Discussion about Legitimacy of HTML5 Mouse Testing

All Unidentified Forum Objects go in this area! Any fun alien talk goes in this U.F.O. Abduction Lounge, even topics other than monitors or computers. Introduce yourself!
User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: Discussion about Legitimacy of HTML5 Mouse Testing

Post by Chief Blur Buster » 25 Nov 2020, 22:01

IMPORTANT: TestUFO Mouse Tester is NOT LAUNCHED YET.
1000WATT wrote:
25 Nov 2020, 21:39
I alone cannot find this test? :lol: TestUFO Mouse Tester
It's a brand new upcoming test. Undergoing internal beta testing.

Not yet launched.

Schizobeyondpills was criticizing a test that they have never seen; which is why this forum thread exists.
Chief Blur Buster wrote:
07 Nov 2020, 23:22
Also, adjusting the level of patience is a strategic move of a researcher to allow time for a researcher to generate peer reviewed content. I suggest you pause further critical commentary until reviewing the actual upcoming tests -- you're criticizing something that hasn't even been published.
I take this extremely seriously. Extremely so, to the point that diplomacy gloves is off in this thread and referree-call somebody else's "engineer snobbery disdain". This person has a history of putting-down the good-intentioned kids in these forums (whether they are wrong or not), while I'm trying to be like a patient friendly Chris Hadfield to them.

It's possible this person is high-pay real-time C++/ASM outfit, like a VxWorks system engineer or big-fund nation-funded esports team (metaphorical 21st century version of govt-funded USSR Olympics teams) or some other ultra-latency snobbery. There's a place for ultra latency, yes. But I can't stand this type of non-inclusiveness of ultra snobbery. I can't make any assumptions, but schizobeyondpills is highly intelligent and clearly low-level oldtimer programmer.

But he blanket-dismisses a not-yet-launched TestUFO Mouse Tester that he has never seen.

I rarely so clearly break my own "Be Nice" forum rules (Administrator Privilege!) and this is one of these exceptions. ;)

My ego on these forums are well known by longtime readers, and I'm fine with it. Most love me as friendly as a protector of esports kids and mainstream users, even if they are still sometimes learning. I think outside the box and shake widely held misbeliefs and data uselessnesses in the refresh rate race to retina refresh rates (and HTML5 uselessness falls so squarely in this sphere).

I love to improve the art of ASM as well as the art of JavaScript/C#/fourthlevel languages, to bring former kiddie scripts into useful realtime operating territory. It's impressive how some HTML5 WebXR pages are able to framepace so perfectly well and near laglessly on my Quest 2 VR headset too ...despite being sheer HTML5/JS.

We gotta help kids program precise software, even if they just do JavaScript or C# or whatever. I'll Chris Hadfield and Neil deGrasse Tyson my way into that, while I ShieldsUp on those ASM/C++ snob purists (even though I cut my teeth on 6502 ASM too).

"I want to be an astronaut!"
"I want to program games when I grow up!"

Today's kids are tomorrow's game programmers, and I will still help them make C#/JS more precise, even if they later move onto C++ when they are impressed by the imprecisions of lesser languages, or they get smart and learn how to jitter-compensate the imperfect languages. Or whatever route. More power, all routes are valid. Rising tides lifts all boats. Convince OS vendors while I'm having fun too. Sometimes it's fluidity-priority rather than latency-priority, since 1-frame stutters are more nauseous than adding +5ms latency. Everybody is different but I like to be inclusive.

But, to my credit, I have a long-known history of being super-reluctant to ban forum members, so even people like schizobeyondpills still continue to be forum members even though he has been banned from some other forums/banned from some Discord channels. The number of permanent-banned forum members on these forums (that are not spambots/adbots) are nearly zero -- less than a full single hand's of count -- in the 7 years that these forums have operated.

Great moderating here is why we're not Twitter/Facebook speil around here. Certainly, these forums are well-moderated but users already knew that when they signed up. Blur Busters Forums is well known (by most) as very responsible/studious in moderating these forums. Occasionally aggressive at moderating, but most of the time, gentle no-eggshell-cracking handling (for majority of forum members). There are certainly some complaints, but by majority, people are largely happy with how geek nirvana Blur Busters Forums is, much more pure (little non-tech discussion, little adbottery, etc).

