summaryrefslogtreecommitdiffstats
path: root/extra/xbmc/xbmc-9.11-shader-upscalers.patch
blob: d4feaa47f705dfe536b7c9eb48ea65d7dff1e88c (plain)
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
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
http://bugs.gentoo.org/306661

backport shader based upscalers from svn trunk

--- language/English/strings.xml
+++ language/English/strings.xml
@@ -1554,16 +1554,17 @@
   <string id="16304">Lanczos2</string>
   <string id="16305">Lanczos3</string>
   <string id="16306">Sinc8</string>
-
   <string id="16307">Bicubic (software)</string>
   <string id="16308">Lanczos (software)</string>
   <string id="16309">Sinc (software)</string>
-
   <string id="16310">(VDPAU)Temporal</string>
   <string id="16311">(VDPAU)Temporal/Spatial</string>
   <string id="16312">(VDPAU)Noise Reduction</string>
   <string id="16313">(VDPAU)Sharpness</string>
   <string id="16314">Inverse Telecine</string>
+  <string id="16315">Lanczos3 optimized</string>
+  <string id="16316">Auto</string>
+
   <string id="17500">Display sleep timeout</string>
 
   <string id="19000">Switch to channel</string>
--- system/shaders/convolution-6x6.glsl
+++ system/shaders/convolution-6x6.glsl
@@ -0,0 +1,69 @@
+uniform sampler2D img;
+uniform float     stepx;
+uniform float     stepy;
+
+#if (HAS_FLOAT_TEXTURE)
+uniform sampler1D kernelTex;
+
+vec3 weight(float pos)
+{
+  return texture1D(kernelTex, pos).rgb;
+}
+#else
+uniform sampler2D kernelTex;
+
+vec3 weight(float pos)
+{
+  //row 0 contains the high byte, row 1 contains the low byte
+  return ((texture2D(kernelTex, vec2(pos, 0.0)) * 256.0 + texture2D(kernelTex, vec2(pos, 1.0)))).rgb / 128.5 - 1.0;
+}
+#endif
+
+vec3 pixel(float xpos, float ypos)
+{
+  return texture2D(img, vec2(xpos, ypos)).rgb;
+}
+
+vec3 line (float ypos, vec3 xpos1, vec3 xpos2, vec3 linetaps1, vec3 linetaps2)
+{
+  vec3  pixels;
+
+  pixels  = pixel(xpos1.r, ypos) * linetaps1.r;
+  pixels += pixel(xpos1.g, ypos) * linetaps2.r;
+  pixels += pixel(xpos1.b, ypos) * linetaps1.g;
+  pixels += pixel(xpos2.r, ypos) * linetaps2.g;
+  pixels += pixel(xpos2.g, ypos) * linetaps1.b;
+  pixels += pixel(xpos2.b, ypos) * linetaps2.b;
+
+  return pixels;
+}
+
+void main()
+{
+  float xf = fract(gl_TexCoord[0].x / stepx);
+  float yf = fract(gl_TexCoord[0].y / stepy);
+
+  vec3 linetaps1   = weight((1.0 - xf) / 2.0);
+  vec3 linetaps2   = weight((1.0 - xf) / 2.0 + 0.5);
+  vec3 columntaps1 = weight((1.0 - yf) / 2.0);
+  vec3 columntaps2 = weight((1.0 - yf) / 2.0 + 0.5);
+
+  vec3 xpos1 = vec3(
+      (-1.5 - xf) * stepx + gl_TexCoord[0].x,
+      (-0.5 - xf) * stepx + gl_TexCoord[0].x,
+      ( 0.5 - xf) * stepx + gl_TexCoord[0].x);
+  vec3 xpos2 = vec3(
+      ( 1.5 - xf) * stepx + gl_TexCoord[0].x,
+      ( 2.5 - xf) * stepx + gl_TexCoord[0].x,
+      ( 3.5 - xf) * stepx + gl_TexCoord[0].x);
+
+  gl_FragColor.rgb  = line((-1.5 - yf) * stepy + gl_TexCoord[0].y, xpos1, xpos2, linetaps1, linetaps2) * columntaps1.r;
+  gl_FragColor.rgb += line((-0.5 - yf) * stepy + gl_TexCoord[0].y, xpos1, xpos2, linetaps1, linetaps2) * columntaps2.r;
+  gl_FragColor.rgb += line(( 0.5 - yf) * stepy + gl_TexCoord[0].y, xpos1, xpos2, linetaps1, linetaps2) * columntaps1.g;
+  gl_FragColor.rgb += line(( 1.5 - yf) * stepy + gl_TexCoord[0].y, xpos1, xpos2, linetaps1, linetaps2) * columntaps2.g;
+  gl_FragColor.rgb += line(( 2.5 - yf) * stepy + gl_TexCoord[0].y, xpos1, xpos2, linetaps1, linetaps2) * columntaps1.b;
+  gl_FragColor.rgb += line(( 3.5 - yf) * stepy + gl_TexCoord[0].y, xpos1, xpos2, linetaps1, linetaps2) * columntaps2.b;
+
+  gl_FragColor.a = gl_Color.a;
+}
+
--- system/shaders/bicubic.glsl
+++ system/shaders/bicubic.glsl
@@ -0,0 +1,47 @@
+uniform sampler2D img;
+uniform float stepx;
+uniform float stepy;
+uniform sampler2D kernelTex;
+
+vec4 cubicFilter(float xValue, vec4 c0, vec4 c1, vec4 c2, vec4 c3)
