CMSIS DSP library
Dependents: KL25Z_FFT_Demo Hat_Board_v5_1 KL25Z_FFT_Demo_tony KL25Z_FFT_Demo_tony ... more
Fork of mbed-dsp by
cmsis_dsp/FilteringFunctions/arm_biquad_cascade_df1_f32.c@3:7a284390b0ce, 2013-11-08 (annotated)
- Committer:
- mbed_official
- Date:
- Fri Nov 08 13:45:10 2013 +0000
- Revision:
- 3:7a284390b0ce
- Parent:
- 2:da51fb522205
Synchronized with git revision e69956aba2f68a2a26ac26b051f8d349deaa1ce8
Who changed what in which revision?
User | Revision | Line number | New contents of line |
---|---|---|---|
emilmont | 1:fdd22bb7aa52 | 1 | /* ---------------------------------------------------------------------- |
mbed_official | 3:7a284390b0ce | 2 | * Copyright (C) 2010-2013 ARM Limited. All rights reserved. |
emilmont | 1:fdd22bb7aa52 | 3 | * |
mbed_official | 3:7a284390b0ce | 4 | * $Date: 17. January 2013 |
mbed_official | 3:7a284390b0ce | 5 | * $Revision: V1.4.1 |
emilmont | 1:fdd22bb7aa52 | 6 | * |
emilmont | 2:da51fb522205 | 7 | * Project: CMSIS DSP Library |
emilmont | 2:da51fb522205 | 8 | * Title: arm_biquad_cascade_df1_f32.c |
emilmont | 1:fdd22bb7aa52 | 9 | * |
emilmont | 2:da51fb522205 | 10 | * Description: Processing function for the |
emilmont | 1:fdd22bb7aa52 | 11 | * floating-point Biquad cascade DirectFormI(DF1) filter. |
emilmont | 1:fdd22bb7aa52 | 12 | * |
emilmont | 1:fdd22bb7aa52 | 13 | * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0 |
emilmont | 1:fdd22bb7aa52 | 14 | * |
mbed_official | 3:7a284390b0ce | 15 | * Redistribution and use in source and binary forms, with or without |
mbed_official | 3:7a284390b0ce | 16 | * modification, are permitted provided that the following conditions |
mbed_official | 3:7a284390b0ce | 17 | * are met: |
mbed_official | 3:7a284390b0ce | 18 | * - Redistributions of source code must retain the above copyright |
mbed_official | 3:7a284390b0ce | 19 | * notice, this list of conditions and the following disclaimer. |
mbed_official | 3:7a284390b0ce | 20 | * - Redistributions in binary form must reproduce the above copyright |
mbed_official | 3:7a284390b0ce | 21 | * notice, this list of conditions and the following disclaimer in |
mbed_official | 3:7a284390b0ce | 22 | * the documentation and/or other materials provided with the |
mbed_official | 3:7a284390b0ce | 23 | * distribution. |
mbed_official | 3:7a284390b0ce | 24 | * - Neither the name of ARM LIMITED nor the names of its contributors |
mbed_official | 3:7a284390b0ce | 25 | * may be used to endorse or promote products derived from this |
mbed_official | 3:7a284390b0ce | 26 | * software without specific prior written permission. |
mbed_official | 3:7a284390b0ce | 27 | * |
mbed_official | 3:7a284390b0ce | 28 | * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS |
mbed_official | 3:7a284390b0ce | 29 | * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT |
mbed_official | 3:7a284390b0ce | 30 | * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS |
mbed_official | 3:7a284390b0ce | 31 | * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE |
mbed_official | 3:7a284390b0ce | 32 | * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, |
mbed_official | 3:7a284390b0ce | 33 | * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, |
mbed_official | 3:7a284390b0ce | 34 | * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; |
mbed_official | 3:7a284390b0ce | 35 | * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER |
mbed_official | 3:7a284390b0ce | 36 | * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT |
mbed_official | 3:7a284390b0ce | 37 | * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN |
mbed_official | 3:7a284390b0ce | 38 | * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE |
mbed_official | 3:7a284390b0ce | 39 | * POSSIBILITY OF SUCH DAMAGE. |
