Mister Laggy: Any data regarding vsync_adjust?

For topics which do not fit in other specific forums.
darkwombat
Posts: 35
Joined: Thu May 30, 2024 2:49 pm
Has thanked: 1 time
Been thanked: 17 times

Mister Laggy: Any data regarding vsync_adjust?

Unread post by darkwombat »

We all know that vsync_adjust=2 has the lowest lag compared to vsync_adjust=1 and vsync_adjust=0.

I am curious if anyone has tested these via their Mister Laggy and has the data. I am curious how bad the lag is because I noticed quite a bit of lag if I am not on vsync_adjust=2

Also, have they tested these with VRR off and on?

In a nutshell I am trying to use vsync_adjust=1 (because I have the syncing when loading a new core) and see what we can do to get that setting as low lag as possible.

Thanks!

rhester72
Top Contributor
Posts: 1321
Joined: Thu Jun 11, 2020 2:31 am
Has thanked: 15 times
Been thanked: 213 times

Re: Mister Laggy: Any data regarding vsync_adjust?

Unread post by rhester72 »

I would think that would depend very much on the specific core and individual title refresh rate, no?

darkwombat
Posts: 35
Joined: Thu May 30, 2024 2:49 pm
Has thanked: 1 time
Been thanked: 17 times

Re: Mister Laggy: Any data regarding vsync_adjust?

Unread post by darkwombat »

rhester72 wrote: Mon Jul 01, 2024 4:03 pm

I would think that would depend very much on the specific core and individual title refresh rate, no?

For sure yes.

I am just curious. I am surprised that with all the people who have purchased one, the only thing we ever get are posts and youtube videos comparing an LCD vs another LCD or CRT. No one has taken the time to do real data, refresh rates, resolutions, mister.ini settings, to drill down and find the absolute best settings.

In the end, I may have to buy one and make a video or something myself.

rhester72
Top Contributor
Posts: 1321
Joined: Thu Jun 11, 2020 2:31 am
Has thanked: 15 times
Been thanked: 213 times

Re: Mister Laggy: Any data regarding vsync_adjust?

Unread post by rhester72 »

Well, given that the testing requires use of the Laggy test core, and I don't think it allows for a lot of variance in refresh rate or resolution, I'm not really sure how that would be possible...it sounds like you want real-world test data from actual cores, which is not at all how Laggy works.

darkwombat
Posts: 35
Joined: Thu May 30, 2024 2:49 pm
Has thanked: 1 time
Been thanked: 17 times

Re: Mister Laggy: Any data regarding vsync_adjust?

Unread post by darkwombat »

rhester72 wrote: Mon Jul 01, 2024 4:36 pm

Well, given that the testing requires use of the Laggy test core, and I don't think it allows for a lot of variance in refresh rate or resolution, I'm not really sure how that would be possible...it sounds like you want real-world test data from actual cores, which is not at all how Laggy works.

So changing the mister.ini settings have no effect on the Laggy?
had no idea.

User avatar
Retro-Nerd
Posts: 279
Joined: Fri Jul 08, 2022 2:47 am
Has thanked: 11 times
Been thanked: 80 times

Re: Mister Laggy: Any data regarding vsync_adjust?

Unread post by Retro-Nerd »

The Mister Laggy core takes every resolution and refresh rate you've set in the Mister.ini video settings. Apart from the built-in core resolutions.

The lag difference between all resolutions and refresh rates isn't very high. Mostly around 0,5ms-1m more/less depending on what you've selected. The important thing is (if you have LG devices e.g.) to activate VRR and vsync_adjust=2 (which works for me in 99% of all cores).

darkwombat
Posts: 35
Joined: Thu May 30, 2024 2:49 pm
Has thanked: 1 time
Been thanked: 17 times

Re: Mister Laggy: Any data regarding vsync_adjust?

Unread post by darkwombat »

Retro-Nerd wrote: Mon Jul 01, 2024 6:04 pm

The Mister Laggy core takes every resolution and refresh rate you've set in the Mister.ini video settings. Apart from the built-in core resolutions.

The lag difference between all resolutions and refresh rates isn't very high. Mostly around 0,5ms-1m more/less depending on what you've selected. The important thing is (if you have LG devices e.g.) to activate VRR and vsync_adjust=2 (which works for me in 99% of all cores).

Does VRR enabled with vsync_adjust=2 help with sync issues and blinking when loading a core?

User avatar
Retro-Nerd
Posts: 279
Joined: Fri Jul 08, 2022 2:47 am
Has thanked: 11 times
Been thanked: 80 times

Re: Mister Laggy: Any data regarding vsync_adjust?

Unread post by Retro-Nerd »

VRR enables the low latency boost mode, at least on LG TVs and monitors. It helps a lot with Arcade games which have weird refresh rates, like 55 or 57Hz. Without VRR such games can add 1-3 frames lag, on top of everything. With VRR i measure only 1-2ms extra lag for it, compared to normal computer and consoles refresh rates.

Does VRR enabled with vsync_adjust=2 help with sync issues and blinking when loading a core?

There is no way to avoid the short black screen sync drops if vsync_adjust=2 is enabled. If it annoys you, only vsync_adjust=1 helps.

darkwombat
Posts: 35
Joined: Thu May 30, 2024 2:49 pm
Has thanked: 1 time
Been thanked: 17 times

Re: Mister Laggy: Any data regarding vsync_adjust?

