That would be great and more over it would be interesting to know if there is any lag in Ps2 controller .
So I got around to testing some PC games. They are now posted above in their own section, and I have also compared two modes of the UFB for PC.
I was holding off on PC for a while for a few reasons. Firstly, the environment is not standardised. I’ve posted my specs, but have never really dived into my graphics cards settings. At least on PS4 we only have to content with PS4 and PS4Pro (unpublished, but my testing suggests it makes no difference to UMvC3). Then, there are individual game settigns as well.
My capture device is an El Gato HD 60 Pro, which is in my PC. Without two monitors it was very difficult to capture and have the game running on the same PC. The solution, unsurprisingly was to buy another monitor.
Also, every time I plug in the capture device, the computer seems to want to output at 1080p30 or 720p60, which is not really useful. I overcame this using a splitter, initially just connected to a monitor to make it output 1080p60 and then plugging an additional HDMI cable into the lag testing setup. I wasn’t sure if the splitter would add input lag itself, but having gone back and tested (PS3 BBCPE split and non-slpit) it seems to change it by about a ms or less, so I’m happy enough.
Finally, SFV PC which I’m sure is the game that people are most interested in, seems to be very buggy on my system. It starts lagging under 60fps apparently after any period of time playing, so I always had to load up, get to training mode quickly, turn command history on each time (thanks capcom) and get the test done asap before the frames dropped below 60fps. Interestingly enough, the one signal triggering two controllers is the only one not affected by frame drops.
Anyway, I overcame those issues and managed to capture most of the relevant games that I had on steam. I think the only one I’ve missed is USF4.
Anyway, I’ll keep testing what I can when I can. I should be able to do injustice 2, Rev 2 and Tekken 7 in the coming weeks, and have also some ideas about testing network lag using this method, but probably need to incubate that one a bit longer. Glad I’ve got most of the work done though, as when Tekken 7 comes out, I’ll likely turn my attention to testing the system and Eddy in that game.
Tested my Panthera with original and updated firmware.
[details=Spoiler][list]
[]Original 1.00 https://youtu.be/GW_1reU0Rxo
[] Updated to FW 1.01 and some Brook Sniper Adapter testing
https://youtu.be/CbeEQ70BigI
[/list][/details]
Just for interest’s sake, I had a quick look at the Mario games I have access to.
SM3DW
NSMB WiiU
Mario Maker
NSMB Wii (via WiiU)
SMW (via WiiU VC)
Yoshi’s Woolly World
A recent video about input lag by NX gamer about input lag came out.
It inspired me to test some 3d first person shooters. I tend to pick games that have demos so I can compare PS4 vs X1, as this was the main point of his post. Haven’t done full analysis on them, just a quick check.
Resident Evil 7 demo (X1 appears ~1F faster) - note, only really had access to blocking animation to test, happens on F12/13).
[details=Spoiler]https://twitter.com/noodalls/status/862761986693578752
https://twitter.com/noodalls/status/862761444600815616[/details]
Doom demo
PS4 (reponse on F6)
[details=Spoiler]https://twitter.com/noodalls/status/862407917512695808
https://twitter.com/noodalls/status/862407993131909120[/details]
X1 (response on F6)
[details=Spoiler]https://twitter.com/noodalls/status/862407771219648516
https://twitter.com/noodalls/status/862407836579397634[/details]
Prey Demo
PS4 (response on F13)
[details=Spoiler]https://twitter.com/noodalls/status/862756946780106752
https://twitter.com/noodalls/status/862757060814884864
https://twitter.com/noodalls/status/862757136182398976
https://twitter.com/noodalls/status/862757185775783937[/details]
X1 (response on F9)
[details=Spoiler]https://twitter.com/noodalls/status/862756623672000512
https://twitter.com/noodalls/status/862756731637583872
https://twitter.com/noodalls/status/862756828848955392[/details]
So, I wouldn’t put too much weight on the RE7 results. Doom appears quite fast (perhaps not compared to fighting games), with the same response time across systems. Prey is quite slow on X1, but 4F slower again on PS4 (note this was the first game I captured in 30fps, so it’s actually only 2F slower as every second frame is repeated. However, given that 4F @ 60fps = 66.66ms, and 2F @ 30ms = 66.66ms, and my capture card still generates 60fps, I’ve left it be).
I guess Injustice 2 is next, eh.
But quick question are the result listed for games that have a 1 frame difference between command history animaton, do the results follow the command history or animation?
The results refer to the animation, with a note regarding where the command appears. I do this as I figure what people are actually interested in is when the animation changes, as this is what they will react to.
Also, there is some strange variations, e.g. SFV there doesn’t seem to be a completely consistent relationship between the command history and the animation.
@noodalls I believe your tests are slightly off.