emilmont | 1:fdd22bb7aa52 | 40 | * -------------------------------------------------------------------- */ |
emilmont | 1:fdd22bb7aa52 | 41 | |
emilmont | 1:fdd22bb7aa52 | 42 | #include "arm_math.h" |
emilmont | 1:fdd22bb7aa52 | 43 | |
emilmont | 1:fdd22bb7aa52 | 44 | /** |
emilmont | 1:fdd22bb7aa52 | 45 | * @ingroup groupFilters |
emilmont | 1:fdd22bb7aa52 | 46 | */ |
emilmont | 1:fdd22bb7aa52 | 47 | |
emilmont | 1:fdd22bb7aa52 | 48 | /** |
emilmont | 1:fdd22bb7aa52 | 49 | * @defgroup BiquadCascadeDF1 Biquad Cascade IIR Filters Using Direct Form I Structure |
emilmont | 1:fdd22bb7aa52 | 50 | * |
emilmont | 1:fdd22bb7aa52 | 51 | * This set of functions implements arbitrary order recursive (IIR) filters. |
emilmont | 1:fdd22bb7aa52 | 52 | * The filters are implemented as a cascade of second order Biquad sections. |
emilmont | 1:fdd22bb7aa52 | 53 | * The functions support Q15, Q31 and floating-point data types. |
emilmont | 1:fdd22bb7aa52 | 54 | * Fast version of Q15 and Q31 also supported on CortexM4 and Cortex-M3. |
emilmont | 1:fdd22bb7aa52 | 55 | * |
emilmont | 1:fdd22bb7aa52 | 56 | * The functions operate on blocks of input and output data and each call to the function |
emilmont | 1:fdd22bb7aa52 | 57 | * processes <code>blockSize</code> samples through the filter. |
emilmont | 1:fdd22bb7aa52 | 58 | * <code>pSrc</code> points to the array of input data and |
emilmont | 1:fdd22bb7aa52 | 59 | * <code>pDst</code> points to the array of output data. |
emilmont | 1:fdd22bb7aa52 | 60 | * Both arrays contain <code>blockSize</code> values. |
emilmont | 1:fdd22bb7aa52 | 61 | * |
emilmont | 1:fdd22bb7aa52 | 62 | * \par Algorithm |
emilmont | 1:fdd22bb7aa52 | 63 | * Each Biquad stage implements a second order filter using the difference equation: |
emilmont | 1:fdd22bb7aa52 | 64 | * <pre> |
emilmont | 1:fdd22bb7aa52 | 65 | * y[n] = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] |
emilmont | 1:fdd22bb7aa52 | 66 | * </pre> |
emilmont | 1:fdd22bb7aa52 | 67 | * A Direct Form I algorithm is used with 5 coefficients and 4 state variables per stage. |
emilmont | 1:fdd22bb7aa52 | 68 | * \image html Biquad.gif "Single Biquad filter stage" |
emilmont | 1:fdd22bb7aa52 | 69 | * Coefficients <code>b0, b1 and b2 </code> multiply the input signal <code>x[n]</code> and are referred to as the feedforward coefficients. |
emilmont | 1:fdd22bb7aa52 | 70 | * Coefficients <code>a1</code> and <code>a2</code> multiply the output signal <code>y[n]</code> and are referred to as the feedback coefficients. |
emilmont | 1:fdd22bb7aa52 | 71 | * Pay careful attention to the sign of the feedback coefficients. |
emilmont | 1:fdd22bb7aa52 | 72 | * Some design tools use the difference equation |
emilmont | 1:fdd22bb7aa52 | 73 | * <pre> |
emilmont | 1:fdd22bb7aa52 | 74 | * y[n] = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] - a1 * y[n-1] - a2 * y[n-2] |
emilmont | 1:fdd22bb7aa52 | 75 | * </pre> |
emilmont | 1:fdd22bb7aa52 | 76 | * In this case the feedback coefficients <code>a1</code> and <code>a2</code> must be negated when used with the CMSIS DSP Library. |
emilmont | 1:fdd22bb7aa52 | 77 | * |
emilmont | 1:fdd22bb7aa52 | 78 | * \par |
emilmont | 1:fdd22bb7aa52 | 79 | * Higher order filters are realized as a cascade of second order sections. |
emilmont | 1:fdd22bb7aa52 | 80 | * <code>numStages</code> refers to the number of second order stages used. |
emilmont | 1:fdd22bb7aa52 | 81 | * For example, an 8th order filter would be realized with <code>numStages=4</code> second order stages. |
emilmont | 1:fdd22bb7aa52 | 82 | * \image html BiquadCascade.gif "8th order filter using a cascade of Biquad stages" |
emilmont | 1:fdd22bb7aa52 | 83 | * A 9th order filter would be realized with <code>numStages=5</code> second order stages with the coefficients for one of the stages configured as a first order filter (<code>b2=0</code> and <code>a2=0</code>). |
