• Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
  • We have made minor adjustments to how the search bar works on ResetEra. You can read about the changes here.
Status
Not open for further replies.

linko9

Member
Oct 27, 2017
437
Because from everything I know on the subject it's absolutely not possible.

I'm actually interested in this, handheld response time isn't something I know much about, just heard it can be extremely low (which checks out from personal experience). Can't seem to find any sources testing lag on earlier handhelds, which would be neat to see. Haven't played blossom tale myself, but that video looks legit, so I don't see how we can't accept it. If you've got any sources on the issue, I'd be interested.

edit: Interesting that OP says BT "barely beats SMO" which obviously has well over 2 frames of input lag. I took OP at his word that it was really 1 frame for BT, but since there's no frame data from the camera displayed in the video, that's hard to confirm. I wonder if OP's method of counting frame delay is off a bit. Perhaps BT is closer to 2 frames and HK is closer to 5? Anyway, would be nice to get clarification from OP.
 
Last edited:

Misterman

Member
Jun 10, 2018
212
I feel the same way. Happens rarely but still annoying to wall right off a ledge when I clearly jumped, especially when coming from Celeste
 

nib95

Contains No Misinformation on Philly Cheesesteaks
Banned
Oct 28, 2017
18,498
Dude, 66.6ms is basically nothing. We're talking 0.067 seconds. It's on the lower end of the scale as far as popular games go.
 
Last edited:

Owarifin

The Fallen
Oct 27, 2017
1,741
Just bought the game on the Switch last night...
Haven't played it yet, but almost regretted my purchase.

Wondering though, is there a difference in input lag between portable and docked?
 

Bomblord

Self-requested ban
Banned
Jan 11, 2018
6,390
I'm actually interested in this, handheld response time isn't something I know much about, just heard it can be extremely low (which checks out from personal experience). Can't seem to find any sources testing lag on earlier handhelds, which would be neat to see. Haven't played blossom tale myself, but that video looks legit, so I don't see how we can't accept it. If you've got any sources on the issue, I'd be interested.

Well to put it simply some of the best TVs and monitors on the market have input lag around 9-10ms and Nintendo isn't using a high end OLED panel.

https://displaylag.com/display-database/

So even if they somehow had the game engine have the least amount of input lag ever the display isn't going to allow that to happen GOOD IPS LCDs are around 20ms. Which is greater than 1 frame.
 

TeenageFBI

One Winged Slayer
Member
Oct 25, 2017
10,240
You guys think he doctored the video or something?
A lot of this goes on whenever someone mentions input lag in Hollow Knight threads. I still haven't seen a video that proves their clim though.

For my part, I notice a slight delay when jumping in the Switch version. I can work around it but I occasionally miss jumps that it feels like I should have nailed.
 
Last edited:

Nacho Papi

Member
Oct 27, 2017
2,338
If people can play Street Fighter 4 at the highest level of competitive play with 4+ frames of input lag you can certainly play Hollow Knight.

Ofc you can but that doesn't negate the fact there is a fair amount of input lag for a game like this. For what it's worth I don't think it's bad at all in handheld mode but when docked and adding extra latency from my receiver + tv it feels like I'm playing 'online' with a crappy connection...
 
Oct 25, 2017
9,105
Ofc you can but that doesn't negate the fact there is a fair amount of input lag for a game like this. For what it's worth I don't think it's bad at all in handheld mode but when docked and adding extra latency from my receiver + tv it feels like I'm playing 'online' with a crappy connection...
"For a game like this" is not going to convince me given that SF4 requires some inputs to be accurate to a single frame. 4 frames of lag is just not an issue for a game like Hollow Knight which is certainly challenging but not nearly that demanding of perfect inputs.
 

DanteMenethil

Member
Oct 25, 2017
8,058
Clearly it is for some people?

Given they are likely in the same pool of people who like retro gaming and how people treat any input lag there it seems fully understandable to me.

its 1 frame more than gaming on a CRT. If 4 frame is an issue for people they wouldn't be able to play 99.99% of post crt games. I don't think you grasp how good 4f actually is.
 

RockmanBN

Visited by Knack - One Winged Slayer
Member
Oct 25, 2017
27,981
Cornfields
Mega Man Legacy Collection 1 on switch undocked had a MUCH worse lag. This was easily tolerable.
 
Last edited:
Oct 25, 2017
4,801
New York City
I noticed severe* input lag in a couple of Switch games, including Vs. Super Mario Bros. and Mega Man Legacy Collection 1 (not 2!). Those games are pretty much unplayable to me, sadly. But I haven't noticed any lag at all in Hollow Knight.

*Severe to me
 

Custódio

Member
Oct 25, 2017
1,901
Brazil, Unaí/MG
After playing on the PC for dozens of hours, I felt the input lag on the Switch on the first jump, and I actually was using the Switch Pro Controller on the PC version. That being said, I have a really good g-sync monitor, so is not a fair comparison.
 
Last edited:
OP
OP
AtomicShroom

AtomicShroom

Tools & Automation
Verified
Oct 28, 2017
3,078
Because from everything I know on the subject it's absolutely not possible. Taking the panel (it's IPS) lag, the time for the controller to send the signal, and the time for the game engine to process it there's just no way the input lag is at 16ms.

Look I don't pretend that my methodology is flawless. I started counting frames at the moment where my finger was fully pressed in, but it's not impossible that the signal was sent earlier when the button was partially pressed in, which could mean it's 2 frames (33ms) and not 1 frame (16ms). In any case I counted the same for both games and what ultimately matters is that Hollow Knight takes 3 more frames to respond to input.
 

Mars People

Comics Council 2020
Member
Oct 25, 2017
18,192
Never noticed any input lag on the PC.

But Unity on consoles is always a nightmare. Hell I've run into problems with that crap engine on pc too.
I hate it. I really do.
 

PaulloDEC

Visited by Knack
Member
Oct 25, 2017
7,418
Australia
I just did a test on the PC version using a wired 360 controller. My phone is a OnePlus 3T shooting slowmo at 120fps (according to the info I could find, the phone itself doesn't tell you this).

hk_inputntjpy.gif


It took 16 frames after my finger had fully depressed the A button before the character reacted (that is to say the 16th frame after hitting the button is when the animation began). Who wants to do the math for me?
 

InfiniDragon

Member
Oct 25, 2017
2,312
I just did a test on the PC version using a wired 360 controller. My phone is a OnePlus 3T shooting slowmo at 120fps (according to the info I could find, the phone itself doesn't tell you this).

hk_inputntjpy.gif


It took 16 frames after my finger had fully depressed the A button before the character reacted. Who wants to do the math for me?

Yeah, that's my experience with the Switch version. It's only jumps, so if it's also on PC i guess it's all versions.
 

Mcjmetroid

Member
Oct 28, 2017
1,843
Limerick, Ireland
I didn't notice any input lag however the Switch analogue on the joy cons are horrendous. I'm now using my PS4 controller with the USB adapter and using the DPAD, it's the only way to get those down slash moves perfectly every time.
 

DanteMenethil

Member
Oct 25, 2017
8,058
I just did a test on the PC version using a wired 360 controller. My phone is a OnePlus 3T shooting slowmo at 120fps (according to the info I could find, the phone itself doesn't tell you this).

hk_inputntjpy.gif


It took 16 frames after my finger had fully depressed the A button before the character reacted (that is to say the 16th frame after hitting the button is when the animation began). Who wants to do the math for me?

Should be 8 frames of delay if this is 120fps. Do you happen to know your monitor model so we can substract that lag from the total?
 

PaulloDEC

Visited by Knack
Member
Oct 25, 2017
7,418
Australia
Should be 8 frames of delay if this is 120fps. Do you happen to know your monitor model so we can substract that lag from the total?

Absolutely, monitor is a DELL U2410 over DisplayPort.

I just did a quick test for the slash too, and it came out at around 14 frames before animation started. So very slightly quicker, but that could be an error in my testing.
 

linko9

Member
Oct 27, 2017
437
So it seems that the input lag is almost certainly higher than 4 frames for the jump in both versions. Should probably update the OP.
 

low-G

Member
Oct 25, 2017
8,144
I just did a test on the PC version using a wired 360 controller. My phone is a OnePlus 3T shooting slowmo at 120fps (according to the info I could find, the phone itself doesn't tell you this).

hk_inputntjpy.gif


It took 16 frames after my finger had fully depressed the A button before the character reacted (that is to say the 16th frame after hitting the button is when the animation began). Who wants to do the math for me?

Don't forget you can probably lock the prerendered frames in your display driver to a lower number than default. Set it to 1 frame render-ahead and test again.

Personally I do 3 frames only for modern taxing games. Anything 2D or slightly old can set to 1 frame.
 

Odrion

Banned
Oct 27, 2017
1,148
I fucking KNEW it. The game feel is terrible and many times I felt like I was somehow too slow to react to a certain enemy.
 

DanteMenethil

Member
Oct 25, 2017
8,058
Absolutely, monitor is a DELL U2410 over DisplayPort.

I just did a quick test for the slash too, and it came out at around 14 frames before animation started. So very slightly quicker, but that could be an error in my testing.

That monitor has 33ms of input lag, so 2 frames. Plus whatever amount of frames pre-rendered in nvidia settings if you have that.

So it seems that the input lag is almost certainly higher than 4 frames for the jump in both versions. Should probably update the OP.

