As mentioned in the Pimax thread, over the weekend I tested Elite Dangerous on my shiny new Pimax 5K+ in a variety of environments, FoV, supersampling and graphical detail levels.
Specs
My machine has a 6700K @4.5 Ghz, 16GB RAM and 1080ti. The GPU is not overclocked.
No other program was running other than PiTool, SteamVR, Elite, a browser window with 3 tabs open (Brave) and the software needed for my joysticks.
Testing methodology
Tests were performed in the LTT 4487 system, flying an Eagle MkII, except the Space Station one that was done in Shinrarta Dezra, in Solo Mode to eliminate network-related performance issues. PiTool was set to have Parallel Projections ON, Normal Brightness, Rendering Quality 1.0.
All results were taken using FpsVR. Since no Reset Average functionality is available, I had to estimate the average values.
Results
The results are here:
https://docs.google.com/spreadsheets/d/1Ui0A6RUNZ13LqJsUg-4JiVsqUvwvaJyiohRKSKTtL1M/edit#gid=0
My comments are as follows:
Specs
My machine has a 6700K @4.5 Ghz, 16GB RAM and 1080ti. The GPU is not overclocked.
No other program was running other than PiTool, SteamVR, Elite, a browser window with 3 tabs open (Brave) and the software needed for my joysticks.
Testing methodology
Tests were performed in the LTT 4487 system, flying an Eagle MkII, except the Space Station one that was done in Shinrarta Dezra, in Solo Mode to eliminate network-related performance issues. PiTool was set to have Parallel Projections ON, Normal Brightness, Rendering Quality 1.0.
- for the Supercruise test I circled around the main star at a distance of about 500ls;
- for the Ice Rings test I boosted around inside the icy rings of LTT 4487 3;
- for the Planetside test I landed on LTT 4487 3 a on a random place and drove around with my SRV; of course, areas with different geometric complexity could be more or less taxing.
- for the Nav Beacon test I flew around on the Compromised Nav Beacon of LTT 4487; this spawned different amounts and types of ship that may have affected test results;
- for the Space Station test inside I stood on a landing pad in Jameson Memorial; the pad was a different one each time due to the requirement of restarting the game under every one of the major test conditions, that may have affected test results. For the test outside I drifted around the station at a distance of about 1Km.
All results were taken using FpsVR. Since no Reset Average functionality is available, I had to estimate the average values.
Results
The results are here:
https://docs.google.com/spreadsheets/d/1Ui0A6RUNZ13LqJsUg-4JiVsqUvwvaJyiohRKSKTtL1M/edit#gid=0
My comments are as follows:
- As noted by Pimax Forum user Milopapa, the performance for Large Fov is oddly out of place, especially considering the insane rendering resolution. It seems that PiTool massively downsamples the scene in Large Fov in order to make rendering bearable; this is immediately apparent simply by looking at the text in the HUD, which is significantly less sharp. It would also seem that Steam SS is completely ignored in this case, as changing it between 72% and 100% produced no appreciable difference.
- Any value above 45 FPS is very much enjoyable, it seems some degree of Reprojection is already implemented. Values below 45 are noticeably less fluid, but not bad enough to give me VR sickness, although by now I am largely immune to it.
- Because of Vsync, values above 90 FPS appear as 90.
- CPU was never a limiting factor, as frame times never exceeded the 11ms threshold even during spikes; average value was ~4ms under most circumstances.
- After trying them all out, my setting of choice is likely going to be Normal Fov, 72% Steam SS, VR Medium.
Last edited:
