AAPL stock: $258.11 ( -2.85 ) |
| Tips and Deals ---- 'Friendly' Political Ranting |
| Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Article Accelerator
Date: January 31, 2010 04:57PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: M A V I C
Date: January 31, 2010 05:07PM
|

| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: DRR
Date: January 31, 2010 05:11PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Psurfer
Date: January 31, 2010 05:13PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: H1N1
Date: January 31, 2010 05:13PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: mattkime
Date: January 31, 2010 05:17PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: M A V I C
Date: January 31, 2010 05:19PM
|
Quote
mattkime
if i have to take a stab at it, i'd guess that html 5 video takes advantage of quicktime and GPU acceleration whereas flash video doesn't. flash is fixing that - on windows.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Doc
Date: January 31, 2010 05:30PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: M A V I C
Date: January 31, 2010 05:37PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: space-time
Date: January 31, 2010 05:55PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Dick Moore
Date: January 31, 2010 06:00PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: M A V I C
Date: January 31, 2010 06:05PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Grumpyguy
Date: January 31, 2010 06:08PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: M A V I C
Date: January 31, 2010 06:37PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Golfer
Date: January 31, 2010 07:10PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Grumpyguy
Date: January 31, 2010 07:51PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Anthony
Date: January 31, 2010 10:00PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: M A V I C
Date: January 31, 2010 10:52PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Winston
Date: January 31, 2010 10:59PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: M A V I C
Date: January 31, 2010 11:21PM
|

| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: DRR
Date: January 31, 2010 11:58PM
|
Quote
M A V I C
Test these:
Flash - [home.comcast.net]
HTML5 - [home.comcast.net]
Also post your system & Flash versions when you post your results.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Winston
Date: February 01, 2010 12:19AM
|
Quote
M A V I C
Winston, that's why I gave the wink, I knew i was too big
The test I setup use .mp4 because it's H.264 which is common format between HTML5 in Webkit and Flash.
Also, I would update to at least the latest release version of the Flash Player, if not the 10.1 beta.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: M A V I C
Date: February 01, 2010 10:09AM
|

| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Winston
Date: February 01, 2010 12:20PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Dick Moore
Date: February 01, 2010 10:19PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Doc
Date: February 01, 2010 10:35PM
|
Quote
Dick Moore
Tried again with the downloads from MAVIC via Comcast. I'm running Flash 10.0.42.34 under 10.6.2/64-bit and Safari 4 latest. Nvidia GT220/1GB, E6750 C2D 2.67GHz, 4GB RAM. QE, CL OK in 64-bit.
...
No chop, no stutter, no probs. Great video.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: M A V I C
Date: February 01, 2010 10:51PM
|
Quote
Doc
Quote
Dick Moore
Tried again with the downloads from MAVIC via Comcast. I'm running Flash 10.0.42.34 under 10.6.2/64-bit and Safari 4 latest. Nvidia GT220/1GB, E6750 C2D 2.67GHz, 4GB RAM. QE, CL OK in 64-bit.
...
No chop, no stutter, no probs. Great video.
Winston is complaining about the quality of streaming HD video on his Ti667 over 802.11b.
Anyone not using an ancient G4 with max 768MB RAM, a small, fragmented hard drive and a painfully slow network connection is not going to be making a fair comparison.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: deckeda
Date: February 01, 2010 10:58PM
|
Quote
Article Accelerator
[www.macobserver.com]
"My results (Macbook Pro 13 inch):
Flash: 80-120% total CPU load (browser plus plugin)
HTML5: 7-8% CPU load"
Whoa...
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: deckeda
Date: February 02, 2010 06:42AM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Winston
Date: February 02, 2010 12:20PM
|
Quote
Doc
Winston is complaining about the quality of streaming HD video on his Ti667 over 802.11b.
Quote
Anyone not using an ancient G4 with max 768MB RAM, a small, fragmented hard drive and a painfully slow network connection is not going to be making a fair comparison.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Doc
Date: February 02, 2010 12:30PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Winston
Date: February 02, 2010 12:35PM
|
Quote
Doc
Quote
Winston
I and others haver rebutted the network connection issue.
No, you really haven't.
Your methods prefer the QuickTime Player. You're not streaming files to the QuickTime Player.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: M A V I C
Date: February 02, 2010 01:09PM
|
Quote
Doc
Quote
Winston
I and others haver rebutted the network connection issue.
No, you really haven't.
Your methods prefer the QuickTime Player. You're not streaming files to the QuickTime Player.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Doc
Date: February 02, 2010 01:47PM
|
Quote
M A V I C
But at this point, IMHO he has done enough to determine that at least on his system there is more overhead for Flash video.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Winston
Date: February 02, 2010 01:56PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Doc
Date: February 02, 2010 02:09PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: M A V I C
Date: February 02, 2010 03:36PM
|
Quote
Doc
He hasn't eliminated QuickTime's HD and RAM cache as a factor
Quote
you don't know what changes you're making to the file when you encode it as Flash and you don't know what optimizations might be in Flash for Intel Macs vs PPC so even if you could perform adequate testing, the results wouldn't be applicable to a modern Mac.
Quote
And he's using a 9 year old computer with 802.11b and he's complaining about the speed.
How can you take that seriously?
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Doc
Date: February 02, 2010 04:10PM
|
Quote
M A V I C
There's no transcoding taking place. It's the exact same file being played back. One just uses HTML5 to embed and another uses a swf wrapper (the one Flash provides.)
Quote
M A V I C
for him, the complaint is serious.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: M A V I C
Date: February 02, 2010 04:35PM
|
Quote
Doc
I don't think it just puts a "wrapper" on the file. You're using what... CS3... CS4... if you're encoding H.264 files.
Quote
The only way to keep H.264 is to encode the file as an F4V and that does change the file embedding cue points and event-data.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Winston
Date: February 02, 2010 05:08PM
|
Quote
Doc
Quote
Winston
Because it's one data point among many which indicate that Flash doesn't work as well as it could.
On a 9 year old slow PPC running an old OS and with a lousy network connection.
Yes, that's a data point. What's it worth?

| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Doc
Date: February 02, 2010 05:28PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: M A V I C
Date: February 02, 2010 05:33PM
|
Quote
Winston
I could connect my TiBook to ethernet, but why would that make any difference if QuickTime Player can play a streaming video and Flash can't?
Quote
Although you did insist that QuickTime Player can't play streaming video...
Quote
Never mind that a decent player would be able to work around a slow network connection.
Quote
I remain critical of Flash for that. It may not matter to anyone else, but I think the criticism is valid as far as my use of Flash is concerned.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Doc
Date: February 02, 2010 05:35PM
|
Quote
Winston
I could connect my TiBook to ethernet, but why would that make any difference
Quote
Winston
Although you did insist that QuickTime Player can't play streaming video...
Quote
Winston
Never mind that a decent player would be able to work around a slow network connection.
Quote
Winston
(And I get similar results on our upgraded Cube
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Winston
Date: February 02, 2010 06:03PM
|
Quote
Doc
Quote
Winston
I could connect my TiBook to ethernet, but why would that make any difference
Flash doesn't cache the same way that QuickTime does and Ethernet has less system-overhead than 802.11b does. If you have a slow Mac and a slow connection then Flash will always be at a disadvantage.
Quote
Quote
Winston
Although you did insist that QuickTime Player can't play streaming video...
No. I'm saying that you're not streaming QuickTime video in your "tests." If you're not getting the file from a streaming media server then it's not streaming.
Quote
Quote
Winston
Never mind that a decent player would be able to work around a slow network connection.
Doctor: "Damn this river-rock!! Why doesn't it cut like a scalpel?!!"
Quote
Quote
Winston
(And I get similar results on our upgraded Cube
AAAAAAAAAAAAAAAAAAAAAAARRRGHH!!!!
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Winston
Date: February 02, 2010 06:19PM
|
Quote
M A V I C
I'm trying to help you Winston, but, uh...
Quote
Quote
Winston
I could connect my TiBook to ethernet, but why would that make any difference if QuickTime Player can play a streaming video and Flash can't?
Has to do with caching. QT may be a bit better at less reliable connections. One thing I've noticed about HTML 5 is that it takes a bit longer to get started than Flash. So I think it's relying more on the initial caching than a steady network connection.
Quote
Quote
Although you did insist that QuickTime Player can't play streaming video...
Technically the method you're using is not real streaming...
Quote
Quote
Never mind that a decent player would be able to work around a slow network connection.
Oy. That's a bag of worms. The players that can do that content creators don't use that much because it takes a fair amount of extra work and understanding. Plus cost. Again, a technical thing.Quote
I remain critical of Flash for that. It may not matter to anyone else, but I think the criticism is valid as far as my use of Flash is concerned.
Yes, but you also have to recognize we're talking about business here. Based on results, if content creators can save costs by not fixing things for a small percentage of users, they do so.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Doc
Date: February 02, 2010 06:29PM
|
Quote
Winston
it sounds like what you are saying is that QuickTime works better than Flash, in this instance, which is all I've been trying to say.
Quote
Winston
when it goes to Flash is streaming but when it goes to QuickTime it's not streaming?
Sorry, that doesn't make any sense to me.
Quote
Winston
...this Flash, why doesn't it work as well as QuickTime Player for streaming the same video?
Quote
Winston
I've said all along that if I could be given a decent reason why video should not play properly on my admittedly old machine, I'd grumble and accept it.
But my experience is that video, from YouTube in my examples, CAN play properly, via streaming, via 802.11b, if I use QuickTime Player, and can't with Flash.
IGH:::Quote
Winston
if there is a way to play a given video which works, I think it is fair to criticize Flash for not being able to do it.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: M A V I C
Date: February 02, 2010 06:45PM
|
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Winston
Date: February 02, 2010 07:35PM
|
Quote
M A V I C
You know enough to be dangerous.
Quote
Doc, I think you must be the only other poster on this forum that knows the actual meaning of the terms streaming and progressive download.
Quote
I fear it's a losing battle. Another technology will get killed off and replaced by an inferior product that everyone will rave about for five years, then complain about for no good reason and it will get replaced by another inferior product. I think that's going to be the saga of web video.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Winston
Date: February 02, 2010 09:46PM
|
Quote
Winston
Do y'all know of a source for information on what types of computers (hardware and software) are accessing the internet? I am sure Google knows really well, but I don't know if anyone else does.
| Re: Interesting discussion on CPU load for HTML5 vs. Flash video Posted by: Doc
Date: February 02, 2010 10:33PM
|
Quote
Winston
Do y'all know of a source for information on what types of computers (hardware and software) are accessing the internet?
Quote
Winston
The more I think about it, the more critical I think the answer to this question is. Do people who design web sites pay close attention to what types of computers are accessing their sites?

