DigitalVideoEditing Adobe Dual: G4 1Ghz vs Dual Athlon 2000+MP vs Single P4 2.53Ghz

[Apex]

http://images.digitalmedianet.com/2002/07_jul/features/cw_macvspc2/macvspc2title.jpg

Digital Video Editing - Professional Production Resources just did a dual in Adobe After Effects 5.5 using QuickTime codec (for both Mac and PC), and Adobe Photoshop 7.0.

Competitors (each with 512mb memory, all using ATA drives):

Dell Precision 340, single processor P4 2.53Ghz, WinXP
Boxx dual processor Athlon 2000+MP, WinXP
Macintosh dual G4 1Ghz, OSX 10.1.5

The editors loved all 3 machines. Here is the comparison test though:

Benchmark Duel: Mac vs. PC, Round II (http://www.digitalvideoediting.com/2002/07_jul/features/cw_macvspc2.htm)

Round I did not include the Dell.

[Speedfreak]

That was no surprize to me. This is a place where the P4 shines over the Athlon. As for the Mac, well, I though it would be more competative. I guess the speed really isn't double or triple that of a PC. It's a 2Ghz total box, yet it doesn't even seem like that.

[ribitch]

well apex, i give you credit, using after effects the PC kicked the apples ass. Why? After Effects isnt written to fully utilize the second mac processor. Photoshop has a few more optimizations for a dual processor ssytem, but for the most part it isnt. Take a look at benchmarks completed by a single CPU 800 MHz G4 and a dual CPU 800 MHz G4, or a sincle 933 and a dual GHz. The daul GHz slighltly edges out the 933. My guess is adobe rewrote most of their apps for OS X instead of porting the apps for OS X.

Now try an app like combustion. That app is written to utilize a second CPU better than photoshop 7 or After Effects 5.5. Any DV edittor knows com

As with most mac vs pc tests this magazine does, the pc always edges out the mac. why? partially because this source is always more pc friendly than mac friendly, and secondly with any multiplatform benchmark, one set of tests would run fatser on one platform than on the other platform. We all know its true. A series of photoshop filters runs faster on a pc than mac, while a second set of filters will run faster on a mac than the pc. Same will go with after effects.

we could both post benchmarks all day long using photoshop and after effects showing one system is better than the other. why? because for that task, one system is faster than the other. but for another task its vice versa.

i would still take a mac running fcp3 over the pc running premiere. I would like to see a set of benchmarks showing that off.

theres a reason that you can walk into almost every major video pordduction house and see more macs than pc's. in the big picture, when all factors are joined together, not jsut one app, the mac is the platform of choice by DV professionals.

[Jeffbx]

It would be interesting to see the same tests run with the Precision 530 w/ dual Xeon processors...

[skynet]

I think it is kind of unfair that the pentium has only one chip.
I agree with JeffBX. And ribitch, Do you think that those programs are all written with the proper P4 or even AMD optimizations like SSE, MMX, 3DNOW and what ever other ones there are?

I had to use a MAC in my photoshop class last year. Our usinversity just bought them brand new and I hated using them. I was open minded but I really felt like it was slower. And not only that but with was like moving a fat woman. You have one mouse button, I don't like the way it clicked and it was just so stiff like. Any one know what I am talking about? The MAC seemed just not as responsive as the PC.

[jase71]

Originally posted by ribitch
well apex, i give you credit, using after effects the PC kicked the apples ass. Why? After Effects isnt written to fully utilize the second mac processor.

Perhaps true... but realistically, just like the Apple vs. Dell server discussion, they test what shipped, not what "might have been".

In the server study, what shipped favored Apple over Dell for the benchmarks. In this case, it didn't.

Ya win some, ya lose some. Maybe this will spur Apple to clean things up and make better use of that second CPU. Or at least, prod Adobe to do it for their own app...

[LegendKiller]

Originally posted by jase71


Perhaps true... but realistically, just like the Apple vs. Dell server discussion, they test what shipped, not what "might have been".

In the server study, what shipped favored Apple over Dell for the benchmarks. In this case, it didn't.

