Technical questions about AM/FM radio quality

Factual Questions
1

Lots of questions on my mind:

If you assume perfect transmission, is it a perfect reproduction? Or does the transform required to transmit the signal inherently lose some of the information? For instance, why is FM considered better for music than AM? Would you get better results with an analog source, since it’s an analog system? Do hardware limitations have a large effect?

This was prompted by seeing a control in an MP3 encoder that described 96 kbps as “radio-quality”. That doesn’t even make sense, so I wondered what “radio-quality” actually means.

2

Stereo (more widely, at least), less interference.

“Radio-quality” MP3s are putting listener expectations as labels on compression rates. 96kbps music will sound warbled, perhaps tinny. While you don’t hear these types of effects on the radio, the difference in compression quality is supposed to be similar to the quality expectations between radios and CDs (ie, bursts of static or clicks and pops may be acceptable artifacts when listening to the radio, but they wouldn’t be tollerated in a CD recording)

3

“Radio Quality”? If anything it means ‘not so good.’
AM or “Audio Modulated” radio signals are an amplitude modulated carrier frequency. The carrier frequency is fixed, only the amplitude is varied depending on the input audio frequency and volume. These signals are subject to interference by all kinds of electrical signals, e.g. lightning, switching of lights, random arcs, etc.
FM or “Frequency Modulation” sometimes termed Phase Modulation varies the frequency of the carrier. This results in a signal that is NOT affected by random interference. Hence FM is a high quality reproduction of the original input. Also it is either picked up by the receiver or it is not. It is not subject to fading as AM is.

4

Frequency modulation and phase modulation are two fundamentally different things, although they share similarities. There’s a pretty good article on the differences here:

Bolding mine.

5

There are two different things involved here. First is the type of signal encoding itself, which is amplitude modulation (AM) and frequency modulation (FM). The second thing is the bandwidth allocations set up by the FCC.

As far as the modulation methods themselves, either can be used to create an accurate reproduction of the sound, at least as far as the human ear can hear. The major disadvantage of AM is that whenever two signals of the same frequency are mixed together, their amplitudes are simply added together, so that it becomes impossible for an AM receiver to discriminate one signal from the other. The effect of this is that any noise at all on the same frequency cannot be removed from AM broadcast signals by the receiver. FM, since it is frequency modulated, just tracks the frequency of the loudest signal. Quieter signals on similar frequencies are ignored by the receiver. This means that FM signals don’t get all that annoying static that AM signals are prone to.

The typical human hearing range roughly 20 Hz to 20 kHz, although some people can hear as high as 24 or 25 kHz (rare). This range tends to decrease with age, so older people rarely hear above 15 kHz. This is where the FCC limits start to come into play. For an AM radio station, how much bandwidth they allow translates into how much radio bandwidth they take up. I hope I’m remembering these numbers correctly, but I believe that AM radio is limited to a single channel (which means no stereo music) with a bandwidth of only 10 kHz. Since this is well below the range of what most people can hear, there is audible signal loss in AM radio. FM does slightly better, with a bandwidth of 15 kHz, but it is still below the 20 kHz that many people can hear.

An MP3 that is “radio quality” will be comperable to an FM broadcast radio signal.

AM and FM are both analog systems, which means that any noise present in all of the pre-amplifiers and signal conditioners, transmitters, receivers, etc. all gets mixed into the final sound and causes noise. Digital systems, just by their nature, are 100 percent faithfully reproduced within the limitations of their sampling rate, due to the fact that any noise that is not large enough to flip a bit is completely ignored by all components throughout the system. So, FM is better than AM, and digital is better than either one.

Note - if the sampling rate of a digital system is too low, it can still produce audible distortion of the signal. A good example of this is phone systems, which are very narrowly limited in bandwidth before being digitized in order to maximize the number of voice channels that can be carried through the system. Digital phone systems tend to have roughly the same signal quality as AM radio as a result of this.

6

Plus, you can implement error-checking and error-correcting protocols in digital, something you can’t do with analog. Simply put, you can devote a handful of bits per sample to recording what should be there, and give the reciever a chance to fix it if it isn’t.

