Quest Pro supports click-display for eye-tracking, but how much better is performance?
If you are not familiar with the term, Eye Tracking Foveal Rendering (ETFR) is a technology where Only the area of the screen you’re currently viewing is displayed in full resolution, which releases performance since the rest is in lower resolution. This additional performance can be used to improve graphics resolution in applications or to obtain a higher base resolution.
You don’t notice the low resolution in the periphery because the human eye itself can only see high resolution in the center – the fovea. This is why you cannot read a page of text without moving your gaze. Believe it or not, this foveal region is only about 3 degrees wide.
ETFR has long been considered the “holy grail” of VR, because if your GPU only has to display 3 degrees of field of view at full resolution, performance gains could be on the order of 20x. This would allow for high-resolution screens or incredibly detailed graphics. But in reality, it requires achieving it perfect Latency-free eye tracking, absurdly high display refresh rate and a high-quality rebuild algorithm so you won’t even notice flickering and flickering.
Quest Pro is Meta’s first expedition helmet with eye tracking. The end-to-end latency of first-generation eye-tracking technology is 50ms, and the screen refresh rate maxes out at 90Hz. The click is away from 20x.
Mita headphones support fixed render foof (FFR) – render edges of a file lens Less accurately — since the Oculus Go program six years ago. In a talk to developers this week, Meta detailed the subtle performance advantages of eye tracking render foof (ETFR) and compare it to FFR.
Both types of pure rendering are enabled by developers on a per-application basis (although it is clear that both types cannot be used at the same time). Developers have three options for reducing edge resolution: Level 1, Level 2, and Level 3. With ETFR Level 1, edge is rendered 4 times fewer pixels, while at Level 3 most of the time with 16 times fewer pixels.
The exact performance benefit of click rendering also depends on the underlying resolution of the application. The higher the accuracy, the greater the savings.
In the benchmark Meta implementation, they found that at default accuracy, FFR provides between 26% and 36% of performance depending on the foveation level, while the new ETFR provides between 33% and 45%.
But at 1.5 times the default accuracy, the savings were greater, with an FFR of 34% to 43% and an ETFR of 36% to 52%. That’s a 2x boost compared to no click at all – but only a small advantage over FFR.
Of course, what really matters is what we don’t know yet: how felt Are all of these ETFR levels? And how does that compare to how clear FFR is? This is what to compare – not a certain level of FFR the same level of ETFR. It’s something we’ll test in detail for our Quest Pro review.
In Mission 2, FFR Level 1 isn’t noticeable at all, but Level 3 is definitely. And from Quest Pro lenses that are sharper both in the center and at the edges, the FFR can be sharper than ever, making ETFR even more useful.
On PlayStation VR2, the performance advantage claimed for pure viewing is greater. Sony claims the FFR provides about 60%, while the ETFR provides about 72%. This is likely due to the very different GPU architectures of the console and PC GPUs compared to mobile GPUs, as well as the higher resolution. It could also be due to differences in eye-tracking technology – Meta is in the company while Sony used Tobii.
Eye tracking on Quest Pro and PlayStation VR2 is optional for privacy reasons. But disabling it will also disable ETFR, so apps will have to fall back to FFR.