PDA

View Full Version : Vista Reportedly 18-24% Slower at Video Encoding


Pookie
31st January 2007, 23:16
than XP, at least with the products they tested.

Xvid 1.2 and Mainconcept h264 Results

http://www.tomshardware.com/2007/01/29/xp-vs-vista/page5.html

zambelli
1st February 2007, 01:14
than XP, at least with the products they tested.

Xvid 1.2 and Mainconcept h264 Results

http://www.tomshardware.com/2007/01/29/xp-vs-vista/page5.html
I find it really interesting that those particular 2 benchmarks faired so poorly in Vista, especially considering the other benchmarks were nearly all on par with XP. I notice in both cases MPEG-2 was used as a source. Could it be that different MPEG-2 decoders were used in the comparisons and the testers didn't notice? Crunching numbers just isn't one of those things where OS should make such a noticeable difference so I wonder if something else was the culprit.

In the conclusion they also point out that no version of Windows has ever been faster than its predecessor, which isn't necessarily a good thing, but at least shows that nobody should be surprised with the slight perf regression in some departments. :)

foxyshadis
1st February 2007, 01:35
They never used NT 3.5 after 3.0, then. (Neither have I, honestly, but the grizzled warriors tell me that there was an enormous perf update, and notes from old MS devs indicate that was the main reason for it.) But I'll accept their general premise.

dragongodz
1st February 2007, 02:50
The article does point out that the much-feared Aero performance hit isn't actually noticable at all - and that's something worth noting.
sorry but i can not find that statement at all.
the closest is
There is good news as well: we did not find evidence that Windows Vista's new and fancy AeroGlass interface consumes more energy than Windows XP's 2D desktop. Although our measurements indicate a 1 W increase in power draw at the plug, this is too little of a difference to draw any conclusions.
this is about power requirements and not about actual performance.

setarip_old
1st February 2007, 03:40
In the conclusion they also point out that no version of Windows has ever been faster than its predecessor, which isn't necessarily a good thingThis may qualify as the understatement of the year - and it's only January!

check
1st February 2007, 04:18
so are any of the regulars here planning on upgrading to a sytem running only vista? I plan to eventually install it when it's needed for games, but to dual boot when that time comes.

ilovejedd
1st February 2007, 04:42
When it comes to new operating systems, I usually never upgrade. Well, that is if I have to pay for it, if I'll be getting it free (http://forum.doom9.org/showthread.php?t=121539), then why not :p. Instead I buy a new PC with the OS pre-installed. Usually after a year or so and most of the bugs have been ironed out... So yeah, I guess I will be upgrading to a system running only Vista, but I'm still keeping my old XP machine. :D

The Scientist
1st February 2007, 07:13
upgrading to a sytem running only vista?Yes I have it, love it.

zambelli
1st February 2007, 09:46
this is about power requirements and not about actual performance.
Hmm... You may be right. I figured power implied CPU performance too, but I guess that's not explicitly stated anywhere.

so are any of the regulars here planning on upgrading to a sytem running only vista?
At home - probably not for a year or so, or until I find that I need it for something specific. But this is pretty typical of my home computer usage - I didn't upgrade to XP until 2005 when I upgraded all of my hardware too.

Sharktooth
1st February 2007, 14:42
i had a similar impression when i tested it. it thought it was a videocard drivers issue but it seems not...
also
Tweak the OS: turn off animations and AeroGlass for maximum system performance.
Disable User Access Control to prevent it from interrupting certain benchmarks.
Have the OS process pending idle tasks
Turn off system restore
Install all applications, and execute them several times (with restarts in between) to make SuperFetch aware that you want them to be available.
Don't use the system after reboots during your SuperFetch training period: this way, Vista gets sufficient idle time to "superfetch" applications.
as you can see the fancy stuff was all disabled, so Vista is definalty slower than XP (more than the predicted 10%!) even without Aero and stuff... that means all the $h!t MS thrown in is definatly hogging the system. Everything that needs CPU power is dramatically slower in Vista (and no, i dont consider syntetic tests as real world applications).

JohnnyMalaria
1st February 2007, 15:20
I ran some tests last night comparing XP Pro SP2, XP x64 and Vista Ultimate (RTM).

I used our Enosoft DV Processor to convert an four minute 16:9 DV clip to letterboxed 4:3 and burn the time and date onto the video.

The tests were run on the same hardware. An external USB 2.0 drive was used for the source clip and a second external USB 2.0 was used for the destination clip.

On all three OS platforms, the CPU use averaged around 50% - I suspect that the process was I/O limited due to the external USB drives.

XP Pro SP2 - 124 seconds (+/- 1 second)
XP x64 - 127 seconds (+/- 1 second)
Vista Ultimate (32-bit) - 133 seconds (+/- 1 second)

Hence, Vista ran 5 to 8% slower than XP Pro SP2 (allowing for the accuracy of the timing).

