120Hz Users: Help convince Microsoft to add 120Hz support to Internet Explorer

[Mark Rejhon]

Help Convince Microsoft To Support 120Hz in Internet Explorer
(Even if you hate Internet Explorer)

Internet Explorer does not support 120fps animations because it has Internet Explorer has a built-in 105fps framecap!!!
All other major browsers successfully do 120fps animations already but Internet Explorer does not support animations at 120 frames per second, due to its framecap.

I want to ask all 120Hz/144Hz monitor users (QNIX, ASUS, BENQ, etc) in HardForum population to do this:

1. See www.testufo.com in IE10 or IE11, or any website that supports 120Hz
   .
2. Observe IE incorrectly detects refresh rate, due to IE's framecap.
   .
3. Go to Microsoft.com 120Hz IE bug entry, and click "Sign In" at top to logon.
   .
4. After signing in, a new link appears. Simply
   Click the "I can too" link
   that appears underneath "User(s) can reproduce this bug" at the upper-right.
   .
   

That's it! It benefits the Web in the future, if the whole web supports 120Hz and 144Hz. Internet Explorer is the ONLY browser that has a built-in framerate cap that cannot be removed. We want all 120Hz and 144Hz users to make sure Microsoft falls in line with all the other web browsers in implementing 120fps@120Hz support like all other major browsers. I am exchanging emails to a Microsoft employee directly in a separate email, that they are reconsidering. What we need is a little extra push, since we have actual 120Hz and 144Hz users.

Appreciated. Thanks!

 

[J Macker]

Why do you care?

Just use a different browser.

 

[IdiotInCharge]

Why do you care?

Just use a different browser.

As a fellow IE hater, I agree with you, but note that it's not always an option.

 

[Mark Rejhon]

Why do you care?
Just use a different browser.

This is why: www.testufo.com/browser.html
Tomorrow's web designers and tomorrow's HTML5 game creators care.
If just ONE browser does not support 120Hz, some websites won't create 120Hz content for your favourite Chrome or Opera or FireFox browser.
Tomorrow's web designers may care. Tomorrow's HTML5 game creators care. Tomorrow's 120Hz users care.
No matter what 120Hz you use, QNIX, Catleap, Overlord, BENQ, ASUS, etc.

Chrome 18+ supports 120Hz VSYNC
Opera 15+ supports 120Hz VSYNC
FireFox 24+ supports 120Hz VSYNC (I'm the one who successfully convinced FireFox, see bugzilla #856427)
Safari 6+ supports 120Hz VSYNC (despite rare use of 120Hz on Mac)

Internet Explorer is the only browser that does not support 120Hz. When JavaScript tries to detect 120Hz, it detects 60Hz -- no way to animate at 120Hz from within IE.
IE10+ animations works perfectly at 100Hz, 101Hz, 102Hz, 103Hz, 104Hz, 105Hz, but suddenly halves framerate at 106Hz, 107Hz, 108Hz... (ToastyX/PowerStrip custom tests)
In private emails, Microsoft confirmed an intentional internal 105fps frame limiting cap, which exists even if animations only use 1 percent of CPU.
The Microsoft email, additionally confirmed, they are considering removing the 105fps cap.
We want 100% royal flush of all 5 major browsers to support 120Hz, so any web developer can create 120Hz animation content that works regardless of browser.

Web developer information: Spreadsheet: 120Hz tests in web browsers ... And matching JavaScript animations to refresh rate is actually a W3C recommendation, bottom of section 5: www.w3.org/TR/animation-timing/ .... And most browsers now follow this, frame caps should be discouraged from any web browser.

FireFox was successfully convinced to support 120Hz (My involvement in Bugzilla #856427). This shows a browser maker can be convinced by grassroots work like this.
Can you help convince Microsoft, too, for the sake of the whole web, and even to benefit your favourite browser?

 

[Morkai]

"Closed
as By Design"

 

[Falkentyne]

Why do you care?

Just use a different browser.

Can you possibly be MORE rude and anal?

 

[Mark Rejhon]

"Closed
as By Design"

They have promised me to reopen it in a private email to me:

Mark,

Thanks for bringing this issue to our attention. Seeing that 120Hz monitors will become more common in time, I think it’s reasonable to remove the clamp. I will reactivate the bug to see if we can have it reconsidered for IE11. If we do not implement this change this release, I think we should definitely consider it for a future version of IE.

Thanks,

Yes, I know a lot of you 120hz users hate Microsoft.
...But I've got a direct channel to Microsoft Internet Explorer team.
...I'm THIS close to convincing Microsoft. They just need a little more convincing.
Please help the WHOLE WEB!

 

[jojo69]

internet what?

 

[Mark Rejhon]

internet what?