emilmont | 1:fdd22bb7aa52 | 84 | * |
emilmont | 1:fdd22bb7aa52 | 85 | * \par |
emilmont | 1:fdd22bb7aa52 | 86 | * The <code>pState</code> points to state variables array. |
emilmont | 1:fdd22bb7aa52 | 87 | * Each Biquad stage has 4 state variables <code>x[n-1], x[n-2], y[n-1],</code> and <code>y[n-2]</code>. |
emilmont | 1:fdd22bb7aa52 | 88 | * The state variables are arranged in the <code>pState</code> array as: |
emilmont | 1:fdd22bb7aa52 | 89 | * <pre> |
emilmont | 1:fdd22bb7aa52 | 90 | * {x[n-1], x[n-2], y[n-1], y[n-2]} |
emilmont | 1:fdd22bb7aa52 | 91 | * </pre> |
emilmont | 1:fdd22bb7aa52 | 92 | * |
emilmont | 1:fdd22bb7aa52 | 93 | * \par |
emilmont | 1:fdd22bb7aa52 | 94 | * The 4 state variables for stage 1 are first, then the 4 state variables for stage 2, and so on. |
emilmont | 1:fdd22bb7aa52 | 95 | * The state array has a total length of <code>4*numStages</code> values. |
emilmont | 1:fdd22bb7aa52 | 96 | * The state variables are updated after each block of data is processed, the coefficients are untouched. |
emilmont | 1:fdd22bb7aa52 | 97 | * |
emilmont | 1:fdd22bb7aa52 | 98 | * \par Instance Structure |
emilmont | 1:fdd22bb7aa52 | 99 | * The coefficients and state variables for a filter are stored together in an instance data structure. |
emilmont | 1:fdd22bb7aa52 | 100 | * A separate instance structure must be defined for each filter. |
emilmont | 1:fdd22bb7aa52 | 101 | * Coefficient arrays may be shared among several instances while state variable arrays cannot be shared. |
emilmont | 1:fdd22bb7aa52 | 102 | * There are separate instance structure declarations for each of the 3 supported data types. |
emilmont | 1:fdd22bb7aa52 | 103 | * |
emilmont | 1:fdd22bb7aa52 | 104 | * \par Init Functions |
emilmont | 1:fdd22bb7aa52 | 105 | * There is also an associated initialization function for each data type. |
emilmont | 1:fdd22bb7aa52 | 106 | * The initialization function performs following operations: |
emilmont | 1:fdd22bb7aa52 | 107 | * - Sets the values of the internal structure fields. |
emilmont | 1:fdd22bb7aa52 | 108 | * - Zeros out the values in the state buffer. |
mbed_official | 3:7a284390b0ce | 109 | * To do this manually without calling the init function, assign the follow subfields of the instance structure: |
mbed_official | 3:7a284390b0ce | 110 | * numStages, pCoeffs, pState. Also set all of the values in pState to zero. |
emilmont | 1:fdd22bb7aa52 | 111 | * |
emilmont | 1:fdd22bb7aa52 | 112 | * \par |
emilmont | 1:fdd22bb7aa52 | 113 | * Use of the initialization function is optional. |
emilmont | 1:fdd22bb7aa52 | 114 | * However, if the initialization function is used, then the instance structure cannot be placed into a const data section. |
emilmont | 1:fdd22bb7aa52 | 115 | * To place an instance structure into a const data section, the instance structure must be manually initialized. |
emilmont | 1:fdd22bb7aa52 | 116 | * Set the values in the state buffer to zeros before static initialization. |
emilmont | 1:fdd22bb7aa52 | 117 | * The code below statically initializes each of the 3 different data type filter instance structures |
emilmont | 1:fdd22bb7aa52 | 118 | * <pre> |
emilmont | 1:fdd22bb7aa52 | 119 | * arm_biquad_casd_df1_inst_f32 S1 = {numStages, pState, pCoeffs}; |
emilmont | 1:fdd22bb7aa52 | 120 | * arm_biquad_casd_df1_inst_q15 S2 = {numStages, pState, pCoeffs, postShift}; |
emilmont | 1:fdd22bb7aa52 | 121 | * arm_biquad_casd_df1_inst_q31 S3 = {numStages, pState, pCoeffs, postShift}; |
emilmont | 1:fdd22bb7aa52 | 122 | * </pre> |
emilmont | 1:fdd22bb7aa52 | 123 | * where <code>numStages</code> is the number of Biquad stages in the filter; <code>pState</code> is the address of the state buffer; |
