FFmpeg
 All Data Structures Namespaces Files Functions Variables Typedefs Enumerations Enumerator Macros Groups Pages
tiny_psnr.c
Go to the documentation of this file.
1 /*
2  * Copyright (c) 2003 Michael Niedermayer <michaelni@gmx.at>
3  *
4  * This file is part of FFmpeg.
5  *
6  * FFmpeg is free software; you can redistribute it and/or
7  * modify it under the terms of the GNU Lesser General Public
8  * License as published by the Free Software Foundation; either
9  * version 2.1 of the License, or (at your option) any later version.
10  *
11  * FFmpeg is distributed in the hope that it will be useful,
12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14  * Lesser General Public License for more details.
15  *
16  * You should have received a copy of the GNU Lesser General Public
17  * License along with FFmpeg; if not, write to the Free Software
18  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19  */
20 
21 #include <stdio.h>
22 #include <stdlib.h>
23 #include <string.h>
24 #include <inttypes.h>
25 #include <math.h>
26 #include <float.h>
27 #include <limits.h>
28 
29 #include "libavutil/intfloat.h"
30 #include "libavutil/intreadwrite.h"
31 
32 #define FFMIN(a, b) ((a) > (b) ? (b) : (a))
33 #define F 100
34 #define SIZE 2048
35 
36 uint64_t exp16_table[21] = {
37  65537,
38  65538,
39  65540,
40  65544,
41  65552,
42  65568,
43  65600,
44  65664,
45  65793,
46  66050,
47  66568,
48  67616,
49  69763,
50  74262,
51  84150,
52  108051,
53  178145,
54  484249,
55  3578144,
56  195360063,
57  582360139072LL,
58 };
59 
60 #if 0
61 // 16.16 fixpoint exp()
62 static unsigned int exp16(unsigned int a){
63  int i;
64  int out= 1<<16;
65 
66  for(i=19;i>=0;i--){
67  if(a&(1<<i))
68  out= (out*exp16_table[i] + (1<<15))>>16;
69  }
70 
71  return out;
72 }
73 #endif
74 
75 // 16.16 fixpoint log()
76 static int64_t log16(uint64_t a)
77 {
78  int i;
79  int out = 0;
80 
81  if (a < 1 << 16)
82  return -log16((1LL << 32) / a);
83  a <<= 16;
84 
85  for (i = 20; i >= 0; i--) {
86  int64_t b = exp16_table[i];
87  if (a < (b << 16))
88  continue;
89  out |= 1 << i;
90  a = ((a / b) << 16) + (((a % b) << 16) + b / 2) / b;
91  }
92  return out;
93 }
94 
95 static uint64_t int_sqrt(uint64_t a)
96 {
97  uint64_t ret = 0;
98  uint64_t ret_sq = 0;
99  int s;
100 
101  for (s = 31; s >= 0; s--) {
102  uint64_t b = ret_sq + (1ULL << (s * 2)) + (ret << s) * 2;
103  if (b <= a) {
104  ret_sq = b;
105  ret += 1ULL << s;
106  }
107  }
108  return ret;
109 }
110 
111 static int16_t get_s16l(uint8_t *p)
112 {
113  union {
114  uint16_t u;
115  int16_t s;
116  } v;
117  v.u = p[0] | p[1] << 8;
118  return v.s;
119 }
120 
121 static float get_f32l(uint8_t *p)
122 {
123  union av_intfloat32 v;
124  v.i = p[0] | p[1] << 8 | p[2] << 16 | p[3] << 24;
125  return v.f;
126 }
127 
128 static double get_f64l(uint8_t *p)
129 {
130  return av_int2double(AV_RL64(p));
131 }
132 
133 static int run_psnr(FILE *f[2], int len, int shift, int skip_bytes)
134 {
135  uint64_t i, j;
136  uint64_t sse = 0;
137  double sse_d = 0.0;
138  uint8_t buf[2][SIZE];
139  int64_t max = (1LL << (8 * len)) - 1;
140  uint64_t size0 = 0;
141  uint64_t size1 = 0;
