Analog FM Audio Processing

[Brian Bowers]

Some of the discussion lately (revealing, constructive, heated, etc.) at times has revolved around FM processing. When adjusting, I personally like to analyze my results on more than one monitoring system (anyone else?), before finalizing the session(s), and walking away happy (if that is ever the case). Those systems would include my vehicle audio system, the main off air in house tuner, the transmitter room tuner, and my home system. In other words, all systems monitoring the analog off air signal.

After being in the broadcast engineering field/department for some 30 plus years, I just lately in the past year made a surprising discovery. After comparing five home tuners, I was amazed at how different each individual FM's analog audio sounded on all of these tuners. As I had mentioned, I also critically analyze my processing adjustments at home. What I am discovering is that my particular station being analyzed sounds different on all of the tuners. I don't think any two sound exactly alike - very close, but not totally alike. So now I am not so quick to make knee jerk processing adjustments.

The tuners utilized at home are:
-Denon TU-850
-McIntosh MR 78
-Pioneer TX-7800
-SONY 730ES
-SONY SA5ES
-Technics ST-9030

All of the home tuners output audio level was adjusted using Dorrough Loudness Meters, so the volume perception is equal. I was mainly listening to spectral balance and texture. Those of you adjusting may want to keep this in mind next time. Granted I'm listening for every nuance I can pick out (as that is what some of us are adjusting on the audio processor), but I was very surprised to hear the difference in tuners, on the same exact setting, or tuning around and comparing the same station on these different tuners.

 

[bossbill]

And outside of "in-car" time, most listeners are hearing the station on cheap mono receivers in their bedrooms, bathrooms, offices, etc. So we would probably want to do comparisons on a few brands of automotive receivers. Then select one that is somewhere in the middle, and use that one for future tweaking.

I've always found that long-term consistency is the most important factor, and when tweaking processing, one should always do it in very minor ways, over long periods of time.

Same thing with format adjustments also.

 

[TomT]

One thing to be aware of is that most radios--car radios, tuners, and table radios--use detectors that are not true FM detectors.

With AM, the greater the modulation, the louder the detected signal. Of course, if modulation is increased to a point where the carrier is pinched off (negative modulation) there is splatter and other forms of distortion. However, positive modulation above 100% makes the station sound louder. Hence the loudness wars of the 60's and 70's; when AM was king. Louder is better... Most PD's in their 50's have this permanently engraved in their brain.

Supposedly, not for for FM. In classical theory, a limiter/discriminator detector will not sound "louder" if you increase the deviation of the transmitted signal. So, in theory, over-modulation gains you nothing.

Unfortunately, it ain't necessarily so. The detectors used in most radios more closely resemble AM detectors, so over-deviation does sound louder--up to a point. Then the signal sounds "swishy", clipped, or otherwise distorted. That point is often determined by the design of the IF strip in the radio, and therefore will be different in every radio.

And the "loudness" of the signal will also depend (again, varying by radios), on how much signal the radio sees.

That's why, after you have the station sounding nice and clean, the sales guy comes in complaining he lost a sale because a local car dealer plopped him down in the new Skoda 2000, turned the radio on --and said "There's Station X--there's station Y, there's you.
You ain't loud enough, your station's no good...."

Some days it don't pay to get out of bed.....

 

[BROADCAST]

My experience has been different regarding deviation greater than normal.
It depends on the linearity of the tuner with excessive deviation,for instance a Sony st-se700 works well with deviation of 120khz,whereas an old Ford radio started to distort at deviation above 90khz.

When i did deviation testing on my tuner i scoped the mpx out of the demodulator,and looked at the symmetry and peak control,this looked the same as the mpx from the processor to the exciter.
Turning the exciter dev up to 120khz still showed the same mpx waveforms on the demod,but increasing the exciter dev up to 140khz started to show added overshoots,the overshoots reminded me of lossy compression of 128kb audio !

Obviously some tuner designs are going to be less tollerant of excess deviation,but the upshot is they should all be designed to work with 75khz deviation.

The ceramic if filters seem to be the main trade off regarding selectivity/bandwidth.

In the 80's Ireland had 2 major unlicensed stations that had multiple transmitters located on mountains and other high points,they were Sunshine and Super q ,these stations has deviation levels at 120khz,and they sounded superb.
From 1986 onwards they addopted the Bill Cunningham Hot Hits format.

At that time ireland had no law for unlicensed staions,and these 2 stations used pro broadcast equipment and also had major advertising,all this came to an end on december 30th 1988 when new laws came in to force to outlaw unlicensed staions.

 

[WNTIRadio]

I use a Denon TU-680NAB, a Yamaha T-80, Yamaha RX-V4600 (the one with HD radio), a Sony portable radio and the stock radio in my car (Subaru, with the optional subwoofer) on the "amateur" side of tuners to check audio processing.

On the "pro" side, there's a Day Sequerra M2.0X, Inovonics 531 and the Belar FMM-2.

Each one of the receivers/tuners imparts its own sonic signature on the audio, as compared to the mod monitors which are all very close in sound given their extremely wide bandwidth and design to be neutral.

One interesting thing is that the RX-V4600 won't go stereo unless the pilot injection is 9% or more. Tried running at 8% and even with a strong signal still no stereo.

Just like in the studio, where the mix is always checked on a crap pair of speakers or a boom box, it's always best to check your processing setup on an "average" radio or two. Naturally, most of the listening public isn't going to have a bunch of audiophile tuners or mod monitors to listen on. If you can get it to sound good on the crappy radios, most times it will sound great on a great system.

 

[radiorob2.0]

Just like in the studio, where the mix is always checked on a crap pair of speakers or a boom box, it's always best to check your processing setup on an "average" radio or two. Naturally, most of the listening public isn't going to have a bunch of audiophile tuners or mod monitors to listen on. If you can get it to sound good on the crappy radios, most times it will sound great on a great system.

This is what used to be called "The Lloyd Theory of Audio" now called "The Coby Theory of Audio" stating no matter how good it sounds through the perfect circumstances the end user hears it through a discount department store radio. The key is to make it sound good on both extreme of receivers.