MPEG-2 source was used for because it represents real-world usage for the purpose of the comparison that was being done. The developers of all codecs of course knew the purpose and significance of the comparison when they submit their settings for it (and I believe one of the goals in the comparison is to achieve a certain reasonable encoding speed, to reflect real-world usage).Wheee_It's_Me! wrote:Yes but the developers are basing the settings on an all around average of encoding speed, quality, compression. Furthermore, I wonder how many of those developers would suggest OTHER settings if they knew they were being used in such a comparison. Again though, the point I made was about compression/quality capability where as that Doom9 comparison is an "all around" comparison and largely subjective (ie based on his personal observations rather than using software to grade the differences between frames). Further, the sources they used were already compressed into MPEG-2 which would have a significant factor on many of those codecs when it comes to yet a THIRD encoding. The best means of testing quality and capability would be to create CLEAN rendered animation sequences of various complexity, outputting in uncompressed RGB. And then rather than rely on human visual testing use a program to actually compare the differences on the pixel level and give a TRULY accurate representation of the differences.Zarxrax wrote:Doom9 wrote:All codecs were tested in a 2 pass setup using the settings suggested by the developers.
VP7 may not have been using the very best settings, but neither were the other codecs. In particular, x264 was using some quite fast settings as well, and could have seen probably a significant gain by using much slower settings.
In addition, x264 has been updated often since the comparison originally took place, and has improved noticeably. VP7 has since only received one or two very minor updates.
But as it stands, your original assertion was that VP6 beats H264. On2 themselves make the claim that VP7 is 50% more efficient than VP6 was. If I were to be gracious and give you that VP7 is equivalent in quality to x264, then by extension, it shows that VP6 is clearly worse than x264.