However, the on XP Pro SP2, the software's preview window was choppy (a common DirectShow "trick"). On Vista, it wasn't.

I will run some more tests that will push the processor harder and will also compare the same external drives but using FireWire instead of USB 2.0.

Sharktooth
1st February 2007, 15:25
well, chances are your encoder is not multithreaded or there was a bottleneck on the USB drives.

HyperHacker
3rd February 2007, 07:33
Yeah, USB drives are going to be a big bottleneck in most cases. Ideally you'd use two different internal hard drives (not on the same IDE channel if using IDE).

Sharktooth
4th February 2007, 15:37
All the unwanted restrictions and checks that were put into the OS are even coming at a hard price for the end user.
one of the biggest hits is data encryption between cpu and gpu. it is made using cpu cycles (that will be stolen to other intensive processes).
Vista is going to be a very low performer...

zambelli
4th February 2007, 21:40
All the unwanted restrictions and checks that were put into the OS are even coming at a hard price for the end user.
one of the biggest hits is data encryption between cpu and gpu. it is made using cpu cycles (that will be stolen to other intensive processes).
Isn't this data encryption used only when Protected Video Path is used (aka HD-DVD/BD playback)? Do you have data to the contrary?

Sharktooth
5th February 2007, 15:04
yeah, right, i meant the encryption will be done during video decoding.... so h.264 decoding (bluray and hd-dvd) will be even harder on vista...

Inventive Software
5th February 2007, 16:12
But not native H.264 or VC-1 decoding... i.e the decrypted streams played back...

Sharktooth
5th February 2007, 18:00
No, but the encryption is an addition to the other CPU-hungry checks and resident stuff that hogg the performance in every situation where cpu power is needed.

zambelli
5th February 2007, 20:45
No, but the encryption is an addition to the other CPU-hungry checks and resident stuff that hogg the performance in every situation where cpu power is needed.
Unless you have data, this sounds a lot like speculation to me. :sly:

JohnnyMalaria
5th February 2007, 21:51
Vista is going to be a very low performer...

For certain things, maybe.

If you are encoding your own material that isn't encrypted, where's the crippling overhead?

Sharktooth
5th February 2007, 22:16
look at the tom's benchmarks... even winrar is slow as hell on vista.
if you look at the graphs everything that needs CPU power and memory bandwidth is crippled (real world applications and even some benchmarks)...

Pookie
6th February 2007, 02:45
Hey Johnny - what specific features of Vista have convinced you to be such a proponent? Even the great zambelli doesn't wave the V flag with such enthusiasm.

Sharktooth
6th February 2007, 14:54
he's a fanboy.

Audionut
6th February 2007, 16:46
he's a fanboy.

And you're an M$ basher.

Go Figure.



Give vista time to mature, or more importantly, give programmers more time to get accustomed to vista before jumping to any conclusions.

Sharktooth
6th February 2007, 18:49
mature or immature it's no longer what an OS should be.
it never happened in the windows history to have a such boost in requirements and get the OS so filled with useless stuff and user restrictions.
with maturation it can only get worse. every analyst say "dont get vista"... "vista is not ready"... "vista il bloated"...
it's not a single opinion and when i tried it i verified it is a pile of crap.

Inventive Software
6th February 2007, 19:31
A BBC reporter has his opinion on a new Vista laptop he bought... http://www.bbc.co.uk/blogs/thereporters/robertpeston/

I can't believe that computer manufacturers are rushing to get Vista installed on every new machine, even though it's not as reliable or compatible as it claims to be...

Pookie
6th February 2007, 20:15
There's usually four reasons to upgrade software - Stability/Reliability, Performance, Compatibility, Ease of Use.

Stability was the compelling reason for businesses to upgrade from Win 98 to 2000. Stability could also be argued as the reason for upgrading from 2000 to XP, as the BSODs were much less frequent in XP. Performance suffered, however. It required a RAM upgrade to maintain the performance of the previous O/S. The cost of the RAM justified the increase in stability.

In a business environment, there are two computing functions that create panic when they don't work - Printing, and E-mail. Would an upgrade enhance the functionality of either ?

Then, is the "period of pain" (adjusting to new features and dealing with unexpected problems) worth the benefit of the upgrade? IT Management walks a fine line in this regard. If you've ever been involved in a big rollout, you'll know what I mean :D

setarip_old
6th February 2007, 20:52
Hi!I can't believe that computer manufacturers are rushing to get Vista installed on every new machineI guess you haven't considered that, if it's "business as usual", Microsoft is now willing to sell the OEMs only Vista licenses...

JohnnyMalaria
6th February 2007, 21:21
Hey Johnny - what specific features of Vista have convinced you to be such a proponent? Even the great zambelli doesn't wave the V flag with such enthusiasm.

