fixing some shaders on the fractional table

Artoriuz · Jan 26, 2024 · 752ff9e · 752ff9e
1 parent d4bbe69
commit 752ff9e
Show file tree

Hide file tree

Showing 23 changed files with 105 additions and 105 deletions.
diff --git a/blog/images/mpv_upscaling/fractional/ArtCNN_C4F16.png b/blog/images/mpv_upscaling/fractional/ArtCNN_C4F16.png
diff --git a/blog/images/mpv_upscaling/fractional/ArtCNN_C4F32.png b/blog/images/mpv_upscaling/fractional/ArtCNN_C4F32.png
diff --git a/blog/images/mpv_upscaling/fractional/ArtCNN_C4F8.png b/blog/images/mpv_upscaling/fractional/ArtCNN_C4F8.png
diff --git a/blog/images/mpv_upscaling/fractional/FSR.png b/blog/images/mpv_upscaling/fractional/FSR.png
diff --git a/blog/images/mpv_upscaling/fractional/FSRCNNX_x2_16-0-4-1.png b/blog/images/mpv_upscaling/fractional/FSRCNNX_x2_16-0-4-1.png
diff --git a/blog/images/mpv_upscaling/fractional/FSRCNNX_x2_8-0-4-1.png b/blog/images/mpv_upscaling/fractional/FSRCNNX_x2_8-0-4-1.png
diff --git a/blog/images/mpv_upscaling/fractional/benchmark_result.txt b/blog/images/mpv_upscaling/fractional/benchmark_result.txt
@@ -6,14 +6,14 @@ FSRCNNX_x2_16-0-4-1, 0.00802905589281976, 30.913782250462077, 0.9833859631769207
 FSRCNNX_x2_8-0-4-1, 0.008000409950457055, 30.957214927852093, 0.9828547585545852, 0.9964430236370497
 bilinear, 0.012745267252808591, 28.33959674383636, 0.9612300787218305, 0.9936028997098988
 lanczos, 0.009131681320214914, 31.11099230924609, 0.9768275660746599, 0.9966263076654822
-nnedi3-nns128-win8x4, 0.00874556730165375, 31.174536630672147, 0.9780784519717719, 0.996582834628736
-nnedi3-nns256-win8x4, 0.00874556730165375, 31.174536630672147, 0.9780784519717719, 0.996582834628736
-nnedi3-nns32-win8x4, 0.00874556730165375, 31.174536630672147, 0.9780784519717719, 0.996582834628736
-nnedi3-nns64-win8x4, 0.00874556730165375, 31.174536630672147, 0.9780784519717719, 0.996582834628736
+nnedi3-nns128-win8x4, 0.01136847742655782, 28.92468632346648, 0.9683903304560925, 0.995767913211886
+nnedi3-nns256-win8x4, 0.011356861000628004, 28.922338307305715, 0.968428636244278, 0.9957581860613082
+nnedi3-nns32-win8x4, 0.011440592073128187, 28.894831828856894, 0.9680317291289462, 0.9957221647282218
+nnedi3-nns64-win8x4, 0.011420968529760647, 28.895740591035096, 0.9681279415158695, 0.9957244189487038
 polar_lanczos4sharpest, 0.008494379317563317, 31.191959542259617, 0.9785178456147021, 0.9966058506497129
 polar_lanczossharp, 0.00958134638196916, 30.82352293521764, 0.9751989570190256, 0.9964678383114579
-ravu-lite-ar-r2, 0.00874556730165375, 31.174536630672147, 0.9780784519717719, 0.996582834628736
-ravu-lite-ar-r3, 0.00874556730165375, 31.174536630672147, 0.9780784519717719, 0.996582834628736
-ravu-lite-ar-r4, 0.00874556730165375, 31.174536630672147, 0.9780784519717719, 0.996582834628736
+ravu-lite-ar-r2, 0.00925968878654665, 30.123854634919972, 0.9766077439094667, 0.9959187791505901
+ravu-lite-ar-r3, 0.009117399344079272, 30.24453763588971, 0.9775096842844977, 0.9959369215835091
+ravu-lite-ar-r4, 0.009063069569464798, 30.21875277731665, 0.977681370934043, 0.995899057457979
 ravu-zoom-ar-r2, 0.007804230339822762, 31.575520380873606, 0.9816706305363946, 0.997001011822425
 ravu-zoom-ar-r3, 0.00759951503733166, 31.732636485811025, 0.9828727442800578, 0.997091151129254
diff --git a/blog/images/mpv_upscaling/fractional/bilinear.png b/blog/images/mpv_upscaling/fractional/bilinear.png
diff --git a/blog/images/mpv_upscaling/fractional/downscaled.png b/blog/images/mpv_upscaling/fractional/downscaled.png
diff --git a/blog/images/mpv_upscaling/fractional/lanczos.png b/blog/images/mpv_upscaling/fractional/lanczos.png
diff --git a/blog/images/mpv_upscaling/fractional/nnedi3-nns128-win8x4.png b/blog/images/mpv_upscaling/fractional/nnedi3-nns128-win8x4.png
