Skip to content

Commit

Permalink
Updated metrics and graphics
Browse files Browse the repository at this point in the history
  • Loading branch information
WyohKnott committed Sep 28, 2017
1 parent 71e1a54 commit ea01bd0
Show file tree
Hide file tree
Showing 36 changed files with 7,651 additions and 16,563 deletions.
3 changes: 1 addition & 2 deletions .gitattributes
Original file line number Diff line number Diff line change
@@ -1,10 +1,9 @@
*.aom -delta
*.bpg -delta
*.flif -delta
*.jp2 -delta
*.jpg -delta
*.jxr -delta
*.ogv -delta
*.png -delta
*.vp10 -delta
*.webp -delta
*.tar.gz -delta

This file was deleted.

2,076 changes: 0 additions & 2,076 deletions Bits per pixel at equivalent quality according to RGB-SSIM.svg

This file was deleted.

This file was deleted.

2,045 changes: 0 additions & 2,045 deletions Bits per pixel at equivalent quality according to VMAF.svg

This file was deleted.

This file was deleted.

2,122 changes: 0 additions & 2,122 deletions Bits per pixel at equivalent quality according to Y-MSSSIM.svg

This file was deleted.

This file was deleted.

2,146 changes: 0 additions & 2,146 deletions Bits per pixel at equivalent quality according to Y-PSNR-HVS-M.svg

This file was deleted.

This file was deleted.

1,620 changes: 0 additions & 1,620 deletions Bits per pixel at equivalent quality according to Y-SSIM.svg

This file was deleted.

