• 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.

***The Official Ray Tracing Thread - Read the First Post before posting***

Alan Wake 2 new DLC:


60fps difference at 4K between Raster and Path Tracing, that is a much smaller fps diff than what I was expecting which shows the level of optimisation remedy have made since AW2 came out it seems. Nice. Can be argued the difference indoors isn't night and day, but a game where shadows and reflections/lighting play a key part, even subtle improvements can play a big role in immersion.

If you can spare the frames, then it's well worth it for sure.

IMO Alan Wake 2 is now the gold standard for what ray/path tracing should be like in a game and all the GPU technologies seem to be implemented perfectly too... If you're on Nvidia :p
 
Last edited:
Alan Wake 2 new DLC:


60fps difference at 4K between Raster and Path Tracing, that is a much smaller fps diff than what I was expecting which shows the level of optimisation remedy have made since AW2 came out it seems. Nice. Can be argued the difference indoors isn't night and day, but a game where shadows and reflections/lighting play a key part, even subtle improvements can play a big role in immersion.

If you can spare the frames, then it's well worth it for sure.

IMO Alan Wake 2 is now the gold standard for what ray/path tracing should be like in a game and all the GPU technologies seem to be implemented perfectly too... If you're on Nvidia :p

It's not on steam this game is it? Will have to wait until it comes to steam and if you wait that long might as well wait a bit more and get it for £20 or something. So 2 years time maybe? :p
 
Last edited:
I found the new HU video on RT pretty interesting, like with everything else depends on implementation rather than just being a guaranteed level of visual effects
 
Yeah that's what some of us have been saying for so long, which is often why we also go back to the same games to show them as demonstrators on how RT should be done.
 
As a whole though, still not convinced on it if there's anvidia tax to pay to sometimes enjoy good implementation in some games.

Guess it is the future though
 
Look how well optioomised path tracing is, 40fps without any upscaling or frame gen at 4K:


And there's no mouse latency etc when you enable FG even without DLSS! It's a shame there is still some traversal stutter when moving between some rooms but the overall optiisation is awesome here and hopefully one day all games are like this.

We have modern games with no ray tracing today that can't even get 60fps and they don't have any RT!
 
  • Like
Reactions: TNA
which shows the level of optimisation remedy have made since AW2 came out it seems
No, it doesn't. It's just a matter of level selection and a lack of vegetation (which hits even harder with RT).

You can even test this out easily in CP2077, go to the Reconciliation Park in The Glen, test, then find a limited in-door area and test there (like what you'd see in AW2 LH); starting area of the Sandra Dorsett rescue is similar, still accessible after completion. You'll see a huge performance difference.
 
Vegetation makes little difference considering the frametime issues in SH2 remake exist even when later in the game, solely inside buildings just like in Alan Wake 2. In the earlier parts of AW2 and its expansions when you are outside and have vegetation, the game runs the same as it does inside too, so once again a sign of good optimisation.

I have zero issues with frametimesi n Cyberpunk regardless of where in the city I am. RED engine has other issues though, but not frametime or optimisation in its current and final state.

This issue being discussed is exclusively optimisation related and the very fact that Bloober sorted most of it with the new patch proves it.

This is exactly my observation testing this stuff back and forth between games.
 
Last edited:
Vegetation makes little difference considering the frametime issues in SH2 remake exist even when later in the game, solely inside buildings just like in Alan Wake 2. In the earlier parts of AW2 and its expansions when you are outside and have vegetation, the game runs the same as it does inside too, so once again a sign of good optimisation.

I have zero issues with frametimesi n Cyberpunk regardless of where in the city I am. RED engine has other issues though, but not frametime or optimisation in its current and final state.

This issue being discussed is exclusively optimisation related and the very fact that Bloober sorted most of it with the new patch proves it.

This is exactly my observation testing this stuff back and forth between games.




There's definitely a significant difference between the forest area and a room.

One thing I hate about AW is the stutter, even panning around. horrible.
 
Last edited:
I'm talking about the main game. It's weird as it doesn't show up in the MSI line graph as I'd expect to be.
Ah I see, I just loaded up an old save in the main game which is just the forest to test and old memories came back and I see what you mean, this was discussed before and the issue is exclusively mouse camera panning, the exact same issue existed with last of us on PC which was eventually fixed. The issue does not exist when using a controller!

I recorded a 4K 120fps video (game locked to 90fps) to demonstrate the issue for others too as it's a good reference point, the expansions do not have this issue even outside so clearly the problem has been patched in the expansions weirdly. Is this what you are seeing?

(make sure to download the video and play locally as it's 120fps then).
 
lol.....

Hm.. yes yes, interesting, yes nice presentation Epic Games.............. *Yoink*

aUdKETi.jpeg
 

Interesting video, and surprising, the average performance loss on the 4090 with RT was 29%, it was 39% on the 7900 XTX, he also made the point that RDNA 3 suffers enormously with Path Tracing, if you take to two or three path traced games out the average performance loss was only 35% to the 4090's 29%.

Why is this surprising? The slide below, on the 6900XT the performance loss for Doom Eternal was 48%, for the 7800 XT it was 39%, so clearly RDNA 3 has better RT than RDNA 2.
More over if you watch back the video what you will see is that Doom Eternal was one of the worst games for performance loss on RDNA 3, its an older game, if you pay attention you should see all of the older games had this larger performance loss on RDNA 3 vs the 4090, newer games not so much, a lot of them near the same loss as the 4090.
So it seems to me more games are now getting properly optimised for AMD and RDNA 3, with that now happening the performance loss vs Nvidia is little different.

IC1eZBJ.png
 
Last edited:
Back
Top Bottom