Permabanned
What competitive games need a 4090?
I am sure even a 3090 can deliver 60 FPS at 4K other than native.
Unless the GPU is for DL2 and CBP2077, more power to you if it is but what a waste.
-
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 RT Related Games, Benchmarks, Software, Etc Thread.
- Thread starter Kaapstad
- Start date
- Status
I am sure even a 3090 can deliver 60 FPS at 4K other than native.
Unless the GPU is for DL2 and CBP2077, more power to you if it is but what a waste.
Same here, though I don’t play competitive or twitch games.
Some games I bump it up to 90-120fps for a bit more smoothness. But generally 60 is enough for me
Always good to get insight from the game developers on vendors tech.
wccftech.com
DLSS 3 Latency Doesn't Impact Player Experience, Says NFS Unbound Dev; FSR 2.2 Is in the Game, XeSS May Be Added
We chatted with Criterion Studio Technical Director Andrei Shires about the DLSS 3 implementation in NFS Unbound. AMD FSR 2.2 is also in.
Because more frames but worse input latency really helps a driving game
Because more frames but worse input latency really helps a driving game
TBF, 3 points for this specific game and implementation:
- FPS is already very high so the base latency will be well within appropriate range
As per tftcentral and latency:
Also, nvidia and developers could have reduced the latency hit with FG further as nixxes did with spiderman miles morale compared to the original
- Alex/DF found latency impact isn't as straight forward as just being tied to what FPS you're getting, it varies depending on the game engine etc.
- Higher FPS will increase fluidity and motion clarity, which is pretty important for racing games
Do agree though that if overall latency is high than 20ms, it is pretty detrimental, more so if this where a proper sim like iracing.
EDIT:
And biggest thing is if you're using M+K over controller where you will notice any latency differences more easily.
Last edited:
Yes, there's a certain amount of optimization that devs can do to reduce input latency of whatever engine they're using - that's nothing new though and any optimizations made here will only relate to the actual framerate.
So far as DLSS frame generation is concerned, it's Nvidia's Reflex (only) that helps here and as we've seen, DLSS frame generation + Reflex results in worse input latency than just enabling DLSS 2.x (even without Reflex).
Motion fluidity is of little use in fast-paced games (like racing games) when your controls don't respond with the same fluidity - I mostly play slower-paced 3rd person stuff but even I can tell the difference in input latency when I'm playing Forza Horizon 5 at (actual) 120fps vs 60.
PR puff-pieces are all the same - they're not going to say anything controversial and they're certainly not going to say anything that annoys Nvidia's marketing department.
Or, as some wag commented on that article:
Last edited:
input lag won't matter when it's arcarde crap anyway, in something like iracing or a rally sim it would make a huge difference to you spinning out or not
Yes, there's a certain amount of optimization that devs can do to reduce input latency of whatever engine they're using - that's nothing new though and any optimizations made here will only relate to the actual framerate.
So far as DLSS frame generation is concerned, it's Nvidia's Reflex (only) that helps here and as we've seen, DLSS frame generation + Reflex results in worse input latency than just enabling DLSS 2.x (even without Reflex).
Motion fluidity is of little use in fast-paced games (like racing games) when your controls don't respond with the same fluidity - I mostly play slower-paced 3rd person stuff but even I can tell the difference in input latency when I'm playing Forza Horizon 5 at (actual) 120fps vs 60.
PR puff-pieces are all the same - they're not going to say anything controversial and they're certainly not going to say anything that annoys Nvidia's marketing department.
Or, as some wag commented on that article:
Well yeah if you're at 60 or less fps then you'll will notice latency difference when you go to 120+ fps, that has always been stated with frame generation across the tech press, even by nvidia themselves, its results will always be best only when you're at >60 fps before enabling FG.
I'm not 100% certain but I don't think this is nvidia sponsored either.
It's a case of pick your poison:
- run lesser graphical settings just to avoid FG latency increase
- run maxed graphical settings and enable FG to keep smooth motion/gameplay with a slight increase in latency, which will depend on how sensitive a person is to this and things like if using controller or M+K
- run maxed graphical settings without FG and have choppy/stuttery motion/gameplay
I've got a 175hz qd-oled screen and love the motion clarity when pushing 100+ fps but especially 140+ so would always use FG when possible as long as latency isn't >25ms especially if it means being able to get better visuals.
