wjm@whuxj.UUCP (MITCHELL) (06/12/84)
<CHOMPOR> In response to the waves made about square wave testing (pun intentional): To Dick Grantges: Thanx for reminding me that most square wave testing is done at 1KHz. However, when checking the treble response of a system, I would think that a 10KHz or so test signal might be more appropriate. I'll agree that the square wave tests of most CD players are better than those of many analog transducers (e.g. phono cartridges and speakers) and further that they are useful in determining the performance of filters on CD players. I don't think that CD's are per se bad, far from it --- however, I am disturbed that a relatively low sampling rate was chosen. As for recommendations on CD players, if I were in the market right now, I'd get the Kyocera DA-01 for two reasons: the phase response of its filters and its error-correction performance. (However, keep in mind that I'm not that interested in fancy programming features - to me sound quality is the crucial parameter.) Regards, Bill Mitchell
rgg@aplvax.UUCP (06/18/84)
I do not understande why the ability of an amplifier to accurately reproduce sqare waves is a good measure of its sound quality. Real music does not contain square waves. Would someone please explain the basis for using square wave response as an indicator of sound quality? -- Richard Greenberg ...decvax!harpo!seismo!umcp-cs!aplvax!rgg ...rlgvax!cvl!umcp-cs!aplvax!rgg
rcd@opus.UUCP (Dick Dunn) (06/20/84)
First, the oops - I made a stupid statement in a posting yesterday that the lowest-frequency component present in a 20 KHz square wave but not in a 20 KHz sine wave is at 40 KHz and therefore pretty hard to hear. That's wrong; it's at 60 KHz (but still hard to hear!) Square waves contain only odd harmonics. Yes, I know better and I found the error myself - but not before sending the foolishness to the edges of the known universe. ...but we still don't have Mr. Pearson straightened out... > But you missed the point... > If the CD logic is sound it must reproduce a perfectly square wave > given a properly generated square wave test disk. Still wrong. A perfect square wave has substantial components well beyond audibility. No audio reproduction system needs this inaudible junk - it can overload amp stages and fry tweeters. The best systems will delicately remove it without removing or phase-shifting the audible spectrum. It's a fortunate side-effect of the way CD's work that they don't (more correctly, shouldn't) produce any components above 22 KHz or so. It's not necessarily intuitive that you can take the HF components out of a square wave without damage - but intuition doesn't help here. I suppose it's even less intuitive that the CD playback circuitry COULD produce a very nearly square wave but will not do so if it's working. > They ring, and you know it. And they're supposed to. The longer-than-infinitesimal risetime and the overshoot at the top of the rise are both CORRECT in the reproduction of an audio-frequency signal from the information on a CD. If you're a reasonably good programmer and you've got some sort of graphical device, try plotting some "square waves", working up from a sine wave adding components. You'll be a little surprised at how many components you need before it looks decently square. > CD get your basics correct first... this 'off by one' attitude will > not be tollerated. Hoo, boy - who's doing the "tollerating" (sic) here, anyway? You can rail against CD's 'til you turn blue. You don't set the standards and you haven't yet stated a problem. -- Dick Dunn {hao,ucbvax,allegra}!nbires!rcd (303)444-5710 x3086 ...Cerebus for dictator!
alex@sdcsvax.UUCP (06/20/84)
Real (which reel? the third reel, of this exciting motion picture!) music mayn't have 20KHz square waves in it, but there's some pretty unreal music out there. Alex
slag@charm.UUCP (Peter Rosenthal) (06/22/84)
I can hear x-rays. There is a little known phenomenon pertinent to this discussion. It is known as the gibbs phenomenon. Basically, it amounts to the following: Given a square wave approximated by additive synthesis, it is generally true that there will be an overshoot of eight percent at the step. This remains true as more and more components are included. As the approximation gets better and better, the overshoot does not improve, it only gets narrower. In the limiting case it becomes infinitely narrow. Josiah Willard Gibbs, the father of statistical mechanics is responsible for this gem, but I don't know the details of the work.
andrew@inmet.UUCP (06/23/84)
#R:aplvax:-66200:inmet:2600082:000:444 inmet!andrew Jun 21 23:55:00 1984 > Would someone please explain the basis for using square wave response > as an indicator of sound quality? The frequency response of a system is directly related to its transient/ impulse response. Deficiencies in response will appear as rounded edges, ringing, or both. For further details, I refer you to any junior-year EE text (very boring) or DCM Time Windows literature. Andrew W. Rogers ...{harpo|ihnp4|ima|esquire}!inmet!andrew
dsn@umcp-cs.UUCP (06/29/84)
I'm *really* getting tired of reading about square waves. Can we talk about
something else for a while?
--
Dana S. Nau
CSNet: dsn@umcp-cs ARPA: dsn@maryland
UUCP: {seismo,allegra,brl-bmd}!umcp-cs!dsn