Yesterday I wanted to revisit the TBM at New Orleans in bad weather situation which gave me a bad time the day before. Here is what I’ve tried and what I’ve been observing.
I wanted to check whether I could get better results with SteamVR, just in case. To cut a story short: no I didn’t. However I noticed something interesting. When comparing WMR/OXR 75% vs SteamVR at 76% (can’t dial 75 in the GUI), the latter was significantly sharper!
I didn’t try comparing other resolutions but it was noticeable enough to me and I really don’t think it was placebo at all. There is a video setting in SteamVR which is called “Advanced Supersample Filtering” which to me is most likely something about using a bicubic like filter instead of a bilinear, and this might be the reason. Unfortunately I didn’t spend time trying this on/off.
Anyhow since I didn’t get better fps with motion smoothing, I reverted back to WMR/OXR and I tried to further adjust the settings for the TBM at New Orleans. I’ve spent quite some time testing different approaches and settings and I’ll spare you the details, except that when comparing TAA/OXR res., although the higher the OXR the better details in the distance, the other graphical issues like line edges or “stair steps” on coastlines or small features are never really eliminated and given these, and since my test hardware can’t do 100/100, I really had to accept the limitations and consider maximizing what can be, not what can’t.
In the end, here is what I’ve been able to push on the 9700K+2070S:
1- Starting with my earlier Base settings…
... then changing only these:
||Surprisingly good enough with TBM EFIS because they are large, when paired with an adequate OXR res (see below)
||I was using Med or High depending
||This only affects the aircraft cockpit textures, not the decals, and this might have helped saving a few fps but I didn’t try High for comparison…
||…or maybe this allowed me trading cockpit texture for trees shadows instead
||This looks cool but when you’re trying to reach the threshold for motion smoothing, Off can make a difference even compared to Low.
||I was using High in the base settings. I was quite puzzled to find the image of the distant objects slightly better with Low or Off here, and this was also a good contributor to saving fps when comparing with High. Furthermore Low isn’t showing the issue with SU5 and lagging reflection as strong as with High!
2- Setting WMR/OXR Custom render scale to 70% (so much for my 75% ) with Motion Always On.
This gives 2632 x 2596 display resolution, and with the in-game TAA 70% a render resolution slightly below the native panel res.
3- Test flights
I was able to fly the TBM without any problem with the bad weather. Motion smoothing almost never failed on me. I then decided to try another aircraft, the KA350 over Los Angeles photogrammetry with a flight from Los Alamitos to Van Nuys. And again, motion smoothing almost never failed on me and looking down to the ground when flying up to about 8000ft was really showing lots of details to me.
I’ll have to try these changes with other aircraft and areas, but I honestly feel better with these than with TAA100+OXR50. However I know TAA70+OXR70 was usable here because these two aircraft are using large EFIS and they remained legible enough on the PFD, and mostly legible on the center MFD.
This is also comforting more and more my idea of prioritizing resolution over settings. This means finding a good set of base settings universally good for any hardware and high enough for good visuals, then adjusting only (mostly) the TAA/OXR ratio depending on your hardware and the type of aircraft. TAA100 for Airliners/EFIS, OXR100 for analogue gauges, and a medium mix of 70/70 (eventually 75/75) as the everyday/reference option.