emilmont | 1:fdd22bb7aa52 | 124 | * <code>pCoeffs</code> is the address of the coefficient buffer; <code>postShift</code> shift to be applied. |
emilmont | 1:fdd22bb7aa52 | 125 | * |
emilmont | 1:fdd22bb7aa52 | 126 | * \par Fixed-Point Behavior |
emilmont | 1:fdd22bb7aa52 | 127 | * Care must be taken when using the Q15 and Q31 versions of the Biquad Cascade filter functions. |
emilmont | 1:fdd22bb7aa52 | 128 | * Following issues must be considered: |
emilmont | 1:fdd22bb7aa52 | 129 | * - Scaling of coefficients |
emilmont | 1:fdd22bb7aa52 | 130 | * - Filter gain |
emilmont | 1:fdd22bb7aa52 | 131 | * - Overflow and saturation |
emilmont | 1:fdd22bb7aa52 | 132 | * |
emilmont | 1:fdd22bb7aa52 | 133 | * \par |
emilmont | 1:fdd22bb7aa52 | 134 | * <b>Scaling of coefficients: </b> |
emilmont | 1:fdd22bb7aa52 | 135 | * Filter coefficients are represented as fractional values and |
emilmont | 1:fdd22bb7aa52 | 136 | * coefficients are restricted to lie in the range <code>[-1 +1)</code>. |
emilmont | 1:fdd22bb7aa52 | 137 | * The fixed-point functions have an additional scaling parameter <code>postShift</code> |
emilmont | 1:fdd22bb7aa52 | 138 | * which allow the filter coefficients to exceed the range <code>[+1 -1)</code>. |
emilmont | 1:fdd22bb7aa52 | 139 | * At the output of the filter's accumulator is a shift register which shifts the result by <code>postShift</code> bits. |
emilmont | 1:fdd22bb7aa52 | 140 | * \image html BiquadPostshift.gif "Fixed-point Biquad with shift by postShift bits after accumulator" |
emilmont | 1:fdd22bb7aa52 | 141 | * This essentially scales the filter coefficients by <code>2^postShift</code>. |
emilmont | 1:fdd22bb7aa52 | 142 | * For example, to realize the coefficients |
emilmont | 1:fdd22bb7aa52 | 143 | * <pre> |
emilmont | 1:fdd22bb7aa52 | 144 | * {1.5, -0.8, 1.2, 1.6, -0.9} |
emilmont | 1:fdd22bb7aa52 | 145 | * </pre> |
emilmont | 1:fdd22bb7aa52 | 146 | * set the pCoeffs array to: |
emilmont | 1:fdd22bb7aa52 | 147 | * <pre> |
emilmont | 1:fdd22bb7aa52 | 148 | * {0.75, -0.4, 0.6, 0.8, -0.45} |
emilmont | 1:fdd22bb7aa52 | 149 | * </pre> |
emilmont | 1:fdd22bb7aa52 | 150 | * and set <code>postShift=1</code> |
emilmont | 1:fdd22bb7aa52 | 151 | * |
emilmont | 1:fdd22bb7aa52 | 152 | * \par |
emilmont | 1:fdd22bb7aa52 | 153 | * <b>Filter gain: </b> |
emilmont | 1:fdd22bb7aa52 | 154 | * The frequency response of a Biquad filter is a function of its coefficients. |
emilmont | 1:fdd22bb7aa52 | 155 | * It is possible for the gain through the filter to exceed 1.0 meaning that the filter increases the amplitude of certain frequencies. |
emilmont | 1:fdd22bb7aa52 | 156 | * This means that an input signal with amplitude < 1.0 may result in an output > 1.0 and these are saturated or overflowed based on the implementation of the filter. |
emilmont | 1:fdd22bb7aa52 | 157 | * To avoid this behavior the filter needs to be scaled down such that its peak gain < 1.0 or the input signal must be scaled down so that the combination of input and filter are never overflowed. |
emilmont | 1:fdd22bb7aa52 | 158 | * |
emilmont | 1:fdd22bb7aa52 | 159 | * \par |
emilmont | 1:fdd22bb7aa52 | 160 | * <b>Overflow and saturation: </b> |
emilmont | 1:fdd22bb7aa52 | 161 | * For Q15 and Q31 versions, it is described separately as part of the function specific documentation below. |
emilmont | 1:fdd22bb7aa52 | 162 | */ |
emilmont | 1:fdd22bb7aa52 | 163 | |
emilmont | 1:fdd22bb7aa52 | 164 | /** |
emilmont | 1:fdd22bb7aa52 | 165 | * @addtogroup BiquadCascadeDF1 |
emilmont | 1:fdd22bb7aa52 | 166 | * @{ |
emilmont | 1:fdd22bb7aa52 | 167 | */ |
emilmont | 1:fdd22bb7aa52 | 168 | |
emilmont | 1:fdd22bb7aa52 | 169 | /** |
emilmont | 1:fdd22bb7aa52 | 170 | * @param[in] *S points to an instance of the floating-point Biquad cascade structure. |