175 changes: 55 additions & 120 deletions comparison.html
Original file line number Diff line number Diff line change
Expand Up @@ -4,147 +4,82 @@
<meta http-equiv="content-type" content="text/html; charset=utf-8">
<meta name="generator" content="ReText 7.0.0">
<link rel="stylesheet" media="screen" href="style.css" type="text/css" />
<title>comparison</title>
<title>Evolution of the AV1 codec between October 2016 and July 2017</title>
</head>
<body>
<h1>Evolution of the AV1 codec between October 2016 and July 2017</h1>
<p>The VMAF metric significatively deteriorated however all other metrics improved. The encoding time increased by a factor of around
7,3.</p>
<h3>Lossless compression ratio and Weissman score:</h3>
<table>
<thead>
<tr>
<th style="border-bottom: 2px solid #000000; border-right: 2px solid #000000;" bgcolor="#DDEBF7" align="left">codec</th>
<th style="border-bottom: 2px solid #000000;" bgcolor="#F8CBAD" align="center">compression ratio</th>
<th style="border-bottom: 2px solid #000000;" bgcolor="#F4B084" align="center">avg. enc. time</th>
<th style="border-bottom: 2px solid #000000;" bgcolor="#A9D08E" align="center">avg. dec. time</th>
<th style="border-bottom: 2px solid #000000;" bgcolor="#8EA9DB" align="center">Reduction gain (from PNG)</th>
<th style="border-bottom: 2px solid #000000;" bgcolor="#8EA9DB" align="center">Weissman score</th>
<th>format</th>
<th align="right">avg_bpp</th>
<th align="right">avg_compression_ratio</th>
<th align="right">avg_space_saving</th>
<th align="right">wavg_encode_time</th>
<th align="right">wavg_decode_time</th>
<th align="right">weissman_score</th>
</tr>
</thead>
<tbody>
<tr>
<td style="border-right: 2px solid #000000;" bgcolor="#DDEBF7" align="left">Daala</td>
<td bgcolor="#F8CBAD" align="center">4,249</td>
<td bgcolor="#F4B084" align="center">0,13</td>
<td bgcolor="#A9D08E" align="center">0,13</td>
<td bgcolor="#8EA9DB" align="center">183,62%</td>
<td bgcolor="#8EA9DB" align="center">3,10</td>
<td>daala</td>
<td align="right">5.791</td>
<td align="right">2.798</td>
<td align="right">0.6426</td>
<td align="right">1.4952</td>
<td align="right">1.2960</td>
<td align="right">3.2145</td>
</tr>
<tr>
<td style="border-right: 2px solid #000000;" bgcolor="#DDEBF7" align="left">AV1 oct2016</td>
<td bgcolor="#F8CBAD" align="center">4,436</td>
<td bgcolor="#F4B084" align="center">1,27</td>
<td bgcolor="#A9D08E" align="center">0,09</td>
<td bgcolor="#8EA9DB" align="center">196,13%</td>
<td bgcolor="#8EA9DB" align="center">2,20</td>
<td>av1-20160930</td>
<td align="right">5.565</td>
<td align="right">2.912</td>
<td align="right">0.6566</td>
<td align="right">31.1063</td>
<td align="right">1.6304</td>
<td align="right">2.3640</td>
</tr>
<tr>
<td style="border-right: 2px solid #000000;" bgcolor="#DDEBF7" align="left">AV1 jul2017 mt</td>
<td bgcolor="#F8CBAD" align="center">4,554</td>
<td bgcolor="#F4B084" align="center">3,98</td>
<td bgcolor="#A9D08E" align="center">0,14</td>
<td bgcolor="#8EA9DB" align="center">203,99%</td>
<td bgcolor="#8EA9DB" align="center">1,94</td>
<td>av1-20170809</td>
<td align="right">5.366</td>
<td align="right">3.020</td>
<td align="right">0.6689</td>
<td align="right">118.6229</td>
<td align="right">1.1642</td>
<td align="right">2.1707</td>
</tr>
<tr>
<td style="border-right: 2px solid #000000;" bgcolor="#DDEBF7" align="left">AV1 jul2017</td>
<td bgcolor="#F8CBAD" align="center">4,565</td>
<td bgcolor="#F4B084" align="center">9,25</td>
<td bgcolor="#A9D08E" align="center">0,13</td>
<td bgcolor="#8EA9DB" align="center">204,73%</td>
<td bgcolor="#8EA9DB" align="center">1,77</td>
<td>mozjpeg</td>
<td align="right">14.252</td>
<td align="right">1.137</td>
<td align="right">0.1205</td>
<td align="right">14.0222</td>
<td align="right">0.6776</td>
<td align="right">1.0000</td>
</tr>
<tr>
<td style="border-right: 2px solid #000000;" bgcolor="#DDEBF7" align="left">Reference (PNG)</td>
<td bgcolor="#F8CBAD" align="center">1,498</td>
<td bgcolor="#F4B084" align="center">0,20</td>
<td bgcolor="#A9D08E" align="center">0,11</td>
<td bgcolor="#8EA9DB" align="center">0,00%</td>
<td bgcolor="#8EA9DB" align="center">1,00</td>
</tr>
<tr>
<td style="border-right: 2px solid #000000;" bgcolor="#DDEBF7" align="left">BPG</td>
<td bgcolor="#F8CBAD" align="center">1,752</td>
<td bgcolor="#F4B084" align="center">0,72</td>
<td bgcolor="#A9D08E" align="center">0,68</td>
<td bgcolor="#8EA9DB" align="center">16,93%</td>
<td bgcolor="#8EA9DB" align="center">0,94</td>
</tr>
</tbody>
</table>
<h3>Speed relatively to PSNR-HVS-M at "large" quality:</h3>
<table>
<thead>
<tr>
<th style="border-bottom: 2px solid #000000; border-right: 2px solid #000000;" bgcolor="#DDEBF7" align="left">codec</th>
<th style="border-bottom: 2px solid #000000;" bgcolor="#F8CBAD" align="center">PSNR-HVS-M</th>
<th style="border-bottom: 2px solid #000000;" bgcolor="#F4B084" align="center">avg. enc. time</th>
<th style="border-bottom: 2px solid #000000;" bgcolor="#A9D08E" align="center">avg. dec. time</th>
<th style="border-bottom: 2px solid #000000;" bgcolor="#8EA9DB" align="center">PSNR-HVS-M gain (from Mozjpeg)</th>
<th style="border-bottom: 2px solid #000000;" bgcolor="#8EA9DB" align="center">Pseudo-Weissman score</th>
</tr>
</thead>
