Must AM Band Sound Bad? A Debate

 

Neil and I have been complaining about AM sound being muffled on the Sangean K-200. Steven writes:

The AM sounds muffled because... it's AM.

Not trying to be snotty, just pointing out that AM sound quality has always been intrinsically inferior to that of FM, and since 1989, transmitted AM audio bandwidth has been further limited (to 10.2 KHz, vs 15 KHz for FM) to allow more transmitters to be accommodated. And that's before factoring in AM's disadvantages in when it comes to rejecting noise from fluorescent lights, electric motors, etc, that tend to mask high frequencies and muffle things even further.

That's why music broadcasters fled AM and it has become mostly the domain of right-wing punditry and sports-talk blather, where the shouting can be heard above the muffled audio quality.

Herculodge Response

I'm not convinced that AM is SUPPOSED to be hopelessly bad. If I use my Panasonic RF-888, for example, the AM sounds supreme, clear, and majestic. It's not just an AM problem. It's a radio problem.

Comments

[Gary]

Just compare the sound of AM of a Sangean WR-2 to that on a PR-D5 and you'll notice how much crisper the WR-2 sounds.

Excessively muffled sound on AM is mainly due to:
1. An IF filter that is too narrow.
2. A poorly designed audio stage, a poor speaker, or inappropriate equalization. Full range bass and treble controls can often help.

It's true that FM has much wider bandwidth than AM, but it's also true that not all radios sound muffled on AM.

[Jeff]

Get a GE Superadio III. Tune it to a strong am signal. Slide the "wide band" switch to "Wide". You will be amazed at how good AM sounds.

[S Patrick]

Nah, it's an AM radio and it's due to it's frequency (540khz to 1700khz) plus AM modulation is inferior to FM modulation which is why all emergency services use FM, even though they use what is called narrow FM. Try using AM inside a building vs FM. No comparison.

Does AM sound better then FM? only someone with faded memories can possibly say that. It's equal to those who claim tv sets were high quality back in the 70s (they aren't) or that shortwaves were better 40 years ago (they weren't)

[Brandon]

AM Stereo on my Delco car receiver sounds pretty good when the signal is strong.

[Ed S.]

AM radio has more than enough quality for voice. A telephone has less than a 3 kHz bandwidth, and yet you can still instantly recognize the voice of any of your friends or family. The trouble is when little speakers are bass-boosted ("voiced") in order to sound a little warmer on music, it winds up making voices sound boomy/muffled.

[Scooby214]

My SSTran AM transmitter broadcasts a rather wide bandwidth signal. When coupled with a good tube radio, such as my 1956 Blaupunkt set, the sound is astonishingly good. The Blaupunkt in question has both AM and FM. I won't deny that the FM still has a very slight edge in fidelity, but the average listener that hears it thinks that the radio is playing FM instead of AM from my SSTran.

The flip side to this is that commercial stations aren't allowed to broadcast using the bandwidth my SSTran uses. Not to mention that the AM is still prone to interference, no matter how wide the broadcast bandwidth or quality of the AM receiver.

[Demetri]

AM's source material's upper end is hard limited at 10khz. With single filter radios, the filter width must balance good adjacent channel rejection with fidelity so some fidelity is sacrificed in most designs. Also, the radios that sound good probably also artificially raise the high end to give that 10khz every chance to sound as good as possible.

FM source material is limited to 15khz for each channel so you get quite a bit more on the high end. Also the channel spacing is much wider on FM so the filters will not limit the fidelity as much.

Remember FM as 108-88=20Mhz to play with but AM only has 1.710-.530=1.2 MHz to fit all those stations.

[Neil Goldstein]

I didn't realize until this morning how much discussion this had sparked. I agree that modern AM radio will not sound as good as FM, but I'm not listening to music on it. For news and talk it is fine. This does not pass as an excuse for a poor-sounding AM radio though. The K-200 discussion that started this whole thread has been a bit exaggerated (possibly by my not clarifying the extent of the problem). Yes, it is muffled, but NOT to an unpleasant level. All in all it is much easier to listen to than some of my cheaper pocket models (Degen DE-205 for instance).

There are some radios that make AM very pleasing to listen to though. Even though having adjustable bandwidth is a big help, my best sounding AM radio would probably be my antique Sears Silvertone tube radio. This is followed by the Redsun RP-2000, and the Tecsun PL390. The 390 is limited by speaker size, but has nice clear audio. My SuperRadio III (currently in my mother's senior apartment) is also pleasant enough to not cause migraines. The Kchibo D96L is actually a bit too bright, and sounds better with a dust cloth over the speaker. There is also a strange, and annoying "pumping" affect on strong AM stations (remedied by off-tuning by a couple kHz, but then that causes the soft-mute to kick in).

The small headphone models that the Ultralight DX group recommend also sound very good when amplified through some basic computer or iPod speakers. The Sony SRF-M37 models are a bit too broad on bandwidth for serious DX-ing at times, but that shortcoming makes for great audio with strong local stations. The SRF-59 is also good amplified or through headphones. The mini that really disappoints in the audio department for me is my Degen DE-1123, which excels on FM for its size, but makes AM a bit murky.

The K-200 will continue to play WCBS 880 every morning here on Long Island, and either the local NPR stations or WXPK-FM from White Plains the rest of the time. As I said, up on the fridge the acoustics are actually pleasant. There is a new FM all-news station in NYC now, but it's pretty awful.

By the way, our other AM all-news station, 1010 WINS, still plays the teletype-ticker-tape sound effect in the background during the newscasts. Anyone else out there have this? I find it pretty hilarious.

