http://ffdshow-tryout.sourceforge.net/wiki/ 2009-12-05T01:22:46+00:00 http://ffdshow-tryout.sourceforge.net/wiki/ http://ffdshow-tryout.sourceforge.net/wiki/lib/images/favicon.ico text/html 2009-08-09T13:42:56+00:00 http://ffdshow-tryout.sourceforge.net/wiki/faq:builds?rev=1249825376&do=diff The main difference is the compiler that is used for compiling ffdshow.ax and thereby resulting CPU requirements: Build Bit Compiler Requirements Generic 32 Microsoft Visual C++ i686-compatible CPU like Pentium Pro/Athlon or higher 64 Microsoft Visual C++ 64-bit CPU and 64-bit Windows ICL10 32 Intel C++ Compiler SSE-capable CPU like Pentium III/Athlon XP or higher text/html 2009-03-30T08:52:52+00:00 http://ffdshow-tryout.sourceforge.net/wiki/faq:multithreading?rev=1238403172&do=diff The following list gives an overview of parts in ffdshow that are multithreaded and thus show an improvement when run on a multi-core CPU: * libavcodec's MPEG1, MPEG2, and H.264 decoders are multithreaded. The number of cores is automatically detected during ffdshow setup but can be configured under “Decoder options” -> “Number of decoding threads”. This setting has no effect on Pentium 4 HT CPUs. * ffmpeg-mt is the multithreaded version of libavcodec and can be used for H.264 decoding. … text/html 2009-03-26T11:28:39+00:00 http://ffdshow-tryout.sourceforge.net/wiki/faq:performance?rev=1238066919&do=diff All builds have in common that the GCC compiler is used for building libavcodec, which is ffdshow's main library for decoding. This library contains a lot of hand-optimized assembly code (which auto-detects the CPU's supported feature and instruction sets). Using additional compiler optimizations or advanced instruction sets (such as SSE/SSE2) doesn't result in any significant or noticeable performance increases. So when comparing pure decoding performance, there is almost no difference between … text/html 2009-03-26T11:51:53+00:00 http://ffdshow-tryout.sourceforge.net/wiki/faq:performance_with_filters?rev=1238068313&do=diff Optimized builds of ffdshow.ax can have an impact on performance when certain filters are used in ffdshow. Below is a list of - so far tested - filters that benefit from optimizations: * xsharpen * denoise3d The following chart illustrates the possible increase in decoding performance when using optimized builds and filters: text/html 2009-06-23T13:23:32+00:00 http://ffdshow-tryout.sourceforge.net/wiki/faq:releases?rev=1245763412&do=diff SVN builds are “bleeding edge” which means they contain the latest features, updates, and - unfortunately - bugs. They are usually released a few times per week. Official releases are - simply put - “based” on these SVN builds. When an acceptable feature and stability level is reached during development, one SVN version is “destined” to become the next official release. Its development continues on another branch independently from the current SVN trunk. Adding new features is stopped (so called… text/html 2009-05-18T21:08:44+00:00 http://ffdshow-tryout.sourceforge.net/wiki/faq:stable?rev=1242680924&do=diff * Official releases are the most stable builds and generally recommended over any other build. * 32-bit generic/ICL10 SVN builds are considered as relatively stable but they don't receive the same thorough testing as official releases. * 64-bit SVN builds are currently in an “experimental” state, thus unstable. For that reason, 32-bit builds are recommended instead. text/html 2009-03-26T11:52:57+00:00 http://ffdshow-tryout.sourceforge.net/wiki/faq:win9x?rev=1238068377&do=diff ANSI and thus Windows 98/ME support have been dropped starting with build 2353. The last officially supported build is 2322 and can be downloaded here.