Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site lanl-a.UUCP
Path: utzoo!watmath!clyde!burl!hou3c!hocda!houxm!ihnp4!zehntel!hplabs!hao!seismo!cmcl2!lanl-a!wkp
From: wkp@lanl-a.UUCP
Newsgroups: net.audio
Subject: Record speeds
Message-ID: <13812@lanl-a.UUCP>
Date: Mon, 24-Sep-84 13:17:06 EDT
Article-I.D.: lanl-a.13812
Posted: Mon Sep 24 13:17:06 1984
Date-Received: Thu, 27-Sep-84 04:14:01 EDT
Organization: Los Alamos National Laboratory
Lines: 44

This article is in reply to the question appearing in net.audio regarding
the reasons for choosing the peculiar speeds of 33-1/3, 45, and 78 rpm on 
those primitive analog recordings called records :-).

1.  The true speed for the old shellac pressings are not 78 rpm,
    but 78.26 rpm.

2.  Analog recordings are a compromise.  On the one hand, a listener
    wants a long playing time; on the other hand, most also desire
    a decent frequency response.

3.  For a given fixed diameter disk (e.g. 10 or 12 inches) one can make
    the tracking speed very very slow so that the playing time will be very
    long.  However, the slower the tracking speed, the poorer the frequency
    response.

4.  A formula can be derived for the maximum playing time of a record as
    a function of the highest frequency f one wishes to resolve and the record
    radius.  The rpm at this playing time is given by the handy-dandy formula:
          rpm = 4*f*h/R 
    where h is the groove waveform size (~1/100 in).
    Thus, for a 10 inch record, and for a frequency band of up to 8000 Hz
    (people in the 1930's were not golden ears) one can derive that the
    rpm necessary is about 78 rpm.  A similar story holds for speeds of 33-1/3.

5.  A corollary to this:  The reason why there is so much wasted space on
    the inside of a record near its center (e.g., where the label is) is that
    near its center the needle goes slower, and hence the higher frequencies
    are not resolvable.

6.  A second corollary:  Higher frequency music should sound better at      
    the beginning of a record, since these frequencies are easily resolved
    at the start of a record.  Towards the end of the record (near the label)
    it is harder to resolve these frequencies.

7.  Any further technical discussion of this belongs on net.physics where
    I can go into more detail.  All flames should be sent to dsl@lanl and 
    alp@lanl who helped write this article; all money should be sent to
    me (wkp@lanl) as I am saving for a CD.

					    bill peter