+{
+  vec4 h = texture2D(kernelTex, vec2(xValue, 0.5));
+  vec4 r = c0 * h.r;
+  r += c1 * h.g;
+  r += c2 * h.b;
+  r += c3 * h.a;
+  return r;
+}
+
+void main()
+{
+  vec2 f = vec2(gl_TexCoord[0].x / stepx , gl_TexCoord[0].y / stepy);
+  f = fract(f);
+  vec4 t0 = cubicFilter(f.x,
+  texture2D(img, gl_TexCoord[0].xy + vec2(-stepx,    -stepy)),
+  texture2D(img, gl_TexCoord[0].xy + vec2(0.0,       -stepy)),
+  texture2D(img, gl_TexCoord[0].xy + vec2(stepx,     -stepy)),
+  texture2D(img, gl_TexCoord[0].xy + vec2(2.0*stepx, -stepy)));
+
+  vec4 t1 = cubicFilter(f.x,
+  texture2D(img, gl_TexCoord[0].xy + vec2(-stepx,    0.0)),
+  texture2D(img, gl_TexCoord[0].xy + vec2(0.0,       0.0)),
+  texture2D(img, gl_TexCoord[0].xy + vec2(stepx,     0.0)),
+  texture2D(img, gl_TexCoord[0].xy + vec2(2.0*stepx, 0.0)));
+
+  vec4 t2 = cubicFilter(f.x,
+  texture2D(img, gl_TexCoord[0].xy + vec2(-stepx,    stepy)),
+  texture2D(img, gl_TexCoord[0].xy + vec2(0.0,       stepy)),
+  texture2D(img, gl_TexCoord[0].xy + vec2(stepx,     stepy)),
+  texture2D(img, gl_TexCoord[0].xy + vec2(2.0*stepx, stepy)));
+
+  vec4 t3 = cubicFilter(f.x,
+  texture2D(img, gl_TexCoord[0].xy + vec2(-stepx,    2.0*stepy)),
+  texture2D(img, gl_TexCoord[0].xy + vec2(0,         2.0*stepy)),
+  texture2D(img, gl_TexCoord[0].xy + vec2(stepx,     2.0*stepy)),
+  texture2D(img, gl_TexCoord[0].xy + vec2(2.0*stepx, 2.0*stepy)));
+
+  gl_FragColor = cubicFilter(f.y, t0, t1, t2, t3);   
+  gl_FragColor.a = gl_Color.a;
+}
+
--- system/shaders/convolution-4x4.glsl
+++ system/shaders/convolution-4x4.glsl
@@ -0,0 +1,60 @@
+uniform sampler2D img;
+uniform float     stepx;
+uniform float     stepy;
+
+#if (HAS_FLOAT_TEXTURE)
+uniform sampler1D kernelTex;
+
+vec4 weight(float pos)
+{
+  return texture1D(kernelTex, pos);
+}
+#else
+uniform sampler2D kernelTex;
+
+vec4 weight(float pos)
+{
+  //row 0 contains the high byte, row 1 contains the low byte
+  return (texture2D(kernelTex, vec2(pos, 0.0)) * 256.0 + texture2D(kernelTex, vec2(pos, 1.0))) / 128.5 - 1.0;
+}
+#endif
+
+vec3 pixel(float xpos, float ypos)
+{
+  return texture2D(img, vec2(xpos, ypos)).rgb;
+}
+
+vec3 line (float ypos, vec4 xpos, vec4 linetaps)
+{
+  vec3  pixels;
+
+  pixels  = pixel(xpos.r, ypos) * linetaps.r;
+  pixels += pixel(xpos.g, ypos) * linetaps.g;
+  pixels += pixel(xpos.b, ypos) * linetaps.b;
+  pixels += pixel(xpos.a, ypos) * linetaps.a;
+
+  return pixels;
+}
+
+void main()
+{
+  float xf = fract(gl_TexCoord[0].x / stepx);
+  float yf = fract(gl_TexCoord[0].y / stepy);
+
+  vec4 linetaps   = weight(1.0 - xf);
+  vec4 columntaps = weight(1.0 - yf);
+
+  vec4 xpos = vec4(
+      (-0.5 - xf) * stepx + gl_TexCoord[0].x,
+      ( 0.5 - xf) * stepx + gl_TexCoord[0].x,
+      ( 1.5 - xf) * stepx + gl_TexCoord[0].x,
+      ( 2.5 - xf) * stepx + gl_TexCoord[0].x);
+
+  gl_FragColor.rgb  = line((-0.5 - yf) * stepy + gl_TexCoord[0].y, xpos, linetaps) * columntaps.r;
+  gl_FragColor.rgb += line(( 0.5 - yf) * stepy + gl_TexCoord[0].y, xpos, linetaps) * columntaps.g;
+  gl_FragColor.rgb += line(( 1.5 - yf) * stepy + gl_TexCoord[0].y, xpos, linetaps) * columntaps.b;
+  gl_FragColor.rgb += line(( 2.5 - yf) * stepy + gl_TexCoord[0].y, xpos, linetaps) * columntaps.a;
+
+  gl_FragColor.a = gl_Color.a;
+}
+
--- xbmc/settings/VideoSettings.h
+++ xbmc/settings/VideoSettings.h
@@ -51,9 +51,10 @@
 {
   VS_SCALINGMETHOD_NEAREST=0,
   VS_SCALINGMETHOD_LINEAR,
-  
+
   VS_SCALINGMETHOD_CUBIC,
   VS_SCALINGMETHOD_LANCZOS2,
+  VS_SCALINGMETHOD_LANCZOS3_FAST,
   VS_SCALINGMETHOD_LANCZOS3,
   VS_SCALINGMETHOD_SINC8,
   VS_SCALINGMETHOD_NEDI,
@@ -61,7 +62,9 @@
   VS_SCALINGMETHOD_BICUBIC_SOFTWARE,
   VS_SCALINGMETHOD_LANCZOS_SOFTWARE,
   VS_SCALINGMETHOD_SINC_SOFTWARE,
-  VS_SCALINGMETHOD_VDPAU_HARDWARE
+  VS_SCALINGMETHOD_VDPAU_HARDWARE,
+
+  VS_SCALINGMETHOD_AUTO
 };
 