For example, you have Street Fighter V at 88.6ms (5.3f), whereas it is at 117ms (7.0f) according to another source.
You also have KoF XIV at 76.5ms (4.6f) whereas another source has it at 100ms (6.0f).
Given that you have PS3 version of TT2 at 94.0ms (5.6f) and the other source has 116.7 (7.0f), the Tekken 7 input lag of 120.1ms (7.2f) is probably closer to 9f.
I always recommend not comparing my results to other results, but rather comparing them within.
I have messaged the person who wrote the front page article and asked him to correct that.
Having said which, when I last checked my results versus displaylag’s results, we seemed to have a fairly consistent difference between results for most games.
The method I use does result in faster results than other methods, but I’m fairly happy to justify why that is the case.
I was under the impression that you were doing these tests from the sake of comparison, but the the resulting articles seem to have created considerable inconsistency due to a lack of context. Thank you for clarifying.
Anyone know if the updated Hrap4 kai (with hayabusa buttons) we get in the west uses the same PCB as the hori used in the test here?
Remember when Sony said the new DS4 would have less input lag when wired?
the same question. What is the reason to use USB wired NEW DS4 if it has more input lag than wireless old DS4?
Very ineteresting to know the input lag for Tekken 7 on PC (Vsync On/Off) in comparison to PS4?
Also, is there any difference in input lag between PS4 and PS4 Pro?
Noodalls has tested input lag on PC to be 105.8ms for Tekken 7.
If you take his BBCPE PC test of 56.2ms (3.4f) and adjust for his methodology it comes out to around 66.7ms (4f), there is a difference of 10.5ms (0.6f). If you adjust his Tekken 7 PC result it comes out to around 116.3ms (7.0f).
If you take his BBCPE PS4 test of 42.9ms (2.6f) and adjust for his methodology it comes out to around 66.7ms (4f), there is a difference of 23.8ms (1.4f). If you adjust his Tekken 7 PS4 result of 120.5ms (7.2f), it comes out to 144.3ms (8.7f).
In other words there is about a 2f input delay reduction on PC. There is also a 1f input delay reduction on XBONE. He also confirmed that V-sync and Pro have no effect on his results.
What do you mean by this.
He has different setups for his tests on PC and PS4. Using his BBCPE tests across systems allows for accurate comparison. Also, as the method he uses gives slightly faster results than other methods it requires a minor adjustment for accurate comparison with other sources.
Yes, I get that. I wanted you to explain the nuts and bolts of how you adjusted. I should have been more clear.
Please disregard anything I posted previously. For starters I made the wrong assumption that BBCPE had 4 frames of input lag/buffer built in like Xrd instead of 3. With that adjustment 6ms would need to be deducted from the PC version results of T7 making it 100ms (6f). Also there was a 8ms (BBCPE) to 28ms (SFV) discrepancy between PS4 results of https://displaylag.com/video-game-input-lag-database/ and those of Noodalls. That would likely make PS4 version of T7 have about 133ms (8f) input lag in that range of error (128ms to 148ms). That would put the XBONE version of T7 at 117ms (7f) of input lag.
Input Lag in Tekken Games:
5f - Tekken 5 (ARC/PS2/PSP/PS3); Tekken 6 (PS3/PSP/X360);
6f - Tekken 6 (ARC/XBONE); Tekken 7 (PC);
7f - Tekken Tag 2 (ARC/PS3/X360); Tekken 7 (XBONE);
8f - Tekken Tag 2 (XBONE); Tekken 7 (PS4);
The difference in the method between the consoles and PC is that the PC means I have to use a splitter. For whatever reason, my graphics card won’t output to my testing device at 1080p60 unless I use a splitter, output to a monitor first and then split off the signal to the testing device. Other than that, the hardware is exactly the same.
I generally don’t enjoy testing things on PC, as I posted on twitter about basically.
- I need to use a splitter, which I don’t for any other tests
- Every time you post any PC result, inevitably the next response will be “what about with vsync on/off? what about with borderless/full screen? what about on a graphics card that I don’t own?”
- Every test takes 20-30mins to do properly. So every variation takes extra time. This is why I’m secretly happy that I haven’t seen any variation with pro/boost mode etc.
- There have been some times where I’ve got bizarre results from the PC games, which doesn’t seem to happen on consoles. I can’t quite explain it, but suspect it’s a combination of dropped frames/recording on the same PC as I’m playing on (even when the frame rate stays at 60fps). Some of the results have been significantly faster than expected. When I tested PC T6, I recorded with a camera as well to ensure that the results were true before posting.
- Web camera doesn’t seem to ever capture at 60fps, no matter what I try. I have both the new logitech and the razer, and in neither can I get 60fps in Obs/El Gato. This would greatly improve things.
- My PC is 3 years old, so might have an effect on results.