Path: utzoo!attcan!uunet!peregrine!elroy!spl1!laidbak!jeq
From: jeq@laidbak.UUCP (Jonathan E. Quist)
Newsgroups: comp.sys.next
Subject: Re: 16-bit Digital Sound: FP or int?
Message-ID: <1795@laidbak.UUCP>
Date: 17 Nov 88 22:20:56 GMT
References: <44@softart.UUCP> <26814@ucbvax.BERKELEY.EDU>
Reply-To: jeq@laidbak.UUCP (Jonathan E. Quist)
Organization: Lachman Associates, Inc. Naperville, Il.
Lines: 29

In article <26814@ucbvax.BERKELEY.EDU> edward@ucbarpa.Berkeley.EDU.UUCP (Edward Wang) writes:
>In article <44@softart.UUCP> riml@softart.UUCP (Research in Motion Limited) writes:
>>The reason I ask these questions is simple.  I have been convinced that
>>the "CD" digital encoding uses floating point.  Why?  If it didn't, low
>>level signals would be hopelessly distorted due to the lack of information
>>content in a LINEAR encoding at low levels.  
>> [blah, blah, blah]
>
>CD's use 16-bit fixed point.
>
>Low-level signals are distorted, but they are low level, as in hard to hear.

Low-level signals are _not_ distorted, at least not "hopelessly",
they just have a lower signal to noise ratio.

In computer sound synthesis, it is common practice to do the numerical
work in floating point, converting to integers for playback.
After the final mix of a piece is completed, the whole thing is scaled
based upon the largest sample, so that the integer version uses the
full scale of the D/A converters.

A floating point digital audio format would have some advantages,
but it will be millions of years before the manufacturers agree on
a format.  :')


jonathan quist
Lachman Associates, Inc.
laidbak!jeq