 class CVideoSettings
--- xbmc/cores/VideoRenderers/VideoShaders/VideoFilterShader.cpp
+++ xbmc/cores/VideoRenderers/VideoShaders/VideoFilterShader.cpp
@@ -21,6 +21,7 @@
 #include "system.h"
 #include "VideoFilterShader.h"
 #include "utils/log.h"
+#include "ConvolutionKernels.h"
 
 #include <string>
 #include <math.h>
@@ -63,60 +64,13 @@
 
 BicubicFilterShader::BicubicFilterShader(float B, float C)
 {
-  string shaderf = 
-    "uniform sampler2D img;"
-    "uniform float stepx;"
-    "uniform float stepy;"
-    "uniform sampler2D kernelTex;"
-    
-    "vec4 cubicFilter(float xValue, vec4 c0, vec4 c1, vec4 c2, vec4 c3)"
-    "{"
-    " vec4 h = texture2D(kernelTex, vec2(xValue, 0.5));"
-    " vec4 r = c0 * h.r;"
-    " r += c1 * h.g;"
-    " r += c2 * h.b;"
-    " r += c3 * h.a;"
-    " return r;"
-    "}"
-    ""
-    "void main()"
-    "{"
-    "vec2 f = vec2(gl_TexCoord[0].x / stepx , gl_TexCoord[0].y / stepy);"
-    "f = fract(f);"
-    "vec4 t0 = cubicFilter(f.x,"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(-stepx, -stepy)),"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(0.0, -stepy)),"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(stepx, -stepy)),"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(2.0*stepx, -stepy)));"
-    ""
-    "vec4 t1 = cubicFilter(f.x,"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(-stepx, 0.0)),"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(0.0, 0.0)),"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(stepx, 0.0)),"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(2.0*stepx, 0.0)));"
-    ""
-    "vec4 t2 = cubicFilter(f.x,"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(-stepx, stepy)),"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(0.0, stepy)),"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(stepx, stepy)),"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(2.0*stepx, stepy)));"
-    ""
-    "vec4 t3 = cubicFilter(f.x,"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(-stepx, 2.0*stepy)),"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(0, 2.0*stepy)),"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(stepx, 2.0*stepy)),"
-    "texture2D(img, gl_TexCoord[0].xy + vec2(2.0*stepx, 2.0*stepy)));"
-    
-    "gl_FragColor = cubicFilter(f.y, t0, t1, t2, t3) ;"    
-    "gl_FragColor.a = gl_Color.a;"
-    "}";
-  PixelShader()->SetSource(shaderf);
+  PixelShader()->LoadSource("bicubic.glsl");
   m_kernelTex1 = 0;
   m_B = B;
   m_C = C;
-  if (B<=0)
+  if (B<0)
     m_B=1.0f/3.0f;
-  if (C<=0)
+  if (C<0)
     m_C=1.0f/3.0f;
 }
 