<tbody>
<tr>
<td style="border-right: 2px solid #000000;" bgcolor="#DDEBF7" align="left">Reference (Mozjpeg)</td>
<td bgcolor="#F8CBAD" align="center">41,10</td>
<td bgcolor="#F4B084" align="center">0,29</td>
<td bgcolor="#A9D08E" align="center">0,03</td>
<td bgcolor="#8EA9DB" align="center">0,00%</td>
<td bgcolor="#8EA9DB" align="center">1,00</td>
</tr>
<tr>
<td style="border-right: 2px solid #000000;" bgcolor="#DDEBF7" align="left">Daala</td>
<td bgcolor="#F8CBAD" align="center">48,90</td>
<td bgcolor="#F4B084" align="center">1,708</td>
<td bgcolor="#A9D08E" align="center">0,10</td>
<td bgcolor="#8EA9DB" align="center">18,96%</td>
<td bgcolor="#8EA9DB" align="center">0,90</td>
</tr>
<tr>
<td style="border-right: 2px solid #000000;" bgcolor="#DDEBF7" align="left">BPG</td>
<td bgcolor="#F8CBAD" align="center">41,29</td>
<td bgcolor="#F4B084" align="center">0,74</td>
<td bgcolor="#A9D08E" align="center">0,39</td>
<td bgcolor="#8EA9DB" align="center">0,46%</td>
<td bgcolor="#8EA9DB" align="center">0,86</td>
</tr>
<tr>
<td style="border-right: 2px solid #000000;" bgcolor="#DDEBF7" align="left">AV1 (oct2016)</td>
<td bgcolor="#F8CBAD" align="center">46,94</td>
<td bgcolor="#F4B084" align="center">3,19</td>
<td bgcolor="#A9D08E" align="center">0,04</td>
<td bgcolor="#8EA9DB" align="center">14,22%</td>
<td bgcolor="#8EA9DB" align="center">0,80</td>
</tr>
<tr>
<td style="border-right: 2px solid #000000;" bgcolor="#DDEBF7" align="left">AV1 (jul2017) mt</td>
<td bgcolor="#F8CBAD" align="center">47,79</td>
<td bgcolor="#F4B084" align="center">39,00</td>
<td bgcolor="#A9D08E" align="center">0,07</td>
<td bgcolor="#8EA9DB" align="center">16,29%</td>
<td bgcolor="#8EA9DB" align="center">0,62</td>
</tr>
<tr>
<td style="border-right: 2px solid #000000;" bgcolor="#DDEBF7" align="left">AV1 (jul2017)</td>
<td bgcolor="#F8CBAD" align="center">47,79</td>
<td bgcolor="#F4B084" align="center">97,69</td>
<td bgcolor="#A9D08E" align="center">0,07</td>
<td bgcolor="#8EA9DB" align="center">16,27%</td>
<td bgcolor="#8EA9DB" align="center">0,57</td>
<td>bpg</td>
<td align="right">14.095</td>
<td align="right">1.150</td>
<td align="right">0.1303</td>
<td align="right">21.4230</td>
<td align="right">5.8602</td>
<td align="right">0.9682</td>
</tr>
</tbody>
</table>
<h4>Bits per pixel at equivalent quality according to VMAF</h4>
<p><img alt="Bits per pixel at equivalent quality according to VMAF" src="http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20VMAF%20-%20comparison%20oct2016-jul2017.svg"></p>
<h4>Bits per pixel at equivalent quality according to Y-PSNR-HVS-M</h4>
<p><img alt="Bits per pixel at equivalent quality according to Y-PSNR-HVS-M" src="http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20Y-PSNR-HVS-M%20-%20comparison%20oct2016-jul2017.svg"></p>
<h4>Bits per pixel at equivalent quality according to Y-MSSSIM</h4>
<p><img alt="Bits per pixel at equivalent quality according to Y-MSSSIM" src="http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20Y-MSSSIM%20-%20comparison%20oct2016-jul2017.svg"></p>
<h4>Bits per pixel at equivalent quality according to Y-SSIM</h4>
<p><img alt="Bits per pixel at equivalent quality according to Y-SSIM" src="http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20Y-SSIM%20-%20comparison%20oct2016-jul2017.svg"></p>
<h4>Bits per pixel at equivalent quality according to RGB-SSIM</h4>
<p><img alt="Bits per pixel at equivalent quality according to RGB-SSIM" src="http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20RGB-SSIM%20-%20comparison%20oct2016-jul2017.svg"></p>
<h3>Encoding times:</h3>
<p><img alt="Encodinq time in function of bits per pixel" src="http://wyohknott.github.io/image-formats-comparison/subset1.encoding_time.(av1-20160930,av1-20170809">.svg)</p>
<h3>Bits per pixel at equivalent quality according to VMAF</h3>
<p><img alt="Bits per pixel at equivalent quality according to VMAF" src="http://wyohknott.github.io/image-formats-comparison/subset1.vmaf.(av1-20160930,av1-20170809">.svg)</p>
<h3>Bits per pixel at equivalent quality according to Y-PSNR-HVS-M</h3>
<p><img alt="Bits per pixel at equivalent quality according to Y-PSNR-HVS-M" src="http://wyohknott.github.io/image-formats-comparison/subset1.psnr-hvs-m.(av1-20160930,av1-20170809">.svg)</p>
<h3>Bits per pixel at equivalent quality according to Y-MSSSIM</h3>
<p><img alt="Bits per pixel at equivalent quality according to Y-MSSSIM" src="http://wyohknott.github.io/image-formats-comparison/subset1.ms-ssim.(av1-20160930,av1-20170809">.svg)</p>
<h3>Bits per pixel at equivalent quality according to Y-SSIM</h3>
<p><img alt="Bits per pixel at equivalent quality according to Y-SSIM" src="http://wyohknott.github.io/image-formats-comparison/subset1.y-ssim.(av1-20160930,av1-20170809">.svg)</p>
<h3>Bits per pixel at equivalent quality according to RGB-SSIM</h3>
<p><img alt="Bits per pixel at equivalent quality according to RGB-SSIM" src="http://wyohknott.github.io/image-formats-comparison/subset1.rgb-ssim.(av1-20160930,av1-20170809">.svg)</p>
</body>
</html>
50 changes: 21 additions & 29 deletions comparison.md
Original file line number Diff line number Diff line change
@@ -1,50 +1,42 @@
#Evolution of the AV1 codec between October 2016 and July 2017

