V4.0.1 of the ARM CMSIS DSP libraries. Note that arm_bitreversal2.s, arm_cfft_f32.c and arm_rfft_fast_f32.c had to be removed. arm_bitreversal2.s will not assemble with the online tools. So, the fast f32 FFT functions are not yet available. All the other FFT functions are available.

Dependents:   MPU9150_Example fir_f32 fir_f32 MPU9150_nucleo_noni2cdev ... more

Committer:
emh203
Date:
Mon Jul 28 15:03:15 2014 +0000
Revision:
0:3d9c67d97d6f
1st working commit.   Had to remove arm_bitreversal2.s     arm_cfft_f32.c and arm_rfft_fast_f32.c.    The .s will not assemble.      For now I removed these functions so we could at least have a library for the other functions.

Who changed what in which revision?

UserRevisionLine numberNew contents of line
emh203 0:3d9c67d97d6f 1 /* ----------------------------------------------------------------------
emh203 0:3d9c67d97d6f 2 * Copyright (C) 2010-2014 ARM Limited. All rights reserved.
emh203 0:3d9c67d97d6f 3 *
emh203 0:3d9c67d97d6f 4 * $Date: 12. March 2014
emh203 0:3d9c67d97d6f 5 * $Revision: V1.4.3
emh203 0:3d9c67d97d6f 6 *
emh203 0:3d9c67d97d6f 7 * Project: CMSIS DSP Library
emh203 0:3d9c67d97d6f 8 * Title: arm_dct4_q31.c
emh203 0:3d9c67d97d6f 9 *
emh203 0:3d9c67d97d6f 10 * Description: Processing function of DCT4 & IDCT4 Q31.
emh203 0:3d9c67d97d6f 11 *
emh203 0:3d9c67d97d6f 12 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
emh203 0:3d9c67d97d6f 13 *
emh203 0:3d9c67d97d6f 14 * Redistribution and use in source and binary forms, with or without
emh203 0:3d9c67d97d6f 15 * modification, are permitted provided that the following conditions
emh203 0:3d9c67d97d6f 16 * are met:
emh203 0:3d9c67d97d6f 17 * - Redistributions of source code must retain the above copyright
emh203 0:3d9c67d97d6f 18 * notice, this list of conditions and the following disclaimer.
emh203 0:3d9c67d97d6f 19 * - Redistributions in binary form must reproduce the above copyright
emh203 0:3d9c67d97d6f 20 * notice, this list of conditions and the following disclaimer in
emh203 0:3d9c67d97d6f 21 * the documentation and/or other materials provided with the
emh203 0:3d9c67d97d6f 22 * distribution.
emh203 0:3d9c67d97d6f 23 * - Neither the name of ARM LIMITED nor the names of its contributors
emh203 0:3d9c67d97d6f 24 * may be used to endorse or promote products derived from this
emh203 0:3d9c67d97d6f 25 * software without specific prior written permission.
emh203 0:3d9c67d97d6f 26 *
emh203 0:3d9c67d97d6f 27 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
emh203 0:3d9c67d97d6f 28 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
emh203 0:3d9c67d97d6f 29 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
emh203 0:3d9c67d97d6f 30 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
emh203 0:3d9c67d97d6f 31 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
emh203 0:3d9c67d97d6f 32 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
emh203 0:3d9c67d97d6f 33 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
emh203 0:3d9c67d97d6f 34 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
emh203 0:3d9c67d97d6f 35 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
emh203 0:3d9c67d97d6f 36 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
emh203 0:3d9c67d97d6f 37 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
emh203 0:3d9c67d97d6f 38 * POSSIBILITY OF SUCH DAMAGE.
emh203 0:3d9c67d97d6f 39 * -------------------------------------------------------------------- */
emh203 0:3d9c67d97d6f 40
emh203 0:3d9c67d97d6f 41 #include "arm_math.h"
emh203 0:3d9c67d97d6f 42
emh203 0:3d9c67d97d6f 43 /**
emh203 0:3d9c67d97d6f 44 * @addtogroup DCT4_IDCT4
emh203 0:3d9c67d97d6f 45 * @{
emh203 0:3d9c67d97d6f 46 */
emh203 0:3d9c67d97d6f 47
emh203 0:3d9c67d97d6f 48 /**
emh203 0:3d9c67d97d6f 49 * @brief Processing function for the Q31 DCT4/IDCT4.
