Firefox 57 (aka Quantum) out now. Huge performance improvements.

[ScratchCat]

18 hours ago, LAwLz said:

That's weird.

It says that webm is disabled in my about:config, but clearly it is working on youtube for me. You sure it's not some old issues that has since then been fixed?

That was how my nightly install was setup , however this was on an Atom Z3735 with 1GB RAM so it may be dependant on the device ,probably if it supports hardware VP8/9 decoding to prevent it being an abysmal experience and putting people off (the number of people selecting 1080p video on a 720p screen because they thing it will provide a noticeable quality improvement 

EDIT: Seems that youtubes compression and bit rate can lead to higher resolutions being beneficial.

 

17 hours ago, Dabombinable said:

Getting around to comparing the way Firefox is now handling Youtube videos to Chrome:

• Videos start a lot faster
• significantly shorter wait when changing video resolution
• No dropped frames at all (stock 4790K+stock GTX 970 G1 Gaming)
• Lower CPU usage, higher GPU usage (I wonder if I could set FF to use my GTX 650ti for decoding).

And overall the user experience has been better.

From the comments on this thread I am fairly certain Firefox has an issue with hardware decoding depending on the codec and platform. My R9 290 used to use ~150W while watching a 720p YouTube video until some combination of updates put an end to that.

[pyrojoe34]

4 hours ago, ScratchCat said:

(the number of people selecting 1080p video on a 720p screen because they thing it will provide a noticeable quality improvement (downscaling isn't that impressive ) is large).

With streaming sites (especially YouTube) it absolutely does make a huge difference... but it has nothing to do with the resolution. It’s because the bitrate served is resolution dependent. You can try it yourself. Use a 1080p screen and compare 1080p YouTube vs 4k YouTube (especially with moving scenes and foliage artifacting). You will notice a massive difference in video quality (even works with 1080p vs 1440p).

 

I always watch YouTube videos at 4k on 1080+p screens and watch in 1080p or 1440p on my laptops (which are 768p and 900p). There is absolutely a massive increase in quality even considering the interpolation.

 

For example: YouTube uses ~8Mbps for 1080p30 and ~35-45Mbps for 4k30. Personally, when rendering 1080p30 h.264 videos I use at the very least 45Mbps to minimize compression artifacts. That’s also why you should upscale 1080p footage to 4k before uploading, you get a noticeable increase in quality compared to uploading in its native 1080p resolution due to YT’s compression metrics.

[Dabombinable]

7 hours ago, ScratchCat said:

From the comments on this thread I am fairly certain Firefox has an issue with hardware decoding depending on the codec and platform. My R9 290 used to use ~150W while watching a 720p YouTube video until some combination of updates put an end to that.

AMD's cards have always been a bit power hungry compared to Nvidia's when it comes to video decoding, can't find the review I saw the figures on, but it was around 50W for an R9 290/290X watching Blueray.

2 hours ago, pyrojoe34 said:

With streaming sites (especially YouTube) it absolutely does make a huge difference... but it has nothing to do with the resolution. It’s because the bitrate served is resolution dependent. You can try it yourself. Use a 1080p screen and compare 1080p YouTube vs 4k YouTube (especially with moving scenes and foliage artifacting). You will notice a massive difference in video quality (even works with 1080p vs 1440p).

 

I always watch YouTube videos at 4k on 1080+p screens and watch in 1080p or 1440p on my laptops (which are 768p and 900p). There is absolutely a massive increase in quality even considering the interpolation.

 

For example: YouTube uses ~8Mbps for 1080p30 and ~35-45Mbps for 4k30. Personally, when rendering 1080p30 h.264 videos I use at the very least 45Mbps to minimize compression artifacts. That’s also why you should upscale 1080p footage to 4k before uploading, you get a noticeable increase in quality compared to uploading in its native 1080p resolution due to YT’s compression metrics.

My internet's only good enough for 1440p

 

However, even 1440p still gives a noticeable quality increase.

[Drak3]

And I'm back on Edge. 

[8uhbbhu8]

I absolutely love the new Firefox Update. I have noticed the speed updates and less resource usage the most. Super happy with it  

[MyName13]

7 hours ago, Drak3 said:

And I'm back on Edge. 

How does Edge work for you when compared to other browsers?It's significantly slower than chrome and firefox on my PC.

[Drak3]

2 hours ago, MyName13 said:

How does Edge work for you when compared to other browsers?It's significantly slower than chrome and firefox on my PC.

Hands down the fastest.

[LAwLz]

9 minutes ago, Drak3 said:

Hands down the fastest.

Fastest in what way?

[Drak3]

28 minutes ago, LAwLz said:

Fastest in what way?

Load, response, and download times.

[LAwLz]

9 minutes ago, Drak3 said:

Load, response, and download times.

When you compared, did you use an old or a new install of Chrome/Firefox? Did you use any extensions for the different browsers?