If just ONE browser does not support 120Hz, some websites won't create 120Hz content for your favourite Chrome or Opera or FireFox browser.
Tomorrow's web designers may care. Tomorrow's HTML5 game creators care. Tomorrow's 120Hz users care.
No matter what 120Hz you use, QNIX, Catleap, Overlord, BENQ, ASUS, etc.

Chrome 18+ supports 120Hz VSYNC
Opera 15+ supports 120Hz VSYNC
FireFox 24+ supports 120Hz VSYNC (I'm the one who successfully convinced FireFox, see bugzilla #856427)
Safari 6+ supports 120Hz VSYNC (despite rare use of 120Hz on Mac)

 

[wonderfield]

We want 100% royal flush of all 5 major browsers to support 120Hz, so any web developer can create 120Hz animation content that works regardless of browser.

No one should not be creating animations at any fixed timestep and calling requestAnimationFrame to animate them. Doing so makes absolutely no sense, as you're never guaranteed when its callback will be fired: you'll invariably wind up animating at the wrong timestep. And if you use a setInterval-based animation loop to do fixed-step animation, you're also doing it wrong.

The agenda to remove the cap is a good thing, but you're mistaken as to the reason why. It's not as important as you suggest.

 

[Mark Rejhon]

Correct, browsers should not depend on framerates matching Hz, but the key problem is it's currently impossible to do so in Internet Explorer on 120Hz monitors.
It is the only web browser completely unable to do 120hz animations, even at 0% CPU.

No one should not be creating animations at any fixed timestep and calling requestAnimationFrame to animate them.

I am the author of the TestUFO motion tests.
I have done fluidity tests and animations are perfect in a lot more browsers than expected if you have a GPU-accelerated Windows/Mac system.

The use of JavaScript performance.now() enables high-precision performance measurements in Javascript now. I also have some animation self-measuring graphs at www.testufo.com/#test=animation-time-graph as well as www.testufo.com/#test=animation-time-graph&measure=rendering (it's interesting how different they look in different browsers, and how the rendergraph slopes upwards in some browsers as the graph gets more complex to render). Using such tools now makes it possible for me to create the above spreadsheet (it took me 4 hours), by looking for fluidity problems in various different www.testufo.com tests on various different systems. This is reliable and repeatable on ATI/nVidia/Intel GPU-accelerated Windows/Mac systems made in the last few years. Systems that are not GPU-accelerated will typically take a long time inside rAF() code. As a general rule of thumb, if code inside rAF() takes less than 2 milliseconds to execute, and it's one of the supported browsers, it's generally pretty certain (on average) it's animating perfectly synchronized to VSYNC. You will see I have fine-tuned the rAF() stutter-detection algorithm, especially when using Chrome.

Chrome 18+ -- confirmd perfect VSYNC-locked at all refresh rates (if everything in chrome://gpu is enabled)
Opera 15+ -- confirmed perfect VSYNC-locked at all refresh rates (uses the Chrome engine)
FireFox 24+ -- confirmed perfect VSYNC-locked at all refresh rates (a bit sensitive to stutters; FireFox knows & is fixing)
IE 10+ -- confirmed perfect VSYNC-locked until 105Hz, then halves framerate at 106Hz plus.
For more information about supported browsers, see www.testufo.com/browser.html
Some edge cases obviously exist (e.g. multimonitor users), but over time, this will resolve itself (some browsers now automatically sync to the refresh rate the window is overlapping the most on)

And there are other use cases:

(1) Perfectly fluid JavaScript-controlled scrolling.
For example, a page-scrolling function may want to scroll a page at the full refresh rate. The effect would be very similiar to this: http://www.testufo.com/#test=framerates-text
Except at 120Hz, this would have half the amount of motion blur (try this in Google Chrome, during 60Hz mode, and 120Hz mode, and you'll understand what I mean).

(2) More fluid photo slideshow transitions
View this: http://www.testufo.com/#test=photo on both 60Hz and 120Hz. A HTML5 app that uses panning transitions, would be able to have far less motion blur than this.

(3) 120fps video using HTML5 <VIDEO>
Many cameras such as the popular GoPro Hero 3 -- the POV camera used in Indy Races, surfing, skydiving, etc, have a high-speed 120fps HDTV mode. This is normally used for slow-motion playback, however it also produces extremely fluid motion when played back in real-time on a 120Hz monitor. It looks absolutely gorgeous. I know of one major video site experimenting with 120fps video (and it already works in at least one browser, since the H.264 / MP4 format supports 120fps video files that are configured to play back in real time instead of slow-motion)

(4) HTML5 gaming applications (e.g. platform scrolling game, space scrolling shoot-em-up)
During horizontal/vertical scrolling motion in games, there would be half as much motion blur on LCD displays, during fast horizontal and vertical scrolling. Scrolling is a very low CPU/GPU activity on modern GPU's taking less than a fraction of millisecond (excluding the time spent drawing new data at the edges of screen).

