Foveated Rendering Documentation

Hi all,

One of the things that makes Pimax special is the combination of performance and visual quality made possible by dynamic foveated rendering. However, tweaking the detailed settings is difficult without more specific documentation on what each field actually does. I’ve attached two screenshots (Fine mode and Quick mode with Custom preset) showing the settings I’m referring to.

Could someone point me to the latest documentation covering these fields in detail? A few specific questions in the meantime:

Preset values: When using one of the default presets (Performance, Balanced, Quality, Extreme), it would be helpful to be able to see what underlying values those map to in the Custom tab — right now switching presets doesn’t seem to reveal the actual numbers being applied.

Transition modes: Documentation on the Transition Mode options (Off, Linear, Alpha) and what Transition Range / Alpha actually control would be appreciated.

Supersampling pipeline: I assume setting Center Resolution above 100% is the same as supersampling (vs variable rate shader clustering for much of DFR). What’s the actual pipeline for how this interacts with the general render quality / resolution setting (e.g. the global render scale elsewhere in Pimax Play or SteamVR)? Are the two multiplicative/additive, or does Center Resolution operate independently on top of whatever the base render resolution is set to?

Thanks in advance for any pointers to docs or clarification on the above.

2 Likes

sadly to most of pimax users knowledge, there is no public documentation for anything. its been all trial and error for all users and us helping each other in the discord server(s). (i hate how many there are)

QV in specific has always been a wild one, especially with the presets. we got no clue how it truly works and why it behaves so differently compared to Mbucchia Quadviews Foveated Rendering + Tally Mouse QV Companion which this is supposedly based off from when it was first implemented yet looks worse and performs worse.

I was so happy to figure that all settings up to now except “fine tune” was showing horisontal and vertical size of foveated box.

Only fine tune shows opposite; value from edge of whole view to the foveated box.

This is not a problem, just not consistent…