Here's my opinion:

http://www.dvinfo.net/conf/showpost.php?p=616552&postcount=7

zambelli
7th February 2007, 02:44
Even the great zambelli doesn't wave the V flag with such enthusiasm.
That's because I'm not here to sell anyone anything. :) I don't even try to sell people on WMV. I only step in in defense of Microsoft technologies when I feel like people are being unfair or uninformed in their opinions. Most technologies we discuss on these forums have their place and an appropriate use - otherwise they probably wouldn't exist in the first place. Even if I feel like promoting one, I have no interest in bashing others.

Audionut
7th February 2007, 02:52
it's no longer what an OS should be.

What options are there, Linux. The so called excellent OS that was supposed to take over years ago.


it never happened in the windows history to have a such boost in requirements and get the OS so filled with useless stuff and user restrictions.

It's no worse than XP. Just like x264, you don't need everything enabled. And most of the restrictions are easily disabled.


with maturation it can only get worse. every analyst say "dont get vista"... "vista is not ready"... "vista il bloated"...
it's not a single opinion and when i tried it i verified it is a pile of crap.

Well I personally like it. But I don't take every opportunity to express my views on others.

Sharktooth
7th February 2007, 03:15
if it's so good why every indipendent review and analyst report are negative?
Vista is not good for anything. incompatibilities, applications and games slowdowns, higher harwdare (CPU, RAM, ecc) requirement, user restrictions, unfair DRM (all the "power" to the content providers), security issues (even if MS said it's the most secure OS ever created... blah blah), encryption and content "crippling" that affects even non protected content...
And what you get? A fancy UI that eats even more RAM, annoying messages that could be useful for the newbie but not for the experieced user and a shitload of useless apps that can be replaced by free software that does the same job (maybe even better).
Thanks but no. I prefer the good and old WinXP or even better Linux. At least i have more free resources and i can do whatever i want and everything runs from 2 to 30% faster (expecially the CPU intensive tasks) with the same hardware.

Audionut
7th February 2007, 03:46
if it's so good why every indipendent review and analyst report are negative?
Vista is not good for anything. incompatibilities, applications and games slowdowns, higher harwdare (CPU, RAM, ecc) requirement

It was the same for winxp when it first arrived.


user restrictions,
Windows key + r, Msconfig,
http://img387.imageshack.us/img387/6781/sshot1az5.png


unfair DRM, encryption and content "crippling" that affects even non protected content...

Have faith, HDDVD and Bluetooth are already broken.
So is driver signing protection.

Update on Driver Signing Bypass

However, it does bypass DRM. As part of the Protected Media Path, (PMP), Windows Vista sets up a number of requirements for A/V software and drivers in order to ensure it complies with the demandes of the media companies. One of these features, which has been heavily criticized as being the actual reason behind driver signing, is that “some premium content may be unavailable” if test signing mode is used. Originally, I assumed that this meant that the kernel would set some sort of variable, but this didn’t make sense: once your unsigned driver could load, it could disable this check.
http://www.alex-ionescu.com/?p=24


security issues (even if MS said it's the most secure OS ever created... blah blah)

Let's be reasonable. Nothing will ever be secure. If your beloved linux eventually becomes mainstream it's security flaws will soon become apparent.

And what you get? A fancy UI that eats even more RAM,

I like the UI. Every new version of an OS needs more ram. And do you really think that you have to use all the new GUI features. Just like xp you can disable all the optional extras.


annoying messages that could be useful for the newbie but not for the experieced user

See the screenshot above.

and a shitload of useless apps that can be replaced by free software that does the same job (maybe even better).

So now your complaining about the bundled software. Really there's not that much more that xp. If you don't like them then disable them or uninstall them. :confused:


Thanks but no. I prefer the good and old WinXP or even better Linux. At least i have more free resources and i can do whatever i want and everything runs from 2 to 30% faster with the same hardware.

And that's your choice as a consumer. I just don't like the way you go around spreading what IMO is your nonsense.

Sharktooth
7th February 2007, 04:07
winxp was another story. it was more stable and more secure than win2k. the firewall was needed and saved a lot of ppl from viruses and malware/spyware illicit activities. the UI was ugly but it only consumes about 30Mbs of system ram and doesnt require pixel shader 2.0 to work...
about user restrictions i didnt mean the user account restrictions, but what you cannot do even as an administrator: license restrictions, DRM, signed drivers, software revokation etc...
and even if it's true that DRM, driver signing, etc was and is being cracked, it's an illegal practice. so if you do it you can as well get a pirated copy of vista...:p
The RAM issue is another thing i cant stand. Why i should DISABLE or uninstall things i dont want? Just make them optional during the installation. If i want them ill enable/install them (as it was in the old good days).
So basically (and this happens in some way with XP too) you install the OS and have to look here and there to uninstall and disable the unwanted features/softwares without even having a full documentation on system services and other stuff. That's quite silly and annoying not only for me, but for all the sane people.
Summing it up, if notebooks and branded PCs wouldnt come with a preinstalled copy of Vista OEM, it wouldnt be so popular...

foxyshadis
7th February 2007, 05:02
if it's so good why every indipendent review and analyst report are negative?

