

Page: 37 / 195

OPTIMOD INTRODUCTION
1-19
unobtrusive regardless of whether the listener is hearing compressed or uncompressed
audio.
We suspect that it is impractical to pass through, without review, dialog normalization
values created by program and commercial providers, because some commercial provid-
ers will inevitably try to game the system to make their commercials excessively loud.
Instead, if dialog normalization is to be actively used in transmission, the broadcaster
must strip its existing value from the program, and must then preview each piece of pro-
gram material and replace the value with one that will ensure consistency from one piece
of program material to the next. We think that very few local stations will want to devote
the necessary resources to this activity. Instead, it’s an obvious thing for the networks to
do.
If the networks have done their job well, they will choose dialog normalization values
that ensure consistency from source to source, and when the viewer changes channels. It
is improbable that this can be done by automation. The best we will be able to do is to
manually identify dialog or other baseline sounds, measure their loudness with a true
loudness meter, and manually adjust the dialog normalization parameter so that these
baseline sounds emerge with a standardized loudness. CBS’s research into this area
showed that no simple meter could do this accurately, including frequency-weighted me-
ters with averaging characteristics. The errors in such measurements were so large that
they were not useful in controlling the levels of commercials well enough to eliminate
viewer complaints.
The CBS loudness meter divides the signal into seven octave bands and weights the
gains of the bands according to the 70-phon equal-loudness curve of the ear. It then av-
erages the output of each band with a 15-millisecond time constant. The averaged out-
puts of the bands are then added and the sum is applied to a 200-millisecond time con-
stant. This is applied to the meter, which is assumed to have instantaneous response so
that it clearly shows the effect of the two previous time constants.
In tests, this meter agreed with average listeners within 2dB. However, it’s important to
note that listeners disagreed amongst themselves by as much as 4dB when asked to as-
sess the subjective loudness of a given piece of program material. So any loudness meter
can only work for an average listener, and may show considerably greater errors when
compared to any given listener.
Advantages of Multiband Compression in DTV Audio
Multiband compression is useful in performing an “automatic equalization” function to
change the frequency balance of the audio on a program-adaptive basis. In a multiband
compressor, frequency bands containing excessive energy are automatically compressed
more than other bands. This results in a re-equalization of the program material towards
some target spectral balance. The 6200’s two-band compressor controls excessive bass,
which can otherwise cause muddy balances. The five-band compressor can perform
more detailed automatic re-equalization that can be very useful for program material
such as live news.