Du kannst nicht mehr als 25 Themen auswählen Themen müssen entweder mit einem Buchstaben oder einer Ziffer beginnen. Sie können Bindestriche („-“) enthalten und bis zu 35 Zeichen lang sein.

vor 21 Jahren
vor 21 Jahren
vor 21 Jahren
vor 21 Jahren
vor 21 Jahren
vor 21 Jahren
vor 21 Jahren
vor 21 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 21 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 21 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 21 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 4 Jahren
vor 21 Jahren
vor 4 Jahren
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306
  1. /*
  2. * This file is part of Aptdec.
  3. * Copyright (c) 2004-2009 Thierry Leconte (F4DWV), Xerbo (xerbo@protonmail.com) 2019-2020
  4. *
  5. * Aptdec is free software: you can redistribute it and/or modify
  6. * it under the terms of the GNU General Public License as published by
  7. * the Free Software Foundation, either version 2 of the License, or
  8. * (at your option) any later version.
  9. *
  10. * This program is distributed in the hope that it will be useful,
  11. * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12. * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13. * GNU General Public License for more details.
  14. *
  15. * You should have received a copy of the GNU General Public License
  16. * along with this program. If not, see <https://www.gnu.org/licenses/>.
  17. *
  18. */
  19. #include <stdlib.h>
  20. #include <stdio.h>
  21. #include <string.h>
  22. #include <math.h>
  23. #include "filter.h"
  24. #include "filtercoeff.h"
  25. // In case your C compiler is so old that Pi hadn't been invented yet
  26. #ifndef M_PI
  27. #define M_PI 3.141592653589793
  28. #endif
  29. // Block sizes
  30. #define BLKAMP 1024
  31. #define BLKIN 1024
  32. #define Fc 2400.0
  33. #define DFc 50.0
  34. #define PixelLine 2080
  35. #define Fp (2 * PixelLine)
  36. #define RSMULT 15
  37. #define Fi (Fp * RSMULT)
  38. extern int getsample(float *inbuff, int count);
  39. static double Fe;
  40. static double offset = 0.0;
  41. static double FreqLine = 1.0;
  42. static double FreqOsc;
  43. static double K1, K2;
  44. // Check the sample rate and calculate some constants
  45. int init_dsp(double F) {
  46. if(F > Fi) return 1;
  47. if(F < Fp) return -1;
  48. Fe = F;
  49. K1 = DFc / Fe;
  50. K2 = K1 * K1 / 2.0;
  51. // Number of samples per cycle
  52. FreqOsc = Fc / Fe;
  53. return 0;
  54. }
  55. /* Fast phase estimator
  56. * Calculates the phase angle of a signal from a IQ sample
  57. */
  58. static inline double Phase(double I, double Q) {
  59. double angle, r;
  60. int s;
  61. if(I == 0.0 && Q == 0.0) return 0.0;
  62. if (Q < 0) {
  63. s = -1;
  64. Q = -Q;
  65. } else {
  66. s = 1;
  67. }
  68. if (I >= 0) {
  69. r = (I - Q) / (I + Q);
  70. angle = 0.25 - 0.25 * r;
  71. } else {
  72. r = (I + Q) / (Q - I);
  73. angle = 0.75 - 0.25 * r;
  74. }
  75. if(s > 0){
  76. return angle;
  77. }else{
  78. return -angle;
  79. }
  80. }
  81. /* Phase locked loop
  82. * https://arachnoid.com/phase_locked_loop/
  83. * Model of this filter here https://www.desmos.com/calculator/m0uadgkoee
  84. */
  85. static double pll(double I, double Q) {
  86. // PLL coefficient
  87. static double PhaseOsc = 0.0;
  88. double Io, Qo;
  89. double Ip, Qp;
  90. double DPhi;
  91. // Quadrature oscillator / reference
  92. Io = cos(PhaseOsc);
  93. Qo = sin(PhaseOsc);
  94. // Phase detector
  95. Ip = I * Io + Q * Qo;
  96. Qp = Q * Io - I * Qo;
  97. DPhi = Phase(Ip, Qp);
  98. // Loop filter
  99. PhaseOsc += 2.0 * M_PI * (K1 * DPhi + FreqOsc);
  100. if (PhaseOsc > M_PI)
  101. PhaseOsc -= 2.0 * M_PI;
  102. if (PhaseOsc <= -M_PI)
  103. PhaseOsc += 2.0 * M_PI;
  104. FreqOsc += K2 * DPhi;
  105. if (FreqOsc > ((Fc + DFc) / Fe))
  106. FreqOsc = (Fc + DFc) / Fe;
  107. if (FreqOsc < ((Fc - DFc) / Fe))
  108. FreqOsc = (Fc - DFc) / Fe;
  109. return Ip;
  110. }
  111. // Convert samples into pixels
  112. static int getamp(double *ampbuff, int count) {
  113. static float inbuff[BLKIN];
  114. static int idxin = 0;
  115. static int nin = 0;
  116. for (int n = 0; n < count; n++) {
  117. double I, Q;
  118. // Get some more samples when needed
  119. if (nin < IQFilterLen * 2 + 2) {
