Subject: Re: [linux-audio-dev] LADSPA 64bit FP support ?
From: James McCartney (asynth_AT_io.com)
Date: Sat Mar 25 2000 - 21:49:16 EET
on 3/25/00 12:59 PM, David Olofson at david_AT_gardena.net wrote:
> the mantissa is
> actually 25 bits, including the one's bit that's an implied 1 in
> non-denormals
No, its 24. 1 sign bit, 8 exponent bits, 23 mantissa bits plus the one
implied.
>> 16 bits is already sufficient if it were just SNR.
>
> No, not unless it's dynamic. Ever heard a CD fade out when playing it
> back at above 100 dB peak?
CDs dont have 16 bits SNR at all amplitudes. That's my point.
FP has 24 bits SNR at all amplitudes. Integer formats have SNR depending on
their amplitude.
Don't confuse dynamic range with SNR.
N
> But what happens if you add a very high amplitude signal to an audio
> signal (to make some other plugin modulate the audio signal for
> example), and then filter that signal away. What happens to the
> mantissa?
I'd say you are making a mathematical mistake if you are formulating your
calculations in such a way.
>
> Two different definitions of "dynamic range":
>
> 1) The difference between the minimum signal level that can
> be registeged, and the clip level.
correct.
> 2) The resolution of a sample at the peak level.
This is signal to noise ratio, not dynamic range.
>
>> With FP the dyanamic range doesn't depend on the amplitude for
>> any practical purpose.
>
> Depends on what you call a "practical purpose". (See above.)
Wrong. For FP dynamic range does not depend on amplitude, unless you have a
denormalized number, period.
--- james mccartney james_AT_audiosynth.com <http://www.audiosynth.com>
SuperCollider - a real time synthesis programming language for the PowerMac.
<ftp://www.audiosynth.com/pub/updates/SC2.2.7.sea.hqx>
This archive was generated by hypermail 2b28 : Sat Mar 25 2000 - 22:31:21 EET