(5) (Speculative) Future Direct3D JavaScript, if Microsoft has plans.
Right now, FireFox running WebGL and JavaScript is running a full screen 3D video game in FireFox (no plugins) at 120 frames per second on a Geforce GTX 680 -- http://www.unrealengine.com/html5/ (This is not TestUFO tests; but full screen WebGL 3D graphics in 100% pure JavaScript and WebGL with zero plugins, executing the whole scene JavaScript 3D rendering at less than 5 milliseconds per frame at 1920x1080 <canvas> on my computer!). This is some interesting news: http://www.tomshardware.com/news/Epic-Citadel-HTML5-asm.js-UE3-Unreal-Engine,22406.html ...

(6) Teaching People -- Vision Science Education
I teach and educate people about motion blur using TestUFO links:
"Here's an educational demo of eye-tracking motion blur at http://www.testufo.com/#test=eyetracking "

But it does not work properly unless framerate=Hz, so I have to tell people:
"Here's an educational demo of eye-tracking motion blur at http://www.testufo.com/#test=eyetracking ... Please use Chrome to view this animation, IE10 does not support 120Hz"
(I've posted this about 30 times already to blogs and discussion forums).

Load this link in modern browsers with ATI/nVidia/recent Intel:
http://www.testufo.com/#test=eyetracking
- Chrome @ 60Hz looks correct
- Chrome @ 120Hz looks correct
- Opera15+ @ 60Hz looks correct
- Opera15+ @ 120Hz looks correct
- FireFox24+ @ 60Hz looks correct
- FireFox24+ @ 120Hz looks correct
- InternetExplorer10 @ 60Hz looks correct
- InternetExplorer10 @ 120Hz looks wrong on several models of 120Hz displays

(7) Detect the user's current refresh rate
-- Counting the number of requestAnimationFrames per second and trusting it (in certain popular browsers) to be usually equal the refresh rate, allows optimization of certain animations to be more optimal for specific framerates.
-- Detection of a 120Hz refresh rate can automatically trigger automatic playback of a higher-bandwidth 120fps video via the HTML5 <VIDEO> element
-- Detection of a 120Hz refresh rate can automatically configure certain special features in a web-browser-based game for 120Hz users, such as software-based black frame insertion (e.g. http://www.testufo.com/#test=blackframes ).

For a good test case of refresh rate detection, see http://www.testufo.com at the bottom of the screen. It waits until the loading time settles down (e.g. variances between rAF() becomes consistent) then begins counting (and becomes surprising more reliable than you'd expect when the timing deviation between rAf() becomes consistent while in foreground)

Believe me, I'm also involved as an Internet Standards writer ( http://www.xmpp.org/extensions/xep-0301.html -- I'm the author) and am considering diving into W3C involvement for JavaScript-based capabilities querying (e.g. querying for actual VSYNC status, querying more accurate refresh rate, etc). I believe I'm uniquely qualified, with my knowledge, to make a difference in the web browser world for 120Hz.

Perhaps HardForum isn't the right audience to bring it up to, but in case any fellow programmers/developers are reading this, then now you know the huge extent of my browser refresh rate research.

 

[Falkentyne]

It's fine, Mark. There are just a lot of trolls here (and everywhere else, TBH).
Remember, everyone has an opinion (and an ass).

 

[rudy]

I added a comment to support this cause.

 

[Mark Rejhon]

I added a comment to support this cause.

Much appreciated!

Don't forget to click the "I can too" link too:

 

[Black Octagon]

Mark, I would support this in a second if I hadn't removed my Microsoft/Live accounts years ago. From what I can tell, one needs such an account to log into this thing.

 

[Snowknight26]

Good luck with that.. Out of the over 80 IE bugs I've submitted, probably half have been closed as "Won't Fix."

 

[Mark Rejhon]

Good luck with that.. Out of the over 80 IE bugs I've submitted, probably half have been closed as "Won't Fix."

I've moved mountains before -- I successfully convinced FireFox employees to support 120Hz by my arguments on their Bugzilla #856427 and now is built into FireFox 24+ Beta. And also, This Microsoft reply is promising. But it is important to remind them on a continual basis, via several channels (internal and publically).

TestUFO (Blur Busters Motion Tests at www.testufo.com) visitor statistics indicate more 120Hz visitors than 60Hz visitors over the last three weeks (it is now possible to reliably detect refresh rate for web statistics with >95% accuracy in most recent version of browsers, using JavaScript rAF() counting technique after waiting for the rAF() deviations to stabilize during page loading), since TestUFO is a website specialized to people who care about calibrating for best motion fluidity (the same type of people who buy 120Hz monitors).

