Okay, time for a little unofficial survey -- for those of you who
engineer AM stations, how much pre-emphasis do you use?

In the late '80s, the NRSC established a standard "modified 75 uS" AM
pre-emphasis curve, but the FCC never officially made its use a
requirement, except for AM Stereo stations. And today, it seems that
most AM stations are using a higher amount of pre-emphasis -- even
many AM Stereo stations.

In fact, the Optimod 9100B offers four different pre-emphasis curves:
"Blue", "Red", "Yellow", and "Green". "Blue" is the NRSC curve, while
the others provide increased amounts of high-frequency boost. The
"Green" curve can be maxed out to provide over _twice_ as much
pre-emphasis as the NRSC curve!

The mono Optimod 9200 offers a similarly wide choice of pre-emphasis
curves, and in fact, it comes shipped from the factory with a default
setting which exceeds the NRSC curve by up to 3 dB at mid-range
frequencies (~4 kHz).

The result of this variation, combined with the overwhelming
predominance of Optimod processors on the AM band, is that we're
basically back to square one, before the NRSC standards were created.
Some AM stations sound dull and muddy, while others sound tinny and
shrill -- and you don't even need a "wideband" receiver to hear the
difference.

Over the decades, adjacent-channel interference is the number-one
complaint which has led to almost universally narrow-bandwidth,
low-fidelity AM receivers. But Optimod-ized AM broadcasters are only
aggravating this problem by using extreme amounts of pre-emphasis.

And now, IBOC proponents want to "solve" this problem by chopping off
everything above 5 kHz. They even claim analog AM radio will sound
_better_ as a result, because receivers can be opened up to 5 kHz
bandwidth, as compared to the 2 - 3 kHz they typically offer today.
But says who? The hundreds of pages of iBiquity, NRSC, and FCC
documentation I have read make absolutely _no_ mention of what kind of
pre-emphasis, if any, IBOC stations should use, or how much bandwidth
the analog portion of IBOC receivers should provide.

Meanwhile, the Omnia 4.5 AM audio processor curiously appears to offer
no user choice of pre-emphasis. It does offer a choice of 4, 6, or 10
kHz audio bandwidth, but the user's manual makes no mention of
pre-emphasis at all. I guess Omnia chose a curve that "sounds good"
to them, and that's that?

In article , (WBRW) wrote:

Okay, time for a little unofficial survey -- for those of you who
engineer AM stations, how much pre-emphasis do you use?

In the late '80s, the NRSC established a standard "modified 75 uS" AM
pre-emphasis curve, but the FCC never officially made its use a
requirement, except for AM Stereo stations. And today, it seems that
most AM stations are using a higher amount of pre-emphasis -- even
many AM Stereo stations.

In fact, the Optimod 9100B offers four different pre-emphasis curves:
"Blue", "Red", "Yellow", and "Green". "Blue" is the NRSC curve, while
the others provide increased amounts of high-frequency boost. The
"Green" curve can be maxed out to provide over _twice_ as much
pre-emphasis as the NRSC curve!

Hi Kevin,

I don't know anything about the Optimod 9100B, do you know what the
various curves look like? It would surprise me if they simply added more
and more "preemphasis". IIRC, and of course I may not, the original
Optimod 9000 had separate equalization controls for peaking the area
around 3,500 Hz where the typical AM radio is rolling off, and a separate
control for the higher frequencies up to 10 kHz or so. If I remember it
correctly, the idea was to be able to boost the frequencies that would
help the typical radio the most, without overdoing it at the higher
frequencies, and causing problems by boosting frequencies that weren't
going to be heard anyway, except on a minuscule number of radios. I
wonder if the four different colors don't mainly provide different amounts
of peaking in that mid frequency area where most radios need some help?

The mono Optimod 9200 offers a similarly wide choice of pre-emphasis
curves, and in fact, it comes shipped from the factory with a default
setting which exceeds the NRSC curve by up to 3 dB at mid-range
frequencies (~4 kHz).

If I follow you, this midrange boost is consistent with the equalization
in the Optimod 9000, and what I would guess the colors might do in the
Optimod 9100B, i.e. provide midrange boost beyond the NRSC curve.

