Avatar: Frontiers of Pandora PC vs PS5 Graphics, Optimised Settings, FSR 3 Breakdown

Ubisoft's menus how you see what your settings are doing, should be a standard in PC games. This is a beautiful looking game so seeing what you can get with the right settings which do not have a big impact on performance is very helpful.

Devs need to have a console settings preset like Horizon/God of War did. Makes it easy to tweak and compare.

Still just shocked to see Ubisoft putting out such an amazing PC port. Massive is easily their best studio.

Good to see Massive's PC developer roots are still there. Wish we could get another World In Conflict though.

I bought this game because of Alex's first video on the graphics and his emphasis on the explorer mode and crysis style gameplay. First Ubi game I've been happy with in a while.

I think what the video is missing is an explanantion of the "Fixed" and "Biased" Upscaling modes. Usually you just change the resolution for DLSS and FSR with the Quality modes (Qualiy, Balanced, Performance, etc), but the "Biased" scaling mode is an additional scaling modifier that depends on the output resolution and is also inconsistent. I need a chart to fully explain it. But summarized, the internal resolution in "Biased" mode is a lot higher than the usual DLSS and FSR screen percentages. For example Biased-Performance is higher res than Fixed-Quality. At least for 1080p and 1440p. At 2160p there is no difference between Fixed and Biased. Also no combination seems to drop below 720p internal resolution. So at 1080p, DLSS Ultra Performance has the same 720p internal resolution as DLSS Quality mode. This probably leads to many users thinking the game performs worse than it does, because "DLSS Performance" doesn't give you nearly as much extra fps in the biased mode as it usually does, and the users don't know that it's actually running at a way higher resolution.

Frohe Weihnachten Alex & dem gesamten DF-Team! Ihr macht einen super Job!

Great work Alex, thanks a bunch for these settings and side-by-side comparisons :)

I'm actually surprised how well the software RT holds up visually, compared to how it performs

I love this games setting page it shows good pictures to show what each setting is and has alot of options. I really like the audio settings too because it plays a sound effect when adjusting the levels so you don’t have to guess is voice or music is how loud you want it.

I am playing on PS5 on Performance mode, on a oled television. It looks absolutely stunning

[20:13] Thanks for testing, analyzing and showing the PC optimized settings. This is going to save a lot of time for so many people.

Picked up Avatar Frontiers of Pandora on PS5 and man the graphics are absolutely stunning! It’s the first game since Horizon Forbidden West that has blown me, the world of Pandora is so lush, the natural environment, lighting is top notch. Flying around on the Ikran is a treat, Far Cry Primal is one of my all time favorite games and Avatar Frontiers of Pandora captures the vibe of that game perfectly.

Great and in-depth analysis from Alex as always . 👍

FSR 3 works flawlessly with my 7900 xtx and variable refresh. I think not using an AMD 7000 series is an issue when evaluating an AMD tech.

Kudos to Massive for the quality work.

I checked like a dragon gaiden with fsr3, it has a similar implementation but i noticed that if u cap your frame rate at something less than your max refresh rate(and hit that target) but within the vrr window u do get smooth frame times. I had jutter with v sync so i capped the frame rate 1 fps below the max refresh and it worked perfectly with fsr3, straight line on the frame times.

Awesome video Alex! Really detailed as usual but a quick thing I noticed: for the judder with FSR3 & VRR there is a simple fix. Just limit the fps to slightly under the max refresh of the screen and the pb is gone. I know you got a lof of work, and this video was surely a lot of work as well since you went through so much stuff, but you can't spend a minute telling us that you can fix DLSS by doing "x and y" and not investigating if FSR3 VRR problem can be solve. It's not optimal of course I know, but the workaround is not the worst thing either.

Solid breakdown thanks Alex!

This is such a great, in depth and informative video. Well done Alex. Great job.