And another the side effect (of not banning most people who try to waste my time) is that a lot of interesting discussion soup spews out from my interactions -- some speil of which actually incubates new businesses (it has happened once already) from discoveries in a strong debate -- so these threads are useful reads to others (much like the Amazing Human Visible Feats of The Milliseconds) and other unintended consequences of the refresh rate race to retina refresh rates.

Enjoy!
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!

1000WATT
Posts: 391
Joined: 22 Jul 2018, 05:44

Re: Discussion about Legitimacy of HTML5 Mouse Testing

Post by 1000WATT » 25 Nov 2020, 22:05

schizobeyondpills wrote:
08 Nov 2020, 02:15
no pre-judging, I block people who play games with browser open. Chrome is well known bloated browser, to the point people make memes out of it. the only thing worse engineered for low latency/high performance and more bloated than Chrome is Windows 10. Chrome comes pretty close on 2nd place.
My chrome when not in the foreground automatically goes into
"pssuspend". You talked so often about optimization and did not guess that this can be done for chrome. It turns out that your system is not optimized. :o
I often do not clearly state my thoughts. google translate is far from perfect. And in addition to the translator, I myself am mistaken. Do not take me seriously.

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

Re: Discussion about Legitimacy of HTML5 Mouse Testing

Post by Chief Blur Buster » 25 Nov 2020, 22:20

1000WATT wrote:
25 Nov 2020, 22:05
My chrome when not in the foreground automatically goes into
"pssuspend". You talked so often about optimization and did not guess that this can be done for chrome. It turns out that your system is not optimized. :o
True. At some point, the bloat (of 200 minimized tabs) starts to intrerfere because of memory starvation, not much memory for other apps to perform well.

Nontheless, a single browser tab (no other open tabs) is a reasonably well-oiled beast (from browser POV) thanks to years of browser optimization, when nothing else is running that surprisingly tells a large story in a single graph screenshot.

Browser noise in a graph is so recognizable to my eyes, that I can mentally filter that data out while looking at other data in the same graphs (e.g. browser-interference jitter by specific browser behaviors, superimposed on other easily interpretable data that isn't hidden by the browser bloat at all.).

Remember, I'm born deaf and I am very vision-focussed. For the hearing person, white noise and tree flutter doesn't erase the sounds of a tooting steamship or revving of a Ferrari. Likewise, for my own eyes, browser noise doesn't drown out other things. It's beautiful how I see things in a graph, being born deaf and super-analytical to a fault. I love interpreting imperfect tools (to stunning success), and I am smart enough that one-click tests are the Way Of The Future, for helping the advocacy goals, and getting industry needles moving faster. The Animation Time Graph is just simple beginner pianist. The Mouse Tester graph is Beethoven/Mozart of complex beautiful music to my eyes that allows me to tell people things.

Just like LIGO and VIRGO has to contend with the vibrations (TestUFO browser noise), it still successful captures useful data like gravity waves (Mouse imperfections). But LIGO/VIRGO is huge engineering feats, not one click away to an automated test that automatically prods the user to move the mouse faster and squeals flaws at the user ("Warning: Graph Has Detected Dropouts"), until the user is happy and finally screenshots some clean looking graph and posts on forums, and then smart discussion ensues. RInse, repeat. TestUFO magic sauce.

Anyway, I can usually recognize when people have too many tabs open, just by looking at a graph screenshot too, but I also immediately recognize when it's non-browser software that's doing the background processing too. To this deafie, I have the equivalent of a TestUFO-emulator in my brain -- I mentally emulate many tests before I invent them -- that's how I know about a lot of the Blur Busters behaviours of many displays -- and that's how I invented pursuit camera test.

I have no problem with upcoming flawed "TestUFO Mouse Tester" screenshots -- I can tell users if that's browser fault, system fault, mouse fault, etc. Eventually I'll probably make a glossary/FAQ of the 10 most common screenshot-patterns and label them ("This was found to be the mouse RGB driver". "This was found to be caused by having too many tabs open". "This was caused by OS not keeping up at 8000Hz") etc. etc. They all show very clear distinct graphpatterns.

