MULTI Linus: Dualshock 4 "is the BEST game controller"

Status
Not open for further replies.

Necrokiller

Expert
Apr 16, 2009
13,594
5,125
129
I figured someone who is renowned on PG for complaining about input lag all the time would at least know how to measure and calculate it. Looks like both are things you just learned since the start of this thread.

Will have to download the game to check, though.

Will have to download that too.
I imagine this list would be growing exponentially from now on lol
 

CerebralTiger

Expert
Apr 12, 2007
19,839
5,868
129
Islamabad
I figured someone who is renowned on PG for complaining about input lag all the time would at least know how to measure and calculate it. Looks like both are things you just learned since the start of this thread.
I've been testing input lag on displays using a camera and timer since over a decade ago lol. Never cared for testing game-related latency (seen such analog stick deflection videos several times on Youtube before) because there's nothing that can be done to fix it besides increasing the frame rate, and that too is usually only an option on PC. Most of the games I enjoy are exclusive to Sony/Nintendo hardware.

P.S. I don't own a 240 or 120fps camera. The one I used to test TLOU2/GoW doesn't belong to me.
 
Last edited:

CerebralTiger

Expert
Apr 12, 2007
19,839
5,868
129
Islamabad
You don't say. These calculations (caught before the edit lol) speak for themselves. Oh well, atleast you found a hobby and learned something from this thread.
The text you quoted is exactly what I wrote initially (some mod can verify my edit history for you). I edited it to ask how you conducted the test because I didn't know the specs of your camera. Edited it back in with the assumption that it was a 60fps camera ¯\_(ツ)_/¯

Hate to disappoint you, but I didn't learn anything new from this thread :ROFLMAO:

You, on the other hand, learned about the DS4's best in class gyro polling rate, and why it's ideal to test the quickest action to determine the approx. input lag that isn't a design consideration.
 
Last edited:

Necrokiller

Expert
Apr 16, 2009
13,594
5,125
129
I edited it to ask how you conducted the test because I didn't know the specs of your camera.
If you knew how its calculated that'd be the first question to ask ¯\_(ツ)_/¯

Anyway...

My point in showing God of War test was that even with a game with such a high input lag (by design), people like you end up praising the combat regardless instead of complaining about the lag. Same goes for TLOU on PS3. Since every game is different in that regard, at the end of the day, the controller choice hardly makes a difference. Other factors like frame rate and display lag have a greater influence.

As for your Last of Us Part 2 test, since I don't have the game, I'll assume that to be correct until I can verify it myself. It looks to me like the movement began at either the 74th or 75th frame, not the one you ended up counting. But that could just be due to the quality of the gif. Also, while you already have someones camera, wouldn't hurt to do other tests as well.
 
Last edited:

CerebralTiger

Expert
Apr 12, 2007
19,839
5,868
129
Islamabad
If you knew how its calculated that'd be the first question to ask ¯\_(ツ)_/¯
Assuming you were using a 60fps camera =/= not knowing how it's calculated. Yes, there's a chance some frames would get missed in capturing the video, but that doesn't mean a 60fps camera can't be used to get comparative results. It's just not as accurate as a 120fps camera would be for capturing 60Hz footage or a 240fps camera would be for capturing 120fps footage. Even with higher fps cameras, there's still the possibility of human error using the analog stick, as the speed of deflection can vary from test to test.

My point in showing God of War test was that even with a game with such a high input lag (by design), people like you end up praising the combat regardless instead of complaining about the lag.
GoW's movement latency isn't indicative of overall latency. Other actions are faster. It's simply a design choice to have his movement feel weighty. Also, it may have higher movement latency than The Witcher 3's alternate controls setting, and it would still be vastly superior. Why? Because other design choices also play a role, such as what level of control you have over your attack patterns, how good or bad the AI in an encounter is, how much overcoming a combat encounter feels earned or satisfying, what kind of hit reactions and hit detection there is. There are several factors to consider. The same is true for TLOU part I.

More importantly, increasing the frame rate considerably improves Kratos' movement latency. The same is not true for all actions in RDR2, as animation priority takes precedence.

Since every game is different in that regard, at the end of the day, the controller choice hardly makes a difference. Other factors like frame rate and display lag have a greater influence.
Nowhere did I say that frame rate doesn't have a greater influence. Display lag, on the other hand, doesn't if the right display is chosen. There are displays that have lower input lag than the differential latency between the DS4 and Steam controller buttons. And like I said earlier, these things aren't mutually exclusive. Game latency, display lag, and controller lag all add up. The aim is to lessen this as much as possible, not disregard one aspect because another makes a bigger difference.