When the only reviews you read/believe are anti-Vista, it might seem as if every one is. For every negative review, there's another positive and another totally ambivalent. (And claiming that everyone who likes it must be a shill or a fankid is just denial.) After reading hundreds of reviews, from offhand blog posts to deep formal writeups, in the past 9 months, especially the last two, all I know is that apparently no two people can agree on what to love and hate about it.

Audionut
7th February 2007, 06:30
it's an illegal practice. so if you do it you can as well get a pirated copy of vista...:p

2 totally different things. You're really grasping at straws for your arguments.
It's called customizing the OS to your needs. People have been doing it for years. See next statement.

Why i should DISABLE or uninstall things i dont want? Just make them optional during the installation. If i want them ill enable/install them (as it was in the old good days).

Because microsoft isn't in the near future gonna design an OS made to your specific specifications.
They (MS) have to cater for millions of customers, not just you.


I don't overly like microsoft any more that most. But for me there are no alternatives that suit my needs. And I'm not goning to bash M$ simply cause I got nothing better to do.
I'm not 100% for vista, but I'm not going to add negative points when you are more than capable.

We are not going to come to any sort of compromise, so I will end my discussion here.

zambelli
7th February 2007, 11:42
winxp was another story. it was more stable and more secure than win2k. the firewall was needed and saved a lot of ppl from viruses and malware/spyware illicit activities.
I will buy and ship a copy of Vista Ultimate to the first person who finds me a forum or newsgroup post from you dated anytime between 2001 and 2004 that calls XP less stable or less secure than Windows 2000 or questions the usefulness of SP2 firewall. :sly: I can't find it myself, but I know it's gotta be somewhere out there! :D

P.S. The winner can opt for a copy of XP instead of Vista if that's their preference. :)

Sharktooth
7th February 2007, 13:35
uhm, no. you cant find it sorry ;)
i got XP Pro as soon as it was available and i was conscious it had initial incompatibilities with very old PCs.
Maybe you can find a post of mine criticizing XP Home coz it has no reason to exist.

@audionut: well im not alone "bashing" M$ and there is a reson for doing it. Here's another one: http://uk.theinquirer.net/?article=37394

jeffy
8th February 2007, 02:56
Make your own opinion, bits and pieces...

Slower than XP?
Vista vs. Windows XP: Head to Head Benchmarks
http://tech.yahoo.com/blogs/null/13487
"In most cases, Vista is slower than XP, but not terribly."
http://www.pcworld.com/article/id,128305-page,1-c,vistalonghorn/article.html

http://www.pcworld.com/article/id,128305-page,1-c,vistalonghorn/article.html
Lab Tests: Vista's Fast If You Have the Hardware

Aero slowdown on integrated graphics (extreme: 16% slower)
http://www.pcworld.com/zoom?id=128305&page=5&zoomIdx=1

"... there was not much of a difference between XP and Vista overall, ..."
http://techgage.com/article/windows_vista_system_performance_reports

Problems:
-- Hardware

More Windows Vista Problems for Korean Firms
- some MP3 players incompatible -
http://english.chosun.com/w21data/html/news/200702/200702060016.html

--Software

Will My Programs Run In Vista?
http://www.lockergnome.com/nexus/windows/2007/02/06/will-my-programs-run-in-vista/

Windows Vista RTM Software Compatibility List
http://www.iexbeta.com/wiki/index.php/Windows_Vista_RTM_Software_Compatibility_List

Intuit Quickbooks incompatible, new version $200
http://blogs.zdnet.com/Berlind/?p=315
http://www.quickbooks.com/Helpcenter/Vista/

(unverified) Compatibility Isssues
http://kb.wisc.edu/helpdesk/page.php?id=5175

Compatibility update for some apps:
http://support.microsoft.com/kb/929427

Nvidia Fails The Vista Test
http://www.informationweek.com/news/showArticle.jhtml?articleID=197004259
"Gamers are up in arms after drivers for the graphics processors prove incompatible with Microsoft's latest Windows operating system."

Apple: None of our products are Vista-compatible
http://www.itnews.com.au/newsstory.aspx?CIaNID=45539&src=site-marq
"According to a document that Apple has posted on its website, none of the software that it's made available for the Windows environment has been updated for Vista compatibility."

http://docs.info.apple.com/article.html?artnum=304854
"Ejecting an iPod from the Windows System Tray using the "Safely Remove Hardware" feature may corrupt your iPod."


