• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Fidelity Super Resolution in 2021

ok same as the leaked games dont own any to test but someone will on youtube today

The following titles will support AMD FSR Driver at launch (June 22):

  1. 22 Racing Series
  2. Anno 1800
  3. Evil Genius 2
  4. Godfall
  5. KingShunt
  6. Terminator Resistance
  7. The Riftbraker
The following titles have AMD FSR Driver support coming soon after :

  1. Asterigos
  2. Baldur's Gate III
  3. DOTA 2
  4. Edge of Eternity
  5. FarCry 6
  6. Farming Simulator 22
  7. Forspoken
  8. Myst
  9. Necromunda: Hired Gun
  10. Resident Evil Villiage
  11. Swordsman Remake
  12. Vampire The Masquerade: Bloodhunt
 
FSR is now live and can be tested
And it looks amazing even on ultra performance :)
But you need to have the games patched to be able to test FSR - it will be an in game option that can also be used on Nvidia cards, probably with less perf without support from Nvidia.
ok same as the leaked games dont own any to test but someone will on youtube today

Add Star Citizen too, it will support FSR as soon as it becomes gold. :)
The games will probably be updated today to include FSR.
 
Does any of them games actually need fsr to play ok at 1440p or is it only 4k users that should fsr apply to? I dont see any of the main games it needs fsr like watchdogs, rdr2, gtav with mods, cyberpunk, witcher 3 with mods probably etc? Big aaa open world games are what really could use fsr at 4k or 1440p even for old computers.

Its a shame they dont use fsr for old consoles like ps2 or xbox 360 with bios updates if its not really gpu dependant and then have those consoles use a upscaler like virtual super res to play old games at higher res without perf loss. Yes u could just use emulators but i bet most would like real hardware versions instead. Would let ppl enjoy old games again.
 
Does any of them games actually need fsr to play ok at 1440p or is it only 4k users that should fsr apply to? I dont see any of the main games it needs fsr like watchdogs, rdr2, gtav with mods, cyberpunk, witcher 3 with mods probably etc? Big aaa open world games are what really could use fsr at 4k or 1440p even for old computers.
I think FSR needs to be proven before it will become widely adopted. AMD does not have Nvidia's resources and even Nvidia was unable for a lot time to put DLSS in a lot of games. It is a good start for AMD, if there is some quality in FSR then we will see it in a lot more games.
 
And it looks amazing even on ultra performance :)
But you need to have the games patched to be able to test FSR - it will be an in game option that can also be used on Nvidia cards, probably with less perf without support from Nvidia.


Add Star Citizen too, it will support FSR as soon as it becomes gold. :)
The games will probably be updated today to include FSR.

Have a friend who is into star citizen on a vega card atm
 
Let's think for a second: i see you didn't like when i called ghosting a huge problem so let's say DLSS had this small problem in some games, even if, if you go and read on Reddit you see a lot of people claiming that changing the DLSS file brings a big improvement to IQ. But let's call it small problem.

The problem is if you claim DLSS 2.1 has a huge problem then you have set up a false narrative and everything else you claim is immediately suspect and you are lead to false conclusions. There is some ghosting in DLSS2.1, most people don;t mind as the performance gain for the image quality is worth the trade off. If you think there was a huge problem that has suddenly dispeared you are false thinking there as a large and fundamental change going from 2.1 to 2.2 when the relaity is there was a small but welcome improvement.

Objectively, the ghosting visible in DLSS2.1 was equal or less than the ghosting seen on most TAA implementations

So DLSS had this problem in CP for example for 7 months. An update comes for another game, people were taking the dlss file and put it inside CP folder and it solves most of the problem with ghosting in CP. Not only there but also in DS and who knows what other games.
What is changed? There were no updates for CP or DS, no new "motion vectors" to be feed into the neural network.
This is precisbley ht epoint, the games were ot changed so the only improvement can be an improvement in the DLSS algorithm. Some people thought the chnage to DS was due to fixing mtion vectors which was a known problem, but this isn't the case.

The most plausible idea is that the temporal data was in part dumped. Especially when we consider the timing of the update.
No, you just pulled that out of your backside. There is zero evidence for that, and logically it makes no sense what so ever because the main improvement going from DLSS 1 to 2 was including temporal accumulation because that is the only way to gain more information. What is actaully happening is the algorithm for temporal re-projection has been improved, and possibly the DL model has been improved, or both.

I don't think it is a bad thing if they have inspired from UE5 and/or FSR but i don't understand why do you put so much faith in the neural network when there are simpler explanations?

See, now you don;t make any sense what so ever. The TSR in UE5 is purely temporal, no spatial scaling at all. SO this is a complete contradiction to your statement that Nvidia dropped temporal accumulation. If you don't understand how these technologies work then you shouldn't be making such wild claims.

Your 2nd blaringly obvious contradiction is you appear not to understand in the power of neural networks, which is what is doing the spatial upscaling without any temporal data, yet you claimed that Nvidia must have dropepd the temporal accumulation and must onyl be doing spatial. So which is it in your opinion?

The relaity is, DLSS 2.x uses a combination of both temporal accumulation and DL based image reconstruction for image upscaling. THe great thing about DLSS 2.x is that it combines the strengths of both approaches and can mitigate the weaknesses of each.

How come the DLSS was unable to recognize "distortions in an image given the motion vector" for 7 months ( or more if we talk about DS ) and suddenly now it can do that? :)

Where is your evidence that the DL didn't (partially) correct for distortions in the image due to temporal re-projection? You again are making a false claim. Furthermore, it is not necessary that the DL model is improved, but the temporal accumulation is improved, or both .
 