Ya win some, ya lose some. Maybe this will spur Apple to clean things up and make better use of that second CPU. Or at least, prod Adobe to do it for their own app...


I like your way of connecting the servers to this. Apple people usually find fault in a benchmark that doesn't favorably agree with them but really does not find fault (especially where glaring ones exist) in benchmarks that DO agree with them.


Even if the second proc was used, do you think it would give it THAT much of an advantage? I doubt it.

As people have pointed out before, do you know why people use Macs for DV? Because there is a very restant bug in the system, its called the Mymacisbestatdvnomatterwhatyousay-itis. People have never really taken a look at which system is better (except for tests done by DV people on fixed systems), although I did find one with a homegrown PC v mac a few years ago. For 5k they configured both systems and guess what? A thunderbird 1ghz beat the snot out of the mac. A week later I had approval from one of the professors to build him more or less the same syste (with my own changes of course), and he loved the thing.

There is a mentality that will not quit, and that is that macs are ALWAYS better at DV, and publishing. Humans are animals of habit, they will flock to what they know and will not dare to try something out of usual, especially in this case...


LK

[jase71]

Originally posted by LegendKiller



I like your way of connecting the servers to this. Apple people usually find fault in a benchmark that doesn't favorably agree with them but really does not find fault (especially where glaring ones exist) in benchmarks that DO agree with them.

Luckily, I'm not an Apple person. ;)

As long as they're done fairly, or as close as possible, on a level playing field, benchmarks don't lie. We can all nitpick each study and benchmark to death, if we want to. None of them are perfect.
We can poke enough holes to discredit any study, if we have an agenda to promote.

But if they get at the basic truth of the matter, then they're valid.
And while we may not like the results of one study or another, that doesn't really change the outcome. It's more productive to accept it and address it than to deny it, and pretend the problem doesn't exist.

Apple did well in the server study. A similarly priced Apple outperformed a competitively priced Dell. That's a coup for Apple.

And in this study, Apple didn't fare so well. Whether it's because the software didn't use the second cpu properly, or a flawed system design, or just slower hardware, it doesn't really matter. As the system shipped, it didn't compete as well.

I didn't see any glaring faults in the first study, so I'll accept the outcomes. And I didn't see any glaring faults in this one, so I'll accept the outcomes.



Even if the second proc was used, do you think it would give it THAT much of an advantage? I doubt it.

I don't know. I won't speculate. Maybe it would, maybe it wouldn't. It doesn't even really matter. As the system shipped, it didn't use it efficiently. Until it does, speculating about how it would change the benchmarks is kind of pointless. If Apple or Adobe fixes the problem, then benchmark 'em again. Until then, the benchmarks stand as accurate.



As people have pointed out before, do you know why people use Macs for DV? Because there is a very restant bug in the system, its called the Mymacisbestatdvnomatterwhatyousay-itis.

This is a rampant problem in the entire industry, both on the PC side and the Apple side.

PCs are better for office apps and games, ya da ya da ya da...

Apples are better for graphics and video, ya da ya da ya da....

And anyone attached to one is very unlikely to switch to the other, or even admit it has strengths over their own choice.

PC versus Mac is a Ford vs. Chevy kinda thing. Both work. Both work pretty well. Sometimes one accomplishes a specific task better than the other. But overall, both are very capable, and will do pretty much everything the other does. It comes down to personal choice, and maybe whether one of those strengths will have a major benefit for you....

[ribitch]

Originally posted by jase71


Luckily, I'm not an Apple person. ;)

As long as they're done fairly, or as close as possible, on a level playing field, benchmarks don't lie. We can all nitpick each study and benchmark to death, if we want to. None of them are perfect.
We can poke enough holes to discredit any study, if we have an agenda to promote.

But if they get at the basic truth of the matter, then they're valid.
And while we may not like the results of one study or another, that doesn't really change the outcome. It's more productive to accept it and address it than to deny it, and pretend the problem doesn't exist.

Apple did well in the server study. A similarly priced Apple outperformed a competitively priced Dell. That's a coup for Apple.