 

Load - Several independent benchmarks actually puts it behind Chrome and Firefox on a lot of sites, but sometimes coming out on top. Seems like all three are trading blows with one another.

 

Response - You mean like in the UI? I don't think I agree but at the same time I don't think there are any objective tests I can point towards to say that you're wrong.

 

Download times - Wait, so what you're saying is that downloading files is faster in Edge than other browsers? That I call BS on, unless you have some proxy or VPN setting enabled in the other browsers which slows them down.

 

Edit:

It might seem like I am attacking you by asking so many questions but I am actually just curious how you arrived at your conclusions because they contradict the tests I have done and looked at. Browsers are one of my big interests so when someone comes with a different POV I wonder how they think and what tests they have done.

[Drak3]

5 hours ago, LAwLz said:

When you compared, did you use an old or a new install of Chrome/Firefox? Did you use any extensions for the different browsers?

Clean FF57 install.

 

5 hours ago, LAwLz said:

Several independent benchmarks actually puts it behind Chrome and Firefox on a lot of sites, but sometimes coming out on top. Seems like all three are trading blows with one another.

FF57 has only been able to load YouTube faster as far as my personal testing goes. But not by a meaningful margin.

 

5 hours ago, LAwLz said:

You mean like in the UI? I don't think I agree but at the same time I don't think there are any objective tests I can point towards to say that you're wrong.

In Windows, FF57 is fairly clunky after letting the browser sit on the home screen for just a few minutes. I only keep pinned sites on home. It's worse than Edge, FF56 on Windows, and FF57 on OSX. FF57 in Ubuntu in HyperV is more responsive.

 

5 hours ago, LAwLz said:

Wait, so what you're saying is that downloading files is faster in Edge than other browsers? That I call BS on, unless you have some proxy or VPN setting enabled in the other browsers which slows them down.

Yes. Not by a huge margin, but Edge has been consistantly faster than 57. 56 was faster than Edge, and Safari on OSX is faster than both. Even when browsing mock sites off of my NAS.

[ScratchCat]

16 hours ago, Drak3 said:

And I'm back on Edge. 

Microsoft will welcome this extra telemetry . The fact that edge constantly tries to set the homepage to MSN + Ads each update should put most people off.

I guess that it is useful on laptops due to a longer battery life, however it's RAM management seems on the level of that of chrome ( or chrome is really good at reducing memory usage on intel ark).

[adman29]

I actually really enjoy it, kinda thinking about fully switching from chrome to firefox. I have had a few random crashes with it though.

[LAwLz]

1 hour ago, Drak3 said:

Clean FF57 install.

That might explain it. Because of how Windows manages memory, as well as browsers cache things, older installs which has frequently been used will often perform better.

For example Windows might preload certain program files into memory before (SuperFetch) you even launch it. Since your FF install was new, Windows will preload the browser you are used to (Edge) and not do the same for Firefox.

As for websites loading faster, it can be explained with Edge already having some content cached since you've used it more.

 

1 hour ago, Drak3 said:

In Windows, FF57 is fairly clunky after letting the browser sit on the home screen for just a few minutes. I only keep pinned sites on home. It's worse than Edge, FF56 on Windows, and FF57 on OSX. FF57 in Ubuntu in HyperV is more responsive.

Does it get "clunkier" just by having it open? That does not sound right at all. FF57 performing worse than FF56 sounds like your install is messed up. I have so far not seen a single test which shows this.

 

1 hour ago, Drak3 said:

Yes. Not by a huge margin, but Edge has been consistantly faster than 57. 56 was faster than Edge, and Safari on OSX is faster than both. Even when browsing mock sites off of my NAS.

Wait, has it been faster downloading things or loading web pages?

Because the way I interpreted your post, it was faster at downloading for example files, and I find that very hard to believe.

 

 

1 hour ago, ScratchCat said:

I guess that it is useful on laptops due to a longer battery life

You really shouldn't trust what Microsoft says. According to Linus' test Edge actually resulted in shorter battery life than Firefox and Chrome, although the difference was very small (using the same test Microsoft used).

Remember, RP statements like "Edge has better battery life!" is only said in an attempt to make your product look better than competitors. It's very often half-truths or flat out lies.

[ScratchCat]

18 minutes ago, LAwLz said:

 

 

You really shouldn't trust what Microsoft says. According to Linus' test Edge actually resulted in shorter battery life than Firefox and Chrome, although the difference was very small (using the same test Microsoft used).

Remember, RP statements like "Edge has better battery life!" is only said in an attempt to make your product look better than competitors. It's very often half-truths or flat out lies.

I am basing this off independent tests, some are below. Microsoft does like to exaggerate about edge's efficiency however this can be beneficial as Chrome was later improved in order to counter this statement.: https://www.theverge.com/2017/3/15/14932718/google-chrome-browser-battery-life-tab-throttling

 

https://www.custompcreview.com/articles/chrome-vs-edge-vs-firefox-best-browser-battery-life/ 

-Firefox is worse than Edge, noticeably in video streaming.