Reviews, opinions, comments:
Opinion: A Vista Upgrade nightmare:
http://www.tgdaily.com/2007/02/01/opinion_vista_upgrade/

EDIT:
Adding funny video link: My Vista Upgrade: A Short Comedy
http://www.tomshardware.com/site/videos.html
END

Microsoft dismisses claims of Vista breaking games, but ...
http://www.tgdaily.com/2007/02/07/microsoft_dismisses_vista_game_claims/

Some bumps on the road to Vista
'Still, Cope wasn't happy with the more than six hours he spent getting to Vista. "I should've bought a Mac," Cope said.'
http://news.com.com/2100-1016_3-6156968.html

Upgrade?
"Microsoft Windows Vista Can Be Rocky Terrain, Warns Computer Troubleshooters"
http://www.prweb.com/releases/2007/2/prweb503343.htm

Vista poses environmental dangers
"10 million PCs could be discarded over the next two years"
http://www.vnunet.com/computing/news/2174400/vista-poses-environmental

Windows Vista 'wide open' to malware???
http://www.computing.co.uk/vnunet/news/2173632/researchers-say-vista-prone
"Windows Defender, the security software included with Vista, failed to catch 84 per cent of spyware and malware."


New Vista firewall fails on outbound security
http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9010661&taxonomyId=16&intsrc=kc_top
"But as shipped, the Windows Firewall offers little outbound protection, and it's not clear how outbound protection can be configured to protect against spyware, Trojan horses and bots."





News:
Microsoft starts work on Vista Service Pack 1
"Microsoft has not yet disclosed a planned release date or a schedule of issues that it plans to tackle with the update."
http://www.computing.co.uk/vnunet/news/2173392/microsoft-starts-preparing

Reviews:
Windows Vista Ultimate, A Hands On Diary: Entry #1
http://www.tomshardware.com/2007/01/02/windows_vista_ultimate_hands_on/

PC Magazine
http://www.pcmag.com/article2/0,1895,2088444,00.asp

PC World:
http://www.pcworld.com/article/id,127629-page,1-c,vistalonghorn/article.html

CNET:
http://reviews.cnet.com/Windows_Vista_Ultimate/4505-3672_7-32013603.html

Five Ways to Bend Windows Vista to Your Will
http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9009205&intsrc=hm_list


Support charges raised:
$35 -> $59 per incident
Microsoft charges more for Vista support but not only for it!
http://www.digitmag.co.uk/news/index.cfm?newsid=7165

"The new, higher prices for support apply to all older software for which Microsoft still offers mainstream or extended support."
http://arstechnica.com/news.ars/post/20070205-8773.html


Do you want more news?
http://news.google.com/news?hl=en&ned=us&ie=UTF-8&q=Windows+Vista&btnG=Search

Sharktooth
8th February 2007, 14:38
Thank you jeffy :)

JohnnyMalaria
8th February 2007, 16:28
Make your own opinion, bits and pieces...

Windows Vista RTM Software Compatibility List
http://www.iexbeta.com/wiki/index.php/Windows_Vista_RTM_Software_Compatibility_List


Great link - thanks. I've added our products to the "Works" list. Not surprised that Pinnacle, Adobe and Ulead show up in the "Has Serious Problems" category. That trio has always been buggy and troublesome.


Intuit Quickbooks incompatible, new version $200
http://blogs.zdnet.com/Berlind/?p=315
http://www.quickbooks.com/Helpcenter/Vista/


That's Quickbooks for you. Believe me, I know. My wife is constantly on the phone to Quickbooks because the files that her clients send her won't open in her latest and greatest, Pro Accountants version.


Nvidia Fails The Vista Test
http://www.informationweek.com/news/showArticle.jhtml?articleID=197004259


Shame on nVidia (I have a 7600GT on Vista). *NO* excuse.


Apple: None of our products are Vista-compatible


Shame on Apple. *NO* excuse. Also, Microsoft has a team specifically for helping Apple get their cr@p to work on Vista.

Sorry, but I stopped using Quicktime aeons ago when, without giving me the option, iTunes got stomped all over my XP. Even after uninstalling it, there are still vestiges of iTunes on that computer.



http://docs.info.apple.com/article.html?artnum=304854
"Ejecting an iPod from the Windows System Tray using the "Safely Remove Hardware" feature may corrupt your iPod."


Wrong link, BTW. The correct link goes on to say "[t]o always safely eject an iPod, choose Eject iPod from the Controls menu within iTunes." Most likely Apple's hardware doesn't honor Windows' PnP power management etc. i.e., cr@p hardware drivers.