Nontheless, schizobeyondpills has caused me to repeatedly rethink of how good a disclaimer I can orate in a 2-line textbox (with a potential link to a bigger disclaimer / FAQ), and to embed more mouse telemetry easter egg data almost steganographically-like (But still recognizably visible to end user) into the TestUFO Mouse Tester graphs. But in the end, they're just minor 1%-2% changes as a bone-throw to cover my bases. Like a new colorcode there, a new pixelshift there (to flag a specific imperfection), a redflag marker there, adding more dimensions (colorcode value, alphablend value, barwidth value, etc) to create some graphs that essentially look six or seven dimensional visualizations along two dimensions, or create multiple parallel overlaid graphs, etc, etc.

There are multiple internal graph designs, I'm just deciding which graph designs will be most useful while not being too complex to look at. There's many optimizations being done to make the graphs visually easier to interpret/troubleshoot. For example, one graph style hides a lot of jitter while a different graph amplifies visibility of jitter but kills a different advantage until I solved it by adding another dimensionality (like colorcoding or variable barwidths to other creativity), cramming more visualizable mouse telemetry into a single graph that allows me to do all the stories I mentioned, from just the same graph design! Blur Busters Einstein Force is very strong in the surgical graph design on its ability to do a lot of storytelling about a user's mouse (and related interference factors, including software), treating the graph as music to my eyes to "hear by eye" of all the common problem patterns.

I wholeheartedly admit to being late on many of my initiatives, running a lot of things on Valve Time. But good people who work with me fully respects that. Yet so was many visionaries years late on delivering. I wanted the Mouse Tester out earlier this month, so there's that. But people pay attention when I release a flagship new TestUFO test. Some of those 7-year-old tests are so simplistic (Blur Trail) but others are advanced paid code writes like the VRR emulators embedded in the DOTA test written for the NVIDIA 360Hz unveil.

This time around, TestUFO Mouse Tester is self-funded passion (like my own business' Google 20% Time of permitted indie projects) to help me support users more easily because I have an ulterior motive to be able to troubleshoot people more easily just by glancing a TestUFO Mouse Tester graph, in addition to the simultaneous goals of pushing mouse innovation in the refresh rate race to retina refresh rates.

Mind you, I'm just a smalltimer, but I'm brilliant in my niche -- nobody has successfully blanket-reproduced TestUFO from scratch (without copying my code) because few have the simultaneous combination of a brain that successfully emulates motion tests in the brain before they're written (including even visualizing the optical illusions and VRR emulation and Pursuit Camera Sync Track in memory.

Longtime readers know that some of this spawned research papers and university papers AFTER I came up with those mental TestUFO breakthroughs. I love turning $30,000 test equipment into $100-or-free sometimes, bringing things into the realm of hobbyist reviewers. They're not perfect, but I'm a strong defender of DIY tests and hobbyist testing, even if there's a fair bit of imperfect data out there)

Anyway, it is all in the knowledge of interpreting the TestUFO mouse diagnostic tool correctly for your needs (for esports, for mainstream, for user experience of high Hz, etc).
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!

1000WATT
Posts: 391
Joined: 22 Jul 2018, 05:44

Re: Discussion about Legitimacy of HTML5 Mouse Testing

Post by 1000WATT » 25 Nov 2020, 22:52

Chief Blur Buster wrote:
25 Nov 2020, 22:20
True. At some point, the bloat (of 200 minimized tabs) starts to intrerfere because of memory starvation, not much memory for other apps to perform well.
I have no complaints about chrome about the amount of memory used. How much ram to buy is decided by everyone for himself. But there are claims to use the processor in minimized mode.
schizobeyondpills wrote: no pre-judging, I block people who play games with browser open. Chrome is well known bloated browser, to the point people make memes out of it. the only thing worse engineered for low latency/high performance and more bloated than Chrome is Windows 10.
When the application is in "pssuspend". No extra load on the processor. And this can be done automatically as soon as the application is out of focus or minimized.

And it seems to me that the problem can be solved with RAM. I'll go and try.
I'm wasting your time too. https://youtu.be/1sDbsj2mbuc
Last edited by 1000WATT on 25 Nov 2020, 23:15, edited 1 time in total.
I often do not clearly state my thoughts. google translate is far from perfect. And in addition to the translator, I myself am mistaken. Do not take me seriously.

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

Re: Discussion about Legitimacy of HTML5 Mouse Testing

Post by Chief Blur Buster » 25 Nov 2020, 22:59

Anyway, it's all beside the point anyway -- browser strawman arguments, with the fallen strawman decaying in a totally different forest than I'm working in. I'm seeing my forest and each tree in my own forest, that other metaphorical missed-forest-for-the-trees is unimportant for now.