Not really, that test is a completely different environment and doesn't say much toward the switch version.
 

Johnny Ryall

Member
Oct 28, 2017
241
I'm a good 16 hours in with lots of lots of the map uncovered and not once have I thought I died due to lag. Like this entire thread is baffling to me, I've been going around thinking the game is really responsive.

Shrug.
 

PaulloDEC

Visited by Knack
Member
Oct 25, 2017
7,418
Australia
Don't forget you can probably lock the prerendered frames in your display driver to a lower number than default. Set it to 1 frame render-ahead and test again.

Okay, new test. Same everything, except max prerendered frames reduced to 1 in Nvidia control panel, and in-game Vsync disabled. This time the character reacted on the 9th frame of animation rather than the 16th, which is a pretty massive difference. Will get a GIF edited into this post in a few minutes.

preren1_novsync1hsqu.gif
 
Last edited:

Skittzo

Member
Oct 25, 2017
41,037
I'm a good 16 hours in with lots of lots of the map uncovered and not once have I thought I died due to lag. Like this entire thread is baffling to me, I've been going around thinking the game is really responsive.

Shrug.

Nah the majority of people agree with you, only a handful seem to actually find this to be an issue. Even though it's apparently pretty good input lag compared to most modern games.
 

SkekesisVR

Member
Jan 2, 2018
476
Pretty sure unity is the engine they use?

A common Unity pitfall is doing your input polling in FixedUpdate. FixedUpdate is the physics update loop, runs independent of the display framerate, usually lower (especially on mobile). A common optimization technique is relaxing the physics engine framerate.

The only reason to update in this way is if it's rigidbody based movement vs transform manipulation, which would be an interesting choice given the game they made, but not unheard of
 

linko9

Member
Oct 27, 2017
437
Not really, that test is a completely different environment and doesn't say much toward the switch version.

I'm referring more to the fact that OP's analysis of 1 frame of input lag for BT is very unlikely, probably at least 2, which would make HK's jump at least 5 frames. Would be interesting to see the PC version tested in an environment with the lowest lag possible for comparison. Seems likely that this issue has nothing to do with the switch port though.

edit: or maybe so, if PC can really get as low as a couple frames. Again, would be nice to see some more tests.
 
Last edited:

Johnny Ryall

Member
Oct 28, 2017
241
Nah the majority of people agree with you, only a handful seem to actually find this to be an issue. Even though it's apparently pretty good input lag compared to most modern games.

I mean holy smokes I couldn't imagine doing the
Mantis Lords
boss with lag.

Or god the
soul master
that made me swear and it was all me, no lag.
 

low-G

Member
Oct 25, 2017
8,144
Okay, new test. Same everything, except max prerendered frames reduced to 1 in Nvidia control panel, and in-game Vsync disabled. This time the character reacted on the 9th frame of animation rather than the 16th, which is a pretty massive difference. Will get a GIF edited into this post in a few minutes.

preren1_novsync1hsqu.gif

That's more like it. With a monitor with better latency you'd probably be looking at 5-6 frames with your kit, or about 41ms of latency. (Game probably is doing something funny to add latency in the first place judging from your results)

74.7 - 33.3 (your display) + some nominal latency
 

Alek

Games User Researcher
Verified
Oct 28, 2017
8,471
In general I find that even in a game with noticable input lag, like Ghost Recon Wildland's, 9/10 people don't notice it. So people saying 'I haven't noticed anything', likely just aren't sensitive to that type of thing.
 

Toad King

Member
Oct 27, 2017
941
Chicago
Did a quick test with my phone recording the screen at 240fps.

I counted the number of frames between me pressing the Y button down all the way and the main character (Hollow Knight? Does he have a name? Haven't heard one from the game yet) doing the first frame of his attack animation where he turns his head away from the screen. This is on a Pro Controller in wireless mode. Here are my results:

Code:
22
23
20
23
23
25
22
21
22
24
21
23
24
23
22
23
N/A (camera angle bad, could not pinpoint when I pressed the button exactly)
21
21
22
19
23
23
23
21
20
20

Average: 22.076923076923077 frames

Average ms at 239.998fps: ~92ms

Video I used for reference (download it to get the full 240fps version): https://drive.google.com/file/d/1sOVRha85zcNFyp20QEc-olFLa4lom_zk/view

Done on an ASUS VS247H which according to displaylag.com has a very good 11ms of input latency. The Switch HDMI is also going through an HDMI matrix but I doubt that's adding any extra latency.

At 60FPS, that's ~5.5 frames of input latency.
 

Mzo

Member
Nov 30, 2017
1,165
Sounds like it's just the jumps and it's part of the game.

I can't stand input lag and have been fine playing it on Switch.
 
Status
Not open for further replies.