Vista poses environmental dangers
"10 million PCs could be discarded over the next two years"
http://www.vnunet.com/computing/news/2174400/vista-poses-environmental


Oh, puh-lease!!!! What about all the 286, 386, 486, Pentium, P2 systems that have been junked. What about all the older Macs that can't run Apple's latest OS offerings?

What about standard definition TVs that are getting junked?

What about cell phones?

WHAT ABOUT DONATING THEM TO CHARITY??????


News:
Microsoft starts work on Vista Service Pack 1
"Microsoft has not yet disclosed a planned release date or a schedule of issues that it plans to tackle with the update."
http://www.computing.co.uk/vnunet/news/2173392/microsoft-starts-preparing


Great bait for the MS-bashers. MS can't win here. Start too early - get bashed. Start too late - get bashed.


PC World:
http://www.pcworld.com/article/id,127629-page,1-c,vistalonghorn/article.html


Make sure you read the readers' comments, too. They help counter PC World's usual nonsense.


Support charges raised:
$35 -> $59 per incident
Microsoft charges more for Vista support but not only for it!
http://www.digitmag.co.uk/news/index.cfm?newsid=7165


Bit like Apple, then:

http://www.apple.com/support/products/payperincident.html

Heaven help you if you need help with the command line interface:

For $695 USD, you may purchase a single-incident of phone support for...assistance with command-line interface tools.

Plus, Apple doesn't even have to solve the problem - merely identify the root cause.

How about Red Hat's prices? https://www.redhat.com/apps/commerce/

Sirber
8th February 2007, 18:01
You can disable some services...

http://www.speedyvista.com/services.html

jeffy
8th February 2007, 20:11
Wrong link, BTW. The correct link goes on to say "[t]o always safely eject an iPod, choose Eject iPod from the Controls menu within iTunes." Most likely Apple's hardware doesn't honor Windows' PnP power management etc. i.e., cr@p hardware drivers.


I'm sorry, the correct link is this: http://docs.info.apple.com/article.html?artnum=305042

Full quote:
"Ejecting an iPod from the Windows System Tray using the "Safely Remove Hardware" feature may corrupt your iPod. To always safely eject an iPod, choose Eject iPod from the Controls menu within iTunes."
BUT, why do you have "Safely Remove Hardware", just for this corruption to happen? It should have been solved before Vista hit the shelves!

JohnnyMalaria
8th February 2007, 20:27
BUT, why do you have "Safely Remove Hardware", just for this corruption to happen? It should have been solved before Vista hit the shelves!

Solved by Apple or Microsoft?

"Safely Remove Hardware" advises the hardware that it is about to be removed. The hardware is required to enter into a known power state. It is up to the hardware vendor's driver to ensure that happens.

I doubt it is Microsoft's fault. Much more likely is that the hardware driver (i.e., from Apple) is at fault.

Even Quicktime and iTunes don't follow the rules on Windows.

jeffy
8th February 2007, 20:28
Analysis: Vista's Ready Boost is no match for RAM
http://www.tgdaily.com/2007/02/08/analysis_vista_ready_boost/

Vista Licensing Hits Virtual Machine Users in the Wallet
http://www.macobserver.com/article/2007/02/08.2.shtml
http://blogs.zdnet.com/Bott/?p=157

Free System Tool Optimizes and Maintains Windows Vista in One Click
http://www.prweb.com/releases/2007/2/prweb503049.htm
http://iobit.com/

Windows Vista not without its flaws
http://www.mlive.com/mbusinessreview/stories/index.ssf?/mbusinessreview/se/stories/20070208_6.html
"A study conducted by International Data Corp. projects that for every dollar in revenue that Microsoft gets from the release of Vista, the information technology industry will get $18. The study predicted that Vista's release would create more than 100,000 jobs."

" McCully recently tried installing Vista on a computer less than two years old. The computer didn't have all the necessary features - including a video card that wasn't sufficient - to run Vista as advertised."

jeffy
8th February 2007, 20:31
Solved by Apple or Microsoft?

"Safely Remove Hardware" advises the hardware that it is about to be removed. The hardware is required to enter into a known power state. It is up to the hardware vendor's driver to ensure that happens.

I doubt it is Microsoft's fault. Much more likely is that the hardware driver (i.e., from Apple) is at fault.

Even Quicktime and iTunes don't follow the rules on Windows.
Sure, it might be Apple's fault. But I'm still wondering, with so many beta testers of Vista and so many iPods sold, why it remained unsolved??

JohnnyMalaria
8th February 2007, 20:36
Windows Vista not without its flaws
http://www.mlive.com/mbusinessreview/stories/index.ssf?/mbusinessreview/se/stories/20070208_6.html
"A study conducted by International Data Corp. projects that for every dollar in revenue that Microsoft gets from the release of Vista, the information technology industry will get $18. The study predicted that Vista's release would create more than 100,000 jobs."