It looks to me like the movement began at either the 74th or 75th frame, not the one you ended up counting. But that could just be due to the quality of the gif. Also, while you already have someones camera, wouldn't hurt to do other tests as well.
I don't have someone's camera. Someone let me use their camera when they were at my place, which is why I couldn't conduct multiple tests.
Spoiler: show




You may be right about it being the 74th (definitely not the 75th) frame instead of the 66th. The weapon model does indicate some movement, but it may be due to the idle animation, which also has some sway. I'll need to test it again, possibly when she sways in the opposite direction.

Either way, movement latency of 133m is still not at all what you were so sure about.

I'm sure even TLOU2 would fall into the higher side of 30fps games, instead of the most responsive feeling ones.
TLOU had more weighty and slower actions compared to Uncharted games. TLOU Remastered at 60fps has even more lag than whatever action that UC4 picture is suppose to show at 30fps. Same engine and all, that doesn't inspire any confidence in TLOU2 being close to UC4. Both being 30fps and all LOL
Lost Legacy on PS4 isn't the most responsive feeling game either. Takes 200ms for any kind of response to 333ms to get any kind of camera movement. Pretty awful imo.

How does TLOU2 hold up?
 
Last edited:

Necrokiller

Expert
Apr 16, 2009
13,594
5,125
129
Either way, movement latency of 133m is still not at all what you were so sure about.
Don't think you were expecting an engine latency of 116.6ms in RDR2 either. Wouldn't it be reasonable to assume that the next GTA or RDR game would exhibit similar lag/priority of animations as the previous games have? GTA games don't get the same hate as RDR2 but they still don't feel that great to play.

I've played all ND games except for TLOU2 and they haven't really been what I'd call the most snappy or responsive controlling games. ND has always leaned towards making the animations look better than making the most reponsive game. Until I play it, even a lower 133ms figure might not be indicative of the overall feel due to other design decisions that play a role that you highlighted in your post.
 
Last edited:

CerebralTiger

Expert
Apr 12, 2007
19,839
5,868
129
Islamabad
ND has always leaned towards making the animations look better than making the most reponsive game.
That's true, but motion matching has helped them improve the quality of animations and lower input latency.

Turns out, my old cell phone can actually record low res videos at 120fps. Hopefully, this gif should eliminate any ambiguity.



Movement latency of 125ms. Testing multiple times leads to latency results between the range of 116ms and 133ms. Motion matching makes it difficult to tell where the idle animation blends with the animation for movement.

Other tests:


As a bonus, also went back to GoW and got the same latency of 258ms.

Spoiler: show


 
Last edited:
  • Like
Reactions: NaNoW

Necrokiller

Expert
Apr 16, 2009
13,594
5,125
129
(something I actually learned thanks to Necrokiller lol)
Among several other things :ROFLMAO:

Bless this thread with TLOU PS3 test. If not that, then the TLOU Remastered at 30fps so we know what kinda lag was passed off as the "best feeling and playing third person shooter" lol
 

Necrokiller

Expert
Apr 16, 2009
13,594
5,125
129
Just another option in the plethora of available customization on the Steam Controller.


 
Last edited:

CerebralTiger

Expert
Apr 12, 2007
19,839
5,868
129
Islamabad
Just another option in the plethora of available customization on the Steam Controller.


It literally just became available on Steam (which is great, I no longer need to use a 3rd party app for it) and works better with an actual analog stick. Hence the name lol.

Not to mention, it works in combination with the gyroscope for the vertical axis, which further makes this option much better on the DS4. The flick stick does away with your gyro related "steering wheel" complaint, because you're often not using it for horizontal aim. That's the whole point of flick stick.

Also, none of these plethora of options will alter the fact that PC games are more responsive latency-wise when played using a DS4 over a Steam controller.
 