Internal tests show better-than-expected storytelling. TestUFO Mouse Tester is a great mouse diagnostic test that tests more than just the mouse too. The graphs I create are beautiful storytells of the user's system.

Upcoming Future TestUFO Mouse Tester
Story Examples Now Achievable (from internal tests)
  • "It's browser related, I see you're using ACMEv77, upgrade to v83"
  • "It's background software related"
  • "The mouse is creating that problem. Are you using sensor #####?"
  • "That looks like the fake 2000Hz overclock"
  • "Lower your poll Hz to 2000Hz. Your system is chopping up those 8000Hz polls with a repeatable 200ms of data gaps between 100ms of successful 0.125us polls, from browser overflow bug. Currently, 8000Hz is usually overkill for your 240Hz monitor until the software vendor upgrades more of the software."
  • "I see you're at 1000Hz poll for 360Hz, that's the cause of your bad mousefeel. Switch to 2000Hz at least and test again"
  • "That looks like OS is not keeping up with those random 0.125/0.250/0.375us dropouts during your full velocity circling. Try again at 2000Hz or 4000Hz and see if your mouse data starts to sing."
  • "That definitely looks like you have too many browser tabs, that spikepattern is consistent of having multiple browser tabs open"
  • "It looks like you have mouse smoothing enabled, is that what you wanted for your mouse? Usually that will add lag, though it can feel smoother."
  • "Try plugging your mouse into a less congested USB port, it looks like USB contention to me"
  • "That's too noisy for me to interpret, can you rerun rebooting without any system tray stuff, and only run one browser tab"
  • "That browser is probably adding Meltdown/Spectre timer fuzz to those PointerEvent timestamps, run again in Browser XXX instead."
  • "It looks like like very bad DPC noise or poll jitter. I'm not 100% sure, but other graphs looking like that was traced to that."
  • "Nice asynchronous compensation of your poll jitter; I even see the 6000Hz sensor behavior in your 1000Hz mouse. I don't often see that punch all the way to the HTML5 graph. May I guess you're using sensor XXXXXX?"
  • "I see some browser performance issues; but ignoring that, I also see the jitter punching through that noisefloor that reminds me of going without a mousepad on a smooth desk. Are you using a mousepad?"
  • "Great mouse performance! If you want to test the latency too, try [the other tool], to see how low latency your great smoothness is"
  • "Let me send that to the mouse manufacturer, that data definitely looks interesting."
  • "That's a lot of background processing that's definitely not caused by the browser. That will probably interfere with your games too."
  • "You didn't move your mouse fast enough during your test. Can you send another screenshot with you moving the mouse faster? What DPI are you currently at?"
  • "Interesting jitter difference between plastic mousepad, melamine desk, and cloth mousepad!"
  • "That's amazing mouse data even though slightly polluted by browser issues at erratic moments. Are your games smooth in mouse movements?"
  • "Your pollrate is too low for your refresh rate"
  • "You've got amazing smoothness compared to others. Great optimized test run. Oh! You were even already running that 3D benchmarker in an adjacent window in that desktop screenshot? I see only two asset-load spikes but otherwise virtually jitter-free relative to others. If so, that's great system performance. Even though TestUFO Mouse Tester doesn't always translates to mouse performance in games;' hopefully it translates well to your games."
  • etc.
  • etc.
  • etc.
Bluntly put, the upcoming TestUFO Mouse Tester is not a mouse lag tester. (But it will indirectly help that innovation too)

But the upcoming TestUFO Mouse Tester is still going to be a kick-ass mouse fluidity diagnostic test (including from non-mouse causes), thanks to brilliantly converged multidimensional visualization of pretty-eye-candy animated graphs, as explained a few forum pages ago. The data I am able to embed into a realtime HTML5 graph just literally carpetbombs the naysayers about the uselessness of HTML5 in helping the refresh rate race to retina refresh rates. It's already several-orders of magnitude more visually useful to MY own eyes (And similarly smart interpretation), than the next-most-useful HTML5-based mouse tester.

Sure. Mind you, it's still a narrowscope mouse tester, but it's still brilliant at what it can do.

Indeed, it will probably quickly set the gold standard in user-friendly easy mouse fluidity diagnostic testing overnight, and simple browser bugs will quickly solve (e.g. source code commit to fix the Chromium PointerEvents array overflow for 8000Hz mice) long before 2000Hz/4000Hz/8000Hz mice becomes widespread.