@@ -209,8 +163,8 @@
   glBindTexture(GL_TEXTURE_2D, m_kernelTex1);
   glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_LINEAR);
   glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR);
-  glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_REPEAT);
-  glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_REPEAT);
+  glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_CLAMP);
+  glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_CLAMP);
   glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA16F_ARB, size, 1, 0, GL_RGBA, GL_FLOAT, img);
 
   glActiveTexture(GL_TEXTURE0);
@@ -254,4 +208,110 @@
   return val;
 }
 
+ConvolutionFilterShader::ConvolutionFilterShader(ESCALINGMETHOD method)
+{
+  m_method = method;
+  m_kernelTex1 = 0;
+
+  string shadername;
+  string defines;
+
+  if (m_method == VS_SCALINGMETHOD_CUBIC ||
+      m_method == VS_SCALINGMETHOD_LANCZOS2 ||
+      m_method == VS_SCALINGMETHOD_LANCZOS3_FAST)
+    shadername = "convolution-4x4.glsl";
+  else if (m_method == VS_SCALINGMETHOD_LANCZOS3)
+    shadername = "convolution-6x6.glsl";
+
+  m_floattex = glewIsSupported("GL_ARB_texture_float");
+
+  if (m_floattex)
+    defines = "#define HAS_FLOAT_TEXTURE 1\n";
+  else
+    defines = "#define HAS_FLOAT_TEXTURE 0\n";
+
+  CLog::Log(LOGDEBUG, "GL: ConvolutionFilterShader: using %s defines: %s", shadername.c_str(), defines.c_str());
+  PixelShader()->LoadSource(shadername, defines);
+}
+
+void ConvolutionFilterShader::OnCompiledAndLinked()
+{
+  // obtain shader attribute handles on successfull compilation
+  m_hSourceTex = glGetUniformLocation(ProgramHandle(), "img");
+  m_hStepX     = glGetUniformLocation(ProgramHandle(), "stepx");
+  m_hStepY     = glGetUniformLocation(ProgramHandle(), "stepy");
+  m_hKernTex   = glGetUniformLocation(ProgramHandle(), "kernelTex");
+
+  CConvolutionKernel kernel(m_method, 256);
+
+  if (m_kernelTex1)
+  {
+    glDeleteTextures(1, &m_kernelTex1);
+    m_kernelTex1 = 0;
+  }
+
+  glGenTextures(1, &m_kernelTex1);
+
+  if ((m_kernelTex1<=0))
+  {
+    CLog::Log(LOGERROR, "GL: ConvolutionFilterShader: Error creating kernel texture");
+    return;
+  }
+
+  glActiveTexture(GL_TEXTURE2);
+
+  //if float textures are supported, we can load the kernel as a 1d float texture
+  //if not, we load it as a 2d texture with 2 rows, where row 0 contains the high byte
+  //and row 1 contains the low byte, which can be converted in the shader
+  if (m_floattex)
+  {
+    glBindTexture(GL_TEXTURE_1D, m_kernelTex1);
+    glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MAG_FILTER, GL_LINEAR);
+    glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_MIN_FILTER, GL_LINEAR);
+    glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_WRAP_S, GL_CLAMP_TO_EDGE);
+    glTexParameteri(GL_TEXTURE_1D, GL_TEXTURE_WRAP_T, GL_CLAMP_TO_EDGE);
+    glTexImage1D(GL_TEXTURE_1D, 0, GL_RGBA16F_ARB, kernel.GetSize(), 0, GL_RGBA, GL_FLOAT, kernel.GetFloatPixels());
+  }
+  else
+  {
+    glBindTexture(GL_TEXTURE_2D, m_kernelTex1);
+    glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_NEAREST);
+    glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
+    glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_CLAMP_TO_EDGE);
+    glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_CLAMP_TO_EDGE);
+    glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA, kernel.GetSize(), 2, 0, GL_RGBA, GL_UNSIGNED_BYTE, kernel.GetIntFractPixels());
+  }
+
+  glActiveTexture(GL_TEXTURE0);
+
+  VerifyGLState();
+}
+
+bool ConvolutionFilterShader::OnEnabled()
+{
+  // set shader attributes once enabled
+  glActiveTexture(GL_TEXTURE2);
+
+  if (m_floattex)
