Digital music off topic:
I promised a note about digital music, so here ‘tis, more a personal rant than a call to arms for a change.
Many years ago I thought I would get a sound system. I like music well enough but had always contented myself with the simplest record players. When I looked into the matter I didn’t actually listen to what the systems would do. Instead I looked up the specs on various systems. What I found was that the then rather new digital equipment was spectacularly superior to the analogue things I had been used to so I went with the new.
But I found that I did not listen to it often, not so much as I had listened to records before … yes, old vinyl records spun on a cheap portable player. Many years passed, and in fact still many years ago, and I took a college course in sound engineering. Alas I remember little, although the course involved a lot of things I wish I knew and used. But the thing that did stay with me was that when music is digitized the small, smooth wiggles of the pressure waves that are the sound are “sampled” periodically. That is the signal from the microphone is recorded at one instant and then recorded at another instant after a brief pause, and then at the next instant after pause of the same length. A typical wave is sampled many times.
Way back when, they decided that if they took the highest frequency the human ear was capable of and sampled at twice that rate then every wave change would be sampled multiple times and the recording would be ideal. That was why the digital systems performed so well at extremes of the audible range, where analogue equipment began to run into trouble, since it was optimized for what you actually listen to.
But there is a problem. For practical purposes computers are perfect. That introduces “beat.” Imagine two speakers, one giving you a tone at 200 cycles per second and another at 201. At some point in time the signals from the speakers match at your ear and the sound is added. A half second later they tend to cancel, muting the sound. A half second later they are louder. What you here is a pulse that is one second long. You will hear this many times if you listen to somebody tune a piano, where multiple strings represent a single note, and where the tuner does not want them all to be in perfect synch, because then the note would be very loud and would decay quickly. The sustained piano note depends on judicious imperfect tuning of the strings.
Well if you were to look at the pattern of the vibrations from a human voice or any traditional instrument, you will find that the tracings are not perfect repeats. There is a sort of detuning, much as the piano tuner introduces.
Now I don’t have the best ears. Even as a child I could not hear the highest frequencies some are capable of hearing. And I have, alas, gone to a few rock concerts and come out with my ears ringing – the sign that your hearing has been damaged because the sound was too loud and you have lost forever your ability to hear at the very highest frequency you were capable of when you went into the show.
I stick my fingers in my ears these days if I find the music is too loud.
But I do hear pretty well. One thing I used to wonder about was why the piano keyboard centered at about G above middle C, and music was written centering around that range. I would have preferred it all to be an octave lower. You see, when you get a note you also get overtones. If you play a C I will hear it as well as a C an octave higher (twice the frequency) and a G above that (three times the frequency) and so forth. This gives the note richness. It doesn’t matter what the top end of your hearing is, you are going to get more tone (or information or band pass or timbre or whatever you want to call it) from a lower note, and this will be true down to a point where your instrument is reaching the limits of its design or your hearing is beginning to be unable to cope.
So when I listen to a digital recording of a piano or human voice it sounds just fine. The imperfections are out of my range, and they are fleeing since the original sound has some variation.
But consider what happens when you record a digital signal digitally. Now the tone of the note is at the perfect frequency, and it stays right there. It has lots of overtones, and they don’t change either, say for the period of a heartbeat. But up at frequencies you can’t here, they are interacting with the fundamental sampling frequency of the digitizing process, and you get beat. Each note has its own beat. And they have nothing to do with the music. They are quite dissonant. Imagine trying to listen to somebody play the violin while somebody else dismantles a grand piano with a chain saw. It might seem cute, but it is definitely going to have an effect on the quality of your experience with the violin music.
Over the past many years, of course, the digitizing process has changed little. Nobody wants to throw out their entire music library. So we’re stuck with the sampling rate.
But for goodness’ sake, we now have computers of enormous power. Surely they can make digital music that introduces the tiny fluctuations of pitch found in the real world. With those there would be no sustained beat produced. You would not hear that grisly screech in the background every time somebody wanted to please you by playing some music that had been produced and recorded digitally.
Please sing or play to me in real life. Don’t be shy. You can’t possibly be as bad as the best maters with state of the art equipment.
There have been 92 pages over the past month.