diff --git a/blog/images/mpv_upscaling/fractional/nnedi3-nns256-win8x4.png b/blog/images/mpv_upscaling/fractional/nnedi3-nns256-win8x4.png
diff --git a/blog/images/mpv_upscaling/fractional/nnedi3-nns32-win8x4.png b/blog/images/mpv_upscaling/fractional/nnedi3-nns32-win8x4.png
diff --git a/blog/images/mpv_upscaling/fractional/nnedi3-nns64-win8x4.png b/blog/images/mpv_upscaling/fractional/nnedi3-nns64-win8x4.png
diff --git a/blog/images/mpv_upscaling/fractional/polar_lanczos4sharpest.png b/blog/images/mpv_upscaling/fractional/polar_lanczos4sharpest.png
diff --git a/blog/images/mpv_upscaling/fractional/polar_lanczossharp.png b/blog/images/mpv_upscaling/fractional/polar_lanczossharp.png
diff --git a/blog/images/mpv_upscaling/fractional/ravu-lite-ar-r2.png b/blog/images/mpv_upscaling/fractional/ravu-lite-ar-r2.png
diff --git a/blog/images/mpv_upscaling/fractional/ravu-lite-ar-r3.png b/blog/images/mpv_upscaling/fractional/ravu-lite-ar-r3.png
diff --git a/blog/images/mpv_upscaling/fractional/ravu-lite-ar-r4.png b/blog/images/mpv_upscaling/fractional/ravu-lite-ar-r4.png
diff --git a/blog/images/mpv_upscaling/fractional/ravu-zoom-ar-r2.png b/blog/images/mpv_upscaling/fractional/ravu-zoom-ar-r2.png
diff --git a/blog/images/mpv_upscaling/fractional/ravu-zoom-ar-r3.png b/blog/images/mpv_upscaling/fractional/ravu-zoom-ar-r3.png
diff --git a/blog/images/mpv_upscaling/fractional/reference.png b/blog/images/mpv_upscaling/fractional/reference.png
diff --git a/blog/mpv_upscaling.html b/blog/mpv_upscaling.html
@@ -529,104 +529,6 @@ <h3 id="fractional_upsampling_results">Fractional Upsampling Results</h3>
             <td></td>
             <td>0.6944</td>
           </tr>
-          <tr>
-            <td>nnedi3-nns128-win8x4</td>
-            <td>8.75E-03</td>
-            <td>31.1745</td>
-            <td>0.9781</td>
-            <td>0.9966</td>
-            <td></td>
-            <td>0.6893</td>
-            <td>0.6170</td>
-            <td>0.6832</td>
-            <td>0.7175</td>
-            <td></td>
-            <td>0.6768</td>
-          </tr>
-          <tr>
-            <td>nnedi3-nns256-win8x4</td>
-            <td>8.75E-03</td>
-            <td>31.1745</td>
-            <td>0.9781</td>
-            <td>0.9966</td>
-            <td></td>
-            <td>0.6893</td>
-            <td>0.6170</td>
-            <td>0.6832</td>
-            <td>0.7175</td>
-            <td></td>
-            <td>0.6768</td>
-          </tr>
-          <tr>
-            <td>nnedi3-nns32-win8x4</td>
-            <td>8.75E-03</td>
-            <td>31.1745</td>
-            <td>0.9781</td>
-            <td>0.9966</td>
-            <td></td>
-            <td>0.6893</td>
-            <td>0.6170</td>
-            <td>0.6832</td>
-            <td>0.7175</td>
-            <td></td>
-            <td>0.6768</td>
-          </tr>
-          <tr>
-            <td>nnedi3-nns64-win8x4</td>
-            <td>8.75E-03</td>
-            <td>31.1745</td>
-            <td>0.9781</td>
-            <td>0.9966</td>
-            <td></td>
-            <td>0.6893</td>
-            <td>0.6170</td>
-            <td>0.6832</td>
-            <td>0.7175</td>
-            <td></td>
-            <td>0.6768</td>
-          </tr>
-          <tr>
-            <td>ravu-lite-ar-r2</td>
-            <td>8.75E-03</td>
-            <td>31.1745</td>
-            <td>0.9781</td>
-            <td>0.9966</td>
-            <td></td>
-            <td>0.6893</td>
-            <td>0.6170</td>
-            <td>0.6832</td>
-            <td>0.7175</td>
-            <td></td>
-            <td>0.6768</td>
-          </tr>
-          <tr>
-            <td>ravu-lite-ar-r3</td>
-            <td>8.75E-03</td>
-            <td>31.1745</td>
-            <td>0.9781</td>
-            <td>0.9966</td>
-            <td></td>
-            <td>0.6893</td>
