I am considering locking my fps but i have heard that you get input lag from that and if that is true and if it is how much input lag.
FPS lock and input lag
posted in
TF2 General Discussion
I am considering locking my fps but i have heard that you get input lag from that and if that is true and if it is how much input lag.
Depends on how you lock your FPS and what you mean by input lag. You have 2 forms of input lag here.
- Higher average input lag if your fps cap is below what your framerate would be in a given situation without it.
- Literal added input lag not related to the actual frame rate.
If you fps cap to say 60fps and you're about to do a midfight where you would normally have say 100fps you would have about ~6.6ms more input lag on average.
You use V-Sync/G-Sync etc to cap your frame rate (in the case of G-sync you use ONLY G-sync to cap your frame rate) you will have +1 frame of input lag constantly (or +2 if using triple buffer) + Input lag from No.1 still applies here as well so if you use my first example you would have about 6.6ms higher average input lag + 16.7ms of inputlag from V-Sync buffer OR 33.3ms of added input lag from triple buffer. G-Sync also applies here if your framerate is constantly at your max refresh rate which should only have +1 frame of input lag of whatever the refresh rate is.
Using RTSS or other software based frame limiting adds at least some input lag not associated with the lower framerate. This is largely due to such software having to react to the resource demands of whatever game you're trying to play. This input lag is usually a good bit lower than an added frame of input lag (especially if you're looking at 60fps/Hz for V-Sync) also note that doing V-Sync with RTSS to achieve the low input lag (i.e. no added frame buffer) can cause horrible micro stutters or "barely" noticeable micro stutters.
In game FPS limiting is usually significantly better than software based limiting due to it being within the program requesting resources and being able to be much more proactive about it. (unless you're trying to limit to 1/10000th of a frame) i.e. 59.999fps rather than 60fps. Now sometimes game engines have bugged fps caps or bugged FPS caps on certain systems.
Here are at least 2 issues I've found with fps_max as a note these were found via trial and error and may be a correlation rather than the actual cause
1. Horrible micro stutters regardless of framerate cap.
I've personally experienced this on 1 out the 6 PCs I've tested this on I'm not sure on the root cause of this but the issue was still occurring with a fps cap of 999 or if I set it to my refresh rate which makes 0 sense to me. The 1 PC I had with this issue I stopped testing because of how bad it was.
2. Mild micro stutters regardless of framerate cap.
This I've seen on 4 of the 5 PCs I've tested on, now this may not be an actual bug since this could have been related to my framerate timing relative to my monitor refreshing
3. Input lag
Now this is a bit tricky cause it wasn't the same on every PC I tried but the one that happened the most was if I did absolutely nothing and then just started to move my mouse around the initial input would take FOREVER (4 of 5 PCs) but otherwise was pretty smooth and very reactive if I changed directions (2 of those 4). Additional info: This only occurred with fps_max values that were relatively low i.e. less than 200 and rarely higher than that. Issue occured on 60Hz/144Hz/240Hz monitors/settings even if I frame capped to my refresh and was legitimately worse than V-Sync input lag at least for the initial movement.
To date I've had 1 PC that I've had basically no issue with fps_max set to non 0 or something absurdly high like 720 or 960.
Depends on how you lock your FPS and what you mean by input lag. You have 2 forms of input lag here.
[olist]
[*] Higher [i]average[/i] input lag if your fps cap is below what your framerate would be in a given situation without it.
[*] Literal added input lag not related to the actual frame rate.
[/olist]
If you fps cap to say 60fps and you're about to do a midfight where you would normally have say 100fps you would have about ~6.6ms more input lag on average.
You use V-Sync/G-Sync etc to cap your frame rate (in the case of G-sync you use ONLY G-sync to cap your frame rate) you will have +1 frame of input lag constantly (or +2 if using triple buffer) + Input lag from No.1 still applies here as well so if you use my first example you would have about 6.6ms higher average input lag + 16.7ms of inputlag from V-Sync buffer OR 33.3ms of added input lag from triple buffer. G-Sync also applies here if your framerate is constantly at your max refresh rate which should only have +1 frame of input lag of whatever the refresh rate is.
Using RTSS or other software based frame limiting adds at least some input lag not associated with the lower framerate. This is largely due to such software having to react to the resource demands of whatever game you're trying to play. This input lag is usually a good bit lower than an added frame of input lag (especially if you're looking at 60fps/Hz for V-Sync) also note that doing V-Sync with RTSS to achieve the low input lag (i.e. no added frame buffer) can cause horrible micro stutters or "barely" noticeable micro stutters.