The VMAF metric significatively deteriorated however all other metrics improved. The encoding time increased by a factor of around
7,3.


###Lossless compression ratio and Weissman score:

| codec | compression ratio | avg. enc. time | avg. dec. time | Reduction gain (from PNG) | Weissman score |
|:------------------|:-----------------:|:--------------:|:--------------:|:-------------------------:|:--------------:|
| Daala | 4,249 | 0,13 | 0,13 | 183,62% | 3,10 |
| AV1 (oct 2016) | 4,436 | 1,27 | 0,09 | 196,13% | 2,20 |
| AV1 (jul 2017) mt | 4,584 | 3,98 | 0,14 | 203,99% | 1,94 |
| AV1 (jul 2017) | 4,565 | 9,25 | 0,13 | 204,73% | 1,77 |
| Reference (PNG) | 1,498 | 0,20 | 0,11 | 0,00% | 1,00 |
| BPG | 1,752 | 0,72 | 0,68 | 16,93% | 0,94 |
| format |avg_bpp|avg_compression_ratio|avg_space_saving|wavg_encode_time|wavg_decode_time|weissman_score|
|------------|------:|--------------------:|---------------:|---------------:|---------------:|-------------:|
|daala | 5.791| 2.798| 0.6426| 1.4952| 1.2960| 3.2145|
|av1-20160930| 5.565| 2.912| 0.6566| 31.1063| 1.6304| 2.3640|
|av1-20170809| 5.366| 3.020| 0.6689| 118.6229| 1.1642| 2.1707|
|mozjpeg | 14.252| 1.137| 0.1205| 14.0222| 0.6776| 1.0000|
|bpg | 14.095| 1.150| 0.1303| 21.4230| 5.8602| 0.9682|

