Gears Forums

Detailed INPUT LAG analysis of Gears of War UE, 4, and 5 (Bad news for Gears 5)

This is a very detailed discussion, so I encourage you to read and re-read before you post. There is A LOT to take in!

The Coalition needs to be aware of this large amount of input lag, and they need to address this before Gears 5 is officially released, and the players need to be aware of the lag difference. This alone I would consider justification for crossplay to be disabled. Gears Ultimate Edition demonstrates that the Xbox One hardware is capable of much lower lag times, so TC should fix this! THEY MADE ULTIMATE EDITION!

As it is now, the Gears 5 Playtest on the Xbox One X has over 4 additional lag frames compared to the PC version. At 60fps, this equals to 70ms of added latency for the console players. I created a chart to show the input lag of several Gears games on the Xbox One X hardware vs PC.

For my lag sampling, I am using a Xbox One X console set to output 1080p@60Hz with Freesync, a PC with an AMD GPU to also use Freesync, and a wireless controller. For Gears 4 and 5, I used Horde mode and Bootcamp. Both of these modes on the Xbox One X run at 60fps. Specifically with Gears 4 using the “performance”" setting to make Horde run at 60Hz. I am using a Samsung Freesync display set to 1080p@60Hz. On my PC, AMD gpu’s do support the new “Anti-Lag” feature. However, I did not use this feature as it would further skew results into PC’s favor.

I use a 240fps camera to record myself pressing the d-pad and count frames until an action is seen on the screen. For UE, it is the main menu icons. For Gears 4 and 5, it is the white weapon icons. The values in my chart from columns B->P are my lag values from the 240fps recording.

If you really want to see my tests, I created a youtube playlist. Unfortunately, youtube down-samples all my 240fps videos into 60fps, but I did create a frame by frame video of how I obtained these samples in the playlist.

Input lag tests for all Xbox One and PC Gears titles - use the “.” key in youtube and it allows you to advance 1 frame at a time.

To get some very annoying responses out of the way, here are several common responses to these kinds of tests:

  1. “Use a wired controller, the lag will go away.” - Wired will not improve latency, and both my Xbox One X and PC use wireless with a controller using the latest firmware, so even if there was any latency (there isn’t), it would exist on both platforms equally. I am using the official Scorpio controller that came with my Xbox One X, and an official USB wireless adapter for PC.

  2. “Your display is causing the lag, not the game. I don’t feel any lag because I use game mode.” - Any lag caused by the display would add equally to both tests, so the overall difference in lag between platforms would still be the same. In order to use Freesync on my Samsung display, game mode must also be active anyways,

  3. “STOP COMPARING PC TO CONSOLES YOU PC ELITIST!” - No. because crossplay is active, it is very important for players to know that this advantage exists. Considering Gears 5 adds 70 extra milliseconds to the player’s reactions, it is substantial. 70ms is higher than most people’s ping in game! It takes longer for the Xbox to render the image to your TV than it does for your controller data to travel 100s of miles to a server.

  4. “The Xbox One hardware can’t have lower input lag because of buffering and vsync.” - Ultimate Edition says otherwise.

  5. “No one can feel that low of lag, so it doesn’t matter” - I made a frame by frame comparison showing how far your reticle can drift in 4 frames. If you can’t feel this amount of lag in your controls, well, that’s your problem.

    If you tried playing a game like Guitar Hero with 70ms added latency, you would miss 1/2 your notes, which is why rhythm games have an input lag slider in its options.

So, what can TC do about this? Well, they can make the lag tolerable like it was in Ultimate Edition with a mere 1 frame added. It could be the implementation of Unreal Engine 1 vs 4 of how it runs on the Xbox One hardware causing extra lag. Who knows. But if TC would add proper graphics options like a “performance mode” to disable vsync, remove any post processing, buffering, and allow Freesync users full control over the framerate, then the experience would be better for everyone. Some people would definitely prefer removing some graphical prettiness to have a more responsive experience.

50 Likes

So you dont think this has to do with just having too many people on the servers? They are still deploying more. I have had matches that felt fine.

Wow

7 Likes

I used Bootcamp for my Gears5 tests, which doesn’t use online. This is specifically why I requested people read before they post. Even online, having tons of high ping players would be irrelevant to an individual’s controller lag.

5 Likes

What do you mean it doesnt use online? You cant do bootcamp until you connect to the server.

Controller lag is an independent thing from server lag. What you do on your own controller will display on screen regardless if a server is present or not.

2 Likes

But there is a lag to the server between when you put an input and when it appears on screen. This is due to ping and packet loss.

Im not 100% sold this is just the game. If so it would be completely broken and there is no way pros would be playing it.

I hope TC sees this. Gears 1 and UE had next to no input delay. And if the PC versions can have comparable input delays, then TC really needs to figure out how to fix this for Xbox players.

2 Likes

That’s not how ping works. If you do doubt my tests, I highly encourage you to do your own. This is a very detailed discussion, and you seem to lack some basic understanding of how controller output and server lag work. The server is not a variable when it comes to your own controller outputting to the screen. Plus, any lag that existed would also be present on my PC tests, which it is not. Plus again, if controller lag was caused by the servers (which it isn’t) then any time people’s pings shoot up in game, they would be reporting unresponsive controls, which no one has done.

7 Likes

It has nothing to do with when it appears on the screen. Has to do with the time it takes to make it to the server then display on your screen.

But i would agree with you if the server wasnt being used and this was an offline mode.

Connection won’t affect your inputs. It would only affect hit detection and character positions (ie. rubber banding). But when you move your crosshair or press a button, the resulting action will be local, not server side. Otherwise you would see a major difference in input based on ping, which doesn’t happen. The inputs don’t change even when your ping changes.

Also the lowest ping you can really hope for is going to be something like 30ms, which is 5 times higher than the input delays as the OP describes.

Sorry but you’re just wrong here.

3 Likes

Just a correction, the “5 times” thing you wrote seems inaccurate. You might be looking at the column for lag frames and not the column for milliseconds. But your description is spot on. With my samples, Gears 5 has 111ms responsiveness, or roughly 2 to 3 times what some people’s ping is.

Yea you’re right xD

But yea if input was based on ping it would change wildly every game, and even during the game. That’s just not true. It doesn’t matter what the ping is, inputs will be consistent. It’s just the results of said inputs may change, like you teleporting around or bullets not registering as hits, etc.

1 Like

They need remove V-SYNC and Triple buffering on xbox-x the results are INSANE.

They need as well cap MULTIPLAYER on PC to 60 FPS to avoid the unfair advantajes.

5 Likes

Ping absolutely impacts my inputs. Im not quite sure what you mean. When my ping is low movement is faster, my shots are faster, everything is faster. I just dont buy this analysis. If this was the case game would be broken. Pros wouldnt touch it.

2 Likes

I would 110% disagree that this is justification to disable crossplay. 4 frames will get you something at high level play, for sure, but over the course of a match that’s going to be VERY negligible. Unless someone here can claim to be a perfect player down to 70ms of difference (in which case I’m calling BS).

Still, watching the videos, the evidence is pretty clear, so good job putting this all together. I agree that options should be given to console to help alleviate this.

1 Like

you obviously dont understand the subject matter.

4 Likes

Whatever you say i have more gaming experience than you.

yikes

2 Likes