RTSS now has new automatic Low-Lag VSYNC ON (raster based)

Everything about displays and monitors. 120Hz, 144Hz, 240Hz, 4K, 1440p, input lag, display shopping, monitor purchase decisions, compare, versus, debate, and more. Questions? Just ask!
User avatar
RealNC
Site Admin
Posts: 3741
Joined: 24 Dec 2013, 18:32
Contact:

Re: RTSS now has new automatic Low-Lag VSYNC ON (raster base

Post by RealNC » 01 Jan 2019, 19:06

Notty_PT wrote:My bad, I read the Gsync article on the website and ignored the fact that without VRR things are different. So to avoid tearing on a non VRR 144hz screen I need Vsync ON and if I want to avoid higher input lag I should cap to 143fps, in game if possible (althi rtss works too but with a bit more lag) right?
The in-game limiter might not be accurate enough. Also, capping to more than 0.1FPS below refresh rate will cause some microstutter because of repeated frames. If you do care, then you need an accurate limiter like RTSS and cap just slightly below your real refresh rate. See this guide.
So simply capping to 143 without vsync will not deliver tear free experience right?
Correct. Tearing is the result of not syncing the GPU's frames to the monitor, not the result of FPS vs Hz mismatch. Without a sync method, you will get tearing. However, tearing can feel less or more annoying depending on the framerate. Very high frame rates are way better.

Note that the scanline sync function of RTSS with vsync OFF is still a sync method. In this case, it's RTSS that does the syncing through software, not the GPU.
SteamGitHubStack Overflow
The views and opinions expressed in my posts are my own and do not necessarily reflect the official policy or position of Blur Busters.

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: RTSS now has new automatic Low-Lag VSYNC ON (raster base

Post by Chief Blur Buster » 03 Jan 2019, 20:43

RealNC wrote:
Notty_PT wrote:My bad, I read the Gsync article on the website and ignored the fact that without VRR things are different. So to avoid tearing on a non VRR 144hz screen I need Vsync ON and if I want to avoid higher input lag I should cap to 143fps, in game if possible (althi rtss works too but with a bit more lag) right?
The in-game limiter might not be accurate enough. Also, capping to more than 0.1FPS below refresh rate will cause some microstutter because of repeated frames. If you do care, then you need an accurate limiter like RTSS and cap just slightly below your real refresh rate. See this guide.
Correct.

For RTSS Scanline Sync, you need microsecond-accurate frame rate capping. So you cannot use the ingame limiter which is usually not that accurate.

Scanline Sync tearingless VSYNC OFF or the ultralow-lag VSYNC ON) requires using RTSS scan line sync as the capper, while leaving the game uncapped.

The "tearingless VSYNC OFF" tricks are extremely difficult and requires perfect framerate=refreshrate match, by hiding the tearlines exactly between refresh cycles (tearlines above top edge / below bottom edge) but is beautiful with blur reduction when successful.

It's a very fiddly mode but has several excellent use cases.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

Joker18
Posts: 4
Joined: 15 Jan 2019, 04:31

Re: RTSS now has new automatic Low-Lag VSYNC ON (raster base

Post by Joker18 » 15 Jan 2019, 04:48

Hi Guys,
As I couldn't find the answers I was looking for anywhere I decided to make my own tests of inputlag with different Sync Options.
I would like to share my results with you, perhaps this info will be useful to other people.

Vsync Off RTTS off
79,8 ms / 96.6 ms / 75.6 ms / 67.2 ms / 92.4 ms / 71.4 ms
~80.5 ms

RTTS on Scanline Sync On
100,8 ms / 113.4 ms / 121.8 ms / 100.8 ms / 105 ms / 117.6 ms
~109.9 ms

RTTS on Scanline Sync off
113.4 ms / 121.8 ms / 88.2 ms / 117.6 ms / 92.4 ms / 109.2 ms
~107.1 ms

Vsync on
147 ms / 134.4 ms / 142.8 ms / 134.4 ms / 147 ms / 159.6 ms
~ 144.2 ms

Vsync On (NVcpl) + Frame cap 0,01 below ingame.
96.6 ms / 100.8 ms / 109.2 ms / 121.8 ms / 96.6 ms / 113.4 ms /
~106.4 ms