###Encoding times:

###Speed relatively to PSNR-HVS-M at "large" quality:

![Encodinq time in function of bits per pixel](http://wyohknott.github.io/image-formats-comparison/subset1.encoding_time.(av1-20160930,av1-20170809).svg)

| codec | PSNR-HVS-M | avg. enc. time | avg. dec. time | PSNR-HVS-M gain (from Mozjpeg) | Pseudo-Weissman score |
|:-------------------|:-----------------:|:--------------:|:--------------:|:-------------------------:|:--------------:|
| Reference (Mozjpeg)| 41,10 | 0,29 | 0,03 | 0,00% | 1,00 |
| Daala | 48,90 | 1,78 | 0,10 | 18,96% | 0,90 |
| BPG | 41,29 | 0,74 | 0,39 | 0,46% | 0,86 |
| AV1 (oct 2016) | 46,94 | 3,19 | 0,04 | 14,22% | 0,80 |
| AV1 (jul 2017) mt | 47,79 | 39,00 | 0,07 | 16,29% | 0,62 |
| AV1 (jul 2017) | 47,79 | 97,69 | 0,07 | 16,27% | 0,57 |

####Bits per pixel at equivalent quality according to VMAF
###Bits per pixel at equivalent quality according to VMAF

![Bits per pixel at equivalent quality according to VMAF](http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20VMAF%20-%20comparison%20oct2016-jul2017.svg)
![Bits per pixel at equivalent quality according to VMAF](http://wyohknott.github.io/image-formats-comparison/subset1.vmaf.(av1-20160930,av1-20170809).svg)

####Bits per pixel at equivalent quality according to Y-PSNR-HVS-M
###Bits per pixel at equivalent quality according to Y-PSNR-HVS-M

![Bits per pixel at equivalent quality according to Y-PSNR-HVS-M](http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20Y-PSNR-HVS-M%20-%20comparison%20oct2016-jul2017.svg)
![Bits per pixel at equivalent quality according to Y-PSNR-HVS-M](http://wyohknott.github.io/image-formats-comparison/subset1.psnr-hvs-m.(av1-20160930,av1-20170809).svg)

####Bits per pixel at equivalent quality according to Y-MSSSIM
###Bits per pixel at equivalent quality according to Y-MSSSIM

![Bits per pixel at equivalent quality according to Y-MSSSIM](http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20Y-MSSSIM%20-%20comparison%20oct2016-jul2017.svg)
![Bits per pixel at equivalent quality according to Y-MSSSIM](http://wyohknott.github.io/image-formats-comparison/subset1.ms-ssim.(av1-20160930,av1-20170809).svg)

####Bits per pixel at equivalent quality according to Y-SSIM
###Bits per pixel at equivalent quality according to Y-SSIM

![Bits per pixel at equivalent quality according to Y-SSIM](http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20Y-SSIM%20-%20comparison%20oct2016-jul2017.svg)
![Bits per pixel at equivalent quality according to Y-SSIM](http://wyohknott.github.io/image-formats-comparison/subset1.y-ssim.(av1-20160930,av1-20170809).svg)

####Bits per pixel at equivalent quality according to RGB-SSIM
###Bits per pixel at equivalent quality according to RGB-SSIM

![Bits per pixel at equivalent quality according to RGB-SSIM](http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20RGB-SSIM%20-%20comparison%20oct2016-jul2017.svg)
![Bits per pixel at equivalent quality according to RGB-SSIM](http://wyohknott.github.io/image-formats-comparison/subset1.rgb-ssim.(av1-20160930,av1-20170809).svg)



