-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathwebql.ispc
309 lines (255 loc) · 10.5 KB
/
webql.ispc
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
export uniform float calculate_radial_spectrumF16(uniform int16 cubeData[], uniform unsigned int stride, uniform int x1, uniform int x2, uniform int y1, uniform int y2, uniform int cx, uniform int cy, uniform int r2, uniform bool average, uniform float cdelt3)
{
float partial_sum = 0.0f ;
unsigned int64 partial_count = 0 ;
foreach (y=y1 ... y2, x=x1 ... x2)
{
float tmp = half_to_float_fast(cubeData[y*stride+x]) ;
bool nan = isnan(tmp) ;
float dist2 = (cx-x)*(cx-x) + (cy-y)*(cy-y) ;
bool exclude = (dist2 > r2) ;
partial_sum += (nan || exclude) ? 0.0f : tmp ;
partial_count += (nan || exclude) ? 0 : 1 ;
} ;
uniform float sum = reduce_add(partial_sum) ;
uniform int64 count = reduce_add(partial_count) ;
if(count > 0)
{
if(average)
return sum / (float) count ;
else
return sum * cdelt3 ;
}
else
return 0.0f ;
} ;
export uniform float calculate_square_spectrumF16(uniform int16 cubeData[], uniform unsigned int stride, uniform int x1, uniform int x2, uniform int y1, uniform int y2, uniform bool average, uniform float cdelt3)
{
float partial_sum = 0.0f ;
unsigned int64 partial_count = 0 ;
foreach (y=y1 ... y2, x=x1 ... x2)
{
float tmp = half_to_float_fast(cubeData[y*stride+x]) ;
bool nan = isnan(tmp) ;
partial_sum += nan ? 0.0f : tmp ;
partial_count += nan ? 0 : 1 ;
} ;
uniform float sum = reduce_add(partial_sum) ;
uniform int64 count = reduce_add(partial_count) ;
if(count > 0)
{
if(average)
return sum / (float) count ;
else
return sum * cdelt3 ;
}
else
return 0.0f ;
} ;
export uniform float calculate_radial_viewport_spectrumF16(uniform int16 cubeData[], uniform unsigned int cubeStride, uniform float pixels[], uniform bool mask[], uniform unsigned int stride, uniform int x1, uniform int x2, uniform int y1, uniform int y2, uniform int cx, uniform int cy, uniform int r2, uniform bool average, uniform float cdelt3)
{
float partial_sum = 0.0f ;
unsigned int64 partial_count = 0 ;
foreach (y=y1 ... y2, x=x1 ... x2)
{
float tmp = half_to_float_fast(cubeData[y*cubeStride+x]) ;
bool nan = isnan(tmp) ;
float dist2 = (cx-x)*(cx-x) + (cy-y)*(cy-y) ;
bool exclude = (dist2 > r2) ;
partial_sum += (nan || exclude) ? 0.0f : tmp ;
partial_count += (nan || exclude) ? 0 : 1 ;
// the still viewport
varying int idx = (y-y1)*stride+(x-x1) ;
pixels[idx] += (nan || exclude) ? 0.0f : tmp ;
mask[idx] |= (nan || exclude) ? false : true ;
} ;
uniform float sum = reduce_add(partial_sum) ;
uniform int64 count = reduce_add(partial_count) ;
if(count > 0)
{
if(average)
return sum / (float) count ;
else
return sum * cdelt3 ;
}
else
return 0.0f ;
} ;
export uniform float calculate_square_viewport_spectrumF16(uniform int16 cubeData[], uniform unsigned int cubeStride, uniform float pixels[], uniform bool mask[], uniform unsigned int stride, uniform int x1, uniform int x2, uniform int y1, uniform int y2, uniform bool average, uniform float cdelt3)
{
float partial_sum = 0.0f ;
unsigned int64 partial_count = 0 ;
foreach (y=y1 ... y2, x=x1 ... x2)
{
float tmp = half_to_float_fast(cubeData[y*cubeStride+x]) ;
bool nan = isnan(tmp) ;
partial_sum += nan ? 0.0f : tmp ;
partial_count += nan ? 0 : 1 ;
// the still viewport
int idx = (y-y1)*stride+(x-x1) ;
pixels[idx] += nan ? 0.0f : tmp ;
mask[idx] |= nan ? false : true ;
} ;
uniform float sum = reduce_add(partial_sum) ;
uniform int64 count = reduce_add(partial_count) ;
if(count > 0)
{
if(average)
return sum / (float) count ;
else
return sum * cdelt3 ;
}
else
return 0.0f ;
} ;
/* without any downsizing
foreach (y=0 ... height, x=0 ... width)
(...)