-            <td>0.6170</td>
-            <td>0.6832</td>
-            <td>0.7175</td>
-            <td></td>
-            <td>0.6768</td>
-          </tr>
-          <tr>
-            <td>ravu-lite-ar-r4</td>
-            <td>8.75E-03</td>
-            <td>31.1745</td>
-            <td>0.9781</td>
-            <td>0.9966</td>
-            <td></td>
-            <td>0.6893</td>
-            <td>0.6170</td>
-            <td>0.6832</td>
-            <td>0.7175</td>
-            <td></td>
-            <td>0.6768</td>
-          </tr>
           <tr>
             <td>lanczos</td>
             <td>9.13E-03</td>
@@ -655,6 +557,48 @@ <h3 id="fractional_upsampling_results">Fractional Upsampling Results</h3>
             <td></td>
             <td>0.5855</td>
           </tr>
+          <tr>
+            <td>ravu-lite-ar-r4</td>
+            <td>9.06E-03</td>
+            <td>30.2188</td>
+            <td>0.9777</td>
+            <td>0.9959</td>
+            <td></td>
+            <td>0.6346</td>
+            <td>0.4090</td>
+            <td>0.6671</td>
+            <td>0.5529</td>
+            <td></td>
+            <td>0.5659</td>
+          </tr>
+          <tr>
+            <td>ravu-lite-ar-r3</td>
+            <td>9.12E-03</td>
+            <td>30.2445</td>
+            <td>0.9775</td>
+            <td>0.9959</td>
+            <td></td>
+            <td>0.6252</td>
+            <td>0.4146</td>
+            <td>0.6602</td>
+            <td>0.5620</td>
+            <td></td>
+            <td>0.5655</td>
+          </tr>
+          <tr>
+            <td>ravu-lite-ar-r2</td>
+            <td>9.26E-03</td>
+            <td>30.1239</td>
+            <td>0.9766</td>
+            <td>0.9959</td>
+            <td></td>
+            <td>0.6007</td>
+            <td>0.3883</td>
+            <td>0.6236</td>
+            <td>0.5576</td>
+            <td></td>
+            <td>0.5426</td>
+          </tr>
           <tr>
             <td>FSR</td>
             <td>1.10E-02</td>
@@ -669,6 +613,62 @@ <h3 id="fractional_upsampling_results">Fractional Upsampling Results</h3>
             <td></td>
             <td>0.4481</td>
           </tr>
+          <tr>
+            <td>nnedi3-nns256-win8x4</td>
+            <td>1.14E-02</td>
+            <td>28.9223</td>
+            <td>0.9684</td>
+            <td>0.9958</td>
+            <td></td>
+            <td>0.2393</td>
+            <td>0.1268</td>
+            <td>0.2919</td>
+            <td>0.5190</td>
+            <td></td>
+            <td>0.2942</td>
+          </tr>
+          <tr>
+            <td>nnedi3-nns128-win8x4</td>
+            <td>1.14E-02</td>
+            <td>28.9247</td>
+            <td>0.9684</td>
+            <td>0.9958</td>
+            <td></td>
+            <td>0.2373</td>
+            <td>0.1273</td>
+            <td>0.2904</td>
+            <td>0.5213</td>
+            <td></td>
+            <td>0.2941</td>
+          </tr>
+          <tr>
+            <td>nnedi3-nns64-win8x4</td>
+            <td>1.14E-02</td>
+            <td>28.8957</td>
+            <td>0.9681</td>
+            <td>0.9957</td>
+            <td></td>
+            <td>0.2282</td>
+            <td>0.1210</td>
+            <td>0.2797</td>
+            <td>0.5108</td>
+            <td></td>
+            <td>0.2850</td>
+          </tr>
+          <tr>
+            <td>nnedi3-nns32-win8x4</td>
+            <td>1.14E-02</td>
+            <td>28.8948</td>
+            <td>0.9680</td>
+            <td>0.9957</td>
+            <td></td>
+            <td>0.2248</td>
+            <td>0.1208</td>
+            <td>0.2758</td>
+            <td>0.5103</td>
+            <td></td>
+            <td>0.2829</td>
+          </tr>
           <tr>
             <td>bilinear</td>
             <td>1.27E-02</td>