142  uint64_t maxdist = 0;
143  double maxdist_d = 0.0;
144  int noseek;
145 
146  noseek = fseek(f[0], 0, SEEK_SET) ||
147  fseek(f[1], 0, SEEK_SET);
148 
149  if (!noseek) {
150  for (i = 0; i < 2; i++) {
151  uint8_t *p = buf[i];
152  if (fread(p, 1, 12, f[i]) != 12)
153  return -1;
154  if (!memcmp(p, "RIFF", 4) &&
155  !memcmp(p + 8, "WAVE", 4)) {
156  if (fread(p, 1, 8, f[i]) != 8)
157  return -1;
158  while (memcmp(p, "data", 4)) {
159  int s = p[4] | p[5] << 8 | p[6] << 16 | p[7] << 24;
160  fseek(f[i], s, SEEK_CUR);
161  if (fread(p, 1, 8, f[i]) != 8)
162  return -1;
163  }
164  } else {
165  fseek(f[i], -12, SEEK_CUR);
166  }
167  }
168 
169  fseek(f[shift < 0], abs(shift), SEEK_CUR);
170 
171  fseek(f[0], skip_bytes, SEEK_CUR);
172  fseek(f[1], skip_bytes, SEEK_CUR);
173  }
174 
175  for (;;) {
176  int s0 = fread(buf[0], 1, SIZE, f[0]);
177  int s1 = fread(buf[1], 1, SIZE, f[1]);
178 
179  for (j = 0; j < FFMIN(s0, s1); j += len) {
180  switch (len) {
181  case 1:
182  case 2: {
183  int64_t a, b;
184  int dist;
185  if (len == 2) {
186  a = get_s16l(buf[0] + j);
187  b = get_s16l(buf[1] + j);
188  } else {
189  a = buf[0][j];
190  b = buf[1][j];
191  }
192  sse += (a - b) * (a - b);
193  dist = llabs(a - b);
194  if (dist > maxdist)
195  maxdist = dist;
196  break;
197  }
198  case 4:
199  case 8: {
200  double dist, a, b;
201  if (len == 8) {
202  a = get_f64l(buf[0] + j);
203  b = get_f64l(buf[1] + j);
204  } else {
205  a = get_f32l(buf[0] + j);
206  b = get_f32l(buf[1] + j);
207  }
208  dist = fabs(a - b);
209  sse_d += (a - b) * (a - b);
210  if (dist > maxdist_d)
211  maxdist_d = dist;
212  break;
213  }
214  }
215  }
216  size0 += s0;
217  size1 += s1;
218  if (s0 + s1 <= 0)
219  break;
220  }
221 
222  i = FFMIN(size0, size1) / len;
223  if (!i)
224  i = 1;
225  switch (len) {
226  case 1:
227  case 2: {
228  uint64_t psnr;
229  uint64_t dev = int_sqrt(((sse / i) * F * F) + (((sse % i) * F * F) + i / 2) / i);
230  if (sse)
231  psnr = ((2 * log16(max << 16) + log16(i) - log16(sse)) *
232  284619LL * F + (1LL << 31)) / (1LL << 32);
233  else
234  psnr = 1000 * F - 1; // floating point free infinity :)
235 
236  printf("stddev:%5d.%02d PSNR:%3d.%02d MAXDIFF:%5"PRIu64" bytes:%9"PRIu64"/%9"PRIu64"\n",
237  (int)(dev / F), (int)(dev % F),
238  (int)(psnr / F), (int)(psnr % F),
239  maxdist, size0, size1);
240  return psnr;
241  }
242  case 4:
243  case 8: {
244  char psnr_str[64];
245  double psnr = INT_MAX;
246  double dev = sqrt(sse_d / i);
247  uint64_t scale = (len == 4) ? (1ULL << 24) : (1ULL << 32);
248 
249  if (sse_d) {
250  psnr = 2 * log(DBL_MAX) - log(i / sse_d);
251  snprintf(psnr_str, sizeof(psnr_str), "%5.02f", psnr);
252  } else
253  snprintf(psnr_str, sizeof(psnr_str), "inf");
254 
255  maxdist = maxdist_d * scale;
256 
257  printf("stddev:%10.2f PSNR:%s MAXDIFF:%10"PRIu64" bytes:%9"PRIu64"/%9"PRIu64"\n",
258  dev * scale, psnr_str, maxdist, size0, size1);
259  return psnr;
260  }
261  }
262  return -1;
263 }
264 
265 int main(int argc, char *argv[])
266 {
267  FILE *f[2];
268  int len = 1;
269  int shift_first= argc < 5 ? 0 : atoi(argv[4]);
270  int skip_bytes = argc < 6 ? 0 : atoi(argv[5]);
