05.02.2025, 19:27
No clue, never really did much high speed QTGMC benchmarking, since when I filter, I'm usually way below real-time encoding.
That said using tons of threads increase RAM usage and isn't always faster. Sometime changing max RAM usage and lowering threads might result in faster encodes. (more threads => more RAM and more synchronization work needed, which can slow things down)
Also depending on your content, not using GPU filters in QTGMC might improve or hurt speed.
For example in Vapoursynth using DFTTest2 with DFTTest instead of QTGMC for me is usually slower than using the CPU based for SD resolution, but faster for HD and higher resolution.
Especially when using Avisynth, check your RAM usage, thread count, and test different memory and thread settings.
Cu Selur
That said using tons of threads increase RAM usage and isn't always faster. Sometime changing max RAM usage and lowering threads might result in faster encodes. (more threads => more RAM and more synchronization work needed, which can slow things down)
Also depending on your content, not using GPU filters in QTGMC might improve or hurt speed.
For example in Vapoursynth using DFTTest2 with DFTTest instead of QTGMC for me is usually slower than using the CPU based for SD resolution, but faster for HD and higher resolution.
Especially when using Avisynth, check your RAM usage, thread count, and test different memory and thread settings.
Cu Selur
----
Dev versions are in the 'experimental'-folder of my GoogleDrive, which is linked on the download page.
Dev versions are in the 'experimental'-folder of my GoogleDrive, which is linked on the download page.