The result of this variation, combined with the overwhelming
predominance of Optimod processors on the AM band, is that we're
basically back to square one, before the NRSC standards were created.
Some AM stations sound dull and muddy, while others sound tinny and
shrill -- and you don't even need a "wideband" receiver to hear the
difference.

I assume that's the intention, the optional mid range boost is there to
help the typical radio.

Over the decades, adjacent-channel interference is the number-one
complaint which has led to almost universally narrow-bandwidth,
low-fidelity AM receivers. But Optimod-ized AM broadcasters are only
aggravating this problem by using extreme amounts of pre-emphasis.

I wonder if it is "extreme amounts of pre-emphasis", or just the extra
midrange range hump in the 3.5 to 4 kHz area that you are hearing?

Why not just equip your radio with a complimentary four color equalizer switch?


Regards,

John Byrns


Surf my web pages at, http://users.rcn.com/jbyrns/

I don't know anything about the Optimod 9100B, do you know what the
various curves look like?

The graph is a bit messy, but here are the 9100B's pre-emphasis
curves:

http://ourworld.compuserve.com/homep...s/opt-emph.gif

The "Red", "Yellow", and "Green" curves are all continuously variable
from 0 dB (flat) to a maximum of 22 dB boost (at 10 kHz). This graph
shows what each of these three curves look like at boost values of 5,
10, 15, and 20 dB. Meanwhile, the "Blue" NRSC curve maxes out at a
boost of 10 dB, as shown, which meets the NRSC pre-emphasis standard
exactly.

The Optimod 9200's factory-configured pre-emphasis is equivalent to
the "Red" curve at 10 dB boost -- in other words, the first curve on
the graph above the NRSC curve.

I wonder if it is "extreme amounts of pre-emphasis", or just the extra
midrange range hump in the 3.5 to 4 kHz area that you are hearing?

The amount of increased boost extends all the way to 10 kHz,
regardless of which non-NRSC curve is selected, so even the higher
frequencies (which only "wideband" receivers can fully discern) are
increased as well.

Why not just equip your radio with a complimentary four color equalizer
switch?

Because I shouldn't have to. FM radio, TV audio, phono records, audio
tapes, and even CDs all use standardized pre-emphasis curves, with no
user interaction required. Why should AM radio be the same? Canada
made the complete NRSC standards, including pre-emphasis, mandatory
for all of its AM stations in 1988. But the FCC chose to keep the
NRSC standards "voluntary" in the USA, except for a rather lenient RF
spectrum occupancy mask which effectively limits transmitted bandwidth
to +/- 10 kHz (but yet has allowed the disastrous +/- 15 kHz IBOC
system to be used, since its sidebands conform to this RF mask which
was never intended to accomodate digital signals).

The primary goal of the NRSC was to encourage the manufacture of
high-quality AM receivers, by reducing adjacent-channel interference
and providing a "level playing field" of how AM stations broadcast
their signal. Today, unfortunately, these high-quality receivers may
be few and far between, but that does _not_ mean that this kind of
"every man for himself (or station for itself)" attitude should
prevail!

p.s. One correction: The Omnia 4.5AM processor does indeed provide
pre-emphasis that conforms to the NRSC curve. However, if desired,
the user may provide extra mid-range or treble boost by using the EQ
controls.

In article , (WBRW) wrote:

I don't know anything about the Optimod 9100B, do you know what the
various curves look like?

The graph is a bit messy, but here are the 9100B's pre-emphasis
curves:

http://ourworld.compuserve.com/homep...s/opt-emph.gif

The "Red", "Yellow", and "Green" curves are all continuously variable
from 0 dB (flat) to a maximum of 22 dB boost (at 10 kHz). This graph
shows what each of these three curves look like at boost values of 5,
10, 15, and 20 dB. Meanwhile, the "Blue" NRSC curve maxes out at a
boost of 10 dB, as shown, which meets the NRSC pre-emphasis standard
exactly.