Especially after my article launches, I'm probably the loudest person in the world to scream the bullhorn of 1000Hz poll rate becoming a weak link for 360Hz+ monitors, making it necessary to upgrade poll rates for higher refresh rates. High-frequency jitter is motion blur. Blur Busting is eliminating motion blur. It's part of my passion. This specific TestUFO Mouse Tester test is currently one of my unfunded passion projects (except for free samples from the first manufacturers willing to provide said requested samples) much like the unpaid Pursuit Camera Sync Track project was. (For those familiar with Google engineers, it's sorta like a 20%-time passion project)

P.S. Please excuse me while I go silent on this thread in my roughly occasional weekly visit to this HTML5-naysayer thread. I'll defend the HTML5-imperfection dissertations for as long as I need to, but at a reduced frequency as a time-management tactic. I shall balance forum vs busy-in-lab stuff.

</Achivement Unlocked for Esports>
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!

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

Re: Discussion about Legitimacy of HTML5 Mouse Testing

Post by Chief Blur Buster » 26 Nov 2020, 08:18

Offtopic post about 120Hz mainstreaming forked to a different thread:
Debate about 120Hz Mainstreaming
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!

User avatar
schizobeyondpills
Posts: 103
Joined: 06 Jun 2020, 04:00

Re: Discussion about Legitimacy of HTML5 Mouse Testing

Post by schizobeyondpills » 26 Nov 2020, 08:20

Not an offtopic if it destroys your argument by looking deeper into it. again with offtopic silencing ? :mrgreen: :mrgreen: :mrgreen: :mrgreen: :mrgreen: :mrgreen: :mrgreen:

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

Re: Discussion about Legitimacy of HTML5 Mouse Testing

Post by Chief Blur Buster » 26 Nov 2020, 08:35

schizobeyondpills wrote:
26 Nov 2020, 08:20
Not an offtopic if it destroys your argument by looking deeper into it. again with offtopic silencing ? :mrgreen: :mrgreen: :mrgreen: :mrgreen: :mrgreen: :mrgreen: :mrgreen:
:roll: :roll: :roll: :roll: :roll: :roll: :roll:

Good own goal, schizobeyondpills.

It wasn't silenced, it was simply moved over here.

Conflating response time to invalidate 120Hz mainstreaming. Great own goal.

Also, your specific snippet of a post just earned another 7-day time-out. See you next week.

Your post is so silly since you already know refresh rate & response are decoupled. 120Hz can remain as screens improve GtG -- i.e. vendors switch to OLED or MicroLED, etc. If you have any experience with 120Hz OLED panels (Galaxy phones), you notice they scale much more pure/linearly on Blur Busters Law since their GtG~99% 256x256 being pretty reliable tiny fraction of a Hz with OLEDs. Not all phones are LCD, some are OLED.

Besides, the 8K 1000Hz display engineering path in question has a pure GtG100% 256x256 heatmap, not being of traditional LCD panel nor OLED panel, and not even DLP. It isn't something you can put on a desk.

Also if you teardown and GtG-heatmap the VR LCDs in HTC Vive and/or Oculus Rift S / Quest 2, it has some of the best LCD GtG heatmaps found of any LCD panels, far better than any desktop LCD. The Oculus Quest 2 LCD, in particular, apparently manages zero strobe crosstalk for entire top/center/bottom -- it is highly likely in the territory of GtG 99.5%-100% completeness during strobing moments, far better than the next best strobed LCD. One of these days, somebody is going to have to teardown a VR LCD, and put a continuous backlight on it (disabling the low persistence) and GtG-heatmap it. It's almost a Holy Grail LCD in how clean its GtG is, zero ghosts in www.testufo.com/crosstalk in Quest 2's in-VR web browser -- I couldn't detect a single pixel of strobe crosstalk anywhere, it was perfect CRT clarity with no hints of ghosts.

And we all know JumboTrons are already using 600Hz-1920Hz 32x32 and 64x64 LED matrixes, that are repeat-refreshing 60Hz frames at darn nigh practically perfect GtG100% 256x256. But can be modded with custom chips to refresh 600-1920 unique images per second, for a retina-Hz JumboTron. You can read all about the retina refresh rate JumboTron concept.
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!

Post Reply