emh203 0:3d9c67d97d6f 50 * @param[in] *S points to an instance of the Q31 DCT4 structure.
emh203 0:3d9c67d97d6f 51 * @param[in] *pState points to state buffer.
emh203 0:3d9c67d97d6f 52 * @param[in,out] *pInlineBuffer points to the in-place input and output buffer.
emh203 0:3d9c67d97d6f 53 * @return none.
emh203 0:3d9c67d97d6f 54 * \par Input an output formats:
emh203 0:3d9c67d97d6f 55 * Input samples need to be downscaled by 1 bit to avoid saturations in the Q31 DCT process,
emh203 0:3d9c67d97d6f 56 * as the conversion from DCT2 to DCT4 involves one subtraction.
emh203 0:3d9c67d97d6f 57 * Internally inputs are downscaled in the RFFT process function to avoid overflows.
emh203 0:3d9c67d97d6f 58 * Number of bits downscaled, depends on the size of the transform.
emh203 0:3d9c67d97d6f 59 * The input and output formats for different DCT sizes and number of bits to upscale are mentioned in the table below:
emh203 0:3d9c67d97d6f 60 *
emh203 0:3d9c67d97d6f 61 * \image html dct4FormatsQ31Table.gif
emh203 0:3d9c67d97d6f 62 */
emh203 0:3d9c67d97d6f 63
emh203 0:3d9c67d97d6f 64 void arm_dct4_q31(
emh203 0:3d9c67d97d6f 65 const arm_dct4_instance_q31 * S,
emh203 0:3d9c67d97d6f 66 q31_t * pState,
emh203 0:3d9c67d97d6f 67 q31_t * pInlineBuffer)
emh203 0:3d9c67d97d6f 68 {
emh203 0:3d9c67d97d6f 69 uint16_t i; /* Loop counter */
emh203 0:3d9c67d97d6f 70 q31_t *weights = S->pTwiddle; /* Pointer to the Weights table */
emh203 0:3d9c67d97d6f 71 q31_t *cosFact = S->pCosFactor; /* Pointer to the cos factors table */
emh203 0:3d9c67d97d6f 72 q31_t *pS1, *pS2, *pbuff; /* Temporary pointers for input buffer and pState buffer */
emh203 0:3d9c67d97d6f 73 q31_t in; /* Temporary variable */
emh203 0:3d9c67d97d6f 74
emh203 0:3d9c67d97d6f 75
emh203 0:3d9c67d97d6f 76 /* DCT4 computation involves DCT2 (which is calculated using RFFT)
emh203 0:3d9c67d97d6f 77 * along with some pre-processing and post-processing.
emh203 0:3d9c67d97d6f 78 * Computational procedure is explained as follows:
emh203 0:3d9c67d97d6f 79 * (a) Pre-processing involves multiplying input with cos factor,
emh203 0:3d9c67d97d6f 80 * r(n) = 2 * u(n) * cos(pi*(2*n+1)/(4*n))
emh203 0:3d9c67d97d6f 81 * where,
emh203 0:3d9c67d97d6f 82 * r(n) -- output of preprocessing
emh203 0:3d9c67d97d6f 83 * u(n) -- input to preprocessing(actual Source buffer)
emh203 0:3d9c67d97d6f 84 * (b) Calculation of DCT2 using FFT is divided into three steps:
emh203 0:3d9c67d97d6f 85 * Step1: Re-ordering of even and odd elements of input.
emh203 0:3d9c67d97d6f 86 * Step2: Calculating FFT of the re-ordered input.
emh203 0:3d9c67d97d6f 87 * Step3: Taking the real part of the product of FFT output and weights.