Last edited:

CerebralTiger

Expert
Apr 12, 2007
19,839
5,868
129
Islamabad
Bless this thread with TLOU PS3 test. If not that, then the TLOU Remastered at 30fps so we know what kinda lag was passed off as the "best feeling and playing third person shooter" lol
After testing TLOU Remastered at 30fps, I've come to the conclusion that there's no fixed movement latency in the game. There appears to be some lag 'penalty' associated with trying to move shortly after coming to a halt, resulting in 300+ms latency when attempting to resume movement. Moving from an otherwise stationary position results in latency of ~180ms.

https://abload.de/img/bestshksa.gif (best case, ~160ms)
https://abload.de/img/worsto2kzr.gif (worst case, ~330ms)
https://abload.de/img/untitled-3n2k7g.gif (~180ms)
https://abload.de/img/untitled-337wkme.gif (~180ms)
https://abload.de/img/untitled-333frkns.gif (~180ms)

It's clear that the game is using physics based locomotion, and therefore is trying to simulate real life movement. Would also explain the variable readings I got when testing TLOU2.
 
Last edited:

Necrokiller

Expert
Apr 16, 2009
13,594
5,125
129
Why are the tests conducted with Ellie and not Joel? Seems like Joel would have a slightly higher readings.

To get a more realistic real world scenario of how an average player would've experienced the game back in 2013, it would be PS3 gamers playing this game on their HDTVs, not monitors. Even the best case Sony W800B was measured at ~24ms (thats not indicative of the average consumer so its the best case scenario). So let's put that estimate at 20ms excluding your current monitor lag.

Add to this the additional 7.3ms lag from DS3 analog.

Your best case (assuming your measurement is correct and that everybody had the best measured lag of W800B TV) would be ~187ms. Worst case ~357ms.

All this is still not factoring that the game's frame rate which was far from stable on the PS3, which would induce more lag.
 
Last edited:

CerebralTiger

Expert
Apr 12, 2007
19,839
5,868
129
Islamabad
Why are the tests conducted with Ellie and not Joel?
For the sake of consistency across both games.

To get a more realistic real world scenario of how an average player would've experienced the game back in 2013, it would be PS3 gamers playing this game on their HDTVs, not monitors. Even the best case Sony W800B was measured at ~24ms (thats not indicative of the average consumer so its the best case scenario). So let's put that estimate at 20ms excluding your current monitor lag.
Monitors were an equally viable option on PS3, so the HDTV factor doesn't really matter here. The issue worth mentioning with the PS3 setup was that it didn't output 720p games at 1080p, and therefore left the scaling to the display, adding some additional latency.
 

Necrokiller

Expert
Apr 16, 2009
13,594
5,125
129
For the sake of consistency across both games.
Part 2's numbers are irrelevant in determining the threshold or "benchmark" for what some here consider the best playing third person shooter.

Monitors were an equally viable option on PS3
I didn't talk about viability, rather the general use case scenario. Even today, some standard ports across devices don't change the average use case scenario i.e. monitors for PCs and TVs for consoles.
 

CerebralTiger

Expert
Apr 12, 2007
19,839
5,868
129
Islamabad
Part 2's numbers are irrelevant in determining the threshold or "benchmark" for what some here consider the best playing third person shooter.
Part 1 (w/ dlc) and part 2 are both played with multiple characters. I just chose the common one out of both for the sake of this comparison. Don't see the issue here. Your argument would've made sense if Ellie was only playable for an insignificant period of the overall playtime in part 1. She isn't.

I didn't talk about viability, rather the general use case scenario. Even today, some standard ports across devices don't change the average use case scenario i.e. monitors for PCs and TVs for consoles.
I played TLOU1 on both a TV and a monitor back in 2013. Not sure what kind of 'realistic' view you're trying to get.

Regardless, the display is irrelevant here provided the same one is used to perform these tests. Not to mention lower latency is now possible as a standard on hdmi 2.1 TVs with compliant devices.
 
Last edited:

Necrokiller

Expert
Apr 16, 2009
13,594
5,125
129
Don't see the issue here.
Joel is the major playable character. I found it odd how he wasn't part of your test.

Not sure what kind of 'realistic' view you're trying to get.
Don't see whats there to argue about. I derived those number from your tests. I presented a more realistic scenario where most console gamers use a TV, not a monitor. Its not suppose to negate your results.

If we were to listen to lord Ledbetter, he'd argue vast majority of console gamers don't even bother switching to Game Mode.
 
Status
Not open for further replies.
General chit-chat
Help Users
We have disabled traderscore and are working on a fix. There was a bug with the plugin | Click for Discord
  • No one is chatting at the moment.
    Necrokiller Necrokiller: Is it just me or people lately seem to defend every bad game design decision made by the devs...
    • Haha
    Reactions: EternalBlizzard