float tmp = half_to_float_fast(src[y*src_stride+x]) ;
unsigned int idx = y * dst_stride + x ;
*/
export void make_video_frameF16_linear(uniform unsigned int16 src[], uniform int width, uniform int height, uniform int src_stride, uniform unsigned int8 dst_luma[], uniform unsigned int8 dst_mask[], uniform int dst_width, uniform int dst_height, uniform int dst_stride, uniform float black, uniform float slope)
{
uniform float scale_x = (float)width / (float)dst_width ;
uniform float scale_y = (float)height / (float)dst_height ;
foreach (y=0 ... dst_height, x=0 ... dst_width)
{
// source
varying int sx = round(scale_x * (float)x) ;
varying int sy = round(scale_y * (float)y) ;
unsigned int src_x = clamp(sx, (varying int32)0, (varying int32)width-1);
unsigned int src_y = clamp(sy, (varying int32)0, (varying int32)height-1);
unsigned int src_idx = src_y*src_stride+src_x ;
float tmp = half_to_float_fast(src[src_idx]) ;
// destination
unsigned int idx = y * dst_stride + x ;
//alternative coherent if
cif (!isnan(tmp))
{
float pixel = 255.0f / ( 1.0f + exp(-6.0f * (tmp - black) * slope) ) ;
dst_luma[idx] = (int) clamp(pixel, 0.0f, 255.0f) ;
dst_mask[idx] = 255 ;
}
else
{
dst_luma[idx] = 0 ;
dst_mask[idx] = 0 ;
}
} ;
} ;
export void make_video_frameF16_logistic(uniform unsigned int16 src[], uniform int width, uniform int height, uniform int src_stride, uniform unsigned int8 dst_luma[], uniform unsigned int8 dst_mask[], uniform int dst_width, uniform int dst_height, uniform int dst_stride, uniform float median, uniform float sensitivity)
{
uniform float scale_x = (float)width / (float)dst_width ;
uniform float scale_y = (float)height / (float)dst_height ;
foreach (y=0 ... dst_height, x=0 ... dst_width)
{
// source
varying int sx = round(scale_x * (float)x) ;
varying int sy = round(scale_y * (float)y) ;
unsigned int src_x = clamp(sx, (varying int32)0, (varying int32)width-1);
unsigned int src_y = clamp(sy, (varying int32)0, (varying int32)height-1);
unsigned int src_idx = src_y*src_stride+src_x ;
float tmp = half_to_float_fast(src[src_idx]) ;
// destination
unsigned int idx = y * dst_stride + x ;
//alternative coherent if
cif (!isnan(tmp))
{
float pixel = 255.0f / ( 1.0f + exp(-6.0f * (tmp - median) * sensitivity) ) ;
dst_luma[idx] = (int) clamp(pixel, 0.0f, 255.0f) ;
dst_mask[idx] = 255 ;
}
else
{
dst_luma[idx] = 0 ;
dst_mask[idx] = 0 ;
}
} ;
} ;
export void make_video_frameF16_ratio(uniform unsigned int16 src[], uniform int width, uniform int height, uniform int src_stride, uniform unsigned int8 dst_luma[], uniform unsigned int8 dst_mask[], uniform int dst_width, uniform int dst_height, uniform int dst_stride, uniform float black, uniform float sensitivity)
{
uniform float scale_x = (float)width / (float)dst_width ;
uniform float scale_y = (float)height / (float)dst_height ;
foreach (y=0 ... dst_height, x=0 ... dst_width)
{
// source
varying int sx = round(scale_x * (float)x) ;
varying int sy = round(scale_y * (float)y) ;
