wjm@lcuxc.UUCP (B. Mitchell) (01/15/85)
<gulp,gulp> I found Bob Niland's comments about CD's quite interesting and agree with them. However, I have some problems with gould!jon's comment about the 44 kHz sampling rate being adequate. He is correct in stating that a 44 kHz sampling rate will perfectly reproduce a 22 kHz signal (the Nyquist sampling theorem). However, that is not the problem. The problem is that, frankly, 22 kHz may not be adequate for *true* high fidelity reproduction. Although the medical texts state that Homo Sapiens can hear frequencies between 20 Hz and 20 kHz, these figures refer to the Fundamental frequencies. (and that isn't totally true either, some people, like myself, can hear 25 kHz tones - which is why ultrasonic alarms in department stores make me reach for the Excedrin). However, musical instruments do not produce pure sine wave tones, but rather complex waveforms with high amounts of harmonics. Since the third harmonic of a 14 kHz cymbal or bell note is 42 kHz, 22 kHz is nowhere near adequate for proper reproduction. Long term readers of net.audio will remember that this has always been one of my main concerns about the current CD standard. This is NOT and should NOT be interpreted as an anti-digital comment, I just want the digital sampling rate to be high enough to reproduce ALL the musical information. Regards, Bill Mitchell ({ihnp4!}lcuxc!wjm)
karn@petrus.UUCP (01/16/85)
> .....Although the medical texts state that Homo Sapiens > can hear frequencies between 20 Hz and 20 kHz, these figures refer to the > Fundamental frequencies. What difference does that make? The upper limit of a person's hearing has nothing to do with whether the high frequency signal is associated with a lower frequency one. If anything, it'd be harder to detect the presence or absence of a very high frequency component along with a lower frequency fundamental because the latter would be much stronger. > ...However, musical instruments do not > produce pure sine wave tones, but rather complex waveforms with high amounts > of harmonics. Since the third harmonic of a 14 kHz cymbal or bell note is > 42 kHz, 22 kHz is nowhere near adequate for proper reproduction. Prove to me that you can HEAR 42 KHz and I'll believe you. Phil
ag4@pucc-h (Angus Greiswald the fourth) (01/17/85)
> ... However, that is not the > problem. The problem is that, frankly, 22 kHz may not be adequate for *true* > high fidelity reproduction. Although the medical texts state that Homo > Sapiens can hear frequencies between 20 Hz and 20 kHz, these figures refer to > the Fundamental frequencies ... Since the third harmonic of a 14 kHz cymbal > or bell note is 42 kHz, 22 kHz is nowhere near adequate for proper > reproduction. I'm not exactly a professional in this area, so I have some questions: 1. What is so special about a harmonic that it could be heard @ 42kHz whilst a Fundamental couldn't? Does the Harmonic affect the lower frequencies in some way through interaction? 2. How many amplifiers have you seen that even claim to respond to frequencies above 40kHz (much less 20kHz)? -- "It's only talk!" Jeff Lewis vvvvvvvvvvvv {decvax|ucbvax|allegra|seismo|harpo|teklabs|ihnp4}!pur-ee!lewie ^^^^^^^^^^^^
mat@hou4b.UUCP (Mark Terribile) (01/17/85)
>> .....Although the medical texts state that Homo Sapiens >> can hear frequencies between 20 Hz and 20 kHz, these figures refer to the >> Fundamental frequencies. >What difference does that make? The upper limit of a person's hearing >has nothing to do with whether the high frequency signal is associated >with a lower frequency one. If anything, it'd be harder to detect the >presence or absence of a very high frequency component along with a >lower frequency fundamental because the latter would be much stronger. The second part first: the fact that timbres are distinct indicate that small amounts of upper harmonic energy over the base frequency (pitch) ARE significant. And the human ear is NOT a linear device. It is an active nonlinear detector whose output is subject to the most comprehensive analysis known to man. I recall reading of evidence that indicated that upper harmonic energy at 28 kHz or greater could be detected over a base tone by many people. It was not heard as a pitch but rather as a change in timbre. There is something to be said for the ``DC-to-daylight'' amp. On the other hand, it seems to me that at these ridiculous frequencies vinyl may not do all that well either, so the CD loss may not be as great as supposed. -- from Mole End Mark Terribile (scrape .. dig ) hou4b!mat ,.. .,, ,,, ..,***_*.
rs55611@ihuxk.UUCP (Robert E. Schleicher) (01/18/85)
With all this talk about the importance of high-frequency response, and of the importance of harmonics to hearing timbre, I thought it would be worthwhile to point out a few points of reference, with regard to fundamental frequencies. - The highest key on a piano produces a fundamental of 4186 Hz. Thus, with a 20 kHz. response, you can hear the fourth (and almost the 5th) harmonic (This is a Csub8, for you music lovers) - Although some stars can go higher, the fundamentals of a soprano voice only go up to about 1050 Hz. or so (Csub6). - The highest piccolo not fundamental is around 3800 Hz. In short, you get a lot of harmonic content out of a 20 kHz. response. (BTW, the lowest piano key has a fundamental of only 27.5 Hz, meaning that you're really only hearing the harmonics of this key) (These figures came out of a human factors text, "Understanding Human Behavior", in the section on hearing. A pretty good intro to this type of stuff. I'm not in this field, it's just something that is of interest to me.) Bob Schleicher ihuxk!rs55611
ed@mtxinu.UUCP (Ed Gould) (01/18/85)
> Prove to me that you can HEAR 42 KHz and I'll believe you.