" McCully recently tried installing Vista on a computer less than two years old. The computer didn't have all the necessary features - including a video card that wasn't sufficient - to run Vista as advertised."


I fail to see how a new OS having more hardware requirements is news.

Every new version of Windows has always meant that older hardware platforms need upgrading/replacing.

What about all the 286 PCs that suddenly became "obsolete" when Windows 3.1 came out (requiring at least a 386)?

And it isn't unique to Microsoft.

foxyshadis
8th February 2007, 21:02
Sure, it might be Apple's fault. But I'm still wondering, with so many beta testers of Vista and so many iPods sold, why it remained unsolved??

Apple has a vested interest in getting people to switch to Macs. Why else would Quicktime have remained the worst peforming media player on Windows for over 10 years? (If you don't count Realplayer's ads.) They can point to Macs and how smooth and lightning fast everything runs, because on Windows they're just running Quicktime on top of a slow Apple-to-Win32 translation layer for portability. So no, they don't necessarily care if it works on Vista until it effects enough people to move them.

zambelli
16th February 2007, 05:36
FYI...

I ran my own test of XviD encoding in Vista and XP.

I tested the xvid.cvs.head.MTK.2006.12.20 build of XviD in VfW mode with VirtualDub as the frontend.

I used an uncompressed YV12 AVI as my source, 1280x720 @ 59.94 fps, 60 seconds in length. I encoded in 1p CBR at 5000 kbps with default settings.

The test machine was a P4 3.4 GHz w/ HT enabled, with 1 GB of RAM. Operating systems installed were XP Pro SP2 and Vista Enterprise x86.

================
Encoding time results:
================

XP: 252 seconds
Vista: 254 seconds

That's less than a 1% difference, and probably within the margin of error of this test.

This result leads me to suspect even more that the difference noted in the Tom's Hardware benchmark was caused by a bottleneck in the MPEG-2 decoding process, not the encoding process. What exactly that bottleneck was - I don't know because the TH article doesn't document their MPEG-2 decoder setup.

Manao
16th February 2007, 07:19
zambelli : why did you use a source that was 1280 x 720 x 1.5 x 60 x 60 ~ 5 GB big ? My computer takes 3 minutes at least to read those 5 GB, and that's being optimistic.

In Tom's test, mainconcept ought to use its decoder both time. And it's directshow. So if you blame the decoder, it means it's in fact directshow that is guilty. And if directshow in vista slows down by 25% an encoding, by how much will it slow down a playback ?

zambelli
16th February 2007, 08:08
zambelli : why did you use a source that was 1280 x 720 x 1.5 x 60 x 60 ~ 5 GB big ? My computer takes 3 minutes at least to read those 5 GB, and that's being optimistic.
Are you suggesting the disk read rate is the new bottleneck?

I used a YV12 source because I wanted to be consistent in my test. At least I know in both cases that the source was the same, which was my main objection to the TH test: their lack of information about the sourcing/decoding process. It's the ambiguity of it that calls into question the validity of their test.

If you have a better suggestion, I can re-run my test again. I can run the test with DGDecode from an MPEG-2 source, for example.

In Tom's test, mainconcept ought to use its decoder both time. And it's directshow. So if you blame the decoder, it means it's in fact directshow that is guilty. And if directshow in vista slows down by 25% an encoding, by how much will it slow down a playback ?
Like I said, the issue is ambiguity. Did they say which decoder they used? Did they say DirectShow was used? No. Their benchmark setup could be perfect and the test result correct for all we know. The problem is that we don't know.

Manao
16th February 2007, 08:48
Are you suggesting the disk read rate is the new bottleneck?Rather that it might be. I always try to use DGDecode + a mpeg2source when doing benchmark. On my computer, it's what gives the most consistent timings.

But, in your case, I honestly don't think using a mpeg2 source will change the results that much : Vista can't interfere in DGDecode & avisynth workflow.

I think the trouble lies in directshow, since, in that case, Vista does interfere. So the tests you ought to perform would be :
- DGDecode + avisynth, just to be sure
- DirectShowSource + avisynth, enforcing that the same decoder / splitter gets used ( ffdshow would be best ).

Mainconcept, afaik, uses directshow, and uses their own decoder whenever possible. However, I don't have time to check right now, so I'll leave it as "afaik".