unsigned int src_x = clamp(sx, (varying int32)0, (varying int32)width-1);
unsigned int src_y = clamp(sy, (varying int32)0, (varying int32)height-1);
unsigned int src_idx = src_y*src_stride+src_x ;
float tmp = half_to_float_fast(src[src_idx]) ;
// destination
unsigned int idx = y * dst_stride + x ;
//alternative coherent if
cif (!isnan(tmp))
{
float pixel = 5.0f * (tmp - black) * sensitivity ;
dst_luma[idx] = pixel > 0.0f ? (int) clamp( 255.0f*pixel/(1.0f + pixel), 0.0f, 255.0f) : 0 ;
dst_mask[idx] = 255 ;
}
else
{
dst_luma[idx] = 0 ;
dst_mask[idx] = 0 ;
}
} ;
} ;
export void make_video_frameF16_square(uniform unsigned int16 src[], uniform int width, uniform int height, uniform int src_stride, uniform unsigned int8 dst_luma[], uniform unsigned int8 dst_mask[], uniform int dst_width, uniform int dst_height, uniform int dst_stride, uniform float black, uniform float sensitivity)
{
uniform float scale_x = (float)width / (float)dst_width ;
uniform float scale_y = (float)height / (float)dst_height ;
foreach (y=0 ... dst_height, x=0 ... dst_width)
{
// source
varying int sx = round(scale_x * (float)x) ;
varying int sy = round(scale_y * (float)y) ;
unsigned int src_x = clamp(sx, (varying int32)0, (varying int32)width-1);
unsigned int src_y = clamp(sy, (varying int32)0, (varying int32)height-1);
unsigned int src_idx = src_y*src_stride+src_x ;
float tmp = half_to_float_fast(src[src_idx]) ;
// destination
unsigned int idx = y * dst_stride + x ;
//alternative coherent if
cif (!isnan(tmp))
{
float pixel = (tmp - black) * sensitivity ;
dst_luma[idx] = pixel > 0.0f ? (int) clamp( 255.0f*pixel*pixel, 0.0f, 255.0f) : 0 ;
dst_mask[idx] = 255 ;
}
else
{
dst_luma[idx] = 0 ;
dst_mask[idx] = 0 ;
}
} ;
} ;
export void make_video_frameF16_legacy(uniform unsigned int16 src[], uniform int width, uniform int height, uniform int src_stride, uniform unsigned int8 dst_luma[], uniform unsigned int8 dst_mask[], uniform int dst_width, uniform int dst_height, uniform int dst_stride, uniform float dmin, uniform float dmax, uniform float lmin, uniform float lmax)
{
uniform float scale_x = (float)width / (float)dst_width ;
uniform float scale_y = (float)height / (float)dst_height ;
foreach (y=0 ... dst_height, x=0 ... dst_width)
{
// source
varying int sx = round(scale_x * (float)x) ;
varying int sy = round(scale_y * (float)y) ;
unsigned int src_x = clamp(sx, (varying int32)0, (varying int32)width-1);
unsigned int src_y = clamp(sy, (varying int32)0, (varying int32)height-1);
unsigned int src_idx = src_y*src_stride+src_x ;
float tmp = half_to_float_fast(src[src_idx]) ;
// destination
unsigned int idx = y * dst_stride + x ;
//alternative coherent if
cif (!isnan(tmp))
{
float pixel = 0.5f + (tmp - dmin) / (dmax - dmin) ;
dst_luma[idx] = pixel > 0.0f ? (int) clamp( 255.0f * (log(pixel) - lmin) / (lmax - lmin), 0.0f, 255.0f) : 0 ;
dst_mask[idx] = 255 ;
}
else
{
dst_luma[idx] = 0 ;
dst_mask[idx] = 0 ;
}
} ;
} ;