  120. // Number of samples read
  121. int res;
  122. memmove(inbuff, &(inbuff[idxin]), nin * sizeof(float));
  123. idxin = 0;
  124. // Read some samples
  125. res = getsample(&(inbuff[nin]), BLKIN - nin);
  126. nin += res;
  127. // Make sure there is enough samples to continue
  128. if (nin < IQFilterLen * 2 + 2)
  129. return n;
  130. }
  131. // Process read samples into a brightness value
  132. iqfir(&inbuff[idxin], iqfilter, IQFilterLen, &I, &Q);
  133. ampbuff[n] = pll(I, Q);
  134. // Increment current sample
  135. idxin++;
  136. nin--;
  137. }
  138. return count;
  139. }
  140. // Sub-pixel offsetting + FIR compensation
  141. int getpixelv(float *pvbuff, int count) {
  142. // Amplitude buffer
  143. static double ampbuff[BLKAMP];
  144. static int nam = 0;
  145. static int idxam = 0;
  146. double mult;
  147. // Gaussian resampling factor
  148. mult = (double) Fi / Fe * FreqLine;
  149. int m = RSFilterLen / mult + 1;
  150. for (int n = 0; n < count; n++) {
  151. int shift;
  152. if (nam < m) {
  153. int res;
  154. memmove(ampbuff, &(ampbuff[idxam]), nam * sizeof(double));
  155. idxam = 0;
  156. res = getamp(&(ampbuff[nam]), BLKAMP - nam);
  157. nam += res;
  158. if (nam < m)
  159. return n;
  160. }
  161. // Gaussian FIR compensation filter
  162. pvbuff[n] = rsfir(&(ampbuff[idxam]), rsfilter, RSFilterLen, offset, mult) * mult * 256.0;
  163. shift = ((int) floor((RSMULT - offset) / mult)) + 1;
  164. offset = shift * mult + offset - RSMULT;
  165. idxam += shift;
  166. nam -= shift;
  167. }
  168. return count;
  169. }
  170. // Get an entire row of pixels, aligned with sync markers
  171. // FIXME: loses sync easily
  172. int getpixelrow(float *pixelv, int nrow, int *zenith, int reset) {
  173. static float pixels[PixelLine + SyncFilterLen];
  174. static int npv;
  175. static int synced = 0;
  176. static double max = 0.0;
  177. static double minDoppler = 100;
  178. if(reset) synced = 0;
  179. double corr, ecorr, lcorr;
  180. int res;
  181. // Move the row buffer into the the image buffer
  182. if (npv > 0)
  183. memmove(pixelv, pixels, npv * sizeof(float));
  184. // Get the sync line
  185. if (npv < SyncFilterLen + 2) {
  186. res = getpixelv(&(pixelv[npv]), SyncFilterLen + 2 - npv);
  187. npv += res;
  188. if (npv < SyncFilterLen + 2)
  189. return 0;
  190. }
  191. // Calculate the frequency offset
  192. ecorr = fir(pixelv, Sync, SyncFilterLen);
  193. corr = fir(&pixelv[1], Sync, SyncFilterLen - 1);
  194. lcorr = fir(&pixelv[2], Sync, SyncFilterLen - 2);
  195. FreqLine = 1.0+((ecorr-lcorr) / corr / PixelLine / 4.0);
  196. // Find the point in which ecorr and lcorr intercept
  197. if(fabs(lcorr - ecorr) < minDoppler && fabs(lcorr - ecorr) > 2){
  198. minDoppler = fabs(lcorr - ecorr);
  199. *zenith = nrow;
  200. }
  201. // The point in which the pixel offset is recalculated
  202. if (corr < 0.75 * max) {
  203. synced = 0;
  204. FreqLine = 1.0;
  205. }
  206. max = corr;
  207. if (synced < 8) {
  208. int mshift;
  209. if (npv < PixelLine + SyncFilterLen) {
  210. res = getpixelv(&(pixelv[npv]), PixelLine + SyncFilterLen - npv);
  211. npv += res;
  212. if (npv < PixelLine + SyncFilterLen)
  213. return 0;
  214. }
  215. // Test every possible position until we get the best result
  216. mshift = 0;
  217. for (int shift = 0; shift < PixelLine; shift++) {
  218. double corr;
  219. corr = fir(&(pixelv[shift + 1]), Sync, SyncFilterLen);
  220. if (corr > max) {
  221. mshift = shift;
  222. max = corr;
  223. }
  224. }
  225. // If we are already as aligned as we can get, just continue
  226. if (mshift == 0) {
  227. synced++;
  228. } else {
  229. memmove(pixelv, &(pixelv[mshift]), (npv - mshift) * sizeof(float));
  230. npv -= mshift;
  231. synced = 0;
  232. FreqLine = 1.0;
  233. }
  234. }
  235. // Get the rest of this row
  236. if (npv < PixelLine) {
  237. res = getpixelv(&(pixelv[npv]), PixelLine - npv);
  238. npv += res;
  239. if (npv < PixelLine)
  240. return 0;
  241. }
  242. // Move the sync lines into the output buffer with the calculated offset
  243. if (npv == PixelLine) {
  244. npv = 0;
  245. } else {
  246. memmove(pixels, &(pixelv[PixelLine]), (npv - PixelLine) * sizeof(float));
  247. npv -= PixelLine;
  248. }
  249. return 1;
  250. }