Vsync On (NVcpl) + Frame cap 0,01 below RTTS
92.4ms / 109.2ms / 88.2ms / 92.4ms / 105ms / 88.2ms
~95,9ms


The mesurement was done Battle(non)sense style. ( LED connected to mouse click ), recorded HI-Speed and counted the frames between LED lightup and actual action ingame. Everything was done in Battlefield V FPS locked to 60 or 59.988 in last 2 cases.
Don't take this as a value because the measurement was done on my LG OLED TV which has Input lag of it's own.
What is important is the difference between different modes.
PS: The margin of error might be a bit high but still the values should give you an ideea of what improvements you could expect vs traditional Vsync
PS2: Also measured Nvidia 3D on vs Off. but for some reasons I forgot to save my results file when I added it and I only remember the average value which was about 17,5ms higher than Vsync Off RTTS off. So I think the correct value would be 1 frame (16,66ms) which is inline with some data I found in the past saying that 3D adds a buffer of 1 frame.

User avatar
RealNC
Site Admin
Posts: 3741
Joined: 24 Dec 2013, 18:32
Contact:

Re: RTSS now has new automatic Low-Lag VSYNC ON (raster base

Post by RealNC » 15 Jan 2019, 08:55

Joker18 wrote:As I couldn't find the answers I was looking for anywhere I decided to make my own tests of inputlag with different Sync Options.
I would like to share my results with you, perhaps this info will be useful to other people.
Thanks for these!

How many times did you repeat each test?
SteamGitHubStack Overflow
The views and opinions expressed in my posts are my own and do not necessarily reflect the official policy or position of Blur Busters.

Joker18
Posts: 4
Joined: 15 Jan 2019, 04:31

Re: RTSS now has new automatic Low-Lag VSYNC ON (raster base

Post by Joker18 » 15 Jan 2019, 09:04

I written the data only for 6 measurements for each, you have the values there.
Where I had more variation I checked aditional instances to see if there was an anomaly or not, but i found that the aditional measurements fall in the same interval so I left them out to have an even no.

os4321
Posts: 5
Joined: 20 Jun 2018, 18:24

Re: RTSS now has new automatic Low-Lag VSYNC ON (raster base

Post by os4321 » 15 Jan 2019, 12:21

Thanks for sharing. Interesting results.

I measured the input lag using an Arduino Due polling a photodiode at 1ms intervals. I found the input lag between Scanline sync vs vsync off (capped 60fps) to be practically identical in my tests although I dont know how valid my results are as I dont really have anything to compare them with.

Here's link to the graphs which also includes some other tests - https://imgur.com/a/4qTnm85
Here's link to the original thread discussion - viewtopic.php?f=10&t=4602

Looking at the graphs, here are some eyeball readings taken on the initial upward transition lines (i.e. when the photodiode first detects a change).

Vsync off + 60fps capped (in-game engine) + ultra settings
min= 25-26ms / max= 41-42ms / avg(median)= 33-34ms

RTSS Scanline sync + Vsync off + ultra settings
min= 25-26ms / max= 41-42ms / avg(median)= 33-34ms

Vsync on + ultra settings
min= 71-72ms / max= 87-89ms / avg(median)= 78-80ms

Vsync off + uncapped + ultra settings (170-180 fps)
min= 16-17ms / max= 33-34ms / avg(median)= 24-25ms

This was tested in BF4 Test Range on a 60hz monitor (Dell U2414H).
Last edited by os4321 on 15 Jan 2019, 13:15, edited 1 time in total.

User avatar
jorimt
Posts: 2481
Joined: 04 Nov 2016, 10:44
Location: USA

Re: RTSS now has new automatic Low-Lag VSYNC ON (raster base

Post by jorimt » 15 Jan 2019, 12:49

os4321 wrote:I measured the input lag using an Arduino Due polling a photodiode at 1ms intervals. I found the input lag between Scanline sync vs vsync off (capped 60fps) to be practically identical in my tests although I dont know how valid my results are as I dont really have anything to compare them with.

[...]