Thanks for posting the curves. It is not completely clear to me which
ones go together in a family, but from what you are saying, it sounds like
the "color controls the high frequency boost at 10 kHz, and then there is
a separate "mid frequency" equalizer control that controls the bump in the
4 to 5 kHz area?

The Optimod 9200's factory-configured pre-emphasis is equivalent to
the "Red" curve at 10 dB boost -- in other words, the first curve on
the graph above the NRSC curve.

That makes sense, if I have the right curve that is the one I would
probably choose, either that or one of the others in the group that ends
with 10 dB boost at 10 kHz.

Why not just equip your radio with a complimentary four color equalizer
switch?

Because I shouldn't have to. FM radio, TV audio, phono records, audio
tapes, and even CDs all use standardized pre-emphasis curves, with no
user interaction required.

Since when was this the case? The equalization for all these media is
adjusted on the source side to suit what the originator thinks is best.
Why should AM radio be any different?

Why should AM radio be the same? Canada
made the complete NRSC standards, including pre-emphasis, mandatory
for all of its AM stations in 1988. But the FCC chose to keep the
NRSC standards "voluntary" in the USA

The FCC doesn't even specify pre emphasis for FM except for a maximum that
is allowed, but I don't even see why they bother with that, how can they
enforce it, to really control it they would have to regulate the entire
audio chain back to the original pickup microphone, and even the acoustics
of the original performance venue.


Regards,

John Byrns


Surf my web pages at, http://users.rcn.com/jbyrns/

Thanks for posting the curves. It is not completely clear to me which
ones go together in a family, but from what you are saying, it sounds like
the "color controls the high frequency boost at 10 kHz, and then there is
a separate "mid frequency" equalizer control that controls the bump in the
4 to 5 kHz area?

The color (Blue, Red, Yellow, or Green) determines the shape of the
pre-emphasis curve, and the user can also select the total maximum
boost (at 10 kHz) to be applied. The Optimod 9100B does not provide
any other audio tone/equalizer controls, except for an optional bass
boost feature.

The 9100B's total audio bandwidth can also be adjusted, of course.
With no "brick wall" filter enabled, it provides a maximum audio
bandwidth of about 12 kHz. Western Hemisphere models can select
either a 5 or 10 kHz filter. European and Asian models have a choice
of a 4.5, 5, 5.5, or 6 kHz filter.

Meanwhile, the mono Optimod 9200's bandwidth is continuously variable
(in 0.5 kHz steps) from 4.5 to 9.5 kHz, and it offers a range of
pre-emphasis curves and boosts very similar to the 9100B's.

Since when was this the case? The equalization for all these media is
adjusted on the source side to suit what the originator thinks is best.
Why should AM radio be any different?

Because there has to be a level playing field. Why do you think the
RIAA curve for phonograph records was established? Because there was
too much variance and too much confusion, and it was inhibiting the
goal of providing the consumer with increased fidelity and
convenience.

This also applies to AM radio. Whether it be today's 10 kHz or the 5
kHz that IBOC proponents want, consumers will never get to experience
analog AM radio at its full fidelity unless there is a level playing
field and a universal standard which manufacturers can design their
receivers to meet.

Actually, the preemphasis on FM nowadays with all the high end
processing can be done away with...to try to process (compress, limit,
clip) the 0-15 kHz spectrum and especially the high end, means you will
overshoot and overmodulate the xmtr unless you clip/limit the hell out
of it..but when you clip, you cause even more problems....
MOST engineers today agree that the PRE AND DE emphasis curves could be
done away with entirely...because the field has changed a lot since
Armstrong's days and the FCC's early development of the FM rules.
With today's processing and EQing, you really donot need PRE
emphasis...and if you donot need it, why have DE-emph in the rcvr???

The RIAA curve for records was done because of the material used & the
freq response in early days...technology has left that behind and that's
a moot point today...CD's donot have such a requirement...and I always
record my cassettes with Dolby C ON and then play them back with Dolby C
OFF so the high end sounds cleaner and brighter

If the rcvr makes had wanted to, they COULD have had flat response to
10-15 kHz in an analog rcvr and put a filter switch on it (like my Sony
XRA-33 AM STEREO car rcvr has)...but they didn't want to listen to car
buyers bitching about the noise and whine while listening to AM adj
channel...so they made the rcvrs narrow....same reason why 50% of the
time your radio FM stereo light is ON, you are actually listening in
blended mono!!!

