At the release of PSVR 2, many demos had allowed players to take an overview of the experiences offered. Each game gave a different rendering, more or less blurry. At the time, we had (through an article) mentioned this problem through the prism of reprojection but, now, we know from reliable sources that it is far from being the only one responsible.
If the reprojection undeniably brings a blurring effect in lateral movements (sony works to reduce the phenomenon), another equally important element must be taken into account: the definition!
Eh yes, whether on Meta Quest, Pico or PSVR headsets (among others), all these beautiful people are housed in the same boat : namely that the graphical quality of the games is in the hands of the developers, much more than we might think. It is much easier at first to rail against the lack of power hardware from our standalone headsets or our home consoles, but the truth may lie elsewhere. Of course, everything is relative: each manufacturer obliges each studio to respect specifications well-defined, or having to use restricted development kits. All in all, this in no way favors the graphical gain of the experiences offered. Not to mention the studios that have neither the money, nor the time, nor the desire to optimize their games with little onions.
If it is quite easy to find out what definition the games run on our standalone headsets and in PCVR, it is very difficult to do so on PlayStation VR 2. That’s why a VR enthusiastresponding to the name of Techlogi, rolled up his sleeves to find a way to achieve this. L‘user Reddit was therefore based on an old video of Digital Foundry, using in particular a reliable method to precisely count the number of pixels of a game:
To arrive at his ends, Techlogi used the S buttonhare PSVR 2 Sense to take screenshots, these are then recovered by a USB key to avoid any compression effect. By displaying them on his computer, he can now count the number of pixels on the height and length, without forgetting this little detail:
The pixel counting technique is quite valid, provided you count the pixels inside the fovea. If there is no eye-tracked foveal rendering, pixel counting should be done at the center of the screen.
The author of the approach says it himself, ” his solution is not 100% sure “, but sufficient to get an idea. Last Saturday, this one published on Reddit a list of PSVR 2 games with their definitions:
If Red Matter 2, Song in the Smoke, Resident Evil 8, Gran Turismo 7 are doing well (in real 4k!), other games (Horizon Call of the Mountain, Beat Saber, Synapse) oscillate between 63 and 66 % of the headset’s native definition, when not completely below (Pavlov, Green Hell VR, No Man’s Sky). If gamers have long criticized the game’s blur effect no man’s skythe graph shows whose fault it is, the same for Green Hell VR who, not content to afford the luxury of being a port of the version Quest (while a PCVR version exists), offers a totally indecent definition, well out of step with the power of PS5.
You will have noticed, Puzzling Places as well as Pavlov are in a position that does them no credit, and yet the two get away with it rather well once the helmet is screwed on their head. It’s normal, the first does not use a background so the player does not have the possibility of noticing the weak definition of the title, as for the second, it uses a system ofanti aliasing quite powerful, allowing to hit the bull’s eye.
What to conclude? This list allows us to get an idea, but at no time is it there to replace a proper test. The foveal rendering, theanti aliasingL’upscalingso many tools that can allow, when the definition is not too low, to save the furniture or create an illusion. However, let’s hope that this awareness pushes the studios to make an effort in this area.and who knows maybe that by adding a “native 4K” logo, gamers might be more easily tempted.