Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site laidbak.UUCP
Path: utzoo!watmath!clyde!cbosgd!ihnp4!laidbak!jeq
From: jeq@laidbak.UUCP (Jonathan E. Quist)
Newsgroups: net.analog
Subject: Re: Synthesizer Recommendations
Message-ID: <174@laidbak.UUCP>
Date: Thu, 15-Aug-85 19:14:30 EDT
Article-I.D.: laidbak.174
Posted: Thu Aug 15 19:14:30 1985
Date-Received: Sun, 18-Aug-85 02:26:20 EDT
References: <3613@decwrl.UUCP>
Reply-To: jeq@laidbak.UUCP (Jonathan E. Quist)
Organization: LAI Chicago
Lines: 29

In article <3613@decwrl.UUCP> williams@kirk.DEC (John Williams 223-3402) writes:
>	The best realistic S/N ratio you should ever want is 95 dB. This
>is the limit for the human ear. This works out to 18 bits. 12 bits will
>yield about 70 dB, which is probably good enough for alot of applications.

Wrong.  The range of the human ear (from just audible to threshold of pain)
is 120dB.
18 bits = (6dB/bit) * 18 = 124 dB.  12 bits is 72dB.
In practice, once you add analog noise on top of everything else,
12 bits is pushing the low side.  Many folks insist that
the 16 bit standard is still too small a word size.
One large box (the Lucasfilm ASP) uses (correct me if I'm wrong,
anyone) 20 bit converters (both D/A and A/D) and a 24 bit data path.

>2) Do some research on the MIDI standard and include it.
>4) Play test it. Get some professional musicians to try it.

Excellent advice.  Some of the best hardware has been
packaged under some of the worst software/control layouts...
(MIDI products are not exempt from this!)
Even giving it to the kid next door for a week or two will
tell you lots about what to improve; a musician who has
torn his/her hair out with a frustrating product may be
able to point out weak spots more quickly....

Jonathan E. Quist
Lachman Associates, Inc.
ihnp4!laidbak!jeq
``I deny this is a disclaimer.''