flood's input lag measurements

Everything about latency. Tips, testing methods, mouse lag, display lag, game engine lag, network lag, whole input lag chain, VSYNC OFF vs VSYNC ON, and more! Input Lag Articles on Blur Busters.
Sparky
Posts: 682
Joined: 15 Jan 2014, 02:29

Re: flood's input lag measurements

Post by Sparky » 21 May 2017, 17:38

flood wrote:idk honestly...
i don't have the time or interest that i used to

i'm still offering to give the rig away to anyone who wants :P (assuming that it still works)
I'd suggest someone that currently does latency testing via high speed camera.

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

Re: flood's input lag measurements

Post by Chief Blur Buster » 24 May 2017, 08:34

Blur Busters will be doing far more input lag related topics in the coming months. We're currently in a preparatory stage. "Keep tuned".
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!

knypol
Posts: 76
Joined: 18 Aug 2016, 03:40

Re: flood's input lag measurements

Post by knypol » 24 May 2017, 12:39

Chief Blur Buster wrote:Blur Busters will be doing far more input lag related topics in the coming months. We're currently in a preparatory stage. "Keep tuned".
Pls do test 60Hz NVCP Vsync ON + maxprerendered 1 vs 60Hz NVCP Vsync ON + RTSS at 60 (or other frame limiter) aswell.

Kulagin
Posts: 37
Joined: 27 Feb 2016, 08:17

Re: flood's input lag measurements

Post by Kulagin » 01 Jul 2017, 09:53

flood wrote:idk honestly...
i don't have the time or interest that i used to
i'm still offering to give the rig away to anyone who wants :P (assuming that it still works)
Woah. :o
Are you serious? Send it to me. I can take it.

I was just browsing and trying to find on how to make my own lag tester, even started few threads today about it. I was thinking about using https://www.humanbenchmark.com/tests/reactiontime and some 4-pin(5v, GND, -in, -out) photodiode, which would react to some specific color.

I wanted to use this method as follows:
Solder -in and -out of photodiode pins to + and - of LMB on the mouse.
Press a button on your main mouse to start the test, program then randomly changes color of the div to some specific color(let it be green), photodiode reacts to the color change and starts to pass current through in-out pins(presses LMB on the test mouse), once LMB is pressed, program measures the time between it changed color to green and the time LMB event happened.
Very simple and does exactly what you want it to do: it measures electron-to-photon-to-electron(aka reaction time test) latency.

And I was thinking about probably writing simpliest C# application later which will help to do both motion-to-photon-to-electron and electron-to-photon-to-electron latency tests yet with better accuracy using a pattern above with slight tweaks.

It isn't even close to perfect, worse than your setup but should require much smaller amount of time to make this thing yet good enough to measure latency changes when you change something like HPET, Line>MSI or timings on your monitor. But considering you say you can give away your whole rig, I am completely up to it :D

I know it is my first post here but I've been lurking around this kind of topics(including blurbusters) for many-many years now, reading and doing stuff(like overclocking, getting rid of cases from my GPUs, CPUs, reballing GPUs, modding cooling systems, soldering stuff, etc.)

Feel free to add me on Steam: http://steamcommunity.com/profiles/76561198026700417/
or on Discord: wF.Hardcore#0098

We could talk more in TeamSpeak(or any other VOIP) about it, I could tell you more about myself in case you want to know if I am worthy(which I am :mrgreen: ) of sending me this lag tester.

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

Re: flood's input lag measurements

Post by Chief Blur Buster » 13 Jul 2017, 15:28

Thanks for a great first post!

A high speed camera can be just as accurate as a photodiode, and so much easier -- sometimes you already have a high speed camera (new sports cameras, certain smartphones, etc). Some very basic 1000fps cameras, point-and-shoots, are available for cheap (~$200ish) albiet very low resolution (e.g. 64 pixel tall video). A photodiode is a 1-pixel camera, so it's often much better if you have a whole shebang of photodiodes, and all in a premade camera. I'm looking forward to the day when even 5000fps or 10000fps is affordable (<$1000), since that will provide great eSports lag testing capability.

Open source software will eventually need to be written to automate analysis of high speed videos (from trigger to display change, etc).

Browsers add a lot of input lag (FireFox, Chrome, Safari, Edge, IE, all have different input lag) so an executable is ideal, preferably a reaction time tester app running using the same game engine used by many games (e.g. copy of Unreal engine, copy of Source engine, etc)

Also, I have been beta-testing a brand new TestUFO Display Lag Tester (SMTT 2.0 style) and I am pleased that it achieves accuracy roughly identical to SMTT (and occasionally better). The TestUFO Display Lag Test has not yet been launched, but it can produce millisecond-accurate differentials between two displays.

Also, human reaction time is definitely part of the lag equation too. Separate tester software is needed for that (to help exclude web browser lag, which is often very different from lag of Direct3D/OpenGL, and all the various VSYNC ON / OFF / GSYNC / FreeSync modes).

