Oh dear, somebody just pulled my finger. Engage tongue lashing.
Let us first talk about the sound quality that is possible via FM. The potential for FM, in all honesty, is quite high and actually does equal or rival that of compact disc, in certain respects (actual data delivered), and actually quite exceeds that theoretically possible with a vinyl LP, in other respects (separation, frequency response, interchannel crosstalk, etc.). Say what? Allow me to explain:
See, when you have a $20,000 FM alignment and testing rig sitting at your disposal, you can learn a great many things about FM. First is this: FM is capable of incredibly low distortion performance insofar as static testing is concerned. I have regularly obtained results below .01% on a fully modulated signal on a tweaked and perfectly aligned, well-designed tuner. A Magnum Dynalab is having a good day if it can hit the high side of .20%. Shocked? Most "audiophile" tuners are genuine pieces of GARBAGE. These are complex devices and require a massive engineering budget many audiophile companies are not going to spend on a niche market product. High end audio has become a market where parts selection and fancy wire unfortunately often supercedes sound engineering practice. That said, that measurement is lower than many tube amplifiers.
Still, all is still not perfect. STEREO FM does have a real, legitimate, and definite problem with intermodulation distortion both on an individual channel and between the channels. This can become particularly pervasive on badly processed pop music with too much high frequency content. HF noise can actually fold over from the stereo subcarrier to the primary channel, to put it into really simplified terms. Fortunately, with modern and properly configured equipment, this isn't so much a problem. Go read some of Bob Orban's whitepapers on the Orban website if you need a technical fix.
The real problem with FM enters in when one considers the effects of the 19kHz stereo subcarrier. Yes, it is at a low level, but it is still there and a constant potential source of intermodulation distortion. You've seen the 19+20kHz IMD test charts in Stereophile. FM has a constant signal at 19kHz for all of the musical data to cope with. This, in turn, can and does give rise to very measurable and audible problems. The sum of these problems in a tuner with a very carefully designed IF strip with incredible care paid to obtaining steep filter side skirts and the lowest possible group delay can be massively diminsished, although never entirely. A proper demultiplexer implementation is also critical. I'm really skimming over the surace here and cutting a lot of corners, but so be it. FWIW, the MPX chip in the MD tuners is a decent piece. Unfortunately, the IF strip is very poorly designed and is not up to even mass market Japanese high end tuner standards. It is just *AWFUL*. I've called them and told them how to fix it, but I bet they haven't.
FM's other major potential problem CAN revolve around stereo separation. A full techincal discussion would be overkill, so I'll cut it short and suffice to say again that IF implementation is the primary critical factor along with proper compensation. I have obtained nearly 60dB stereo separation from 40Hz to ~14,000kHz, which is more than enough for virtually any recorded program material. Most tuners, of course, won't do this well. Yamaha and Tandberg tuners tend to do the best here, but I have also seen some vintage designs to shockingly well, at 45-50dB or above across the frequency range. Magnum Dynalab, OTOH, and curiously, leaves out the compensation adjustment, and it's gotta be a cold day in hell to squeak 25-30dB out of one of those misable things as frequency increases. Incidentally, however, reduced stereo separation can produce a decrease in noise-level, another major FM problem.
Mono FM can broadcast with truly incredible fidelity. Stereo, on the other hand, requires a whopper of a signal to get the noise level down. That's where the primary objectively identifiable problem with FM comes in at a theoretical level, and it is one that cannot be overcome. A good antenna, and a properly aligned front end are, of course, the best remedies.
Now, onto compression. In reality, *ALL* FM stations use some level of compression. Uncompressed FM is simply too soft. A good compressor/limiter won't be driven too hard and too close to full modulation. If it is, you'll get that awful seasickness feeling where the music is pulsing and changing in volume. Very annoying.
Ergo, in sum, while FM has some theoretical problems, the primary issue is compression IN THE MODERN ERA. Old FM modulators had a *HOST* of other problems. Suffice to say that if you have a favorite station and cash to burn, offer to buy them a new digital setup for a few grand. The difference can be uncanny. Many of the FM stations I listen to around here are not highly overmodulated/compressed, and do, in fact, sound very good. On a live broadcast, from a technical perspective, more actual audio "data" may be making it over the airwaves than CD will allow.
Audio is often about the music, and not just getting the best sound quality possible, unless you're one of those contemptuous boobs who just listens to "audiophile approved" rubbish. Many jazz, classical, and live performances just are not available elsewhere.
And also do pardon my habitual tearing down of an audio industry sacred tuner cow. I know you probably own one, dear audiophile reader. I've talked to the incompetant twit they pass off as their engineer. I told them how to fix their product. I have (partially) fixed their product personally. Want a revelation? MD, for the cost of about $10 in parts could produce tuners with *TEN TIMES* less distortion and *THIRTY DB* more channel separation. I have never seen another high end company, particularly in the tuner realm, make the same omissions and errors these people make, and I, frankly, find it inexcusable, and I'm not even an engineer or an expert. A good tuner can truly be an eye opener, and *CAN* have objective specifications as good as or better than many pre and power amplifiers, if you feed it with a good signal. Not surprisingly, when you eliminate a whole bunch of musically unrelated IM distortion, the component also sounds a heckuva lot better. Have a nice day.