 

https://www.avg.com/en/signal/chrome-battery-life-vs-edge-and-firefox

- Edge now noticeably ahead, this was FF51 though.

 

https://www.digitaltrends.com/computing/web-browers-battery-life-tested/

-Firefox 55 worse than Edge.
 

-Firefox seems to be doing a slightly worse than edge, not much worse though.

 

Overall from what I have tested myself and the above tests Edge can be and mostly is better than Firefox in the battery life department. However this is not Firefox 57+, all these tests used older versions. Mozilla have done an impressive job and are continuing to work on this. From my testing if the plugin container process can be reigned in (mainly by disabling flash for all websites) there would be an improvement.

[chriscoolzap]

Neat I might have to try ff  esp for laptop use....that would be great for more battery life 

[Drak3]

3 hours ago, LAwLz said:

That might explain it. Because of how Windows manages memory, as well as browsers cache things, older installs which has frequently been used will often perform better.

For example Windows might preload certain program files into memory before (SuperFetch) you even launch it. Since your FF install was new, Windows will preload the browser you are used to (Edge) and not do the same for Firefox.

As for websites loading faster, it can be explained with Edge already having some content cached since you've used it more.

It would, if it weren't for the fact that I've used FF56 almost exclusively since my last clean install of Win10 on my main machine, and the only sites I've used Edge for have been a few downloads and PDFs that I've downloaded.

I've also disabled SuperFetch and the page file. Don't really need either when I'm running 32GB of RAM and approx 1.25TB of SSD storage just for OS and programs.

3 hours ago, LAwLz said:

Does it get "clunkier" just by having it open? That does not sound right at all. FF57 performing worse than FF56 sounds like your install is messed up. I have so far not seen a single test which shows this.

The issue with going off of tests is that it's practically impossible to test all configurations. I don't have these issues with the OSX or Ubuntu counterparts of 56. Only in Windows 10.

 

3 hours ago, LAwLz said:

Wait, has it been faster downloading things or loading web pages?

Because the way I interpreted your post, it was faster at downloading for example files, and I find that very hard to believe.

Downloading files.

[Bigbootyjudy]

I just got a new laptop that only has 2 gigs of ram and a celeron b800 and I must say that firefox loads and runs much faster.

[vanished]

57 minutes ago, Drak3 said:

[...]

Downloading files.

I still don't see how this could be possible.  Rendering pages, sure, every browser does that differently and there's a ton of room for optimization but downloading files is just a simple bulk transfer... I can't imagine how this would differ between browsers

[Drak3]

Just now, Ryan_Vickers said:

I still don't see how this could be possible.  Rendering pages, sure, every browser does that differently and there's a ton of room for optimization but downloading files is just a simple bulk transfer... I can't imagine how this would differ between browsers

My working hypothesis is that Edge is capable of getting more access to network resources due to its integration with Win10. It's also the only browser that doesn't slow down to a crawl when  downloading games on steam or anime through the Movies & TV app.

[vanished]

Just now, Drak3 said:

My working hypothesis is that Edge is capable of getting more access to network resources due to its integration with Win10. It's also the only browser that doesn't slow down to a crawl when  downloading games on steam or anime through the Movies & TV app.

They are probably giving it priority then, that would explain the Steam thing, but just like boosting the priority of an app when there's nothing else competing doesn't really affect performance, I can't see how that would impact download speeds without anything else running

[Drak3]

Just now, Ryan_Vickers said:

They are probably giving it priority then, that would explain the Steam thing, but just like boosting the priority of an app when there's nothing else competing doesn't really affect performance, I can't see how that would impact download speeds without anything else running

With no interest in the inner working of any of the current browsers, I don't really care about the why, only the what. And the what, is that FF57 doesn't play as nicely with my system as FF56 and Edge do.

[vanished]

14 minutes ago, Drak3 said:

With no interest in the inner working of any of the current browsers, I don't really care about the why, only the what. And the what, is that FF57 doesn't play as nicely with my system as FF56 and Edge do.

Hm... it is odd.  You've got me curious enough to run some tests myself

 

Edit: I ran speedtest on FF57, then Edge, and then back to FF57 and I can't say there's any noticeable difference.  The variance between each test is greater than any variance between the browsers.

Edited November 28, 2017 by Ryan_Vickers

[xAcid9]

3 hours ago, Drak3 said:

With no interest in the inner working of any of the current browsers, I don't really care about the why, only the what. And the what, is that FF57 doesn't play as nicely with my system as FF56 and Edge do.

Same here, Edge is personally the fastest browser i've ever used but it lack some features to be my daily driver.

[Drak3]

7 minutes ago, xAcid9 said:

Same here, Edge is personally the fastest browser i've ever used but it lack some features to be my daily driver.

Beyond loading web pages, initiating downloads, and not crashing/freezing, I have no additional need or expectation of my web browsers.