Chris
WB5ITT
Houston

WBRW wrote:


Because there has to be a level playing field. Why do you think the
RIAA curve for phonograph records was established? Because there was
too much variance and too much confusion, and it was inhibiting the
goal of providing the consumer with increased fidelity and
convenience.

This also applies to AM radio. Whether it be today's 10 kHz or the 5
kHz that IBOC proponents want, consumers will never get to experience
analog AM radio at its full fidelity unless there is a level playing
field and a universal standard which manufacturers can design their
receivers to meet.


--
Replace NOSPAM with 1st initial and last name for direct reply!

The reason for pre-emphasis has as much to do with the modualtion technique and the
inherent limitations with the method.

If you're familiar with the sideband model for frequency modulation (the one
represented on a spectrum analyzer), you may remember that 100% modulation (+/-
75kHz) will produce many more sidebands at 60Hz than it will at 6kHz. Because the
amount of power in the sidebands for lower-frequency modulation is so much greater,
the low frequencies are inherently quieter (because the limiter stages in FM
demodulators have more to work with to strip out AM noise). Pre-emphasis was added
to provide a modicum of noise improvement (by increasing the amount of power in the
sidebands) for modulation frequencies which have a relative deficiency in modulation
sidebands.

It's arguable whether the North American standard of 75uS is too much, and as I've
never heard European-standard FM radio, I don't know if their 25uS preemp is too
little or still too much, but the adoption of preemphasis compensation has nothing to
do with the technology of the time - it's simply trying to circumvent the laws of
physics, and those haven't changed since the beginnings of frequency modulation
techniques.

Frankly, Chris, if you're having trouble dealing with high-frequency clipping with
your processing set up, you've got it set up wrong, anyway. There's no way that the
high end should be slamming against the pre-emp limit. No wonder you think Texars
are trash, since it's so easy to use too much high end. Finesse is the key.
--
For direct replies, take out the contents between the hyphens. -Really!-



"Chris Boone" wrote in message
...
Actually, the preemphasis on FM nowadays with all the high end
processing can be done away with...to try to process (compress, limit,
clip) the 0-15 kHz spectrum and especially the high end, means you will
overshoot and overmodulate the xmtr unless you clip/limit the hell out
of it..but when you clip, you cause even more problems....
MOST engineers today agree that the PRE AND DE emphasis curves could be
done away with entirely...because the field has changed a lot since
Armstrong's days and the FCC's early development of the FM rules.
With today's processing and EQing, you really donot need PRE
emphasis...and if you donot need it, why have DE-emph in the rcvr???

The RIAA curve for records was done because of the material used & the
freq response in early days...technology has left that behind and that's
a moot point today...CD's donot have such a requirement...and I always
record my cassettes with Dolby C ON and then play them back with Dolby C
OFF so the high end sounds cleaner and brighter

If the rcvr makes had wanted to, they COULD have had flat response to
10-15 kHz in an analog rcvr and put a filter switch on it (like my Sony
XRA-33 AM STEREO car rcvr has)...but they didn't want to listen to car
buyers bitching about the noise and whine while listening to AM adj
channel...so they made the rcvrs narrow....same reason why 50% of the
time your radio FM stereo light is ON, you are actually listening in
blended mono!!!

Chris
WB5ITT
Houston

WBRW wrote:


Because there has to be a level playing field. Why do you think the
RIAA curve for phonograph records was established? Because there was
too much variance and too much confusion, and it was inhibiting the
goal of providing the consumer with increased fidelity and
convenience.

This also applies to AM radio. Whether it be today's 10 kHz or the 5
kHz that IBOC proponents want, consumers will never get to experience
analog AM radio at its full fidelity unless there is a level playing
field and a universal standard which manufacturers can design their
receivers to meet.


--
Replace NOSPAM with 1st initial and last name for direct reply!