emh203 0:3d9c67d97d6f 88 * (c) Post-processing - DCT4 can be obtained from DCT2 output using the following equation:
emh203 0:3d9c67d97d6f 89 * Y4(k) = Y2(k) - Y4(k-1) and Y4(-1) = Y4(0)
emh203 0:3d9c67d97d6f 90 * where,
emh203 0:3d9c67d97d6f 91 * Y4 -- DCT4 output, Y2 -- DCT2 output
emh203 0:3d9c67d97d6f 92 * (d) Multiplying the output with the normalizing factor sqrt(2/N).
emh203 0:3d9c67d97d6f 93 */
emh203 0:3d9c67d97d6f 94
emh203 0:3d9c67d97d6f 95 /*-------- Pre-processing ------------*/
emh203 0:3d9c67d97d6f 96 /* Multiplying input with cos factor i.e. r(n) = 2 * x(n) * cos(pi*(2*n+1)/(4*n)) */
emh203 0:3d9c67d97d6f 97 arm_mult_q31(pInlineBuffer, cosFact, pInlineBuffer, S->N);
emh203 0:3d9c67d97d6f 98 arm_shift_q31(pInlineBuffer, 1, pInlineBuffer, S->N);
emh203 0:3d9c67d97d6f 99
emh203 0:3d9c67d97d6f 100 /* ----------------------------------------------------------------
emh203 0:3d9c67d97d6f 101 * Step1: Re-ordering of even and odd elements as
emh203 0:3d9c67d97d6f 102 * pState[i] = pInlineBuffer[2*i] and
emh203 0:3d9c67d97d6f 103 * pState[N-i-1] = pInlineBuffer[2*i+1] where i = 0 to N/2
emh203 0:3d9c67d97d6f 104 ---------------------------------------------------------------------*/
emh203 0:3d9c67d97d6f 105
emh203 0:3d9c67d97d6f 106 /* pS1 initialized to pState */
emh203 0:3d9c67d97d6f 107 pS1 = pState;
emh203 0:3d9c67d97d6f 108
emh203 0:3d9c67d97d6f 109 /* pS2 initialized to pState+N-1, so that it points to the end of the state buffer */
emh203 0:3d9c67d97d6f 110 pS2 = pState + (S->N - 1u);
emh203 0:3d9c67d97d6f 111
emh203 0:3d9c67d97d6f 112 /* pbuff initialized to input buffer */
emh203 0:3d9c67d97d6f 113 pbuff = pInlineBuffer;
emh203 0:3d9c67d97d6f 114
emh203 0:3d9c67d97d6f 115 #ifndef ARM_MATH_CM0_FAMILY
emh203 0:3d9c67d97d6f 116
emh203 0:3d9c67d97d6f 117 /* Run the below code for Cortex-M4 and Cortex-M3 */
emh203 0:3d9c67d97d6f 118
emh203 0:3d9c67d97d6f 119 /* Initializing the loop counter to N/2 >> 2 for loop unrolling by 4 */
emh203 0:3d9c67d97d6f 120 i = S->Nby2 >> 2u;
emh203 0:3d9c67d97d6f 121
emh203 0:3d9c67d97d6f 122 /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
emh203 0:3d9c67d97d6f 123 ** a second loop below computes the remaining 1 to 3 samples. */
emh203 0:3d9c67d97d6f 124 do
emh203 0:3d9c67d97d6f 125 {
emh203 0:3d9c67d97d6f 126 /* Re-ordering of even and odd elements */
emh203 0:3d9c67d97d6f 127 /* pState[i] = pInlineBuffer[2*i] */
emh203 0:3d9c67d97d6f 128 *pS1++ = *pbuff++;
emh203 0:3d9c67d97d6f 129 /* pState[N-i-1] = pInlineBuffer[2*i+1] */
emh203 0:3d9c67d97d6f 130 *pS2-- = *pbuff++;
emh203 0:3d9c67d97d6f 131