emilmont | 1:fdd22bb7aa52 | 171 | * @param[in] *pSrc points to the block of input data. |
emilmont | 1:fdd22bb7aa52 | 172 | * @param[out] *pDst points to the block of output data. |
emilmont | 1:fdd22bb7aa52 | 173 | * @param[in] blockSize number of samples to process per call. |
emilmont | 1:fdd22bb7aa52 | 174 | * @return none. |
emilmont | 1:fdd22bb7aa52 | 175 | * |
emilmont | 1:fdd22bb7aa52 | 176 | */ |
emilmont | 1:fdd22bb7aa52 | 177 | |
emilmont | 1:fdd22bb7aa52 | 178 | void arm_biquad_cascade_df1_f32( |
emilmont | 1:fdd22bb7aa52 | 179 | const arm_biquad_casd_df1_inst_f32 * S, |
emilmont | 1:fdd22bb7aa52 | 180 | float32_t * pSrc, |
emilmont | 1:fdd22bb7aa52 | 181 | float32_t * pDst, |
emilmont | 1:fdd22bb7aa52 | 182 | uint32_t blockSize) |
emilmont | 1:fdd22bb7aa52 | 183 | { |
emilmont | 1:fdd22bb7aa52 | 184 | float32_t *pIn = pSrc; /* source pointer */ |
emilmont | 1:fdd22bb7aa52 | 185 | float32_t *pOut = pDst; /* destination pointer */ |
emilmont | 1:fdd22bb7aa52 | 186 | float32_t *pState = S->pState; /* pState pointer */ |
emilmont | 1:fdd22bb7aa52 | 187 | float32_t *pCoeffs = S->pCoeffs; /* coefficient pointer */ |
emilmont | 1:fdd22bb7aa52 | 188 | float32_t acc; /* Simulates the accumulator */ |
emilmont | 1:fdd22bb7aa52 | 189 | float32_t b0, b1, b2, a1, a2; /* Filter coefficients */ |
emilmont | 1:fdd22bb7aa52 | 190 | float32_t Xn1, Xn2, Yn1, Yn2; /* Filter pState variables */ |
emilmont | 1:fdd22bb7aa52 | 191 | float32_t Xn; /* temporary input */ |
emilmont | 1:fdd22bb7aa52 | 192 | uint32_t sample, stage = S->numStages; /* loop counters */ |
emilmont | 1:fdd22bb7aa52 | 193 | |
emilmont | 1:fdd22bb7aa52 | 194 | |
mbed_official | 3:7a284390b0ce | 195 | #ifndef ARM_MATH_CM0_FAMILY |
emilmont | 1:fdd22bb7aa52 | 196 | |
emilmont | 1:fdd22bb7aa52 | 197 | /* Run the below code for Cortex-M4 and Cortex-M3 */ |
emilmont | 1:fdd22bb7aa52 | 198 | |
emilmont | 1:fdd22bb7aa52 | 199 | do |
emilmont | 1:fdd22bb7aa52 | 200 | { |
emilmont | 1:fdd22bb7aa52 | 201 | /* Reading the coefficients */ |
emilmont | 1:fdd22bb7aa52 | 202 | b0 = *pCoeffs++; |
emilmont | 1:fdd22bb7aa52 | 203 | b1 = *pCoeffs++; |
emilmont | 1:fdd22bb7aa52 | 204 | b2 = *pCoeffs++; |
emilmont | 1:fdd22bb7aa52 | 205 | a1 = *pCoeffs++; |
emilmont | 1:fdd22bb7aa52 | 206 | a2 = *pCoeffs++; |
emilmont | 1:fdd22bb7aa52 | 207 | |
emilmont | 1:fdd22bb7aa52 | 208 | /* Reading the pState values */ |
emilmont | 1:fdd22bb7aa52 | 209 | Xn1 = pState[0]; |
emilmont | 1:fdd22bb7aa52 | 210 | Xn2 = pState[1]; |
emilmont | 1:fdd22bb7aa52 | 211 | Yn1 = pState[2]; |
emilmont | 1:fdd22bb7aa52 | 212 | Yn2 = pState[3]; |
emilmont | 1:fdd22bb7aa52 | 213 | |
emilmont | 1:fdd22bb7aa52 | 214 | /* Apply loop unrolling and compute 4 output values simultaneously. */ |
emilmont | 1:fdd22bb7aa52 | 215 | /* The variable acc hold output values that are being computed: |
emilmont | 1:fdd22bb7aa52 | 216 | * |
emilmont | 1:fdd22bb7aa52 | 217 | * acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] |
emilmont | 1:fdd22bb7aa52 | 218 | * acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] |
emilmont | 1:fdd22bb7aa52 | 219 | * acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] |
emilmont | 1:fdd22bb7aa52 | 220 | * acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] |
emilmont | 1:fdd22bb7aa52 | 221 | */ |
emilmont | 1:fdd22bb7aa52 | 222 | |
emilmont | 1:fdd22bb7aa52 | 223 | sample = blockSize >> 2u; |
emilmont | 1:fdd22bb7aa52 | 224 | |
emilmont | 1:fdd22bb7aa52 | 225 | /* First part of the processing with loop unrolling. Compute 4 outputs at a time. |
emilmont | 1:fdd22bb7aa52 | 226 | ** a second loop below computes the remaining 1 to 3 samples. */ |
emilmont | 1:fdd22bb7aa52 | 227 | while(sample > 0u) |
emilmont | 1:fdd22bb7aa52 | 228 | { |
