Why is Reclock not using yesgrey's latest resampler?

[Jong]

James,

What is your reason for not incorporating this (http://forum.slysoft.com/showpost.php?p=131333&postcount=430) into your build.

From my brief testing it seems good. CPU usage reported by Reclock is the same, as is the average CPU usage across all cores reported by Task Manager, but it seems to support multi-core. With Blu-ray and GPU assisted HA this gives me very nicely balanced CPU usage (circa 30% on each) even when using "excellent". The current build with Reclock is very heavy on the first core.

You probably have good reasons. Any chance you could share? Is there a problem that I have yet to spot?

[James]

James,

What is your reason for not incorporating this (http://forum.slysoft.com/showpost.php?p=131333&postcount=430) into your build.

From my brief testing it seems good. CPU usage reported by Reclock is the same, as is the average CPU usage across all cores reported by Task Manager, but it seems to support multi-core. With Blu-ray and GPU assisted HA this gives me very nicely balanced CPU usage (circa 30% on each) even when using "excellent". The current build with Reclock is very heavy on the first core.

You probably have good reasons. Any chance you could share? Is there a problem that I have yet to spot?
It does use the version from yesgrey3, but not the one compiled with ICL 10.1, as we don't have a license for this compiler.

[leeperry]

I guess he meant "why don't you include the ICL10.1 compiled DLL from yesgrey3" ?

is it like you can break DMCA in the west indies, but Intel will cum getcha if you release a compiled DLL w/o having the proper commercial licence ? :D

[Jong]

It does use the version from yesgrey3, but not the one compiled with ICL 10.1, as we don't have a license for this compiler.OK, that's clear. Thank you.

[James]

I guess he meant "why don't you include the ICL10.1 compiled DLL from yesgrey3" ?


Quite simple: I want all binaries shipped to be compiled here.

[Jong]

I guess he meant "why don't you include the ICL10.1 compiled DLL from yesgrey3" ?Actually I didn't; I genuinely thought an older version of the code may be included. It's interesting, though, the difference a compiler can make!

[leeperry]

@James : makes sense, thanks for the clarification.

@Jong : as you know, I use a lot of post-processsing in both A/V ffdshow.
the ICL10 versions are much more responsive than the generic ones.

[Jong]

I was just surprised to see a simple compiler change unlock multi-core support. My ignorance, clearly!

[leeperry]

I dunno if you ever use ffdshow, but feel free to try these builds :
http://sourceforge.net/project/showfiles.php?group_id=173941&package_id=199416&release_id=439938

[Jong]

I sure do use ffdshow, for all SD. After today's lesson I might just try them!

[Klaus_1250]

Quite simple: I want all binaries shipped to be compiled here.

I can understand that, but isn;t there someone you trust who can provide you with ICL10 binaries? AFAIK, the dfference between ICL-compiled build vs most other compilers (MS, GCC) is rather large; e.g. 5 to 30% better performance.

[Mark_A_W]

I dunno if you ever use ffdshow, but feel free to try these builds :
http://sourceforge.net/project/showfiles.php?group_id=173941&package_id=199416&release_id=439938


leeperry, do you know if the ICL10 compliled Ffdshow is the "same" as the latest tryouts version?

Do you know if it will decode TrueHD and DTS-MA (or DD+?)?

I can't keep up with the ffdshow builds anymore...

[leeperry]

leeperry, do you know if the ICL10 compliled Ffdshow is the "same" as the latest tryouts version?

Do you know if it will decode TrueHD and DTS-MA (or DD+?)?
audio builds are ICL10, all I can tell ya
http://sourceforge.net/project/showfiles.php?group_id=173941&package_id=229162&release_id=502313