In game FPS limiting is usually significantly better than software based limiting due to it being within the program requesting resources and being able to be much more proactive about it. (unless you're trying to limit to 1/10000th of a frame) i.e. 59.999fps rather than 60fps. Now sometimes game engines have bugged fps caps or bugged FPS caps on certain systems.
Here are at least 2 issues I've found with fps_max [i][b]as a note these were found via trial and error and may be a correlation rather than the actual cause[/b][/i]
1. Horrible micro stutters regardless of framerate cap.
[i]I've personally experienced this on 1 out the 6 PCs I've tested this on I'm not sure on the root cause of this but the issue was still occurring with a fps cap of 999 or if I set it to my refresh rate which makes 0 sense to me. The 1 PC I had with this issue I stopped testing because of how bad it was.[/i]
2. Mild micro stutters regardless of framerate cap.
[i]This I've seen on 4 of the 5 PCs I've tested on, now this may not be an actual bug since this could have been related to my framerate timing relative to my monitor refreshing[/i]
3. Input lag
[i]Now this is a bit tricky cause it wasn't the same on every PC I tried but the one that happened the most was if I did absolutely nothing and then just started to move my mouse around the initial input would take FOREVER (4 of 5 PCs) but otherwise was pretty smooth and very reactive if I changed directions (2 of those 4).[/i] [i]Additional info: This only occurred with fps_max values that were relatively low i.e. less than 200 and rarely higher than that. Issue occured on 60Hz/144Hz/240Hz monitors/settings even if I frame capped to my refresh and was legitimately worse than V-Sync input lag at least for the initial movement.[/i]
To date I've had 1 PC that I've had basically no issue with fps_max set to non 0 or something absurdly high like 720 or 960.
[olist]
[*] Higher [i]average[/i] input lag if your fps cap is below what your framerate would be in a given situation without it.
[*] Literal added input lag not related to the actual frame rate.
[/olist]
If you fps cap to say 60fps and you're about to do a midfight where you would normally have say 100fps you would have about ~6.6ms more input lag on average.
You use V-Sync/G-Sync etc to cap your frame rate (in the case of G-sync you use ONLY G-sync to cap your frame rate) you will have +1 frame of input lag constantly (or +2 if using triple buffer) + Input lag from No.1 still applies here as well so if you use my first example you would have about 6.6ms higher average input lag + 16.7ms of inputlag from V-Sync buffer OR 33.3ms of added input lag from triple buffer. G-Sync also applies here if your framerate is constantly at your max refresh rate which should only have +1 frame of input lag of whatever the refresh rate is.
Using RTSS or other software based frame limiting adds at least some input lag not associated with the lower framerate. This is largely due to such software having to react to the resource demands of whatever game you're trying to play. This input lag is usually a good bit lower than an added frame of input lag (especially if you're looking at 60fps/Hz for V-Sync) also note that doing V-Sync with RTSS to achieve the low input lag (i.e. no added frame buffer) can cause horrible micro stutters or "barely" noticeable micro stutters.
In game FPS limiting is usually significantly better than software based limiting due to it being within the program requesting resources and being able to be much more proactive about it. (unless you're trying to limit to 1/10000th of a frame) i.e. 59.999fps rather than 60fps. Now sometimes game engines have bugged fps caps or bugged FPS caps on certain systems.
Here are at least 2 issues I've found with fps_max [i][b]as a note these were found via trial and error and may be a correlation rather than the actual cause[/b][/i]
1. Horrible micro stutters regardless of framerate cap.
[i]I've personally experienced this on 1 out the 6 PCs I've tested this on I'm not sure on the root cause of this but the issue was still occurring with a fps cap of 999 or if I set it to my refresh rate which makes 0 sense to me. The 1 PC I had with this issue I stopped testing because of how bad it was.[/i]
2. Mild micro stutters regardless of framerate cap.
[i]This I've seen on 4 of the 5 PCs I've tested on, now this may not be an actual bug since this could have been related to my framerate timing relative to my monitor refreshing[/i]
3. Input lag
[i]Now this is a bit tricky cause it wasn't the same on every PC I tried but the one that happened the most was if I did absolutely nothing and then just started to move my mouse around the initial input would take FOREVER (4 of 5 PCs) but otherwise was pretty smooth and very reactive if I changed directions (2 of those 4).[/i] [i]Additional info: This only occurred with fps_max values that were relatively low i.e. less than 200 and rarely higher than that. Issue occured on 60Hz/144Hz/240Hz monitors/settings even if I frame capped to my refresh and was legitimately worse than V-Sync input lag at least for the initial movement.[/i]
To date I've had 1 PC that I've had basically no issue with fps_max set to non 0 or something absurdly high like 720 or 960.
Please
sign in through STEAM
to post a comment.