🚀 go-pugleaf

> 
> Yes..  It seems like the Motorola is well suited to FIR audio filters
> with 24bit fixed point maths and 56bit accumulator.  (IIRC)
> 
> I also Think that 24bit is plenty good enough for audio.  But 32bit
> bit floating point is a lot like 24bit fixed point with an 8-bit
> exponent term.  Not sure how bit the SHARC accumulator is, but I  will
> check it out..The SHARC processors are a lot faster than the Motorola
> 56300 series.  So its not a problem for me that they do 32bit maths
> insted of 24bit.
> 

Just in case it isn't clear - The Sharc is a 32 bit processor. It can be used 
for 32 bit fixed point math (80 bit accumulator) or as a floating point 
processor. At least half of all Sharc applications are fixed point. You can 
also easily convert from float to fixed and fixed to float.

Motorola used to own the digital audio market with their 24 bit fixed point 
processors. In the last few years, the Sharc has been capturing market share 
primarily due to improved interfacing with digital audio data converters and 
the demand for increased bit resolution. 24 bit math is great for 16 bit 
converters but may be a bit short for 24 bit converters (which are really 
only 20 bit).

ADI has two chips that were designed for digital audio applications: the 
ADSP-21065L and the newer ADSP-21161. IMHO, the 21161 is probably the best 
DSP for digital audio applications.

Of course, you can use many different DSPs for this purpose, especially if 
you use IIR filtering. 16 bit DSPs can perform 32 bit fixed point math. It 
just takes more cycles. The advantage of processors like the Mot 56K and 
Sharc is that the math precision is sufficient in a single cycle. 



-- 
Al Clark
Danville Signal Processing, Inc.
--------------------------------------------------------------------
Purveyors of Fine DSP Hardware and other Cool Stuff
Available at http://www.danvillesignal.com