emilmont | 1:fdd22bb7aa52 | 229 | /* Read the first input */ |
emilmont | 1:fdd22bb7aa52 | 230 | Xn = *pIn++; |
emilmont | 1:fdd22bb7aa52 | 231 | |
emilmont | 1:fdd22bb7aa52 | 232 | /* acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] */ |
emilmont | 1:fdd22bb7aa52 | 233 | Yn2 = (b0 * Xn) + (b1 * Xn1) + (b2 * Xn2) + (a1 * Yn1) + (a2 * Yn2); |
emilmont | 1:fdd22bb7aa52 | 234 | |
emilmont | 1:fdd22bb7aa52 | 235 | /* Store the result in the accumulator in the destination buffer. */ |
emilmont | 1:fdd22bb7aa52 | 236 | *pOut++ = Yn2; |
emilmont | 1:fdd22bb7aa52 | 237 | |
emilmont | 1:fdd22bb7aa52 | 238 | /* Every time after the output is computed state should be updated. */ |
emilmont | 1:fdd22bb7aa52 | 239 | /* The states should be updated as: */ |
emilmont | 1:fdd22bb7aa52 | 240 | /* Xn2 = Xn1 */ |
emilmont | 1:fdd22bb7aa52 | 241 | /* Xn1 = Xn */ |
emilmont | 1:fdd22bb7aa52 | 242 | /* Yn2 = Yn1 */ |
emilmont | 1:fdd22bb7aa52 | 243 | /* Yn1 = acc */ |
emilmont | 1:fdd22bb7aa52 | 244 | |
emilmont | 1:fdd22bb7aa52 | 245 | /* Read the second input */ |
emilmont | 1:fdd22bb7aa52 | 246 | Xn2 = *pIn++; |
emilmont | 1:fdd22bb7aa52 | 247 | |
emilmont | 1:fdd22bb7aa52 | 248 | /* acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] */ |
emilmont | 1:fdd22bb7aa52 | 249 | Yn1 = (b0 * Xn2) + (b1 * Xn) + (b2 * Xn1) + (a1 * Yn2) + (a2 * Yn1); |
emilmont | 1:fdd22bb7aa52 | 250 | |
emilmont | 1:fdd22bb7aa52 | 251 | /* Store the result in the accumulator in the destination buffer. */ |
emilmont | 1:fdd22bb7aa52 | 252 | *pOut++ = Yn1; |
emilmont | 1:fdd22bb7aa52 | 253 | |
emilmont | 1:fdd22bb7aa52 | 254 | /* Every time after the output is computed state should be updated. */ |
emilmont | 1:fdd22bb7aa52 | 255 | /* The states should be updated as: */ |
emilmont | 1:fdd22bb7aa52 | 256 | /* Xn2 = Xn1 */ |
emilmont | 1:fdd22bb7aa52 | 257 | /* Xn1 = Xn */ |
emilmont | 1:fdd22bb7aa52 | 258 | /* Yn2 = Yn1 */ |
emilmont | 1:fdd22bb7aa52 | 259 | /* Yn1 = acc */ |
emilmont | 1:fdd22bb7aa52 | 260 | |
emilmont | 1:fdd22bb7aa52 | 261 | /* Read the third input */ |
emilmont | 1:fdd22bb7aa52 | 262 | Xn1 = *pIn++; |
emilmont | 1:fdd22bb7aa52 | 263 | |
emilmont | 1:fdd22bb7aa52 | 264 | /* acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] */ |
emilmont | 1:fdd22bb7aa52 | 265 | Yn2 = (b0 * Xn1) + (b1 * Xn2) + (b2 * Xn) + (a1 * Yn1) + (a2 * Yn2); |
emilmont | 1:fdd22bb7aa52 | 266 | |
emilmont | 1:fdd22bb7aa52 | 267 | /* Store the result in the accumulator in the destination buffer. */ |
emilmont | 1:fdd22bb7aa52 | 268 | *pOut++ = Yn2; |
emilmont | 1:fdd22bb7aa52 | 269 | |
emilmont | 1:fdd22bb7aa52 | 270 | /* Every time after the output is computed state should be updated. */ |
emilmont | 1:fdd22bb7aa52 | 271 | /* The states should be updated as: */ |
emilmont | 1:fdd22bb7aa52 | 272 | /* Xn2 = Xn1 */ |
emilmont | 1:fdd22bb7aa52 | 273 | /* Xn1 = Xn */ |
emilmont | 1:fdd22bb7aa52 | 274 | /* Yn2 = Yn1 */ |
emilmont | 1:fdd22bb7aa52 | 275 | /* Yn1 = acc */ |
emilmont | 1:fdd22bb7aa52 | 276 | |
emilmont | 1:fdd22bb7aa52 | 277 | /* Read the forth input */ |
emilmont | 1:fdd22bb7aa52 | 278 | Xn = *pIn++; |
emilmont | 1:fdd22bb7aa52 | 279 | |
emilmont | 1:fdd22bb7aa52 | 280 | /* acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] */ |
emilmont | 1:fdd22bb7aa52 | 281 | Yn1 = (b0 * Xn) + (b1 * Xn1) + (b2 * Xn2) + (a1 * Yn2) + (a2 * Yn1); |
emilmont | 1:fdd22bb7aa52 | 282 | |
emilmont | 1:fdd22bb7aa52 | 283 | /* Store the result in the accumulator in the destination buffer. */ |
emilmont | 1:fdd22bb7aa52 | 284 | *pOut++ = Yn1; |
emilmont | 1:fdd22bb7aa52 | 285 | |
emilmont | 1:fdd22bb7aa52 | 286 | /* Every time after the output is computed state should be updated. */ |
emilmont | 1:fdd22bb7aa52 | 287 | /* The states should be updated as: */ |