+    glBindTexture(GL_TEXTURE_1D, m_kernelTex1);
+  else
+    glBindTexture(GL_TEXTURE_2D, m_kernelTex1);
+
+  glActiveTexture(GL_TEXTURE0);
+  glUniform1i(m_hSourceTex, m_sourceTexUnit);
+  glUniform1i(m_hKernTex, 2);
+  glUniform1f(m_hStepX, m_stepX);
+  glUniform1f(m_hStepY, m_stepY);
+  VerifyGLState();
+  return true;
+}
+
+void ConvolutionFilterShader::Free()
+{
+  if (m_kernelTex1)
+    glDeleteTextures(1, &m_kernelTex1);
+  m_kernelTex1 = 0;
+  BaseVideoFilterShader::Free();
+}
+
 #endif
--- xbmc/cores/VideoRenderers/VideoShaders/ConvolutionKernels.cpp
+++ xbmc/cores/VideoRenderers/VideoShaders/ConvolutionKernels.cpp
@@ -0,0 +1,226 @@
+/*
+ *      Copyright (C) 2005-2008 Team XBMC
+ *      http://www.xbmc.org
+ *
+ *  This Program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2, or (at your option)
+ *  any later version.
+ *
+ *  This Program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with XBMC; see the file COPYING.  If not, write to
+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ *  http://www.gnu.org/copyleft/gpl.html
+ *
+ */
+#ifdef _WIN32
+  #define _USE_MATH_DEFINES
+#endif
+
+#include "ConvolutionKernels.h"
+#include "MathUtils.h"
+
+#define SINC(x) (sin(M_PI * (x)) / (M_PI * (x)))
+
+CConvolutionKernel::CConvolutionKernel(ESCALINGMETHOD method, int size)
+{
+  m_size = size;
+  m_floatpixels = new float[m_size * 4];
+
+  if (method == VS_SCALINGMETHOD_LANCZOS2)
+    Lanczos2();
+  else if (method == VS_SCALINGMETHOD_LANCZOS3_FAST)
+    Lanczos3Fast();
+  else if (method == VS_SCALINGMETHOD_LANCZOS3)
+    Lanczos3();
+  else if (method == VS_SCALINGMETHOD_CUBIC) 
+    Bicubic(1.0 / 3.0, 1.0 / 3.0);
+
+  ToIntFract();
+}
+
+CConvolutionKernel::~CConvolutionKernel()
+{
+  delete [] m_floatpixels;
+  delete [] m_intfractpixels;
+}
+
+//generate a lanczos2 kernel which can be loaded with RGBA format
+//each value of RGBA has one tap, so a shader can load 4 taps with a single pixel lookup
+void CConvolutionKernel::Lanczos2()
+{
+  for (int i = 0; i < m_size; i++)
+  {
+    double x = (double)i / (double)m_size;
+
+    //generate taps
+    for (int j = 0; j < 4; j++)
+      m_floatpixels[i * 4 + j] = (float)LanczosWeight(x + (double)(j - 2), 2.0);
+
+    //any collection of 4 taps added together needs to be exactly 1.0
+    //for lanczos this is not always the case, so we take each collection of 4 taps
+    //and divide those taps by the sum of the taps
+    float weight = 0.0;
+    for (int j = 0; j < 4; j++)
+      weight += m_floatpixels[i * 4 + j];
+
+    for (int j = 0; j < 4; j++)
+      m_floatpixels[i * 4 + j] /= weight;
+  }
+}
+
+//generate a lanczos3 kernel which can be loaded with RGBA format
+//each value of RGBA has one tap, so a shader can load 4 taps with a single pixel lookup
+//the two outer lobes of the lanczos3 kernel are added to the two lobes one step to the middle
+//this basically looks the same as lanczos3, but the kernel only has 4 taps,
+//so it can use the 4x4 convolution shader which is twice as fast as the 6x6 one
+void CConvolutionKernel::Lanczos3Fast()
+{
+  for (int i = 0; i < m_size; i++)
+  {
+    double a = 3.0;
+    double x = (double)i / (double)m_size;
+
+    //generate taps
+    m_floatpixels[i * 4 + 0] = (float)(LanczosWeight(x - 2.0, a) + LanczosWeight(x - 3.0, a));
+    m_floatpixels[i * 4 + 1] = (float) LanczosWeight(x - 1.0, a);
+    m_floatpixels[i * 4 + 2] = (float) LanczosWeight(x      , a);
+    m_floatpixels[i * 4 + 3] = (float)(LanczosWeight(x + 1.0, a) + LanczosWeight(x + 2.0, a));
+
+    //any collection of 4 taps added together needs to be exactly 1.0
+    //for lanczos this is not always the case, so we take each collection of 4 taps