Sharktooth
16th February 2007, 19:21
Clearly the hard drive was a bottleneck unless zambelli had a hard drive capable of a transfer rate higher than 80Mb/s...
But maybe he had a multiple RAID stack with 15000rpm drives and some uberfast SCSI controller directly connected to the CPU bus..................
That test smells like a mere try to hide the Vista CPU inefficiency by forcing the CPU to NOT work at full rate...

migrena
16th February 2007, 21:18
ok, even if that hdd was fast enough, there is ~20% slowdown in decoding using directshow, think how many programs will be affected... and what with playback speed?
zambelli could you please test also decodnig speeds?

zambelli
17th February 2007, 00:30
That test smells like a mere try to hide the Vista CPU inefficiency by forcing the CPU to NOT work at full rate...
I don't give a damn whether Vista wins out or not in this test. I simply want to know the correct result of the benchmark, as I have already asserted that TH's test setup seemed flawed.

zambelli
17th February 2007, 02:59
Rather that it might be. I always try to use DGDecode + a mpeg2source when doing benchmark. On my computer, it's what gives the most consistent timings.
Fair enough.

I re-ran the test, this time with MPEG-2, a longer source, and repeated it for each OS to get an even more accurate measurement.

The rest of the setup was the same as first time around:


I tested the xvid.cvs.head.MTK.2006.12.20 build of XviD in VfW mode with VirtualDub as the frontend.

I used a 20-minute long MPEG-2 as the source, 720x480 @ 29.97 fps, progressive. DGDecode version 1.4.8. AviSynth version 2.57. I encoded in 1-pass CBR mode, at 1000 kbps, with default settings.

The test machine was a P4 3.4 GHz w/ HT enabled, with 1 GB of RAM. Operating systems installed were XP Pro SP2 and Vista Enterprise x86.


===========================================
Encoding time results:
===========================================
| OS | 1st run | 2nd run | Avg time |
|-----------------------------------------|
| XP | 1085 | 1084 | 1084.5 |
| Vista | 1086 | 1082 | 1084 |
===========================================


As you can see, encoding times were pretty much exactly the same.

Therefore, it is my conclusion that Vista/XP have no effect on encoding speeds whatsoever. I don't know what Tom's Hardware did in their test setup to make Vista transcoding 20% slower, but I sure am not seeing the bottleneck being in the encoding process.

If anybody wants to perform tests on DirectShow decoding to find if DirectShow MPEG-2 decoding is somehow slower on Vista, I would strongly encourage them. I have proven my point, which was that Vista vs XP have no effect on encoding speed.

Manao
17th February 2007, 05:04
I have proven my point, which was that Vista vs XP have no effect on encoding speed.There I disagree.

What you have proven is that a standalone program isn't slower by itself on Vista. That is, when you don't give Vista any opportunity to interfere, it's not slower ( you may argue that avisynth / virtual dub uses vfw, but in that case they only use the set of interfaces, nothing more ).

Though your conclusion is true for xvid & x264, I'm afraid it will be another story for directshow-based encoders ( Nero, and - still under the "afaik" asumption - mainconcept/moonlight ).

jeffy
17th February 2007, 07:36
Zambelli, could you please try to repeat the test with a short source and AutoGK 2.40? Tom's Hardware used a 182 MB VOB source for AutoGK, target size 100 MB.

More info:
http://www.tomshardware.com/2007/01/29/xp-vs-vista/page3.html

Thank you for your tests.

zambelli
17th February 2007, 09:57
Zambelli, could you please try to repeat the test with a short source and AutoGK 2.40? Tom's Hardware used a 182 MB VOB source for AutoGK, target size 100 MB.
Why, may I ask? My source was longer than theirs, therefore more accurate, and I used Avisynth+DGDecode, which I believe is exactly the same as what AutoGK uses.

jeffy
17th February 2007, 10:11
Why, may I ask? My source was longer than theirs, therefore more accurate, and I used Avisynth+DGDecode, which I believe is exactly the same as what AutoGK uses.
It will use two pass encoding like their test (target size specified).
Regarding bottlenecks, they used two Raptors Western Digital WD1500ADFD, one for system, one for data. May I ask if you used a standard 7,200 rpm or a better hard drive?

zambelli
17th February 2007, 11:21
It will use two pass encoding like their test (target size specified).
One or two passes won't matter. It's all number crunching throughout the whole process, without much interference from the OS.
Regarding bottlenecks, they used two Raptors Western Digital WD1500ADFD, one for system, one for data. May I ask if you used a standard 7,200 rpm or a better hard drive?
My HDD was just a 7200 rpm, but I really don't think the disk read rate was the bottleneck in the MPEG-2 test.

jeffy
17th February 2007, 11:29
Thanks for comments, Alex. I really am wondering why their test was so much slower... so thank you for the time you've already put into testing.

However, they were not the only ones, this time slowdown of WinRAR and Lame MP3:
http://www.bit-tech.net/bits/2007/01/30/Windows_Vista_review/4.html
For a fair Xvid MPEG-2 comparison, they said:
"The Vista system was slower here by 14 seconds, a very small amount and a negligible percentage."

Before Vista went RTM, this test showed also AutoGK slowdown:
http://www.anandtech.com/printarticle.aspx?i=2850