 

[Snowknight26]

It's certianly easier to push changes to Firefox/Chrome as they're open-source. IE, not so much. I've talked to a few Senior Escalation Engineers, Dean Hachamovitch and a few IE General Managers before but even having done that it's like pulling teeth to get the smallest changes implemented.

So again, good luck.

 

[Mark Rejhon]

It's certianly easier to push changes to Firefox/Chrome as they're open-source. IE, not so much. I've talked to a few Senior Escalation Engineers, Dean Hachamovitch and a few IE General Managers before but even having done that it's like pulling teeth to get the smallest changes implemented.

True, but I'm not one to give up easily. Do us a favour and next time you talk to one of them for any other reason, please also mention the 120Hz thing or contribute your experience. The odds slightly increase when multiple independent parties communicate to a single software developer team, that makes a specific team realize there's a wider issue widespread issue.

Eventually TestUFO could support < EMBED/ > so monitor reviews can embed live motion demos and tests (with play/pause) in monitor review webpages, and I could theoretically modify the JavaScript to display a honking "Internet Explorer does not support 120Hz" everytime a review reader views it in Internet Explorer. Warning even more high-end computer users away from Internet Explorer. Ouch for Microsoft. Obviously I don't want to do that, but it'll become an option if IE11 isn't fixed to support 120Hz, and this gives me the theoretical power to cause thousands of people to switch away from Internet Explorer (Blur Busters Blog = 11,000 unique visitors/week). I keep contact with some reviewer sites such as Adam, owner of pcmonitors.info as well Simon, owner of TFTCentral.co.uk (who I assisted in writing an article for), as well as others, who could embed TestUFO motion tests via < EMBED />, and their visitors, too, will also be warned away from Internet Explorer by the message displayed inside the TestUFO player embedded in their review. That thought is enough cause for Microsoft to reconsider.

So again, good luck.

Thanks, I'll need that too!

I write in other channels (e.g. bug tracking sites, W3C mailing lists, direct contacts, etc), but I also post here on HardForum chiefly because it is a major 120Hz audience, being the home of the Internet's most popular LightBoost thread (almost 200,000 views).

 

[rudy]

The least anyone can do is just follow it, report it etc, takes like 5 minutes, it lets MS know 120hz has people paying attention, many of the problems we enthusiasts encounter on any given platform or software are because no one knows we exist, because people say stuff like this, oh well they wont change , or I will just use chrome, but what happens if a patch goes through and google decides to cut 120hz support. If others see IE continues to be the most popular browser not needing 120hz support they could go the other way and say its more efficient to drop down to even 60hz.

 

[Torr Samaho]

Internet Explorer is the only browser that does not support 120Hz. When JavaScript tries to detect 120Hz, it detects 60Hz -- no way to animate at 120Hz from within IE.
IE10+ animations works perfectly at 100Hz, 101Hz, 102Hz, 103Hz, 104Hz, 105Hz, but suddenly halves framerate at 106Hz, 107Hz, 108Hz... (ToastyX/PowerStrip custom tests)
In private emails, Microsoft confirmed an intentional internal 105fps frame limiting cap, which exists even if animations only use 1 percent of CPU.

intentional cap? what is this supposed to do, save resources?

if true, that was useful years ago. some things seem to move really slowly at microsoft. it's funny that apple's browser supports 120 hz although their monitors don't have that feature afaik, yet microsoft's browser is the only one that doesn't, despite most gaming rigs running windows.

i don't use IE but i'll sign that.

 

[Mark Rejhon]

Sucessful Victory!

I've managed to convince Microsoft to re-open this bug.
It's now an "Active" bug, rather than "Closed As Designed"

Microsoft IE Bug Tracking Entry.

Coincidentially, FireFox 24 public just got released, no longer a beta. FireFox can now run the TestUFO motion tests at 120 frames per second on a 120Hz monitor.
It may not be until IE12 before this is fixed, but hopefully this will be fixed during an IE11 Update (fingers crossed).

 

[Chad_Thunder]

*claps*

I wonder what the overlap is, as in how many 120hz fanatics use IE?

 

[Mark Rejhon]

*claps*

I wonder what the overlap is, as in how many 120hz fanatics use IE?

TestUFO web analytics show that more than 75% of visitors use a non-IE browser when viewing TestUFO. However, this is primarily because I've been recommending people to use Chrome when trying TestUFO.

Fortunately, we're now nearly at the point where we got a royal flush of five browsers capable of 120Hz VSYNC -- www.testufo.com/browser.html

Right now, more than 85% of TestUFO visitors, are visiting using a VSYNC-capable browser (IE10+, FF24+, Opera15+, Safari6+, Chrome18+). This month is the month that there is now a royal flush of release browsers at least capable of 60Hz VSYNC on the most common platforms (PC, Mac, iOS), so gradually I hope I can eventually stop recommending specific browsers for TestUFO. For 120Hz, IE is the sole remaining problem browser.