+    //and divide those taps by the sum of the taps
+    float weight = 0.0;
+    for (int j = 0; j < 4; j++)
+      weight += m_floatpixels[i * 4 + j];
+
+    for (int j = 0; j < 4; j++)
+      m_floatpixels[i * 4 + j] /= weight;
+  }
+}
+
+//generate a lanczos3 kernel which can be loaded with RGBA format
+//each value of RGB has one tap, so a shader can load 3 taps with a single pixel lookup
+void CConvolutionKernel::Lanczos3()
+{
+  for (int i = 0; i < m_size; i++)
+  {
+    double x = (double)i / (double)m_size;
+
+    //generate taps
+    for (int j = 0; j < 3; j++)
+      m_floatpixels[i * 4 + j] = (float)LanczosWeight(x * 2.0 + (double)(j * 2 - 3), 3.0);
+
+    m_floatpixels[i * 4 + 3] = 0.0;
+  }
+
+  //any collection of 6 taps added together needs to be exactly 1.0
+  //for lanczos this is not always the case, so we take each collection of 6 taps
+  //and divide those taps by the sum of the taps
+  for (int i = 0; i < m_size / 2; i++)
+  {
+    float weight = 0.0;
+    for (int j = 0; j < 3; j++)
+    {
+      weight += m_floatpixels[i * 4 + j];
+      weight += m_floatpixels[(i + m_size / 2) * 4 + j];
+    }
+    for (int j = 0; j < 3; j++)
+    {
+      m_floatpixels[i * 4 + j] /= weight;
+      m_floatpixels[(i + m_size / 2) * 4 + j] /= weight;
+    }
+  }
+}
+
+//generate a bicubic kernel which can be loaded with RGBA format
+//each value of RGBA has one tap, so a shader can load 4 taps with a single pixel lookup
+void CConvolutionKernel::Bicubic(double B, double C)
+{
+  for (int i = 0; i < m_size; i++)
+  {
+    double x = (double)i / (double)m_size;
+
+    //generate taps
+    for (int j = 0; j < 4; j++)
+      m_floatpixels[i * 4 + j] = (float)BicubicWeight(x + (double)(j - 2), B, C);
+  }
+}
+
+double CConvolutionKernel::LanczosWeight(double x, double radius)
+{
+  double ax = fabs(x);
+
+  if (ax == 0.0)
+    return 1.0;
+  else if (ax < radius)
+    return SINC(ax) * SINC(ax / radius);
+  else
+    return 0.0;
+}
+
+double CConvolutionKernel::BicubicWeight(double x, double B, double C)
+{
+  double ax = fabs(x);
+
+  if (ax<1.0)
+  {
+    return ((12 - 9*B - 6*C) * ax * ax * ax +
+            (-18 + 12*B + 6*C) * ax * ax +
+            (6 - 2*B))/6;
+  }
+  else if (ax<2.0)
+  {
+    return ((-B - 6*C) * ax * ax * ax + 
+            (6*B + 30*C) * ax * ax + (-12*B - 48*C) * 
+             ax + (8*B + 24*C)) / 6;
+  }
+  else
+  {
+    return 0.0;
+  }
+}
+
+
+//convert float to high byte/low byte, so the kernel can be loaded into an 8 bit texture
+//with height 2 and converted back to real float in the shader
+//it only works when the kernel texture uses nearest neighbour, but there's almost no difference
+//between that and linear interpolation
+void CConvolutionKernel::ToIntFract()
+{
+  m_intfractpixels = new uint8_t[m_size * 8];
+
+  for (int i = 0; i < m_size * 4; i++)
+  {
+    int value = MathUtils::round_int((m_floatpixels[i] + 1.0) / 2.0 * 65535.0);
+    if (value < 0)
+      value = 0;
+    else if (value > 65535)
+      value = 65535;
+    
+    int integer = value / 256;
+    int fract   = value % 256;
+
+    m_intfractpixels[i] = (uint8_t)integer;
+    m_intfractpixels[i + m_size * 4] = (uint8_t)fract;
+  }
+
+#if 0
+  for (int i = 0; i < 4; i++)
+  {
+    for (int j = 0; j < m_size; j++)
+    {
+      printf("%i %f %f\n",
+          i * m_size + j,
+          ((double)m_intfractpixels[j * 4 + i] + (double)m_intfractpixels[j * 4 + i + m_size * 4] / 255.0) / 255.0 * 2.0 - 1.0,
+          m_floatpixels[j * 4 + i]);
+    }
+  }
+#endif
+}
+
--- xbmc/cores/VideoRenderers/VideoShaders/VideoFilterShader.h
+++ xbmc/cores/VideoRenderers/VideoShaders/VideoFilterShader.h
@@ -4,6 +4,7 @@
 #ifdef HAS_GL
 