emh203 0:3d9c67d97d6f 132 *pS1++ = *pbuff++;
emh203 0:3d9c67d97d6f 133 *pS2-- = *pbuff++;
emh203 0:3d9c67d97d6f 134
emh203 0:3d9c67d97d6f 135 *pS1++ = *pbuff++;
emh203 0:3d9c67d97d6f 136 *pS2-- = *pbuff++;
emh203 0:3d9c67d97d6f 137
emh203 0:3d9c67d97d6f 138 *pS1++ = *pbuff++;
emh203 0:3d9c67d97d6f 139 *pS2-- = *pbuff++;
emh203 0:3d9c67d97d6f 140
emh203 0:3d9c67d97d6f 141 /* Decrement the loop counter */
emh203 0:3d9c67d97d6f 142 i--;
emh203 0:3d9c67d97d6f 143 } while(i > 0u);
emh203 0:3d9c67d97d6f 144
emh203 0:3d9c67d97d6f 145 /* pbuff initialized to input buffer */
emh203 0:3d9c67d97d6f 146 pbuff = pInlineBuffer;
emh203 0:3d9c67d97d6f 147
emh203 0:3d9c67d97d6f 148 /* pS1 initialized to pState */
emh203 0:3d9c67d97d6f 149 pS1 = pState;
emh203 0:3d9c67d97d6f 150
emh203 0:3d9c67d97d6f 151 /* Initializing the loop counter to N/4 instead of N for loop unrolling */
emh203 0:3d9c67d97d6f 152 i = S->N >> 2u;
emh203 0:3d9c67d97d6f 153
emh203 0:3d9c67d97d6f 154 /* Processing with loop unrolling 4 times as N is always multiple of 4.
emh203 0:3d9c67d97d6f 155 * Compute 4 outputs at a time */
emh203 0:3d9c67d97d6f 156 do
emh203 0:3d9c67d97d6f 157 {
emh203 0:3d9c67d97d6f 158 /* Writing the re-ordered output back to inplace input buffer */
emh203 0:3d9c67d97d6f 159 *pbuff++ = *pS1++;
emh203 0:3d9c67d97d6f 160 *pbuff++ = *pS1++;
emh203 0:3d9c67d97d6f 161 *pbuff++ = *pS1++;
emh203 0:3d9c67d97d6f 162 *pbuff++ = *pS1++;
emh203 0:3d9c67d97d6f 163
emh203 0:3d9c67d97d6f 164 /* Decrement the loop counter */
emh203 0:3d9c67d97d6f 165 i--;
emh203 0:3d9c67d97d6f 166 } while(i > 0u);
emh203 0:3d9c67d97d6f 167
emh203 0:3d9c67d97d6f 168
emh203 0:3d9c67d97d6f 169 /* ---------------------------------------------------------
emh203 0:3d9c67d97d6f 170 * Step2: Calculate RFFT for N-point input
emh203 0:3d9c67d97d6f 171 * ---------------------------------------------------------- */
emh203 0:3d9c67d97d6f 172 /* pInlineBuffer is real input of length N , pState is the complex output of length 2N */
emh203 0:3d9c67d97d6f 173 arm_rfft_q31(S->pRfft, pInlineBuffer, pState);
emh203 0:3d9c67d97d6f 174
emh203 0:3d9c67d97d6f 175 /*----------------------------------------------------------------------
emh203 0:3d9c67d97d6f 176 * Step3: Multiply the FFT output with the weights.
emh203 0:3d9c67d97d6f 177 *----------------------------------------------------------------------*/
emh203 0:3d9c67d97d6f 178 arm_cmplx_mult_cmplx_q31(pState, weights, pState, S->N);
emh203 0:3d9c67d97d6f 179
emh203 0:3d9c67d97d6f 180 /* The output of complex multiplication is in 3.29 format.