emilmont | 1:fdd22bb7aa52 | 288 | /* Xn2 = Xn1 */ |
emilmont | 1:fdd22bb7aa52 | 289 | /* Xn1 = Xn */ |
emilmont | 1:fdd22bb7aa52 | 290 | /* Yn2 = Yn1 */ |
emilmont | 1:fdd22bb7aa52 | 291 | /* Yn1 = acc */ |
emilmont | 1:fdd22bb7aa52 | 292 | Xn2 = Xn1; |
emilmont | 1:fdd22bb7aa52 | 293 | Xn1 = Xn; |
emilmont | 1:fdd22bb7aa52 | 294 | |
emilmont | 1:fdd22bb7aa52 | 295 | /* decrement the loop counter */ |
emilmont | 1:fdd22bb7aa52 | 296 | sample--; |
emilmont | 1:fdd22bb7aa52 | 297 | |
emilmont | 1:fdd22bb7aa52 | 298 | } |
emilmont | 1:fdd22bb7aa52 | 299 | |
emilmont | 1:fdd22bb7aa52 | 300 | /* If the blockSize is not a multiple of 4, compute any remaining output samples here. |
emilmont | 1:fdd22bb7aa52 | 301 | ** No loop unrolling is used. */ |
emilmont | 1:fdd22bb7aa52 | 302 | sample = blockSize & 0x3u; |
emilmont | 1:fdd22bb7aa52 | 303 | |
emilmont | 1:fdd22bb7aa52 | 304 | while(sample > 0u) |
emilmont | 1:fdd22bb7aa52 | 305 | { |
emilmont | 1:fdd22bb7aa52 | 306 | /* Read the input */ |
emilmont | 1:fdd22bb7aa52 | 307 | Xn = *pIn++; |
emilmont | 1:fdd22bb7aa52 | 308 | |
emilmont | 1:fdd22bb7aa52 | 309 | /* acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] */ |
emilmont | 1:fdd22bb7aa52 | 310 | acc = (b0 * Xn) + (b1 * Xn1) + (b2 * Xn2) + (a1 * Yn1) + (a2 * Yn2); |
emilmont | 1:fdd22bb7aa52 | 311 | |
emilmont | 1:fdd22bb7aa52 | 312 | /* Store the result in the accumulator in the destination buffer. */ |
emilmont | 1:fdd22bb7aa52 | 313 | *pOut++ = acc; |
emilmont | 1:fdd22bb7aa52 | 314 | |
emilmont | 1:fdd22bb7aa52 | 315 | /* Every time after the output is computed state should be updated. */ |
emilmont | 1:fdd22bb7aa52 | 316 | /* The states should be updated as: */ |
emilmont | 1:fdd22bb7aa52 | 317 | /* Xn2 = Xn1 */ |
emilmont | 1:fdd22bb7aa52 | 318 | /* Xn1 = Xn */ |
emilmont | 1:fdd22bb7aa52 | 319 | /* Yn2 = Yn1 */ |
emilmont | 1:fdd22bb7aa52 | 320 | /* Yn1 = acc */ |
emilmont | 1:fdd22bb7aa52 | 321 | Xn2 = Xn1; |
emilmont | 1:fdd22bb7aa52 | 322 | Xn1 = Xn; |
emilmont | 1:fdd22bb7aa52 | 323 | Yn2 = Yn1; |
emilmont | 1:fdd22bb7aa52 | 324 | Yn1 = acc; |
emilmont | 1:fdd22bb7aa52 | 325 | |
emilmont | 1:fdd22bb7aa52 | 326 | /* decrement the loop counter */ |
emilmont | 1:fdd22bb7aa52 | 327 | sample--; |
emilmont | 1:fdd22bb7aa52 | 328 | |
emilmont | 1:fdd22bb7aa52 | 329 | } |
emilmont | 1:fdd22bb7aa52 | 330 | |
emilmont | 1:fdd22bb7aa52 | 331 | /* Store the updated state variables back into the pState array */ |
emilmont | 1:fdd22bb7aa52 | 332 | *pState++ = Xn1; |
emilmont | 1:fdd22bb7aa52 | 333 | *pState++ = Xn2; |
emilmont | 1:fdd22bb7aa52 | 334 | *pState++ = Yn1; |
emilmont | 1:fdd22bb7aa52 | 335 | *pState++ = Yn2; |
emilmont | 1:fdd22bb7aa52 | 336 | |
emilmont | 1:fdd22bb7aa52 | 337 | /* The first stage goes from the input buffer to the output buffer. */ |
emilmont | 1:fdd22bb7aa52 | 338 | /* Subsequent numStages occur in-place in the output buffer */ |
emilmont | 1:fdd22bb7aa52 | 339 | pIn = pDst; |
emilmont | 1:fdd22bb7aa52 | 340 | |
emilmont | 1:fdd22bb7aa52 | 341 | /* Reset the output pointer */ |
emilmont | 1:fdd22bb7aa52 | 342 | pOut = pDst; |
emilmont | 1:fdd22bb7aa52 | 343 | |
emilmont | 1:fdd22bb7aa52 | 344 | /* decrement the loop counter */ |
emilmont | 1:fdd22bb7aa52 | 345 | stage--; |
emilmont | 1:fdd22bb7aa52 | 346 | |
emilmont | 1:fdd22bb7aa52 | 347 | } while(stage > 0u); |
emilmont | 1:fdd22bb7aa52 | 348 | |
emilmont | 1:fdd22bb7aa52 | 349 | #else |
emilmont | 1:fdd22bb7aa52 | 350 | |
emilmont | 1:fdd22bb7aa52 | 351 | /* Run the below code for Cortex-M0 */ |
emilmont | 1:fdd22bb7aa52 | 352 | |
emilmont | 1:fdd22bb7aa52 | 353 | do |
emilmont | 1:fdd22bb7aa52 | 354 | { |
emilmont | 1:fdd22bb7aa52 | 355 | /* Reading the coefficients */ |
emilmont | 1:fdd22bb7aa52 | 356 | b0 = *pCoeffs++; |
