Before I start I want to preface this by saying this has taken over three weeks of intermittent testing and swapping out various USB cards, so don’t light me on fire until you try some of these for yourself and determine whether they’ll help you. I’m starting from things that probably impact most people to those that impact the least, so work your way from top to bottom. Do not expect all changes to influence spell accuracy equally. The first is the biggest.
Supersampling and Reprojection Rate
So the biggest one you’re going to find is Supersampling. It was introduced quietly on SteamVR and I’m going to assume Oculus and WMR do something similar since it doesn’t cost them anything on a hardware level and it’s a “free” graphical improvement. Free as in Free Beer. There is no Free Beer.
The first settings you need to change are here:
The biggest offender is Enable Advanced Supersample Filtering. Now if you play a Warrior, Archer, Musketeer? Enjoy your shiny pretty VR. We Runemages have tracking issues if reprojection gets too high. More precision with tracking points during spell casting means smoother spells being written in the air. Try this setting and the next one. If you’re satisfied at that point, you can stop there.
(I turn off interleaved reprojection, it’s a relatively older reprojection and I find my comfort level higher with high performance settings and asynchronous reprojection only. YMMV)
The second change you need to make is here:
Now, this is relative to your HMD resolution. Set it as close as possible to your headset per-eye resolution.
HTC Vive Pro: 1400x1600 per eye
HTC Vive: 1080x1200 per eye
You can of course render lower if you wish, but you’ll be surprised what the default settings are. The supersampling performance cost is ridiculous.
I’m not sure what the settings required for the Oculus Rift or WMR headset are so if anyone uses those and feels like adding to this thread with screenshots for your HMD feel free. I found this very helpful for my spells and I’d just like others to have any easier time of it as well.
The rest of it concerns connectivity and may not be as relevant to other users. If you are comfortable with this level of performance, feel free to stop here. You have been warned, it’s going to get more techy from this point.
The simplest change you can make for the HTC Vive is rendering straight from HDMI to HDMI and not using a Micro DisplayPort adapter. I found Micro DisplayPort to be a bottleneck and had a much better time of it going with HDMI>HDMI passthrough.
A major issue that was with my personal experience was my onboard USB 3.0 onboard chipset. It was AsMedia. My experience with this USB chipset alone has been so bad as to look on future motherboards to ensure I never buy another USB chipset from this company ever again.
If you have Intel or Fresco Logic chipset you should be fine. I’ve tested both of those with the Vive and Vive Pro, no issues. I can’t speak for every single USB chipset but AsMedia WILL give you problems. I’ve found tons of complaints about high bandwidth products like Video Capture cards, VR HMDs, etc.
To find out, we need to go to the Device Manager.
On Windows 10 you can just go to the search bar and type Device Manager, run from there. In older versions of Windows go to Control Panel>Administrative Tools>Computer Management>Device Manager.
As I said, AsMedia is RIGHT OUT. If you have that, here’s a fairly cheap option that’s recommended by the Oculus Rift devs for USB issues. https://www.amazon.com/gp/product/B00B6ZCNGM/ref=ask_ql_qh_dp_hza
Someone said you may need two for both sensors since the sensors are USB. I don’t know. If an Oculus Rift user could chime in that would be great.
If you’re not sure which of your USB hubs your HMD is using, you can check here:
Then go to View, change to Devices by Connection.
If you had your HMD selected before switching views, it’ll stay on that selection and show you what USB hub your HMD is connected through.
I’ll check back here from time to time in case anyone needs help or posts instructions for Oculus Rift or WMR.