emh203 0:3d9c67d97d6f 181 * Hence changing the format of N (i.e. 2*N elements) complex numbers to 1.31 format by shifting left by 2 bits. */
emh203 0:3d9c67d97d6f 182 arm_shift_q31(pState, 2, pState, S->N * 2);
emh203 0:3d9c67d97d6f 183
emh203 0:3d9c67d97d6f 184 /* ----------- Post-processing ---------- */
emh203 0:3d9c67d97d6f 185 /* DCT-IV can be obtained from DCT-II by the equation,
emh203 0:3d9c67d97d6f 186 * Y4(k) = Y2(k) - Y4(k-1) and Y4(-1) = Y4(0)
emh203 0:3d9c67d97d6f 187 * Hence, Y4(0) = Y2(0)/2 */
emh203 0:3d9c67d97d6f 188 /* Getting only real part from the output and Converting to DCT-IV */
emh203 0:3d9c67d97d6f 189
emh203 0:3d9c67d97d6f 190 /* Initializing the loop counter to N >> 2 for loop unrolling by 4 */
emh203 0:3d9c67d97d6f 191 i = (S->N - 1u) >> 2u;
emh203 0:3d9c67d97d6f 192
emh203 0:3d9c67d97d6f 193 /* pbuff initialized to input buffer. */
emh203 0:3d9c67d97d6f 194 pbuff = pInlineBuffer;
emh203 0:3d9c67d97d6f 195
emh203 0:3d9c67d97d6f 196 /* pS1 initialized to pState */
emh203 0:3d9c67d97d6f 197 pS1 = pState;
emh203 0:3d9c67d97d6f 198
emh203 0:3d9c67d97d6f 199 /* Calculating Y4(0) from Y2(0) using Y4(0) = Y2(0)/2 */
emh203 0:3d9c67d97d6f 200 in = *pS1++ >> 1u;
emh203 0:3d9c67d97d6f 201 /* input buffer acts as inplace, so output values are stored in the input itself. */
emh203 0:3d9c67d97d6f 202 *pbuff++ = in;
emh203 0:3d9c67d97d6f 203
emh203 0:3d9c67d97d6f 204 /* pState pointer is incremented twice as the real values are located alternatively in the array */
emh203 0:3d9c67d97d6f 205 pS1++;
emh203 0:3d9c67d97d6f 206
emh203 0:3d9c67d97d6f 207 /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
emh203 0:3d9c67d97d6f 208 ** a second loop below computes the remaining 1 to 3 samples. */
emh203 0:3d9c67d97d6f 209 do
emh203 0:3d9c67d97d6f 210 {
emh203 0:3d9c67d97d6f 211 /* Calculating Y4(1) to Y4(N-1) from Y2 using equation Y4(k) = Y2(k) - Y4(k-1) */
emh203 0:3d9c67d97d6f 212 /* pState pointer (pS1) is incremented twice as the real values are located alternatively in the array */
emh203 0:3d9c67d97d6f 213 in = *pS1++ - in;
emh203 0:3d9c67d97d6f 214 *pbuff++ = in;
emh203 0:3d9c67d97d6f 215 /* points to the next real value */
emh203 0:3d9c67d97d6f 216 pS1++;
emh203 0:3d9c67d97d6f 217
emh203 0:3d9c67d97d6f 218 in = *pS1++ - in;
emh203 0:3d9c67d97d6f 219 *pbuff++ = in;
emh203 0:3d9c67d97d6f 220 pS1++;
emh203 0:3d9c67d97d6f 221
emh203 0:3d9c67d97d6f 222 in = *pS1++ - in;
emh203 0:3d9c67d97d6f 223 *pbuff++ = in;
emh203 0:3d9c67d97d6f 224 pS1++;
emh203 0:3d9c67d97d6f 225
emh203 0:3d9c67d97d6f 226 in = *pS1++ - in;
emh203 0:3d9c67d97d6f 227 *pbuff++ = in;