Hence this comment:
EDIT:
Should be noted that given the already good performance of the game, FG isn't really needed in NFSU anyway, was just good to see another developers insight.
Last edited:
Well let's be clear here - what you call 'gameplay' includes the relationship between a player's inputs and what's happening on screen so all we're gaining is a perceived increase in visual fluidity - nothing else.
Also, what you're calling a 'slight' increase in latency is actually *double* the input latency you should be getting if the framerate were actually what frame generation is putting out - I don't consider that to be 'slight'.
I'll leave this here to demonstrate "motion clarity" at different frame rates.
Blur Busters TestUFO Motion Tests
Blur Busters UFO Motion Tests with ghosting test, 30fps vs 60fps, 120hz vs 144hz vs 240hz, PWM test, motion blur test, judder test, benchmarks, and more.
If you're sensitive to any form of stutter, ghosting or/and poor motion (especially on lcd displays), you will appreciate a higher frame rate over latency increase hence why I played games on my "60HZ" OLED with Vsync over my 144HZ freesync premium IPS monitor... Also, largely because again, I use controller over M+K more often which is where the higher input latency is further minimised by using controller (in terms of perception/feel). @Raiden85 is in a similar boat:
Again, it depends entirely on the base frame rate and the engine/game, I suggest watching Alex's video on frame generation to get a very good in depth explanation into all these points where he has provided examples/evidence to backup the points:
Last edited:
Thanks for the links - although I've already watched Alex's analysis (and used some of DF's examples to illustrate the latency issue with frame generation) - there's nothing in there contradicts my comments above.
No one is denying that latency is increased with FG, that is well known, the only point is that it's a trade of if you want higher fps to get smoother motion, which for some like myself and raiden is just as, if not more important than the increased latency (within reason), if you don't want the latency associated with fg then you have to sacrifice graphical settings to hit your fps target, which again, can depend on the game and what settings you are sacrificing, could be no difference or it could be a huge difference in visuals, the very last bit of the video above shows this point.
Sure - there's no such thing as a free lunch and frame generation has a trade-off - and that trade-off isn't a minor one. Whether it bothers someone or not is going to be a very subjective thing but it's still there. At least Digital Foundry investigated the actual impact - that interview you posted just handwaved it and I *hate* that - which is why I commented.
Guess let's wait and see actual tech comparisons as well as user feedback before we can really comment/judge.
This was a good watch too:
Soldato
What have you used DLSS 3 on, a 4080 or 4090?
Telling that in that video, the testers were easily able to tell the difference between DLSS 2.x and 3.0 at 60fps (with the majority/all preferring DLSS 2.x or native) - this technology really isn't going to scale well on the lower-end cards (which ironically, are the ones that could benefit from it the most).
Last edited:
Did you watch the rest of the video? Particularly the 4k30 vs 4k60 dlss 3 bit?
As is always the case, you pick your poison, if you value latency more than anything then I guess you'll have to make do with sacrificing visual settings, using a lower res. and/or just accept a lower frame rate thus worse motion clarity/stuttery gameplay, which may not bother you if you're not using a high refresh rate display.
I know which one I rather have but again, that's just me, playing on a 4k60 oled tv with vsync (which is absolute worst case scenario for latency even compared to FG) never bothered me and still doesn't put me of enough even though I have my 175hz qd oled here as well.
I think we're just on a merry go round now so lets just agree to disagree as there is no right or wrong and this is entirely subjective.
RT news:
Ray Tracing On Android Devices? Qualcomm’s Ziad Asghar Explains…
Blender Ray-Tracing: Intel Aiming For oneAPI RT In 3.6, AMD HIP-RT Working Internally
Last edited:
I did - although I hardly think it's surprising they preferred a perceived 60fps image over a native 30fps presentation - that doesn't invalidate the fact the all the testers observed and preferred real 60fps latency with native/DLSS 2.x over DLSS 3.0
There's a reason we dial back settings to try to get games to run at 60fps+ rather than just run them at 30 no? (and it's not just the fluidity of the frame rate).
Last edited:
- Status