MAXIMUM_FRAME_LATENCY does it still work

Thanks, great results.
Also confirms that 0 (default) is 3.
But man, your latency seems to be super bad, even at 1.

 

Remember it's recorded in slow motion.

Στάλθηκε από το m2 note μου χρησιμοποιώντας Tapatalk

 

Yeah, could be just optical illusion.

 

Not trying to hide. I know I have some lag.
It is lower than video though.

Στάλθηκε από το m2 note μου χρησιμοποιώντας Tapatalk

 

It would be great to have the same test without vsync

 

So it was not my placebo? anyway for me from 1.8 difference is less noticeable than before. Liakjim try also playing with vsync off and if you are really hardcore enthusiast with different state of Windows Hpet mode

 

Agree with the others, good job @liakjim! Would also appreciate without v-sync.

 

Thank you Jim! I'm not a huge input-lag elitist but I had the very strong impression that going from 0 to 1 significantly decreased lag with VSYNC. I call it confirmed now.
So much for the comparison.

But as for the absolute values: doesn't 27 frames delay mean 0.432 seconds at 60 FPS (~16 ms*27 frames)?!? Because that would be undriveable. In your vid it's a bit hard to tell due to the slow motion but it looks like virtual wheel is attached to real one with a rubber band.

 

Actually it's recorded at 100fps . I will try to also record a vsync off video.

 

He's counting the 100fps video footage frames, not the 60fps game output.
Also, why is Michael always such a pain in the ***?

 

Thank you for that demonstration.

I can't remember who now, but they were going on about latency and lag being potatoes.

Potatoes indeed!!

All these little improvements, do add up.

I drove enough wore out real cars and hwy tractors to know what latency in the wheel feels like.

 

Ok, here is the video , with vsync off added.



To summarize :



Results
MAXIMUM_FRAME_LATENCY=0 (38frames)
MAXIMUM_FRAME_LATENCY=1 ( 27 frames)
MAXIMUM_FRAME_LATENCY=2 ( 34 frames)
vync 0ff , fps uncapped (~103fps) : ( 21 frames)

So it seems that MAXIMUM_FRAME_LATENCY=1, it's not really bad since ~6frames are added over vsync on = 60msecs. It's not that bad.

 

So v-sync and 1 gives pretty much the same results as v-sync off. Thats really nice!

 

Michael is, but he was right about prime. And he was there for me when I thought I was in the hoosegow.

As for the footage count. The second one looked faster to me.

 

Thx for clarifying. So it's 270 ms for VSYNC and MAXIMUM_FRAME_LATENCY = 1.

I would conclude that these 60 ms over uncapped FPS (if you can get 100 FPS) can be accepted for smoothness' sake.

Still I'm impressed it's so much. A fifth to a quarter of a second is a lot.

 

The only difference I notice with v-sync is half my fps are gone.

Each system much react differently?

 

Has anyone else noticed how quick the AC staff answers when the word "Dev" comes up? It kind of makes me feel like I'm at home and my mother always knew when I was being bad.

If you guys haven't noticed, I'm new to computers and software and at the butt of some of Michaels jokes.

Are all software development teams this efficient?

 

Great I think tomorrow will check with wife's iphone... anyway crucial question here is what latency was set to last test with vsync off

Vegaguy 5555 - It's first time in my life when I see so good comunication beetwen devs/gamers - thx Kunos

 

1

Yep, it's a lot. My TV monitor seems to be awful in this area. Funny thing is that i have it already in "game" mode...

I need a 1ms @144hz monitor ASAP...

 

Input lag is not exactly connected with refresh. I'm using 60hz old Dell which input lag is about 10ms, and my TV - Toshiba has input lag 9ms as remember (120hz, but only 60zh because Hdmi limitation). Best is to check online before buy. From my side hate TN and 6bit panels - so need to wait for good 144hz IPS 21/9