271  int shift_last = shift_first + (argc < 7 ? 0 : atoi(argv[6]));
272  int shift;
273  int max_psnr = -1;
274  int max_psnr_shift = 0;
275 
276  if (shift_last > shift_first)
277  shift_first -= shift_last - shift_first;
278 
279  if (argc > 3) {
280  if (!strcmp(argv[3], "u8")) {
281  len = 1;
282  } else if (!strcmp(argv[3], "s16")) {
283  len = 2;
284  } else if (!strcmp(argv[3], "f32")) {
285  len = 4;
286  } else if (!strcmp(argv[3], "f64")) {
287  len = 8;
288  } else {
289  char *end;
290  len = strtol(argv[3], &end, 0);
291  if (*end || len < 1 || len > 2) {
292  fprintf(stderr, "Unsupported sample format: %s\nSupported: u8, s16, f32, f64\n", argv[3]);
293  return 1;
294  }
295  }
296  }
297 
298  if (argc < 3) {
299  printf("tiny_psnr <file1> <file2> [<elem size>|u8|s16|f32|f64 [<shift> [<skip bytes> [<shift search range>]]]]\n");
300  printf("WAV headers are skipped automatically.\n");
301  return 1;
302  }
303 
304  f[0] = fopen(argv[1], "rb");
305  f[1] = fopen(argv[2], "rb");
306  if (!f[0] || !f[1]) {
307  fprintf(stderr, "Could not open input files.\n");
308  return 1;
309  }
310 
311  for (shift = shift_first; shift <= shift_last; shift++) {
312  int psnr = run_psnr(f, len, shift, skip_bytes);
313  if (psnr > max_psnr || (shift < 0 && psnr == max_psnr)) {
314  max_psnr = psnr;
315  max_psnr_shift = shift;
316  }
317  }
318  if (max_psnr < 0)
319  return 2;
320 
321  if (shift_last > shift_first)
322  printf("Best PSNR is %3d.%02d for shift %i\n", (int)(max_psnr / F), (int)(max_psnr % F), max_psnr_shift);
323  return 0;
324 }
const char * s
Definition: avisynth_c.h:768
static int shift(int a, int b)
Definition: sonic.c:82
#define F
Definition: tiny_psnr.c:33
static int64_t log16(uint64_t a)
Definition: tiny_psnr.c:76
const char * b
Definition: vf_curves.c:113
uint64_t exp16_table[21]
Definition: tiny_psnr.c:36
static int sse(MpegEncContext *s, uint8_t *src1, uint8_t *src2, int w, int h, int stride)
uint64_t_TMPL AV_RL64
Definition: bytestream.h:87
static av_always_inline double av_int2double(uint64_t i)
Reinterpret a 64-bit integer as a double.
Definition: intfloat.h:60
uint8_t
static av_cold int end(AVCodecContext *avctx)
Definition: avrndec.c:90
static double psnr(double d)
Definition: ffmpeg.c:1344
#define u(width, name, range_min, range_max)
Definition: cbs_h2645.c:344
static av_unused const uint8_t * skip_bytes(CABACContext *c, int n)
Skip n bytes and reset the decoder.
int main(int argc, char *argv[])
Definition: tiny_psnr.c:265
#define s0
Definition: regdef.h:37
static uint64_t int_sqrt(uint64_t a)
Definition: tiny_psnr.c:95
uint32_t i
Definition: intfloat.h:28
static float get_f32l(uint8_t *p)
Definition: tiny_psnr.c:121
void * buf
Definition: avisynth_c.h:690
#define SIZE
Definition: tiny_psnr.c:34
static int16_t get_s16l(uint8_t *p)
Definition: tiny_psnr.c:111
#define s1
Definition: regdef.h:38
#define snprintf
Definition: snprintf.h:34
static int run_psnr(FILE *f[2], int len, int shift, int skip_bytes)
Definition: tiny_psnr.c:133
int len
FILE * out
Definition: movenc.c:54
static double get_f64l(uint8_t *p)
Definition: tiny_psnr.c:128
#define FFMIN(a, b)
Definition: tiny_psnr.c:32