Let me reiterate, Black Shire, that what you call that “over-crisp sound” is not an innate characteristic of FM but an artifact of bad analog processing. And yes, an analog of this form is distortion (if you’ll pardon the pun) is also an artifact of most digital processing.
The following comment really surprised me: “When an FM Disc Jockey would crumple a piece of paper within range of the microphone, it sounded like a crackling flame from a wooden match--that's an example of the "over-crisp" FM sound that I find grating. The same sound on an AM station sounded...well, normal, like a piece of paper being crumpled and nothing more or different.”
In fact, with ungimmicked FM, the opposite is true. As Lawrence Lessing put it in his biography of Maj. Armstrong, the inventor’s first public demonstration of FM in Nov. 1935 made the difference very clear. One of Armstrong’s colleagues, broadcasting in ungimmicked mono FM to the lecture hall over his amateur station, didn’t stop with demonstrating the superiority of FM for both music and voice. He also demonstrated that two sound effects widely used in network radio drama – crinkling cellophane to simulate the sound of a raging fire, and pouring a glass of water close to a mike to created the sound of a waterfall, were far from convincing over FM.
So why did you hear what you heard in the early Seventies, and why do you still hear something wrong with FM today?
By 1970, most FM stations had both converted to stereo and converted to solid state technology, at least in the studio. What’s more many records were being mastered with solid state electronics by that time. Let’s consider the later factor first.
In Sept. 1972, Russell Hamm delivered his landmark paper, “Tubes vs Transistors: Is There an Audible Difference?” at the Audio Engineering Society convention in New York. He found two chief differences: (1) the spectral distribution of harmonics, and (2) innately less linear transistor amplifiers (and even more non-linear IC op-amps) used excessively large amount of inverse feedback. The latter yielded good results on steady-sate test signals, but allowed the amplifiers to produce gross amounts of short-term “transient intermodulation distortion,” or TIM, making the reproduction harsh and grating to the early on real-world musical sounds, especially those with lots of transient energy. (Hamm’s paper appeared the the Journal of the AES, Vol. XXI, No. 4, May 1973.)
Solid state sound began to improve markedly in the 1980’s. But I surmise that your early exposure to the poor quality of an early solid state amplifier handling high frequency content made a lasting impression.
As for the continuing problem of harsh FM sound, that’s a result of most stations’ unenlightened approach to audio processing.
Some observers were noting this back in the Seventies. I’ll just quote below from a source I think you’d have trouble finding: a guest editorial in “The Audio Amateur” (VIII: 4; Dec. 1977) by Roy H. Trumbull, under the title, “The Decline and Fall of FM.”
A chief engineer at an FM station, Trumbull notes that the problem began when stereo both made FM stations suddenly look financially viable, and forced the stations to cut back slightly on the modulation levels to accommodate the stereo subcarrier. This gave rise to all sorts of non-linear limiters, and even exciters with compromised frequency response curves that barely met the existing lax FCC audio standards.
He concluded his essay with this paragraph:
“One morning at 4 a.m. I was driving along on a trip to get some parts so I could get my 10kW stereo transmitter back on the air before 6 a.m. As I rolled along I was listening to our old unlimited auxiliary transmitter with its exceptional low end response. The music sounded so good and so much like the sound I’d grown up with that I began to wonder: ‘Is this trip really necessary?’”
NOTE TO SUPERCASTER: Nobody is more opposed to iNiquity’s trash technology that I am, let’s be fair about one thing: The capture effect that you cited allows almost any FM receiver to reject a desired station’s own I-BUZZ signals, since these are more than 20 dB below the level of the analog FM signal (and even a terrible FM receiver has a capture ratio of 6 dB or less). The problem comes when you’re trying to get a distant, weak signal on a receiver with very good selectivity, and can’t because the I-BUZZ from a nearby first-adjacent is actually jamming the station you want.
I’ve never heard of I-BUZZ interfering with mono reception of the host analog signal, though it can ruin stereo reception on certain solid-state tuners with very wide IF’s and keyed oscillators that produce a replacement 38-kc carrier with lots of harmonic content (see “HD Radio Self-Noise,”
http://users.tns.net/~bb/hdrsn.htm).