Unread post by darkwombat »

Retro-Nerd wrote: Mon Jul 01, 2024 6:47 pm

VRR enables the low latency boost mode, at least on LG TVs and monitors. It helps a lot with Arcade games which have weird refresh rates, like 55 or 57Hz. Without VRR such games can add 1-3 frames lag, on top of everything. With VRR i measure only 1-2ms extra lag for it, compared to normal computer and consoles refresh rates.

Does VRR enabled with vsync_adjust=2 help with sync issues and blinking when loading a core?

There is no way to avoid the short black screen sync drops if vsync_adjust=2 is enabled. If it annoys you, only vsync_adjust=1 helps.

What would you suggest to combat the lag that vsync_adjust=1 introduces?
This is really what I am getting at. How can I avoid that sync drop while still having minimal input lag?

User avatar
Retro-Nerd
Posts: 279
Joined: Fri Jul 08, 2022 2:47 am
Has thanked: 11 times
Been thanked: 80 times

Re: Mister Laggy: Any data regarding vsync_adjust?

Unread post by Retro-Nerd »

You can't a far as i see it (what Mister or your display device allows). It's either very low latency or around 1 frame of lag without sync drops. VRR could help on some monitors to minimize the frequency of sync drops, but i don't think you can eliminate them completely in vsync_adjust=2 mode.

KennyL
Posts: 121
Joined: Sun May 24, 2020 11:40 pm
Has thanked: 73 times
Been thanked: 50 times

Re: Mister Laggy: Any data regarding vsync_adjust?

Unread post by KennyL »

I tested my LG 42C2 after updating firmware to 13.30.80. Numbers are top, middle, and bottom boxes. Middle number matters the most since most of actions are happening in the middle of the screen.

lg42c2-lag.png
lg42c2-lag.png (159.73 KiB) Viewed 2680 times

Game Optimizer mode in HDR has extremely aggressive ABL so I also tested Filmmaker mode. It has better color and quick flashes are much brighter. The setting with lowest lag adds about 1 frame of lag at 60hz, which is not bad I think especially for modern 3d games. I can't tell even for 2d game since I'm not a lag freak heh heh. I'd say definitely try Filmmaker mode for modern HDR games. YMMV.

User avatar
Retro-Nerd
Posts: 279
Joined: Fri Jul 08, 2022 2:47 am
Has thanked: 11 times
Been thanked: 80 times

Re: Mister Laggy: Any data regarding vsync_adjust?

Unread post by Retro-Nerd »

Middle number matters the most since most of actions are happening in the middle of the screen.

I'm far from beeing an expert in that matter. But i don't think it's true. It's actually the bottom number that matters. 16,7ms takes every CRT (and OLED too, which uses rolling scan) to render a full frame @60Hz, everything above that value (e.g. 17.5-18ms bottom = 0.8-1.3ms lag) is the actual lag. It make sense to me. A CRT is much faster at the scan-in compared to an OLED (top number), catches up fast and at the scan-out (bottom number) they are pretty much equal.

And it confirms what Wickerwaka told about the "Low Latency Boost Mode" (via VRR mode enabled), and the expected lag on modern LGs displays.

Your table actually proofs what i've measured too with my C9. Nearly no display lag with vsync_adjust=2 and VRR. Rejoice. :D

KennyL
Posts: 121
Joined: Sun May 24, 2020 11:40 pm
Has thanked: 73 times
Been thanked: 50 times

Re: Mister Laggy: Any data regarding vsync_adjust?

Unread post by KennyL »

Ah ok thanks for the clarification. Tv reviewers all measure center of screen so I assumed middle is correct.

KennyL
Posts: 121
Joined: Sun May 24, 2020 11:40 pm
Has thanked: 73 times
Been thanked: 50 times

Re: Mister Laggy: Any data regarding vsync_adjust?

Unread post by KennyL »

LG C2 4:3 lag problem is still there with the latest firmware.

lg42c2-lag-02.png
lg42c2-lag-02.png (143.13 KiB) Viewed 2513 times

I asked Wickerwaka about which box matters the most:

center is probably the best, just for the sake of consistency, but they are all interesting. Top and bottom are the most useful measurements because they tell you how long before the TV starts displaying a frame and how long it takes the TV to show the whole frame.

Middle is probably the best approximation of what you eye will actually see.

User avatar
Retro-Nerd
Posts: 279
Joined: Fri Jul 08, 2022 2:47 am
Has thanked: 11 times
Been thanked: 80 times

Re: Mister Laggy: Any data regarding vsync_adjust?

Unread post by Retro-Nerd »

Yeah, C2 or later models still have massive lag in 4:3 modes, which is a shame. A firmware update could fix it easily. It's only a problem if you use this one 1920x1440p resolution directly via Mister. If you use any other output resolution via Mister, or a scaler (like the RT4k or the Morph 4K which both have HDMI-in) then it uses the 16:9 native display resolution of your TV anyway (and frames the aspect ratio you want inside this resolution, which is mostly 4:3 or 1:1 for something like Game Boy). And probably direct video over HDMI. Which is the best combi you could hope for. Believe me: Mister and Retrotink 4K via OLED is a dream. :D

Mister Laggy measurements via my Sony Trinitron CRT TV (60Hz)

top = 0,750ms
middle = 6,88ms
bottom = 13,74ms

The bottom value isn't 16,7ms probably due to the missing lines/not fully visible overscan area. Which is probably normal for the most (if not all) consumer TVs.

btw: My results with VRR on and off for 1080p@55Hz, LG OLED C9 (which is important for arcade games with weird refresh rates e.g. R-Type). As you can see, VRR on helps here a lot.

bottom VRR on+vsync adjust_2 enabled =18,6ms
bottom VRR off+vsync adjust_2 enabled =30,7ms

Post Reply