Digital Foundry Article Technical Discussion [2024]

I swear to god I don't have this issue with Horizon FW. It's perfectly smooth when I rotate the camera.

I haven't bought Ghost of Tsushima yet since I've got so many other things I want to finish playing first.. but I have no reason to expect it wont be the same thing. Maybe it's a 60hz display issue?

Possibly, as that's all I have. If you're bored you could set your display to 60hz sometime and slowly pan with a gamepad to see if you get it. :)

edit: Checked with latest patch. Still there.
 
Last edited:
@Dictator What about load times? From what little I could find, it seems to be much faster on PS5. PC seems to take 3-6 seconds depending on the SSD but the PS5 takes about 1.5 seconds. It doesn't seem possible to match it on PC because DirectStorage isn't in use and even if it were, it doesn't seem to provide many tangible benefits so far.
 
Possibly, as that's all I have. If you're bored you could set your display to 60hz sometime and slowly pan with a gamepad to see if you get it. :)
You might have a point there ... that the game is handling chunks of data or whatever based on an arbitrary period with console specs.

And when it runs on PC you get edit: jank.
 
Last edited:
Great video @Dictator. Another example of pointless ultra settings. Those were some of the best shots of GoT that I’ve seen. It’s a really good looking game. The quality is inconsistent though. Some scenes and characters look very unimpressive.
 
Great video @Dictator. Another example of pointless ultra settings. Those were some of the best shots of GoT that I’ve seen. It’s a really good looking game. The quality is inconsistent though. Some scenes and characters look very unimpressive.
It was designed to run on essentially mid-range hardware from 11 years ago (PS4/2013)..

But I really love when art style and clever programming triumphs after years. Halo3 to me for example has water many times more impressive than Hellblade 2
Bungie in 2005-2007 when they were developing the game didn’t have access to unreal engine 5
 
It was designed to run on essentially mid-range hardware from 11 years ago (PS4/2013)..

But I really love when art style and clever programming triumphs after years. Halo3 to me for example has water many times more impressive than Hellblade 2
Bungie in 2005-2007 when they were developing the game didn’t have access to unreal engine 5

Weird take on water effects when the likes of SoT runs on UE! :)
 
PCIE thing that nixxes keeps doing is weird. The most common 8 GB GPUs in near future will be 6600xts and 4060s and they have PCI 4 8x which will run into same issues as PCIE 3 16x does. It does not even make sense. It is just not sustainable. I hoped a pure PS4 game would be spared of this weird system but I guess I was wrong. Don't even start with "but it's a game designed for UMA" thing. Uncharted 4 and God of War were also PS4 exclusive "UMA" optimized games that run perfectly on 8 GB cards at 4K, at least on my end. This game having troubles with mediocre textures with dynamic resolution at 1440p just proves that Nixxes is creating problems that there really shouldn't be.
 
Last edited:
0:01:15 News 01: Immortals of Aveum FSR3 frame-gen tested!
0:18:16 News 02: Hellblade 2 review reaction
0:45:34 News 03: IGN buys Gamer Network... but what about DF?
0:57:44 News 04: TopSpin 2K25 impressions
1:08:14 News 05: Snapdragon X Elite processors launch
1:21:37 News 06: New Paper Mario: The Thousand Year Door details
1:30:23 News 07: Ark: Survival Ascended features console menu on consoles
1:41:12 Supporter Q1: Should ‘artistic’ effects always be optional in games?
1:47:21 Supporter Q2: Should a frame-rate cap option be added to Ghost of Tsushima?
1:52:55 Supporter Q3: Future compatibility enhancements are very helpful - so should they be a standard feature in console games?
2:01:32 Supporter Q4: Is VRR becoming a crutch for console games?
2:09:00 Supporter Q5: Are full UE5 titles finally becoming common?
2:12:35 Supporter Q6: Which game should get an RTX Remix treatment?
2:17:15 Supporter Q7: By 2030, will 8K gaming be ‘a thing’?

 
I wonder if this acquisition is why there's barely been any new videos from Micheal (NXG) over the recent months, and what the acquisition means for him.
 
Was relatively impressed with FSR3's frame generation quality when I tried it with Aveum on my PS5, even though it doesn't really work properly with a fixed refresh 60hz display, I can at least judge the resolve and the artifacts were difficult to discern in motion. Latency of course was very noticeable, but again this is not the type of intended display to use it with.