And in this study, Apple didn't fare so well. Whether it's because the software didn't use the second cpu properly, or a flawed system design, or just slower hardware, it doesn't really matter. As the system shipped, it didn't compete as well.

I didn't see any glaring faults in the first study, so I'll accept the outcomes. And I didn't see any glaring faults in this one, so I'll accept the outcomes.



I don't know. I won't speculate. Maybe it would, maybe it wouldn't. It doesn't even really matter. As the system shipped, it didn't use it efficiently. Until it does, speculating about how it would change the benchmarks is kind of pointless. If Apple or Adobe fixes the problem, then benchmark 'em again. Until then, the benchmarks stand as accurate.



This is a rampant problem in the entire industry, both on the PC side and the Apple side.

PCs are better for office apps and games, ya da ya da ya da...

Apples are better for graphics and video, ya da ya da ya da....

And anyone attached to one is very unlikely to switch to the other, or even admit it has strengths over their own choice.

PC versus Mac is a Ford vs. Chevy kinda thing. Both work. Both work pretty well. Sometimes one accomplishes a specific task better than the other. But overall, both are very capable, and will do pretty much everything the other does. It comes down to personal choice, and maybe whether one of those strengths will have a major benefit for you....

great points jase. I did admit that the apple lost in this benchmark. One thing that you can do with video/photo editting benchmarks is to pick the test s that your system will perform best in. Being the fact that after effects and photoshop has so many differnt capabilities, one could come up with a slew of tests and people will argue all day long on what is better.

The server benhmark was different. it was done by the developer, so it didnt show any bias on platform. It simply displayed the raw perfromance of the system when coupled with their software. it was not a pick and choose type situation.

[jase71]

Originally posted by ribitch


great points jase. I did admit that the apple lost in this benchmark. One thing that you can do with video/photo editting benchmarks is to pick the test s that your system will perform best in. Being the fact that after effects and photoshop has so many differnt capabilities, one could come up with a slew of tests and people will argue all day long on what is better.


Sure, you can pick the tests that'll show off your system. You can selectively choose tests your preferred machine will win, and eliminate tests where it won't. And you can publish that as benchmarks. And the other guys can do the same.

And you know what? It doesn't even really matter, as long as the individual benchmarks were run fairly. It's not great journalism, but the individual benchmarks are still valid. They may not represent the entire picture, but the numbers still hold true.

For example, the tests run in this study that showed the PCs winning, were those tests representative of the types of things end users would do with the system? Were the same filters run on each machine? Did they run filters that people actually use in the real world?

If so, then for the tests that were run the numbers are acceptable, and valid. And they can be used for comparison's sake between the systems to weigh their strengths and weaknesses.

However, you have to read between the lines. These numbers are perfectly valid for the tests that were run. They may not indicate comparitive system performance for other tests or applications. PCs might not hold such a huge lead if a different battery of benchmarks were run. Or, they might widen the gap even more. But it doesn't matter. Because all those other tests weren't run. Just the ones discussed were run.

Too many people try and take a study like this one, or the server study, and extrapolate it to mean that PCs are always faster, or Macs are always faster. That's a crock. You can't paint broad pictures like that from one set of benchmarks, or even several.

All a study like this proves is that under these circumstances, with these machines, performing these tests, the performance rankings came out like this. It only applies to this scenario. You have to weigh each test, and consider it a piece of the puzzle, not the whole thing.

Under a different set of rules, the rankings might be different. Or they might not be. And if someone comes out with a similar study, and the tests show Apples to be faster, as long as the tests were run fairly, those numbers are valid too. PCs are faster at one set of tasks. Apples might be faster at another. That shouldn't come as a great surprise to anyone.

It's just a set of benchmarks. Interesting, no doubt. But not material to make a case that one system is always better or faster than the other. Same applies to the server test. Very interesting results, but not the final word.

[Ladogaboy]

Two questions:

1) Would it actually cost $4,000 to put a Dual MP system together? :confused: I would think I could do it for cheaper...

2) ribitch, is there a way to get FCP to work on a PC?

[ribitch]

Originally posted by Ladogaboy
Two questions:

1) Would it actually cost $4,000 to put a Dual MP system together? :confused: I would think I could do it for cheaper...

