正在加载视频...

视频加载失败

CS2 end to end input latency results for Hardware: Legacy Flip vs Hardware: Independent Flip in stressed scenario with donk settings + reflex +fps_max 0. Legacy flip: 7.7 ms Independent flip: 7.8ms More data in comments. The FPS testing results were very interesing (soon)

14,524 次观看 • 1 个月前 •via X (Twitter)

12 条评论

CS2 Kitchen 的头像
CS2 Kitchen1 个月前

1000 samples each. Now dfso leads to legacy flip without registry tweaks.

JOCO 的头像
JOCO1 个月前

Would also be interesting to see a comparison between composed flip and independent flip on a multi-monitor setup 🤔

Murzynator 的头像
Murzynator1 个月前

BRO TURN OFF CHROME WHEN YOU TESTING SOMETHING!!!!!!!!!!!

CS2 Kitchen 的头像
CS2 Kitchen1 个月前

no effect on performance

Andrilaz 的头像
Andrilaz1 个月前

There will be a test of Fullscreen (Legacy flip) vs Windowed fullscreen(IFlip)?

Михаил 的头像
Михаил1 个月前

you best🙏🙏🙏🙏🙏🙏

CS2 Kitchen 的头像
CS2 Kitchen1 个月前

I appreciate your kind words

luke o äiä 的头像
luke o äiä1 个月前

ZOMG!!

Tulantro 的头像
Tulantro1 个月前

Shouldn't matter much for this specific comparison, but having the python GUI loading while testing instead of post-data collection, vs code (chromium) and chrome open, steam UI open, dual monitor running, this is just a poor testing setup. Good to know the results are within margin of error

CS2 Kitchen 的头像
CS2 Kitchen1 个月前

The result on the screen is updated after an ADC read. It doesn’t interfere between t press -> threshold cross. Regarding chrome, if you are not doing anything heavy, that also has zero affect on measurement. Thirdly, the recording is only done for visuals, the real testing is done again during validation testing.

Tulantro 的头像
Tulantro1 个月前

Fair enough

CS2 Kitchen 的头像
CS2 Kitchen1 个月前

no issues, valid concern

相关视频