ted@dgbt.doc.ca (Ted Grusec) (04/10/91)
In <1991Apr7.100059.1489@urz.unibas.ch> Iaci writes: >>In article <1991Apr6.004426.24266@dgbt.doc.ca>, ted@dgbt.doc.ca (Ted Grusec) >>writes: >> One comment on this general thread on perfect pitch. Some people seem >> to think that way that we normally deal with visual color is analogous >> to perfect pitch. I don't think this is so. If I show you a "red", and >> then a slightly different shade of "red" at some time later, then you >> are not likely to be able to detect the difference between these two >> different "reds" without having both to compare. >Maybe not me, but maybe Monet could have done it... I doubt if Monet, or anyone could do it. If you present a given colour, rigidly specified by wavelength, in one colour context (e.g. surrounded by a light shade of gray), and the identical wavelength in a different colour context (e.g. surrounded by a dark shade of gray), the two "identical" colours look quite different, even if the colours surrounded by these contexts are presented for simultaneous viewing, side-by-side. This kind of demonstration is often presented graphically in introductory psychology texts. By contrast, a person with perfect pitch will identify the pitch of a given note no matter what very different chords that pitched note is presented in. >>With perfect pitch, a person can detect a slight difference in pitch >>between two notes presented at different times without needing to have both >>present for comparison. Pigeons, and other birds, however, do have a "perfect >> color" sense that IS like perfect pitch, but that's another story. >I think it's very hard to compare perfect pitch to perfect colour sense, since >the ear and the eye have very different capabilities. >As someone already has pointed out the ear is capable to hear a range of up to >10 octaves, the eye about one! >This makes it possible to us to hear all the overtones that enable us to >distinguish vowels (understand language). >When you play a chord to a person with a trained ear he/she will be able to >tell you all the notes you played. If you mix yellow with blue colour then >the eye can only see a green and is not able to see the two previous colours. >On the other hand the eye is much better in geometrical resolution. >And then a question I always was wondering about: if I see a red and you >see the same red, who tells me if we see it the same way? I certainly agree that colour and pitch are very different in many ways. But it is sensible to compare them in this one respect, namely absolute versus relative identification of frequency (i.e. pitch frequency and colour wavelength). It seems to me that the analytic analogy you draw is not quite accurate. You can analyze a chord into constituent notes but you can't analyze a given note into constituent fundamentals + overtones, any more than you can analyze green into yellow and blue, visually. As for understanding language, this is based mostly on a very narrow band of frequencies (centering around about 6K hertz). So people with hearing losses that restrict their perception to below 10K hertz and above, say 4K Hertz have no problems understanding language. That one and one-half octaves is all you need for accurate speech perception, and you don't need the 10 octaves for that task. As for your last point, it's an old philosophical chestnut as to how the subjective perception of any two people may be compared. All we really know is that we can agree on the names we use, we can compare discriminabilities, and we can compare other relevant bits of behaviour. If two people are identical in these and other behavioural characteristics and measures, we have no reason for suspecting that they differ in their perception. Where behavioural differences between people ARE found, we can say they must differ systematically in their perceptions. However, even if behaviourally identical, there is no way we can really say that they perceive identically in the fully subjectively sense. The physiological psychologists might be able to shed further light on this, someday.
barrett@jhunix.HCF.JHU.EDU (Dan Barrett) (04/11/91)
In article <1991Apr10.031342.27656@dgbt.doc.ca> ted@dgbt.doc.ca (Ted Grusec) writes: >I doubt if Monet, or anyone could do it. [That is, have "perfect color" the way some people have "perfect pitch".] I know of one person who does indeed (it is claimed) have "perfect color". But this is hearsay; I have not seen it demonstrated itself. But I have no reason to disbelieve it exists. >If you present a given colour, rigidly specified by wavelength, in one >colour context (e.g. surrounded by a light shade of gray), and the identical >wavelength in a different colour context (e.g. surrounded by a dark shade of >gray), the two "identical" colours look quite different, even if the colours >surrounded by these contexts are presented for simultaneous viewing, >'side-by-side. Similar tricks can be played with sound. Read on. >By contrast, a person with perfect pitch will identify the pitch of >a given note no matter what very different chords that pitched note is >presented in. No, the ear can EASILY be fooled. When several pitches are played at exactly the same time, the ear perceives them as a single sound. It might sound like a single pitch, or even be pitchless. I refer you to the work of Stephen McAdams in this area. Or just listen to a pipe organ. BTW, I have perfect pitch. Dan //////////////////////////////////////\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ | Dan Barrett, Department of Computer Science Johns Hopkins University | | INTERNET: barrett@cs.jhu.edu | | | COMPUSERVE: >internet:barrett@cs.jhu.edu | UUCP: barrett@jhunix.UUCP | \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\/////////////////////////////////////
rkl@cbnewsh.att.com (kevin.laux) (04/11/91)
In article <1991Apr10.031342.27656@dgbt.doc.ca>, ted@dgbt.doc.ca (Ted Grusec) writes: > > As for understanding language, this is based mostly on a very narrow band > of frequencies (centering around about 6K hertz). So people with hearing > losses that restrict their perception to below 10K hertz and above, say 4K > Hertz have no problems understanding language. That one and one-half octaves > is all you need for accurate speech perception, and you don't need the 10 > octaves for that task. Am I missing something here? Speech in the range you mention is going to be harmonics, not fundamental frequencies. Telephone bandwidth cuts off at 3.5 KHz. Where do you get that understanding spoken language centers around 6 KHz? -- ________________________________________________________________________________ R. Kevin Laux Email: rkl1@hound.att.com AT&T Bell Labs Voice: (908) 949-1160 Holmdel, NJ 07733 Fax: (908) 949-0959