Target Support Package™ TC6

C64x Block Exponent - Minimum number of extra sign bits in each input channel

Library

C64x DSP Library — Math and Matrices

Description

The C64x Block Exponent block first computes the number of extra sign bits of all values in each channel of an input signal, and then returns the minimum number of sign bits found in each channel. The number of elements in each input channel must be a multiple of eight. Input elements must be 32-bit signed fixed-point data types. The output is a vector of 16-bit integers — one integer for each channel of the input signal.

This block is useful for determining whether every sample in a channel is using extra sign bits. If so, you can scale your signal by the minimum number of extra sign bits to eliminate the common extra bits. This increases the representable precision and decreases the representable range of the signal.

Block Exponent blocks support both continuous and discrete sample times. This block supports little-endian code generation only.

Dialog Box

Algorithm

In simulation, the Block Exponent block is equivalent to the TMS320C64x DSP Library assembly code function DSP_bexp. During code generation, this block calls the DSP_bexp routine given to produce optimized code.

C64x Bit Reverse

C64x Complex FIR