Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!linus!decvax!harpo!seismo!hao!hplabs!sri-unix!KFL@MIT-MC From: KFL@MIT-MC@sri-unix.UUCP Newsgroups: net.physics Subject: none Message-ID: <4686@sri-arpa.UUCP> Date: Tue, 30-Aug-83 03:49:00 EDT Article-I.D.: sri-arpa.4686 Posted: Tue Aug 30 03:49:00 1983 Date-Received: Thu, 1-Sep-83 20:55:19 EDT Lines: 30 From: Keith F. Lynch Date: 27 August 1983 10:27 EDT From: David C Plummer How many people believe in the Nyquist theory when applied to music? Example: is 44.1 Ksamples per second really sufficient for digital audio disks? (I say no. This gives a Nyquist frequency of 22K and with filters the upper range is diminished to around 20K. People will say that you can't hear above 20K anyway. So what? You can't hear below 16Hz either. You can FEEL below 16Hz, so there is no reason there are not some subtle interactions above 20K.) I can hear up to 23K, and I know people who can hear up to 25. How good do we want our music to sound? If we really want it to be indistinguishable from real life, then we will have to record 50 or 60 Ksamples per second. Ultrasonics, like subsonics, can be felt (it feels just like a loud noise, only without any sound). Why all this attention to audio? What's good for the ear is good for the eye. How much longer will we be stuck with a 50 year old video standard with a few 30 year old patches? What resolution must a TV have to be indistinuishable from reality? How many frames per second? Are three colors really enough? What about reproducing UV and infrared? If we want many really good digitial video and audio channels we are going to rapidly run out of bandwidth. We need realtime compression/ decompression to make things managable. Is anyone working on this?