Does any of them games actually need fsr to play ok at 1440p or is it only 4k users that should fsr apply to? I dont see any of the main games it needs fsr like watchdogs, rdr2, gtav with mods, cyberpunk, witcher 3 with mods probably etc? Big aaa open world games are what really could use fsr at 4k or 1440p even for old computers.

Its a shame they dont use fsr for old consoles like ps2 or xbox 360 with bios updates if its not really gpu dependant and then have those consoles use a upscaler like virtual super res to play old games at higher res without perf loss. Yes u could just use emulators but i bet most would like real hardware versions instead. Would let ppl enjoy old games again.


you are right and the big AAA first person shooters
 
It is possible that previously DLSS wasn't making as efficient use of the motion data the game was presenting it with compared to with the new files. DLSS is a work in progress so it isn't that noteworthy to see changes - there is at least one data set they've not released to the public which seems to be optimised for ray tracing titles and probably more stuff beside.


While projecting picel data based on the motion vector is intuitive at a hgh level and doesn;t need DL or any non-linear system, the reality is the motion vectors are incomplete and can be sparse. Dealing with edge cases, and optimizing how to integrate data from earlier frames whne the motion ectors have a lomger magnitude due to the increased time window leaves many choices in algorithm.

This temporal accumulation is just an extension of TAA. Just as there are many implementations of TAA, some betetr and other worse, the temporal upscaling can be done in different ways to get different results with different trade offs.
 
FSR is now live and can be tested

however, interestingly none of the games with FSR also have DLSS. Coincidence? Maybe I can go with that, until I saw the AMD review guidelines

AMD has sent out these guidelines to reviewers to ensure it doesn't get compared to DLSS, hopefully those sites call out AMD for it

5318453c8001f3d16577600a2f95b772e863787597bbf409be0debfc73ee1bf3.png

Straight form the horses mouth. FSR should not be compared to DLSS and it is obvious why if there is no temporal accumulation and no DL based spatial upscaling.

Will all the posters claiming that FSR will kill DLSS take back their comments when even AMD says they are not comparable?
 
While projecting picel data based on the motion vector is intuitive at a hgh level and doesn;t need DL or any non-linear system, the reality is the motion vectors are incomplete and can be sparse. Dealing with edge cases, and optimizing how to integrate data from earlier frames whne the motion ectors have a lomger magnitude due to the increased time window leaves many choices in algorithm.

This temporal accumulation is just an extension of TAA. Just as there are many implementations of TAA, some betetr and other worse, the temporal upscaling can be done in different ways to get different results with different trade offs.

why thank you Mr Data

you lost me on all of that :) I feel like Patrick Stewart listening to a star trek script

lets just hope this one works and the other one runs with no problems and all the willy waggling ends aboud who's car is bigger
 
While projecting picel data based on the motion vector is intuitive at a hgh level and doesn;t need DL or any non-linear system, the reality is the motion vectors are incomplete and can be sparse. Dealing with edge cases, and optimizing how to integrate data from earlier frames whne the motion ectors have a lomger magnitude due to the increased time window leaves many choices in algorithm.

This temporal accumulation is just an extension of TAA. Just as there are many implementations of TAA, some betetr and other worse, the temporal upscaling can be done in different ways to get different results with different trade offs.
It could have been done 5 months ago. Or one month ago on Metro Exodus. It was not that hard to notice the ghosting and a lot of people were reporting it.
 
From the review rules i bet a lot of reviewers are going "wait so i cant have any split screens showing fsr on the left and dlss on the right with afterburner stats". If thats true, then, reviews be limited as i bet most of em wanted to directly compare em head to head.
 
From the review rules i bet a lot of reviewers are going "wait so i cant have any split screens showing fsr on the left and dlss on the right with afterburner stats". If thats true, then, reviews be limited as i bet most of em wanted to directly compare em head to head.

And this is why I hate youtube reviews

nothing beats a good written article
 
From the review rules i bet a lot of reviewers are going "wait so i cant have any split screens showing fsr on the left and dlss on the right with afterburner stats". If thats true, then, reviews be limited as i bet most of em wanted to directly compare em head to head.
But how can you do that if there are no games with both features? I think that is a reasonable request, to not compare the FSR from Godfall with the DLSS from Metro Exodus but as usual, Nvidia cheerleaders will complain about anything..
 
FSR is now live and can be tested

however, interestingly none of the games with FSR also have DLSS. Coincidence? Maybe I can go with that, until I saw the AMD review guidelines

AMD has sent out these guidelines to reviewers to ensure it doesn't get compared to DLSS, hopefully those sites call out AMD for it

5318453c8001f3d16577600a2f95b772e863787597bbf409be0debfc73ee1bf3.png

Does it really matter if it is or isn't? Don't be losing sleep over it.
The most important goal here is does it work what it's set out to do.

Will it be perfect? Definitely not for me I will be using the ultra option if I can get back some performance from 4k while still looking half decent I'll be happy with that.

The biggest killer for me will be if its easy noticeable while playing for example lots of glitches or increase blur.

But if I need to stop and take pictures to see the downside then in my opinion it's doing well.

Far cry 6 will be my game for testing, still have resident evil village installed.
 
But how can you do that if there are no games with both features? I think that is a reasonable request, to not compare the FSR from Godfall with the DLSS from Metro Exodus but as usual, Nvidia cheerleaders will complain about anything..

I dunno, it depends if a game did have both, im not sure if any do as of yet given the game list, tho one day some will have both features and reviewers wil probably compare em then. For day 1 it doesnt look like they can tho.


I wonder if we can test a game for free to check fsr out, like maybe the res evil village demo might get a update so we can test that without spending money to "try" fsr to see what the fuss is about.
 
Back
Top Bottom