emilmont | 1:fdd22bb7aa52 | 357 | b1 = *pCoeffs++; |
emilmont | 1:fdd22bb7aa52 | 358 | b2 = *pCoeffs++; |
emilmont | 1:fdd22bb7aa52 | 359 | a1 = *pCoeffs++; |
emilmont | 1:fdd22bb7aa52 | 360 | a2 = *pCoeffs++; |
emilmont | 1:fdd22bb7aa52 | 361 | |
emilmont | 1:fdd22bb7aa52 | 362 | /* Reading the pState values */ |
emilmont | 1:fdd22bb7aa52 | 363 | Xn1 = pState[0]; |
emilmont | 1:fdd22bb7aa52 | 364 | Xn2 = pState[1]; |
emilmont | 1:fdd22bb7aa52 | 365 | Yn1 = pState[2]; |
emilmont | 1:fdd22bb7aa52 | 366 | Yn2 = pState[3]; |
emilmont | 1:fdd22bb7aa52 | 367 | |
emilmont | 1:fdd22bb7aa52 | 368 | /* The variables acc holds the output value that is computed: |
emilmont | 1:fdd22bb7aa52 | 369 | * acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] |
emilmont | 1:fdd22bb7aa52 | 370 | */ |
emilmont | 1:fdd22bb7aa52 | 371 | |
emilmont | 1:fdd22bb7aa52 | 372 | sample = blockSize; |
emilmont | 1:fdd22bb7aa52 | 373 | |
emilmont | 1:fdd22bb7aa52 | 374 | while(sample > 0u) |
emilmont | 1:fdd22bb7aa52 | 375 | { |
emilmont | 1:fdd22bb7aa52 | 376 | /* Read the input */ |
emilmont | 1:fdd22bb7aa52 | 377 | Xn = *pIn++; |
emilmont | 1:fdd22bb7aa52 | 378 | |
emilmont | 1:fdd22bb7aa52 | 379 | /* acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] */ |
emilmont | 1:fdd22bb7aa52 | 380 | acc = (b0 * Xn) + (b1 * Xn1) + (b2 * Xn2) + (a1 * Yn1) + (a2 * Yn2); |
emilmont | 1:fdd22bb7aa52 | 381 | |
emilmont | 1:fdd22bb7aa52 | 382 | /* Store the result in the accumulator in the destination buffer. */ |
emilmont | 1:fdd22bb7aa52 | 383 | *pOut++ = acc; |
emilmont | 1:fdd22bb7aa52 | 384 | |
emilmont | 1:fdd22bb7aa52 | 385 | /* Every time after the output is computed state should be updated. */ |
emilmont | 1:fdd22bb7aa52 | 386 | /* The states should be updated as: */ |
emilmont | 1:fdd22bb7aa52 | 387 | /* Xn2 = Xn1 */ |
emilmont | 1:fdd22bb7aa52 | 388 | /* Xn1 = Xn */ |
emilmont | 1:fdd22bb7aa52 | 389 | /* Yn2 = Yn1 */ |
emilmont | 1:fdd22bb7aa52 | 390 | /* Yn1 = acc */ |
emilmont | 1:fdd22bb7aa52 | 391 | Xn2 = Xn1; |
emilmont | 1:fdd22bb7aa52 | 392 | Xn1 = Xn; |
emilmont | 1:fdd22bb7aa52 | 393 | Yn2 = Yn1; |
emilmont | 1:fdd22bb7aa52 | 394 | Yn1 = acc; |
emilmont | 1:fdd22bb7aa52 | 395 | |
emilmont | 1:fdd22bb7aa52 | 396 | /* decrement the loop counter */ |
emilmont | 1:fdd22bb7aa52 | 397 | sample--; |
emilmont | 1:fdd22bb7aa52 | 398 | } |
emilmont | 1:fdd22bb7aa52 | 399 | |
emilmont | 1:fdd22bb7aa52 | 400 | /* Store the updated state variables back into the pState array */ |
emilmont | 1:fdd22bb7aa52 | 401 | *pState++ = Xn1; |
emilmont | 1:fdd22bb7aa52 | 402 | *pState++ = Xn2; |
emilmont | 1:fdd22bb7aa52 | 403 | *pState++ = Yn1; |
emilmont | 1:fdd22bb7aa52 | 404 | *pState++ = Yn2; |
emilmont | 1:fdd22bb7aa52 | 405 | |
emilmont | 1:fdd22bb7aa52 | 406 | /* The first stage goes from the input buffer to the output buffer. */ |
emilmont | 1:fdd22bb7aa52 | 407 | /* Subsequent numStages occur in-place in the output buffer */ |
emilmont | 1:fdd22bb7aa52 | 408 | pIn = pDst; |
emilmont | 1:fdd22bb7aa52 | 409 | |
emilmont | 1:fdd22bb7aa52 | 410 | /* Reset the output pointer */ |
emilmont | 1:fdd22bb7aa52 | 411 | pOut = pDst; |
emilmont | 1:fdd22bb7aa52 | 412 | |
emilmont | 1:fdd22bb7aa52 | 413 | /* decrement the loop counter */ |
emilmont | 1:fdd22bb7aa52 | 414 | stage--; |
emilmont | 1:fdd22bb7aa52 | 415 | |
emilmont | 1:fdd22bb7aa52 | 416 | } while(stage > 0u); |
emilmont | 1:fdd22bb7aa52 | 417 | |
mbed_official | 3:7a284390b0ce | 418 | #endif /* #ifndef ARM_MATH_CM0_FAMILY */ |
emilmont | 1:fdd22bb7aa52 | 419 | |
emilmont | 1:fdd22bb7aa52 | 420 | } |
emilmont | 1:fdd22bb7aa52 | 421 | |
emilmont | 1:fdd22bb7aa52 | 422 | |
emilmont | 1:fdd22bb7aa52 | 423 | /** |
emilmont | 1:fdd22bb7aa52 | 424 | * @} end of BiquadCascadeDF1 group |
emilmont | 1:fdd22bb7aa52 | 425 | */ |