Anyone could share the acceptable latency for smooth flight basing on your experince?

[kerlcat]

Managed the setting to scure FPS ~50-55... looks great other than ghosting which I wish MT stable version might solve someday.

But the latency seems pretty high ~50-70ms via VD -ultra mode -H.264 -no ASW/buffer, even each of "game" "encoding" "decoding" "networking" all close or under 10ms. Yes wireless via dedicated band @stable 1200Mbps as it's so conveniance v.s. USB tethering. 

I know it would be pefect if latency less than 20ms for VR experince. 

Wonder what's acceptable latency for smooth flight, before it's too laggy to enjoy. Appreciate if anyone should share your experince on this. 

Currently using Pico4 which works good for me comparing with Rift S. Definitely is still looking for next gen PCVRset w/ DP connection so this topic won't bother me any more. 

Edited April 24, 2023 by kerlcat

[Qcumber]

A few things I have noticed but not really measured properly is that the overall latency appears to increase by about 10ms when I enable the "enhanced contrast setting" in VD. I think openxr toolkit might add a few ms on too. The latency is about 45ms at altitude, 55ms at med altitude and about 60-70 at treetop level. This is on the Normandy 2 map and does not include London or Paris. The FPS generally maintains around 45-55 at high setting on VD. 

Pico 4; 5600x; rx6700

 

Edited April 24, 2023 by Qcumber

[kerlcat]

@Qcumber thanks for your sharing. I have 60-70 at high altitude while total those 4 items add up to only ~35. Wonder where those additional latency comes from so I might address with focus. When I set cloud to ultra, saw 6-8ms of increase meanwhile fps dropped by 5-7 fps.

Let me try without “enhanced contrast setting” today to see if the image quality still acceptable.

Edited April 24, 2023 by kerlcat

[Qcumber]

8 hours ago, kerlcat said:

I have 60-70 at high altitude while total those 4 items add up to only ~35

VD say that the individual latencies don't always add up. There must be some hidden latency somewhere which might be accounted for by additional "layers" such as openxr toolkit etc. I'm not sure if this is a correct assumption so would appreciate the opinion of someone with more technical expertise.