emh203 0:3d9c67d97d6f 228 pS1++;
emh203 0:3d9c67d97d6f 229
emh203 0:3d9c67d97d6f 230 /* Decrement the loop counter */
emh203 0:3d9c67d97d6f 231 i--;
emh203 0:3d9c67d97d6f 232 } while(i > 0u);
emh203 0:3d9c67d97d6f 233
emh203 0:3d9c67d97d6f 234 /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
emh203 0:3d9c67d97d6f 235 ** No loop unrolling is used. */
emh203 0:3d9c67d97d6f 236 i = (S->N - 1u) % 0x4u;
emh203 0:3d9c67d97d6f 237
emh203 0:3d9c67d97d6f 238 while(i > 0u)
emh203 0:3d9c67d97d6f 239 {
emh203 0:3d9c67d97d6f 240 /* Calculating Y4(1) to Y4(N-1) from Y2 using equation Y4(k) = Y2(k) - Y4(k-1) */
emh203 0:3d9c67d97d6f 241 /* pState pointer (pS1) is incremented twice as the real values are located alternatively in the array */
emh203 0:3d9c67d97d6f 242 in = *pS1++ - in;
emh203 0:3d9c67d97d6f 243 *pbuff++ = in;
emh203 0:3d9c67d97d6f 244 /* points to the next real value */
emh203 0:3d9c67d97d6f 245 pS1++;
emh203 0:3d9c67d97d6f 246
emh203 0:3d9c67d97d6f 247 /* Decrement the loop counter */
emh203 0:3d9c67d97d6f 248 i--;
emh203 0:3d9c67d97d6f 249 }
emh203 0:3d9c67d97d6f 250
emh203 0:3d9c67d97d6f 251
emh203 0:3d9c67d97d6f 252 /*------------ Normalizing the output by multiplying with the normalizing factor ----------*/
emh203 0:3d9c67d97d6f 253
emh203 0:3d9c67d97d6f 254 /* Initializing the loop counter to N/4 instead of N for loop unrolling */
emh203 0:3d9c67d97d6f 255 i = S->N >> 2u;
emh203 0:3d9c67d97d6f 256
emh203 0:3d9c67d97d6f 257 /* pbuff initialized to the pInlineBuffer(now contains the output values) */
emh203 0:3d9c67d97d6f 258 pbuff = pInlineBuffer;
emh203 0:3d9c67d97d6f 259
emh203 0:3d9c67d97d6f 260 /* Processing with loop unrolling 4 times as N is always multiple of 4. Compute 4 outputs at a time */
emh203 0:3d9c67d97d6f 261 do
emh203 0:3d9c67d97d6f 262 {
emh203 0:3d9c67d97d6f 263 /* Multiplying pInlineBuffer with the normalizing factor sqrt(2/N) */
emh203 0:3d9c67d97d6f 264 in = *pbuff;
emh203 0:3d9c67d97d6f 265 *pbuff++ = ((q31_t) (((q63_t) in * S->normalize) >> 31));
emh203 0:3d9c67d97d6f 266
emh203 0:3d9c67d97d6f 267 in = *pbuff;
emh203 0:3d9c67d97d6f 268 *pbuff++ = ((q31_t) (((q63_t) in * S->normalize) >> 31));
emh203 0:3d9c67d97d6f 269
emh203 0:3d9c67d97d6f 270 in = *pbuff;
emh203 0:3d9c67d97d6f 271 *pbuff++ = ((q31_t) (((q63_t) in * S->normalize) >> 31));
emh203 0:3d9c67d97d6f 272
emh203 0:3d9c67d97d6f 273 in = *pbuff;
emh203 0:3d9c67d97d6f 274 *pbuff++ = ((q31_t) (((q63_t) in * S->normalize) >> 31));
emh203 0:3d9c67d97d6f 275
emh203 0:3d9c67d97d6f 276 /* Decrement the loop counter */
emh203 0:3d9c67d97d6f 277 i--;
emh203 0:3d9c67d97d6f 278 } while(i > 0u);