Hearing, per se, isn't the real issue. The fact that I can't hear
a 44 KHz tone isn't relevant. The fact is that the presence of the high-
frequency harmonics *does* have a well-established effect on the
overall sound perception. How important that connection is varies by peoples'
taste and experience.
Another thing to look at when considering a 44 KHz sample rate
is that many people can hear signals with frequency higher than
22 KHz. So even if the sampling theorem that states that
double-the-high-frequency is enough were relevant, 44K isn't high
enough.
I've compared an analog pressing of a recording made digitally
with a 50K sample rate to a CD of the same thing at 44K, and
the vinyl was *definitely* a better sound. What I don't know,
unfortunately, is whether the CD was reprocessed from the 50K
master or was recorded in parallel at 44K.
--
Ed Gould mt Xinu, 739 Allston Way, Berkeley, CA 94710 USA
{ucbvax,decvax}!mtxinu!ed +1 415 644 0146
ag4@pucc-h (Angus Greiswald the fourth) (01/20/85)
Perhaps someone has some better information (like what Bob Schleicher posted) on the frequencies emitted by your average cymbal. I think Bill Mitchell commented about a 14kHz cymbal crash or bell note and its third harmonic being at 42kHz. My ear would tend to disagree with that. When I play a cymbal, I hear a low fundamental (well below 1kHz) which to most people is drowned out by the much stronger higher harmonics. So a cymbal has a chance to give off plenty of harmonics below 22Khz. Can someone verify this. I'm also curious about your average bell note, or any similar seemingly high frequency instrument (i.e. are we really missing all that much by limiting ourselves to 22kHz). -- "blah, blah, blah, blublublah blah blah" Jeff Lewis vvvvvvvvvvvv {decvax|ucbvax|allegra|seismo|harpo|teklabs|ihnp4}!pur-ee!lewie ^^^^^^^^^^^^
jlg@lanl.ARPA (01/21/85)
> > ...However, musical instruments do not > > produce pure sine wave tones, but rather complex waveforms with high amounts > > of harmonics. Since the third harmonic of a 14 kHz cymbal or bell note is > > 42 kHz, 22 kHz is nowhere near adequate for proper reproduction. > > Prove to me that you can HEAR 42 KHz and I'll believe you. More to the point - even if some people CAN hear sounds with frequencies significantly higher than 20kHz (I don't believe 42kHz either), few stereo components and NO speakers (that I'm aware of) are available to reproduce such sounds accurately. The roll-off of even the most expensive speakers above 20kHz is rapid, and most will introduce phase and other distortions above this level as well. If you're listening for a 42kHz harmonic you can keep listening - the 28kHz second is probably not even there. J. Giles
emjej@uokvax.UUCP (01/22/85)
/***** uokvax:net.audio / lcuxc!wjm / 11:12 am Jan 15, 1985 */ [gould!jon] is correct in stating that a 44 kHz sampling rate will perfectly reproduce a 22 kHz signal (the Nyquist sampling theorem). However, that is not the problem. The problem is that, frankly, 22 kHz may not be adequate for *true* high fidelity reproduction....Since the third harmonic of a 14 kHz cymbal or bell note is 42 kHz, 22 kHz is nowhere near adequate for proper reproduction. /* ---------- */ I applaud the non-inflammatory approach of the poster of this note, but I must respectfully state that it's not clear to me how one reaches the conclusion from the admittedly correct premise (that 3 * 14 = 42), nor is it clear that any existing high-fidelity equipment can reproduce a 42 KHz signal with any accuracy. I would think that one need not reproduce anything that humans (or maybe even 1-epsilon fraction of humans) could not perceive (which perhaps includes feeling the shockwaves from the cannon in some pieces)--if a bat wants to spend money on hi-fi, it can spend its own :-). Comments on both issues (especially from folks knowledgeable about speakers and cartridges, which seem to me the obvious weak links in correctly transmitting such information--not the people, the devices :-) are invited. "Have you heard the Magnestatic? It's phenomenal!" James Jones
hrs@houxb.UUCP (H.SILBIGER) (01/22/85)
The fact that you can't hear above 20 kHz, or 15 kHz if you are Organization: AT&T Information Systems, Holmdel NJ Lines: 13 over 25, does mean that you cant hear harmonics above those frequencies. If you hear somethinvg, it is ipso facto withoin the range of your auditory mechanism. There are very few non-auditory effects of ultrasound, and the ultrasound must be of very high intensity. There is some correlation with body size, i.e. your cat can hear higher frequency harmonics than you and may scratch you if displeased, and your elephant may stomp if it hears distorted infrasound. Herman Silbiger, Certified Golden Ears "20 Years in thd Business of Sound Quality"