[Ron]

I'm curious to the ages of the people here. I'm 41. As a kid in the 70's, about all we listened to was AM. That was what was in the car, that was the portable transistors had, that was pretty much it. AM didn't sound bad at all. I have a suspicion that if I was to tune into AM today on any of these radios, it still wouldn't sound bad.

I think the fact is, the 'younger generation' has never owned a radio that has acceptable AM. My experience is that most radios have AM as an afterthought, and do a pretty poor job of it. My truck's stereo has almost impossibly bad AM. It's aftermarket. The original stereo (broken) had fantastic AM.

One of the reasons I like CCrane radios is that whichever tabletop you buy (EP, +, SW) it's going to have good AM. Can't say that about most radios, and when you pay $200+ for something that sounds like crap on a band you want, your heart breaks into little pieces. This is also why I tend to look at multiband portables instead of just AM/FM. If it's got shortwave, it's probably going to have decent sounding AM.

I had this same discussion with some of my coworkers awhile back. They were mainly in their 20's. None of them had ever heard a decent AM radio and were surprised at the clarity of the CCRadio-SW I'd brought in. Then again, the ones that actually owned a radio, it was about 4" x 2", FM digital, and used the headphones as the antenna.

My thinking is, poor technological choices have given AM a bad rap, and FM doesn't sound much better to a lot of people. Most people don't understand why you'd spend $200 on a radio when they cost $12 at the drugstore, I don't find it hard to believe that they think AM is supposed to sound muddled and bad.

[Steve Waldee]

The "NRSC mask" that limits the top frequency response of AM signals these days is not, in practicality, much different from the glory days of AM radio in the forties and fifties--for the FCC always insisted that stations not interfere with adjacent channels. And, for the most part, they did NOT do so if they kept the negative modulation under 100%. For, normal program material does not concentrate an extremely high, consistent amount of energy from 10kHz to 15kHz. Sure--traces of sibilance; the edgy tone of muted trumpets; cymbal crashes; and the like *might* generate substantial HF signals up to the limit of human hearing. But, most AM stations in the old days had an 8 kHz studio- to- transmitter telephone line link (one I engineered until the late 1980s still did!) and old musical recordings up to about the mid-fifties had, when examined on a spectrum analyzer (as I've done in my design work on audio processors) had somewhat diminished response in the highest octave. It was after the rock and roll revolution, using much smaller instrumental ensembles with heavy emphasis on close-miked drum kits and singers and guitars, that HF energy above 10k became stronger on records.

Even so, there is NOT a huge difference to the ear, if you use an actual FLAT playback system, when you switch in an accurate, phase linear 10k lowpass filter, compared to a 15k bandpass. You will notice a very slight difference in the coloration of SOME aspects of the overall musical signal, and virtually no change in anybody's voice.

This has been tested thoroughly, and I for one participated in some of the tests of the system in the 1970s. During some of the tests, wideband program material from high speed original master tapes were played, not merely commercial records. My own voice, for instance, was recorded at 15 ips using a Neumann U47 mike (a classic instrument!) and this recording was used for some specialized tests of asymmetry and bandwidth control by a very large audio processing company.

So, believe me: AM does not 'sound dull' because of the NRSC mask. It sounds dull if the receiver is deficient, with a narrow IF strip, and a poor and nonlinear detector and crummy audio amp and speaker complete the damage.

In a radio station, the signal is almost always monitored from a wideband modulation monitor receiver. To be able to use the power bandwidth effectively, the NRSC pre-emphasis increases the proportion of signal in the upper mid and higher frequencies BEFORE the peak clipping and bandwidth filtering; so if anything, NRSC stations sound "brighter" on average, with greater clarity, detail, and audio crispness, than old fashioned AM stations that used simply a flat peak limiter like an RCA BA-6, no audio pre-emphasis, but yet COULD -- if you tried hard enough with an audio oscillator! -- squeeze a signal up to maybe 12 to 13 k out of the transmitter!

Furthermore, most antenna systems do NOT have a completely linear bandwidth +/- 15 kHz. They understandably roll off at the higher frequencies. There are a lot of inductive, reactive matching components between the output of the transmitter's PA/modulator, and the base of the AM antenna (or feedpoint); so there is an intrinsic PHYSICAL limit of bandwidth, and you can push and shove and burn out tubes and overdrive and pin meters--and get little or nothing for your effort!

That being said, a terribly mis-matched system, with flat modulator tubes and mistuned final, driven by square-wave audio, COULD radiate crap out to 15k and beyond. I remember how bad the Mexican AM border-blaster 50+ KW stations sounded; woe if your station was a near-adjacent!

Finally: strange things happen that cause signals that are INAUDIBLE to be transmitted, causing havoc and interference. When I was engineer for KIBE-AM, 1220, Palo Alto, we had a strange upper sideband whistle that would just DRILL into your brain; it wasn't coming out of our rig, for sure! I traced it to a tiny 1 kW station many tens of miles away in Vallejo. Their ancient Gates transmitter was oscillating erratically; you could NOT hear it unless you got to a particular point between the two stations, when the beat frequency became audible. The engineer of that station panicked when I brought it to his attention and took it down in the middle of the day (though I was actually quite casual and easy going about it, suggesting it might be something he should eventually look into!) The signal spur coming out of his transmitter, and somehow making it thru his single stick ND antenna, was probably around 20-21 kHz and must have modulated the rig with only a very few watts of power. But you could hear this ANNOYING howl, forty to fifty miles away when his actual station audio was virtually imperceptible!

REAL program audio, constrained to 10.2 k bandwidth, can sound FANTASTIC when properly processed. Don't blame the NRSC mask or the FCC rules or the station; use Occam's Razor and BLAME THE RADIO FIRST!

Steve Waldee--retired audio processing specialist and AM/FM station engineer