We've got more lag test articles coming, keep tuned!

I'm a C# programmer myself too. Contact squad[at]blurbusters.com
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!

Sparky
Posts: 682
Joined: 15 Jan 2014, 02:29

Re: flood's input lag measurements

Post by Sparky » 13 Jul 2017, 17:39

Chief Blur Buster wrote:A high speed camera can be just as accurate as a photodiode
There are advantages and disadvantages, A high speed camera lets you see the pixel transitions and verify the accuracy of a photodiode setup(how long does it take between the monitor starting to transition and the diode triggering?) The microcontroller + photodiode setup makes it possible to isolate USB polling and eliminate debounce latency, makes gathering large amounts of data practical, and it is a whole lot cheaper than a high speed camera fast enough to get the same timing precision.

The microcontroller setup is also way less tedious. Getting a couple thousand data points is a matter of letting your monitor try to give your chair a seizure while you go grab a snack, and coming back 10 minutes later to save the csv.

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

Re: flood's input lag measurements

Post by Chief Blur Buster » 15 Jul 2017, 00:22

Sparky wrote:
Chief Blur Buster wrote:A high speed camera can be just as accurate as a photodiode
There are advantages and disadvantages, A high speed camera lets you see the pixel transitions and verify the accuracy of a photodiode setup(how long does it take between the monitor starting to transition and the diode triggering?) The microcontroller + photodiode setup makes it possible to isolate USB polling and eliminate debounce latency, makes gathering large amounts of data practical, and it is a whole lot cheaper than a high speed camera fast enough to get the same timing precision.

The microcontroller setup is also way less tedious. Getting a couple thousand data points is a matter of letting your monitor try to give your chair a seizure while you go grab a snack, and coming back 10 minutes later to save the csv.
That's all true of course, with goals of isolating certain parts of the chain.

It's good to isolate stages of lag testing -- e.g. display, or the cable, or game engine only -- but for a complete human button-to-pixels chain -- it is extremely difficult to automate lag testing of the actual lag of a real complete, true, bona-fide "mouse+USB+game engine/OS+GPU/drivers+cable transmission+display processing+panel" chain.

Executing actual hardware and actual games through the complete chain is extremely hard to automate, which is why we are VERY interested in finding ways to automate high speed video lag testing of actual games better -- including writing software to automate processing of analyzing high speed videos of gameplay videos.

SMTT tests can isolate display differentials, photodiode tests can automate quite a lot of things, Leo Bodnar can test absolute display lag
(Leo Bodnar -- top edge is roughly akin to SMTT differential relative to CRT. However, DisplayLag uses center rather than top -- which results in different Leo Bodnar measurements -- which makes it hard to directly compare DisplayLag.com to TFTCentral.co.uk). On the other hand, if your goal is the complete eSports button-to-pixels chain that includes *everything* that creates lag, we've found it is hard to beat good high speed camera tests (at this time).

The lag testing method depends on if you're trying to isolate certain parts of the input lag chain, or the whole chain all at once.

One of our diagrams (semi-simplified), which I'll reveal now;

Image

That said, we definitely have more lag features and lag testing techniques (including photodiodes). We're reluctant to reveal too much yet, to spoil surprises, but "Keep tuned". :D
[That said, if you're already working on lag testing techniques, please contact me mark[at]blurbusters.com ...]
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!

Sparky
Posts: 682
Joined: 15 Jan 2014, 02:29

Re: flood's input lag measurements

Post by Sparky » 15 Jul 2017, 01:47

Chief Blur Buster wrote: [That said, if you're already working on lag testing techniques, please contact me mark[at]blurbusters.com ...]
I'm not actively working on it at the moment, but I did write up my latency tester and post the code a while ago: http://forums.blurbusters.com/viewtopic ... =50#p22634

I'm happy to answer any questions about it.

flood
Posts: 929
Joined: 21 Dec 2013, 01:25

Re: flood's input lag measurements

Post by flood » 25 Jul 2017, 22:52

lol haven't checked this thread in a while
Kulagin wrote:
flood wrote:idk honestly...
i don't have the time or interest that i used to
i'm still offering to give the rig away to anyone who wants :P (assuming that it still works)
Woah. :o
Are you serious? Send it to me. I can take it.
k added you on steam (asdf)
flood wrote: too busy with stuff to work on this for the rest of the year probably :P
anyone want to take over? i'd be willing to give this rig away to anyone who
1. is curious and has more free time than i do
2. has a crt monitor
3. understands psychological biases and is careful with drawing conclusions from data
4. is reasonably knowledgeable in basic programming, electronics, data analysis, etc...
5. is not epileptic (during measurements, the screen flashes black/white 10-40 times a second in a dark room)

if i forget about this, pm spacediver and he can email me :P

spacediver
Posts: 505
Joined: 18 Dec 2013, 23:51

Re: flood's input lag measurements

Post by spacediver » 25 Jul 2017, 23:22

:)

Post Reply