Of course, eventually you can’t reconstruct the signal in any useful way. This leads to odd static (utterly different from anything analog produces, due to how digital signals are `chunked’ during compression) and commonly a complete dropout of signal with no static in-between (anyone who uses a satellite radio set around large buildings is familiar with this).

7

Any comment on AM stereo?

There were several New Orleans AM stations that advertised AM Motorola Stereo broadcast back in the late 80s. My '85 Lebaron had a MoPar factory radio that received AM stereo and when tuning to the stations advertising it, the stereo LED would light up. Never could really tell if it was actually in stereo because those stations were mostly talk radio.

8

AM stereo has a lot of the same problems that AM mono has, namely that its succeptible to noise and fading. There was talk at one time of converting the AM band to stereo since the majority of listeners had switched to FM and it was starting to look like there might not be any AM radio stations left soon. The problem with this is that if you don’t make the band allocation wider, you only get half as many channels available. You get the same problem if you try and increase the bandwidth of each station to get better sound quality. If you increase each station from 10 kHz to 20 kHz, you can only have half as many stations in the entire AM band (or you have to make the entire band wider). But, a fewer stations are better than no stations, so they were considering it. Then talk radio became more popular and listeners started moving back to the AM band, which removed the need to do anything to fix the “AM is dead” problem. If they are still pushing for AM stereo I haven’t heard about it in a long time.

9

Frequency and phase modulation are related to each other through the derivative. Frequency is the time derivative of phase, i.e. frequency is the rate of change of phase.

The effect of this in practice is that a phase modulator preemphasises the high modulation frequencies causing them to result in more frequency deviation than do the low frequencies. This is desirable in FM broadcasting in order to get the desired noise reduction at high modulating frequencies. If a frequency modulator is used the applied voltage of the modulating signal has to be boosted (presemphasized) at the high frequencies as compared to the amplitude at low frequencies in order to achieve the same noise reduction at both high an low modulating frequencies.

FM quality is better than AM for two reasons. One is the noise reduction possible in FM as compared to AM and the other is that the FM bandwidth allocation allows the stations to use a modulating signal with a broader bandwidth than is the case with AM. AM stations are restricted as to the modulation bandwidth they can transmit in order to prevent interference with adjacent channels.

10

Perhaps confusingly, the large bandwith of FM signals is relevant to the noise reduction as well as the ability to convey high audio frequencies. The bandwith of FM radio signals is much larger than that of the audio being carried. More bandwith allows for more noise immunity; this is the essence of the Shannon-Hartley theorem. There is such a thing as narrow-band FM, which uses the same bandwith as AM; this offers no theoretical minimum noise advantage (I don’t know how this plays out with the kinds of receivers actually used).

11

Narrow band FM is certainly possible, however, commercial FM broadcasters don’t use it because one of the main purposes of FM was to provide an almost noise-free, high quality audio output. As stated in the post, narrow band FM (deviation ratio of about 1) has the same noise performance as AM.

An additional reason for FM noise reduction is that the transmission is in a frequency range where there is little or no atmospheric noise, or static.

From Radio Engineering, F.E. Terman: “Very little static is observed at frequencies too high to be reflected by the ionosphere [Note: This includes frequencies in the US FM broadcast band.] This is in considerable measure due to the very limited range of such frequencies, and in part because there is very little static energy generated at such high frequencies.”

12

No it’s not a perfect reproduction of the original. The big limitation is the loudspeaker. I haven’t paid any attention to speakers for many years now and I suppose there have been some improvements, but loudspeakers are the weak link in the chain.

Yes there are inherent limitations. Even a concert hall has an effect on the quality of the sound you hear from the orchestra or chorus/soloist. You do not hear the same music as the conductor nor does the tympanist in the back. You don’t hear the same orchestra as the patron sitting on the other side of the hall. What you both hear might be very good but it isn’t the same for both and it isn’t the same as what started out from the proscenium.

For instance, why is FM considered better for music than AM? Would you get better results with an analog source, since it’s an analog system? Do hardware limitations have a large effect?

This was prompted by seeing a control in an MP3 encoder that described 96 kbps as “radio-quality”. That doesn’t even make sense, so I wondered what “radio-quality” actually means.
[/QUOTE]