 #include "../../../../guilib/Shader.h"
+#include "../../../settings/VideoSettings.h"
 
 using namespace Shaders;
 
@@ -35,7 +36,7 @@
   class BicubicFilterShader : public BaseVideoFilterShader
   {
   public:
-    BicubicFilterShader(float B=0.0f, float C=0.0f);
+    BicubicFilterShader(float B=-1.0f, float C=-1.0f);
     void OnCompiledAndLinked();
     bool OnEnabled();
     void Free();
@@ -55,6 +56,25 @@
     float m_C;
   };
 
+  class ConvolutionFilterShader : public BaseVideoFilterShader
+  {
+  public:
+    ConvolutionFilterShader(ESCALINGMETHOD method);
+    void OnCompiledAndLinked();
+    bool OnEnabled();
+    void Free();
+
+  protected:
+    // kernel textures
+    GLuint m_kernelTex1;
+
+    // shader handles to kernel textures
+    GLint m_hKernTex;
+
+    ESCALINGMETHOD m_method;
+    bool           m_floattex; //if float textures are supported
+  };
+
 } // end namespace
 
 #endif
--- xbmc/cores/VideoRenderers/VideoShaders/ConvolutionKernels.h
+++ xbmc/cores/VideoRenderers/VideoShaders/ConvolutionKernels.h
@@ -0,0 +1,55 @@
+/*
+ *      Copyright (C) 2005-2008 Team XBMC
+ *      http://www.xbmc.org
+ *
+ *  This Program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2, or (at your option)
+ *  any later version.
+ *
+ *  This Program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with XBMC; see the file COPYING.  If not, write to
+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ *  http://www.gnu.org/copyleft/gpl.html
+ *
+ */
+
+#ifndef CONVOLUTIONKERNELS
+#define CONVOLUTIONKERNELS
+
+#include "system.h"
+#include "../../../settings/VideoSettings.h"
+
+class CConvolutionKernel
+{
+  public:
+    CConvolutionKernel(ESCALINGMETHOD method, int size);
+    ~CConvolutionKernel();
+
+    int      GetSize()           { return m_size; }
+    float*   GetFloatPixels()    { return m_floatpixels; }
+    uint8_t* GetIntFractPixels() { return m_intfractpixels; }
+
+  private:
+
+    void Lanczos2();
+    void Lanczos3Fast();
+    void Lanczos3();
+    void Bicubic(double B, double C);
+
+    double LanczosWeight(double x, double radius);
+    double BicubicWeight(double x, double B, double C);
+
+    void ToIntFract();
+
+    int      m_size;
+    float*   m_floatpixels;
+    uint8_t* m_intfractpixels;
+};
+
+#endif //CONVOLUTIONKERNELS
--- xbmc/cores/VideoRenderers/VideoShaders/Makefile
+++ xbmc/cores/VideoRenderers/VideoShaders/Makefile
@@ -1,5 +1,5 @@
 INCLUDES=-I. -I.. -I../../ -I../../../ -I../../../linux -I../../../../guilib
-SRCS=YUV2RGBShader.cpp VideoFilterShader.cpp
+SRCS=YUV2RGBShader.cpp VideoFilterShader.cpp ConvolutionKernels.cpp
 
 LIB=VideoShaders.a
 
--- xbmc/cores/VideoRenderers/LinuxRendererGL.cpp
+++ xbmc/cores/VideoRenderers/LinuxRendererGL.cpp
@@ -886,6 +886,19 @@
 
   VerifyGLState();
 
+  if (m_scalingMethod == VS_SCALINGMETHOD_AUTO)
+  {
+    bool scaleSD = (int)m_sourceWidth < m_upscalingWidth && (int)m_sourceHeight < m_upscalingHeight &&
+                   m_sourceHeight < 720 && m_sourceWidth < 1280;
+
+    if (Supports(VS_SCALINGMETHOD_VDPAU_HARDWARE))
+      m_scalingMethod = VS_SCALINGMETHOD_VDPAU_HARDWARE;
+    else if (Supports(VS_SCALINGMETHOD_LANCZOS3_FAST) && scaleSD)
+      m_scalingMethod = VS_SCALINGMETHOD_LANCZOS3_FAST;
+    else
+      m_scalingMethod = VS_SCALINGMETHOD_LINEAR;
+  }
+
   switch (m_scalingMethod)
   {
   case VS_SCALINGMETHOD_NEAREST:
@@ -898,13 +911,10 @@
     m_renderQuality = RQ_SINGLEPASS;
     return;
 
+  case VS_SCALINGMETHOD_LANCZOS2:
+  case VS_SCALINGMETHOD_LANCZOS3_FAST:
+  case VS_SCALINGMETHOD_LANCZOS3:
   case VS_SCALINGMETHOD_CUBIC:
-    if(!glewIsSupported("GL_ARB_texture_float"))
-    {
-      CLog::Log(LOGERROR, "GL: hardware doesn't support GL_ARB_texture_float");
-      break;
-    }
-
     if (!m_fbo.Initialize())
     {
       CLog::Log(LOGERROR, "GL: Error initializing FBO");
@@ -917,7 +927,7 @@
       break;
     }
 