emh203 0:3d9c67d97d6f 279
emh203 0:3d9c67d97d6f 280
emh203 0:3d9c67d97d6f 281 #else
emh203 0:3d9c67d97d6f 282
emh203 0:3d9c67d97d6f 283 /* Run the below code for Cortex-M0 */
emh203 0:3d9c67d97d6f 284
emh203 0:3d9c67d97d6f 285 /* Initializing the loop counter to N/2 */
emh203 0:3d9c67d97d6f 286 i = S->Nby2;
emh203 0:3d9c67d97d6f 287
emh203 0:3d9c67d97d6f 288 do
emh203 0:3d9c67d97d6f 289 {
emh203 0:3d9c67d97d6f 290 /* Re-ordering of even and odd elements */
emh203 0:3d9c67d97d6f 291 /* pState[i] = pInlineBuffer[2*i] */
emh203 0:3d9c67d97d6f 292 *pS1++ = *pbuff++;
emh203 0:3d9c67d97d6f 293 /* pState[N-i-1] = pInlineBuffer[2*i+1] */
emh203 0:3d9c67d97d6f 294 *pS2-- = *pbuff++;
emh203 0:3d9c67d97d6f 295
emh203 0:3d9c67d97d6f 296 /* Decrement the loop counter */
emh203 0:3d9c67d97d6f 297 i--;
emh203 0:3d9c67d97d6f 298 } while(i > 0u);
emh203 0:3d9c67d97d6f 299
emh203 0:3d9c67d97d6f 300 /* pbuff initialized to input buffer */
emh203 0:3d9c67d97d6f 301 pbuff = pInlineBuffer;
emh203 0:3d9c67d97d6f 302
emh203 0:3d9c67d97d6f 303 /* pS1 initialized to pState */
emh203 0:3d9c67d97d6f 304 pS1 = pState;
emh203 0:3d9c67d97d6f 305
emh203 0:3d9c67d97d6f 306 /* Initializing the loop counter */
emh203 0:3d9c67d97d6f 307 i = S->N;
emh203 0:3d9c67d97d6f 308
emh203 0:3d9c67d97d6f 309 do
emh203 0:3d9c67d97d6f 310 {
emh203 0:3d9c67d97d6f 311 /* Writing the re-ordered output back to inplace input buffer */
emh203 0:3d9c67d97d6f 312 *pbuff++ = *pS1++;
emh203 0:3d9c67d97d6f 313
emh203 0:3d9c67d97d6f 314 /* Decrement the loop counter */
emh203 0:3d9c67d97d6f 315 i--;
emh203 0:3d9c67d97d6f 316 } while(i > 0u);
emh203 0:3d9c67d97d6f 317
emh203 0:3d9c67d97d6f 318
emh203 0:3d9c67d97d6f 319 /* ---------------------------------------------------------
emh203 0:3d9c67d97d6f 320 * Step2: Calculate RFFT for N-point input
emh203 0:3d9c67d97d6f 321 * ---------------------------------------------------------- */
emh203 0:3d9c67d97d6f 322 /* pInlineBuffer is real input of length N , pState is the complex output of length 2N */
emh203 0:3d9c67d97d6f 323 arm_rfft_q31(S->pRfft, pInlineBuffer, pState);
emh203 0:3d9c67d97d6f 324
emh203 0:3d9c67d97d6f 325 /*----------------------------------------------------------------------
emh203 0:3d9c67d97d6f 326 * Step3: Multiply the FFT output with the weights.
emh203 0:3d9c67d97d6f 327 *----------------------------------------------------------------------*/
emh203 0:3d9c67d97d6f 328 arm_cmplx_mult_cmplx_q31(pState, weights, pState, S->N);
emh203 0:3d9c67d97d6f 329
emh203 0:3d9c67d97d6f 330 /* The output of complex multiplication is in 3.29 format.