2) ribitch, is there a way to get FCP to work on a PC?

nope. apple only software. apple also has dvd studio pro, which is apple only. a new version of shake will be coming out in august (used in ice age), and that is cross platform. pc and linus versions are double the apple price. I also think apple has cheap upgrades from nonapple to apple with that one.

as a side not, shake requires a 3 button mouse. apple owns shake, so i think apple may have a mouse coming out soon.

apple has aquired several big audio/video companies in the past 6 months, and still has serverl billion dollars (4 i think) in the bank. I wouldnt be suprised if apple released so major a/v app or purchases a few more companies in the coming months.

[LegendKiller]

I find this pretty funny...

http://www.theregister.co.uk/content/35/26331.html


Especially this email...

The difference between the Reverend Jim Jones and the People's Temple crowd and Job's and the Mac Zealots as far as I can see is this:

Jones gave his people the Kool Aid and they drank it expecting to live forever in Heaven..

Jobs would SELL them the Kool Aid -- presumably in little white and blue polystyrene disposable cups with OS (skull and crossbones) logo on the outside. The Mac Zealots would drink it expecting to live in Heaven on Earth.

Murray McDowall



HAH!


Ribitch, what do you think of .mac?

[jase71]

How much was the WindowsXP upgrade? What did it really offer over Win2K?

Seems to me there's a parallel here between the 2K->XP path and the OSX->OSX.2 path... in almost every respect.

One more reason to stick to Linux, methinks... :P

[LegendKiller]

There is no denying that they are charging a crapload for few extra's. I just like the whole .mac thing, well, but then there is also the 130 price tag for a little bit. The Win98 upgrade was only 89, xp home upgrade is ~100, XP pro upgrade is ~130.


LK

[jase71]

Originally posted by LegendKiller
...XP pro upgrade is ~130.



And since XP Pro is the closest equivalent to X.2, it seems like a wash to me...

(Mandrake 8.2 -> $free. :P )

Apple's just using tactics they learned from Microsoft. We wouldn't be laughing at Apple now if Microsoft hadn't led the way for this kind of thing a while back. They're companies. Their bottom line matters more to them than your computing experience. You're a revenue source to them, and their concern for you is solely based on your continutation as a revenue source. So to show them too much loyalty, when they show none to us, seems a little silly.

[LegendKiller]

I used linux about 4 years ago, then my GF started using my computer and she always complained about it being harder to use, so me, being a pussy, switched back to windows.

Which distro can run games (natively)? Run DirectX games (Neverwinter nights, Warcraft3), have StarOffice, and have an easy to use window manager (I used to be good at command line, but no longer)?


LK

[ribitch]

i have mixed feelings on .mac. I dont like how they are charging for what was free, but then they also give you a whole lot more than they used to. for 100.00, you get free backing up of your system, and cds and dvd's, virex antivirus from mcafee, 100mb iDisk storage, 15 MB email mailbox, and a homepage. for that amount, it isnt that bad of a price. What would you pay for a backup program and antivirus program together? So the price isnt too bad.

I DO NOT like the name at all. the reasoning behind the name change though is Tenon offers an iTools program. nobody has the .mac name. I think its cheap that they ripped off ms off on it(.net).

the OS price isnt that bad. from the beta i have been using, there ar emany many more features. You will also be getting iSync and iCal with that 130.00, or 20.00 upgrade fee. Apple offered a free upgrade to 10.1, but it was cracked and widely distibuted across the net. The crack allowed the upgrade CD to be a full install cd. They do not require a serial number for the OS though, and they have loose liscensing on the OS when compared to MS.

I will be getting the update for free since I am an ADC member. I have recieved 9.0, 9.1 9.2, 10, 10.1, and all minor individual updates on cd's for my 99.00 membership fee. I also have 9.x in every language imaginable (something like 7 versions). X has all languages on 1 CD. 10.2 will be a 2CD install to allow for all languages, and all apps to be included on it.