As it can now separate from the reconstruction tech, I expect to see pretty widespread use on the PS5 Pro enhanced titles if Sony doesn't have their own in-house frame generation solution.
 
Was relatively impressed with FSR3's frame generation quality when I tried it with Aveum on my PS5, even though it doesn't really work properly with a fixed refresh 60hz display, I can at least judge the resolve and the artifacts were difficult to discern in motion. Latency of course was very noticeable, but again this is not the type of intended display to use it with.

As it can now separate from the reconstruction tech, I expect to see pretty widespread use on the PS5 Pro enhanced titles if Sony doesn't have their own in-house frame generation solution.
My main experience with FSR frame generation has been with the mod that works with CP2077 that replaces DLSSFG with FSRFG, and it's very impressive. There are, of course, artifacts, but I wish more games would support it.
 
I would like to see Alex from DF compare the quality of FSR3 FG with DLSS as the input vs FSR3 to see how the FG algorithm itself performs with higher quality inputs.
 
Last edited:
Was relatively impressed with FSR3's frame generation quality when I tried it with Aveum on my PS5, even though it doesn't really work properly with a fixed refresh 60hz display, I can at least judge the resolve and the artifacts were difficult to discern in motion. Latency of course was very noticeable, but again this is not the type of intended display to use it with.

As it can now separate from the reconstruction tech, I expect to see pretty widespread use on the PS5 Pro enhanced titles if Sony doesn't have their own in-house frame generation solution.
What do you mean? Did you notice increased latency compared to FSR3 OFF?
 
I wonder if this acquisition is why there's barely been any new videos from Micheal (NXG) over the recent months, and what the acquisition means for him.
I imagine they'll keep running with NXG's analysis vids on IGN. They've always been kind of sporadic rather than any regularized content.

I dont think it's a big deal anymore than having IGN and Eurogamer both reviewing the same game.
 
I imagine they'll keep running with NXG's analysis vids on IGN. They've always been kind of sporadic rather than any regularized content.

I dont think it's a big deal anymore than having IGN and Eurogamer both reviewing the same game.
I guess it depends on how they use DF content. Since DF is autonomous, and their content appeared on Eurogamer because of the shared ownership, it's possible that DF content could now appear on IGN because of this new shared ownership. And if that's the case, I would wonder if NXG's work would be affected. Though, I think, an easy solution would be to just co-ordinate coverage so NXG's analysis wouldn't overlap DFs. Or they could just run both pieces if they want. Content is content, after all, and content is king.
 
What do you mean? Did you notice increased latency compared to FSR3 OFF?

Yes, considerably so. Again though, this is not how it's meant to be used on a fixed refresh display so it's not a reflection of the actual latency you will percieve on a proper VRR 120hz TV, I just wanted to see if artifacting was noticeable.

I guess it depends on how they use DF content. Since DF is autonomous, and their content appeared on Eurogamer because of the shared ownership, it's possible that DF content could now appear on IGN because of this new shared ownership. And if that's the case, I would wonder if NXG's work would be affected. Though, I think, an easy solution would be to just co-ordinate coverage so NXG's analysis wouldn't overlap DFs. Or they could just run both pieces if they want. Content is content, after all, and content is king.

Wonder if this will force him to reduce the not-so-subtle potshots he occasionally takes at them in his videos. :)
 
Yes, considerably so. Again though, this is not how it's meant to be used on a fixed refresh display so it's not a reflection of the actual latency you will percieve on a proper VRR 120hz TV, I just wanted to see if artifacting was noticeable.



Wonder if this will force him to reduce the not-so-subtle potshots he occasionally takes at them in his videos. :)
That's not good. Sure but can't you also use 120hz VRR without FSR3? The difference of latency should be the same.
 
That's not good. Sure but can't you also use 120hz VRR without FSR3? The difference of latency should be the same.

It depends how that FSR3 is being generated on a fixed refresh display. With a 60fps cap it's possible that in order for frame generation to work at all, the game has to run internally at 30fps - that's the only way I can get FSR3 or the Lossless utility on PC working with framegen on my PC for example. So potentially you've got the latency reduction going from ~50 fps to ~30, then the added reduction of frame gen on top of it.

It's also a case of the perceptible latency reductions/increases as you go up and down. The latency reduction going from 60fps to 120fps can be significant in the metrics, but the actual 'feel' of it, especially on a controller, will be far less than than the reduction going from 30 to 60. As no one is playing Aveum at a "real" 100fps anyways, there's nothing gauge the potential latency increase regardless.
 
Back
Top