I’ve done a bit of audio recording in my day, and am familiar with the problems of overdriving amps/inputs and the resulting clipping (when the wavform amplitude exceeds the available bandwidth, and the sine wave’s peaks and troughs get “clipped” off) that occurs.
But I sometimes hear older audio types complain about “overmodulation”; is this just an older term for clipping? Or is it something else?
In analog recording, say on magnetic tape, there was a “zone” where you could record over 0dB and the sound would take on a warm characteristic. While the sound was overmodulated, it was not quite distorted, and did not sound displeasing. There are digital modules now that try to replicate that effect.
In digital recording, 0dB is as far as you can go, after that, there is only clipping, and digital clipping sounds nasty. At 44,100 Hz, there are 32,768 samples per channel. There cannot be any more. If you go past 0dB, there is no more headroom, and the waveform gets chopped off on the top - that’s clipping.
I’ve just made a waveform of a 440 Hz tone at 0 dB to analyze the sample count. If you amplify that tone by any amount, it gets louder due to density, but the maximum level won’t go beyond 0, and the sample count remains the same. The waveform just gets clipped.
I hope that helps to explain it. Is this what you wanted to know?
Modulation is changing one signal or medium to carry the pattern of another.
For example, an AM radio signal with a frequency of 1000 kHz can be controlled so that its broadcast power goes up and down slightly in time with an audio signal that ranges from 100 Hz to 15 Khz. They call changing the broadcast power modulating the amplitude, or Amplitude Modulation, or AM.
In FM Radio, it’s similar, except instead of changing the power of the signal so that it pulses in time with the audio signal, the frequency is varied (or modulated) by a small amount.
Similar things happen with television signals. And when doing analog audio recording, magnetic mediums (tape or disk) have maximum amounts of signal they can store.
In all these cases, the modulation or change to the signal is supposed to be restricted to a certain range, both for purely physical and practical reasons. You can’t change an AM radio signal’s power so much that it goes below zero, that’s impossible. You don’t want to vary the frequency of an FM transmission so much it interferes with a station broadcasting on the next base frequency over.
So the receiving or playback equipment (radios, TVs, players, etc.) is built to decode signals that have been modulated within these expected ranges. If a signal is “over modulated” it has been changed too much from the baseline. This can have various effects depending on the severity of the error, and the receiving or playback equipment.
Some of the effects are similar to clipping, but some can be very different. An over modulated TV signal for example, will often cause a varying “buzz” in the audio, because the video portion of the signal is being pushed into the audio portion.
On edit: I see fishbicycle has given a much better explanation of the term in how it is normally used in context of analog tape recording. I’ll leave my two bits in just for general explanation.
Would the term “square wave” also apply?
I’m not an audio expert, just an old audio buff, but I don’t think it’s quite the same, even though in some cases the difference may not really matter.
A (perfect) square wave is like an infinitely high frequency signal that has been clipped. That is, the signal takes zero time to rise from the base line to the maximum output, stays at maximum for some period of time, then returns instantly to zero.
A clipped signal in the normal range (approx 20 Hz to 20 kHz in audio) will take some amount of time to reach maximum level, so the “rise time” (and fall time, for that matter) is always more than zero. Not much more, in some cases, but always more. However, since there is no such thing as perfect, practically speaking a clipped signal is sometimes so close to a square wave that it makes no difference.
An over modulated signal wouldn’t necessarily look “square”. Depending on what it is you’re over modulating, it could look more like a continuous high level. But his could vary depending on many factors. I think under some conditions an over modulated signal could look like a square wave, but certainly not always.
Generally, the waveforms will be more or less whatever their normal shape is up to some point, beyond which, they’re lopped off.
Usually, the result is to just sound nasty, but there are a couple times where overdriving/overmodulating or clipping is desirable. As **fishbicycle ** said, there’s a little sliver of “excess” signal strength on analog tape that will put a nice happy mellowness to the sound. It is distortion, but it’s a “good” distortion, and one that’s been fairly resistant to being duplicated by digital effects.
Guitar amps are another example of intentional overmodulation - the player can dial in anything from a sort of edge or hardness to the sound, on up to thick crunchy sound that sounds nothing at all like a guitar.