-    m_pVideoFilterShader = new BicubicFilterShader(0.3f, 0.3f);
+    m_pVideoFilterShader = new ConvolutionFilterShader(m_scalingMethod);
     if (!m_pVideoFilterShader->CompileAndLink())
     {
       CLog::Log(LOGERROR, "GL: Error compiling and linking video filter shader");
@@ -928,8 +938,6 @@
     m_renderQuality = RQ_MULTIPASS;
     return;
 
-  case VS_SCALINGMETHOD_LANCZOS2:
-  case VS_SCALINGMETHOD_LANCZOS3:
   case VS_SCALINGMETHOD_SINC8:
   case VS_SCALINGMETHOD_NEDI:
     CLog::Log(LOGERROR, "GL: TODO: This scaler has not yet been implemented");
@@ -1895,16 +1903,19 @@
 bool CLinuxRendererGL::Supports(ESCALINGMETHOD method)
 {
   if(method == VS_SCALINGMETHOD_NEAREST
-  || method == VS_SCALINGMETHOD_LINEAR)
+  || method == VS_SCALINGMETHOD_LINEAR
+  || method == VS_SCALINGMETHOD_AUTO)
     return true;
 
-
-  if(method == VS_SCALINGMETHOD_CUBIC 
-  && glewIsSupported("GL_ARB_texture_float")
-  && glewIsSupported("GL_EXT_framebuffer_object")
-  && m_renderMethod == RENDER_GLSL)
-    return true;
-
+  if(method == VS_SCALINGMETHOD_CUBIC
+  || method == VS_SCALINGMETHOD_LANCZOS2
+  || method == VS_SCALINGMETHOD_LANCZOS3_FAST
+  || method == VS_SCALINGMETHOD_LANCZOS3)
+  {
+    if (glewIsSupported("GL_EXT_framebuffer_object") && (m_renderMethod & RENDER_GLSL))
+      return true;
+  }
+ 
   if (g_advancedSettings.m_videoHighQualityScaling != SOFTWARE_UPSCALING_DISABLED)
   {
     if(method == VS_SCALINGMETHOD_BICUBIC_SOFTWARE
--- xbmc/GUIDialogVideoSettings.cpp
+++ xbmc/GUIDialogVideoSettings.cpp
@@ -103,6 +103,7 @@
     entries.push_back(make_pair(VS_SCALINGMETHOD_LINEAR           , 16302));
     entries.push_back(make_pair(VS_SCALINGMETHOD_CUBIC            , 16303));
     entries.push_back(make_pair(VS_SCALINGMETHOD_LANCZOS2         , 16304));
+    entries.push_back(make_pair(VS_SCALINGMETHOD_LANCZOS3_FAST    , 16315));
     entries.push_back(make_pair(VS_SCALINGMETHOD_LANCZOS3         , 16305));
     entries.push_back(make_pair(VS_SCALINGMETHOD_SINC8            , 16306));
 //    entries.push_back(make_pair(VS_SCALINGMETHOD_NEDI             , ?????));
@@ -110,6 +111,7 @@
     entries.push_back(make_pair(VS_SCALINGMETHOD_LANCZOS_SOFTWARE , 16308));
     entries.push_back(make_pair(VS_SCALINGMETHOD_SINC_SOFTWARE    , 16309));
     entries.push_back(make_pair(VS_SCALINGMETHOD_VDPAU_HARDWARE   , 13120));
+    entries.push_back(make_pair(VS_SCALINGMETHOD_AUTO             , 16316));
 
     /* remove unsupported methods */
     for(vector<pair<int, int> >::iterator it = entries.begin(); it != entries.end();)
--- xbmc/Settings.cpp
+++ xbmc/Settings.cpp
@@ -772,7 +772,7 @@
     GetInteger(pElement, "interlacemethod", interlaceMethod, VS_INTERLACEMETHOD_NONE, VS_INTERLACEMETHOD_NONE, VS_INTERLACEMETHOD_INVERSE_TELECINE);
     m_stSettings.m_defaultVideoSettings.m_InterlaceMethod = (EINTERLACEMETHOD)interlaceMethod;
     int scalingMethod;
-    GetInteger(pElement, "scalingmethod", scalingMethod, VS_SCALINGMETHOD_LINEAR, VS_SCALINGMETHOD_NEAREST, VS_SCALINGMETHOD_CUBIC);
+    GetInteger(pElement, "scalingmethod", scalingMethod, VS_SCALINGMETHOD_LINEAR, VS_SCALINGMETHOD_NEAREST, VS_SCALINGMETHOD_AUTO);
     m_stSettings.m_defaultVideoSettings.m_ScalingMethod = (ESCALINGMETHOD)scalingMethod;
 
     GetInteger(pElement, "viewmode", m_stSettings.m_defaultVideoSettings.m_ViewMode, VIEW_MODE_NORMAL, VIEW_MODE_NORMAL, VIEW_MODE_CUSTOM);