I did a lot of performance related testing lately and have come across a very curious thing. The Pitool generated resolutions seem to be all over the place and not necessarily simply a factor of FOV.
Basically, the reason enabling Parallel Projections (PP) for compatibility seems to have a performance impact is not just because they need a different render path, but also the resolution goes up. Also, this explains why PP on large FOV looked blurry to some - there is probably an incorrect configuration in Pitool for large FOV!
I tested all the different combinations and uploaded the screenshots here.
I also created a summary table to show the various resolutions, along with a calculated increase in render load for the GPU, as well as a perceived sharpness value (based on vertical resolution).
Btw, I strongly recommend keeping Pitool quality level at 1.0 and simply setting supersampling per app in SteamVR - makes for a much more controlled environment and you don’t need to reboot SteamVR every time you change it.
Can you please confirm my findings? Might be something wrong with my installation of Pitool.
I am really interested to know if the PiTool supersampling does anything else or is any way different to the SteamVR SS setting. Clearly it is more easily managed in SteamVR where it doesn’t require restarts and is also per game, but then why would PiTool contain a setting for it.
Yep I remember that. The problem is with that answer is that we, the users of this tool, need to know which setting we are meant to be changing and is likely to give us the best results. We don’t need to know the ins and outs of the algorithm necessarily but we do need recommendations on what to do with SS and the multitude of ways we have of setting it currently. It isn’t OK to give us nothing.
@SweViver, this sounds like a great idea for a video. This info really should be provided by Pimax, but I think we will need to depend on you for this.
I know you’ve done some of this before, but the drivers seem to have changed a lot. Pick 1-3 popular games (maybe only 1 per video) and determine and explain to us what the various settings seem to interact and what settings you would recommend based on your laptop and main systems. The idea is to provide a starting point and tweaking guide.
P.S. If you decide to test Elite Dangerous as one of the games, just do the “take off and landing” tutorial and do a “return to surface”. That will display the interior of a station, one of the most demanding areas (graphically).
I feel like it has to be objective as possible as well, which is going to be immensely difficult. But I guess if the impact of final render resolutions is minimal as is the performance difference then it doesn’t matter if one is slightly better than the other. If there is an important difference between a higher PiTool setting and lower SS with similar final resolutions then it ought to be obvious from an FPS point of view or an image quality point of view.