Looking at the graphs, here are some eyeball readings taken on the initial upward transition lines (e.g. when the photophiode first detects a change).
Yup, for 60Hz, those numbers looks as expected...

1) Scanline sync vs. V-SYNC OFF + FPS within refresh rate = virtually the same input lag (at least with middle screen readings; similar to G-SYNC vs. V-SYNC OFF in same scenario).
2) V-SYNC on + FPS above refresh rate = roughly 3 frames of added input lag (in line with the known range of between 2 and 6 that V-SYNC typically adds in that scenario).
3) V-SYNC OFF + 170-180 FPS above refresh rate = lower input lag than 1 & 2, primarily due to the increase of multiple frame scans in a single scanout cycle.
(jorimt: /jor-uhm-tee/)
Author: Blur Busters "G-SYNC 101" Series

Displays: ASUS PG27AQN, LG 48CX VR: Beyond, Quest 3, Reverb G2, Index OS: Windows 11 Pro Case: Fractal Design Torrent PSU: Seasonic PRIME TX-1000 MB: ASUS Z790 Hero CPU: Intel i9-13900k w/Noctua NH-U12A GPU: GIGABYTE RTX 4090 GAMING OC RAM: 32GB G.SKILL Trident Z5 DDR5 6400MHz CL32 SSDs: 2TB WD_BLACK SN850 (OS), 4TB WD_BLACK SN850X (Games) Keyboards: Wooting 60HE, Logitech G915 TKL Mice: Razer Viper Mini SE, Razer Viper 8kHz Sound: Creative Sound Blaster Katana V2 (speakers/amp/DAC), AFUL Performer 8 (IEMs)

Joker18
Posts: 4
Joined: 15 Jan 2019, 04:31

Re: RTSS now has new automatic Low-Lag VSYNC ON (raster base

Post by Joker18 » 15 Jan 2019, 15:21

As far as I know RTTS limiter ads a bit of inputlag compared to ingame limiter and this is one of the reasons it can achieve a much smoother experience.

Battle(non)sense also found a small difference between ingame limiter and RTTS https://youtu.be/rs0PYCpBJjc?t=244

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: RTSS now has new automatic Low-Lag VSYNC ON (raster base

Post by Chief Blur Buster » 15 Jan 2019, 15:54

Joker18 wrote:Vsync On (NVcpl) + Frame cap 0,01 below RTTS
92.4ms / 109.2ms / 88.2ms / 92.4ms / 105ms / 88.2ms
~95,9ms
That's very new!

Once upon a time, NVCPL was laggier than RTSS. But now it beats RTSS?

Time for a big retest of the framerate cappers!

P.S. Blur Busters is looking for people to run new tests like these. Freelance pay available. I rely on assistants as I have other Blur Busters things I need to work on ...like inventing other new testing techniques. If you have a modified LED mouse and a millisecond-accurate capture (e.g. via high speed camera or via photodiode) - then I want to pay you to do speciallized tests like these for Blur Busters articles. Inquire within: [email protected]. Cheers!
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

Joker18
Posts: 4
Joined: 15 Jan 2019, 04:31

Re: RTSS now has new automatic Low-Lag VSYNC ON (raster base

Post by Joker18 » 15 Jan 2019, 16:37

Chief Blur Buster wrote:
Joker18 wrote:Vsync On (NVcpl) + Frame cap 0,01 below RTTS
92.4ms / 109.2ms / 88.2ms / 92.4ms / 105ms / 88.2ms
~95,9ms
That's very new!

Once upon a time, NVCPL was laggier than RTSS. But now it beats RTSS?

Time for a big retest of the framerate cappers!
I this instance only Vsync was enabled in NVCpl and Framelimit was set in RTTS. this was " Low-Lag VSYNC ON"

The setup I have is not worthy of processional tests, this values make sense for me as I don't need ms accuracy. I just wanted to make a decision on what's best to use.
The camera is not 1 ms accurate it is 4,2 ms "accurate" so I don't always get the exact time when the LED turn on so already from here there could be 1-3 ms measurement errors. On the screen side (60Hz) there should not be much of an error considering there is a new image every 16.6ms so I can capture almost 4 frames for each displayed.

Post Reply