So for 130.00, you are getting iDVD, iTunes, iSync, iMovie, iCal, iChat, iPhoto, the main OS, and several bundled programs.

peoplel will bitch about any price increase. Its the way life is. When you look at the big picture, most people who use OS X have a coupon that came with the OS or their computer entitling them to a 19.99 s&h upgrade. many of those coupons have been tosses out by the consumer. why is it apples fault they threw out their discount? I have mine, my gf has hers, and i have a few from when i was with apple.

from what i have used in 10.2, the upgrades are more than alittle bit. There is a great deal that has changed.

[jase71]

Originally posted by LegendKiller

Which distro can run games (natively)?

Technically, any of them. Though you'll have the least amount of work involved if you pick a standard one with a good install process, like Mandrake, RedHat, or Suse. I play Railroad Tycoon II natively, and used to play Quake3, Unreal Tournament, and Myst II. There's a limited selection of games out there for Linux if you need a gaming fix now and then. Just not a huge number of them to choose from.



Run DirectX games (Neverwinter nights, Warcraft3)

Neverwinter Nights has, or at least will very shortly, a Linux-native version. Was supposed to be released in the box with the Windows version, but fell through at the last minute.

You can play some DirectX games using Wine or WineX, an emulator of sorts. Speed of the games, and whether they'll run depends largely on whether enough people like the game to make it work, and make it work well. And handful of games actually run faster in WineX than Windows. But that's unusual. Most games that will run, run slower. It's essentially an emulator, after all. And a good portion won't run at all. DirectX is a MS standard, after all, not an open source one, and they're not keen about releasing the API info so people can easily emulate it.

I don't use WineX, and haven't kept up with it, becuase I don't play many games anymore. There are enough native games to keep me happy on the rare occasion I feel like playing a game.

But while you can play some games on Linux, it's not a gaming OS. Basically, it's pretty lousy for gaming. That doesn't bother me, because if I spend two hours a month playing games, it's more than usual. Last game I was hooked on was BGII, and a little of Medal of Honor. Nice games, and games are fun, but I usually tend to do something more constructive if I have spare time... so lack of games doesn't bother me at all anymore.



, have StarOffice, and have an easy to use window manager (I used to be good at command line, but no longer)?

There's StarOffice available, Open Office, the KDE Office Suite (my favorite), WordPerfect... and a couple of other alternatives I know I'm forgetting.

For Windows managers or desktop environments, there's Gnome 2.0 which just came out, and KDE 3 which just came out. Both are complete environments, much like Windows. Both, especially KDE, function much like Windows. I prefer KDE, but I haven't tried Gnome 2 yet...

There are, of course, all the other countless little windows managers like FVWM, Windowmaker, Sawfish, and others that offer fewer features than a full environment like KDE or Gnome, but also have a correspondingly lighter resource footprint.

The GUIs are extremely customizable now, with a ton of different options. And the nice thing about having so many to choose from is that you can tailor your GUI to your hardware. Running on older hardware? Choose a light window manager like FVWM. Low on eye-candy, but quick and light for an older pc. Got cpu and memory to burn? Play with something like Enlightenment and Gnome. Gorgeous, lots of eye candy, and stiffer requirements as well. Rather than being stuck with one GUI with just a few options to accomodate your hardware level, you have a ton of options that let you run as light, or as flashy as you want.

The downside to Linux, obviously, is still the learning curve. I've tried it off and on over the last four years, and it wasn't until the recent versions, specifically Mandrake 8.2, that I was satsified enough to stick with it. Prior to that there was always something I couldn't do with Linux that made me slink back to Windows. But now, it does everything I want, and a few things I didn't know I wanted until I had them, so that now I couldn't live without it... Windows seems too limiting when I boot my other pc to it.

It's not for everyone, that's for sure... I don't recommend people switch, because most people are more concerned about buying KiddiePhoto+ from Best Buy and having it run than they are about the strengths of linux. Linux is not ready to be Joe Ditchdigger's home pc OS. It requires too much thought on the part of the user, where Windows does all the thinking for you, so he can sit and happily play DeerHunterVII. But if Linux meets your needs and you're willing to put in the effort, it can be a killer OS....