Expand Down
46 changes: 23 additions & 23 deletions lossy_results.html
Original file line number Diff line number Diff line change
@@ -1,23 +1,23 @@
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=utf-8">
<meta name="generator" content="ReText 6.0.2">
<title>Image formats comparison, lossy metrics</title>
<link rel="stylesheet" media="screen" href="style.css" type="text/css" />
</head>
<body>
<h2>Lossy metrics</h2>
<p>For each comparison algorithms, we plot the quality in dB in function of the mean bits per pixel on a logarithmic scale. We can then visualize which codec gives the best quality at a given bit per pixel (top left is better).</p>
<h3>Bits per pixel at equivalent quality according to VMAF</h3>
<p><img class="graph" alt="Bits per pixel at equivalent quality according to VMAF" src="http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20VMAF.svg"></p>
<h3>Bits per pixel at equivalent quality according to Y-PSNR-HVS-M</h3>
<p><img class="graph" alt="Bits per pixel at equivalent quality according to Y-PSNR-HVS-M" src="http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20Y-PSNR-HVS-M.svg"></p>
<h3>Bits per pixel at equivalent quality according to Y-MSSSIM</h3>
<p><img class="graph" alt="Bits per pixel at equivalent quality according to Y-MSSSIM" src="http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20Y-MSSSIM.svg"></p>
<h3>Bits per pixel at equivalent quality according to Y-SSIM</h3>
<p><img class="graph" alt="Bits per pixel at equivalent quality according to Y-SSIM" src="http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20Y-SSIM.svg"></p>
<h3>Bits per pixel at equivalent quality according to RGB-SSIM</h3>
<p><img class="graph" alt="Bits per pixel at equivalent quality according to RGB-SSIM" src="http://wyohknott.github.io/image-formats-comparison/Bits%20per%20pixel%20at%20equivalent%20quality%20according%20to%20RGB-SSIM.svg"></p>
</body>
</html>
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=utf-8">
<meta name="generator" content="ReText 7.0.0">
<link rel="stylesheet" media="screen" href="style.css" type="text/css" />
<title>Lossy metrics</title>
</head>
<body>
<h1>Lossy metrics</h1>
<p>For each comparison algorithms, we plot the quality in dB in function of the mean bits per pixel on a logarithmic scale. We can then visualize which codec gives the best quality at a given bit per pixel (top left is better).</p>
<h2>Bits per pixel at equivalent quality according to VMAF</h2>
<p><img alt="Bits per pixel at equivalent quality according to VMAF" src="http://wyohknott.github.io/image-formats-comparison/subset1.vmaf.(openjpeg,flif,vp9,daala,jxr,bpg,mozjpeg,webp,kdu,av1-20170809">.svg)</p>
<h2>Bits per pixel at equivalent quality according to Y-PSNR-HVS-M</h2>
<p><img alt="Bits per pixel at equivalent quality according to Y-PSNR-HVS-M" src="http://wyohknott.github.io/image-formats-comparison/subset1.psnr-hvs-m.(openjpeg,flif,vp9,daala,jxr,bpg,mozjpeg,webp,kdu,av1-20170809">.svg)</p>
<h2>Bits per pixel at equivalent quality according to Y-MSSSIM</h2>
<p><img alt="Bits per pixel at equivalent quality according to Y-MSSSIM" src="http://wyohknott.github.io/image-formats-comparison/subset1.ms-ssim.(openjpeg,flif,vp9,daala,jxr,bpg,mozjpeg,webp,kdu,av1-20170809">.svg)</p>
<h2>Bits per pixel at equivalent quality according to Y-SSIM</h2>
<p><img alt="Bits per pixel at equivalent quality according to Y-SSIM" src="http://wyohknott.github.io/image-formats-comparison/subset1.y-ssim.(openjpeg,flif,vp9,daala,jxr,bpg,mozjpeg,webp,kdu,av1-20170809">.svg)</p>
<h2>Bits per pixel at equivalent quality according to RGB-SSIM</h2>
<p><img alt="Bits per pixel at equivalent quality according to RGB-SSIM" src="http://wyohknott.github.io/image-formats-comparison/subset1.rgb-ssim.(openjpeg,flif,vp9,daala,jxr,bpg,mozjpeg,webp,kdu,av1-20170809">.svg)</p>
</body>
</html>
Loading

0 comments on commit ea01bd0

Please sign in to comment.