emh203 0:3d9c67d97d6f 331 * Hence changing the format of N (i.e. 2*N elements) complex numbers to 1.31 format by shifting left by 2 bits. */
emh203 0:3d9c67d97d6f 332 arm_shift_q31(pState, 2, pState, S->N * 2);
emh203 0:3d9c67d97d6f 333
emh203 0:3d9c67d97d6f 334 /* ----------- Post-processing ---------- */
emh203 0:3d9c67d97d6f 335 /* DCT-IV can be obtained from DCT-II by the equation,
emh203 0:3d9c67d97d6f 336 * Y4(k) = Y2(k) - Y4(k-1) and Y4(-1) = Y4(0)
emh203 0:3d9c67d97d6f 337 * Hence, Y4(0) = Y2(0)/2 */
emh203 0:3d9c67d97d6f 338 /* Getting only real part from the output and Converting to DCT-IV */
emh203 0:3d9c67d97d6f 339
emh203 0:3d9c67d97d6f 340 /* pbuff initialized to input buffer. */
emh203 0:3d9c67d97d6f 341 pbuff = pInlineBuffer;
emh203 0:3d9c67d97d6f 342
emh203 0:3d9c67d97d6f 343 /* pS1 initialized to pState */
emh203 0:3d9c67d97d6f 344 pS1 = pState;
emh203 0:3d9c67d97d6f 345
emh203 0:3d9c67d97d6f 346 /* Calculating Y4(0) from Y2(0) using Y4(0) = Y2(0)/2 */
emh203 0:3d9c67d97d6f 347 in = *pS1++ >> 1u;
emh203 0:3d9c67d97d6f 348 /* input buffer acts as inplace, so output values are stored in the input itself. */
emh203 0:3d9c67d97d6f 349 *pbuff++ = in;
emh203 0:3d9c67d97d6f 350
emh203 0:3d9c67d97d6f 351 /* pState pointer is incremented twice as the real values are located alternatively in the array */
emh203 0:3d9c67d97d6f 352 pS1++;
emh203 0:3d9c67d97d6f 353
emh203 0:3d9c67d97d6f 354 /* Initializing the loop counter */
emh203 0:3d9c67d97d6f 355 i = (S->N - 1u);
emh203 0:3d9c67d97d6f 356
emh203 0:3d9c67d97d6f 357 while(i > 0u)
emh203 0:3d9c67d97d6f 358 {
emh203 0:3d9c67d97d6f 359 /* Calculating Y4(1) to Y4(N-1) from Y2 using equation Y4(k) = Y2(k) - Y4(k-1) */
emh203 0:3d9c67d97d6f 360 /* pState pointer (pS1) is incremented twice as the real values are located alternatively in the array */
emh203 0:3d9c67d97d6f 361 in = *pS1++ - in;
emh203 0:3d9c67d97d6f 362 *pbuff++ = in;
emh203 0:3d9c67d97d6f 363 /* points to the next real value */
emh203 0:3d9c67d97d6f 364 pS1++;
emh203 0:3d9c67d97d6f 365
emh203 0:3d9c67d97d6f 366 /* Decrement the loop counter */
emh203 0:3d9c67d97d6f 367 i--;
emh203 0:3d9c67d97d6f 368 }
emh203 0:3d9c67d97d6f 369
emh203 0:3d9c67d97d6f 370
emh203 0:3d9c67d97d6f 371 /*------------ Normalizing the output by multiplying with the normalizing factor ----------*/
emh203 0:3d9c67d97d6f 372
emh203 0:3d9c67d97d6f 373 /* Initializing the loop counter */
emh203 0:3d9c67d97d6f 374 i = S->N;
emh203 0:3d9c67d97d6f 375
emh203 0:3d9c67d97d6f 376 /* pbuff initialized to the pInlineBuffer(now contains the output values) */
emh203 0:3d9c67d97d6f 377 pbuff = pInlineBuffer;
emh203 0:3d9c67d97d6f 378
emh203 0:3d9c67d97d6f 379 do
emh203 0:3d9c67d97d6f 380 {
emh203 0:3d9c67d97d6f 381 /* Multiplying pInlineBuffer with the normalizing factor sqrt(2/N) */
emh203 0:3d9c67d97d6f 382 in = *pbuff;
emh203 0:3d9c67d97d6f 383 *pbuff++ = ((q31_t) (((q63_t) in * S->normalize) >> 31));
emh203 0:3d9c67d97d6f 384
emh203 0:3d9c67d97d6f 385 /* Decrement the loop counter */
emh203 0:3d9c67d97d6f 386 i--;
emh203 0:3d9c67d97d6f 387 } while(i > 0u);
emh203 0:3d9c67d97d6f 388
emh203 0:3d9c67d97d6f 389 #endif /* #ifndef ARM_MATH_CM0_FAMILY */
emh203 0:3d9c67d97d6f 390
emh203 0:3d9c67d97d6f 391 }
emh203 0:3d9c67d97d6f 392
emh203 0:3d9c67d97d6f 393 /**
emh203 0:3d9c67d97d6f 394 * @} end of DCT4_IDCT4 group
emh203 0:3d9c67d97d6f 395 */