Experiences with the AT&T DSP32 Digital Signal Processor in Digital Audio Applications
The DSP32 signal processor developed at Bell Labs finds more and more applications in the digital audio field. Due to its 32 bit floating-point architecture with a 24 bit mantissa and an 8 bit exponent the DSP32 is well suited for processing signals in professional digital audio quality. The floating-point format also greatly eases the programming of algorithms particularly useful in digital audio, such as recursive filters, dynamics processors and math routines in general. The DSP32's pipelined architecture implies some strange programming techniques which must be taken care of. Examples for interfacing the DSP32 to fixed-point formats as well as programming examples for several applications are given.
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is temporarily free for AES members.