My informal MacBook iTunes rip speed tests
Below are preliminary test results. The final MacBook Pro Performance Analysis can be found on craigtheguru.com.
Today I tested iTunes on the new MacBooks as well as older PowerBook G4 and PowerMac G5 systems. On each Mac I converted songs using the "Convert Selection to AAC..." feature and noted the rip speed as displayed in iTunes. The sources were unprotected AAC files and they were re-encoded using AAC. Unfortunately, identical files were not used on all machines. The machines were updated and had plenty of RAM. Here's the results:
Machine Model - iTunes Rip Speed
MacBook Pro 2x1.83 (Macworld demo unit) - 4.5x (Actually 20x)
PowerBook G4 1.67 (15" 1280x854 model) - 12.5x
PowerMac G5 2x2.0 (rev A) - 25x
Now I expected the PowerPC processors to win, but not by so much! I'm sure the G4 and G5 processors outperform the Core Duo chip using their vector processing units. I am surprised how far behind the new MacBook is, especially since it is dual core. However, the MacBook's testing environment was the least controlled and the media file was relatively small, so perhaps it was just getting warmed up.
Aside from this test, my other experiences with the MacBook are positive and favorable, including using Photoshop under Rosetta. The GPU in the new MacBooks played HD video very well, far better than G4 PowerBook I'd expect. And the boot speed of the intel systems is insane!
I could write more about using the MacBook but this concludes my test.
--
Followup Notes: My initial Macworld MacBook test produced erroneous results and an encoding speed of 4.5x. Later I returned to Macworld and performed a much more structured test and got a far better result: 20x. Since much of this discussion is based around the original results I'm leaving this post as-is, however MacBook performance is far better than indicated. Please read the rest of this thread for more information.
Also, I will post fully revised results later this weekend.
Today I tested iTunes on the new MacBooks as well as older PowerBook G4 and PowerMac G5 systems. On each Mac I converted songs using the "Convert Selection to AAC..." feature and noted the rip speed as displayed in iTunes. The sources were unprotected AAC files and they were re-encoded using AAC. Unfortunately, identical files were not used on all machines. The machines were updated and had plenty of RAM. Here's the results:
Machine Model - iTunes Rip Speed
MacBook Pro 2x1.83 (Macworld demo unit) - 4.5x (Actually 20x)
PowerBook G4 1.67 (15" 1280x854 model) - 12.5x
PowerMac G5 2x2.0 (rev A) - 25x
Now I expected the PowerPC processors to win, but not by so much! I'm sure the G4 and G5 processors outperform the Core Duo chip using their vector processing units. I am surprised how far behind the new MacBook is, especially since it is dual core. However, the MacBook's testing environment was the least controlled and the media file was relatively small, so perhaps it was just getting warmed up.
Aside from this test, my other experiences with the MacBook are positive and favorable, including using Photoshop under Rosetta. The GPU in the new MacBooks played HD video very well, far better than G4 PowerBook I'd expect. And the boot speed of the intel systems is insane!
I could write more about using the MacBook but this concludes my test.
--
Followup Notes: My initial Macworld MacBook test produced erroneous results and an encoding speed of 4.5x. Later I returned to Macworld and performed a much more structured test and got a far better result: 20x. Since much of this discussion is based around the original results I'm leaving this post as-is, however MacBook performance is far better than indicated. Please read the rest of this thread for more information.
Also, I will post fully revised results later this weekend.
Comments
More importantly, Rosetta runs fast Photoshop you say?
Originally posted by BRussell
Hmm, odd. You used the term rip as if it was from a CD, but they were already AAC files? I don't understand why the new machine would be so much slower, unless it was the optical drive.
True, "ripping" generally implies two aspects: reading data from an audio CD and encoding it. However, to eliminate the CD mechanism as a bottleneck, I skipped that step by reading music off the HD and only encoded the music files. So yes, technically I should probably say "encode speed", but I thought I was clear in the testing methods and besides, the key part of ripping music is the encoding process anyhow.
ripping speed is important for me .... hey the speed of the macbook pro is the worst i've heard about.. my old old emac do the job at 7,5x .. i can't believe what i read!!!
me, who wanted this new macbook pro.. never never never...
Originally posted by Mr. H
There's something very wrong with the OS X intel version of iTunes. My 1.4 GHz Duron with 384 MB RAM running Windows 2000 and iTunes 6.0.2 converts aiff to 128 Kbps AAC at 10 x speed. Hopefully Apple will fix this quickly.
The guys at the booth said they were prototypes, not production models. Chances are the that these problems will be dealt with (or have been dealt with) by the release date.....I hope
Originally posted by taliesin
The guys at the booth said they were prototypes, not production models. Chances are the that these problems will be dealt with (or have been dealt with) by the release date.....I hope
Except the iMac, which has actually been released, has the same problem:
see Xool's post in this thread
The problem is with iTunes, not the hardware.
You should test it on something other than itunes. You even said yourself that other things "were insane"
Trust me the new notebooks a "freaky" fast.
Originally posted by Mr. H
There's something very wrong with the OS X intel version of iTunes. My 1.4 GHz Duron with 384 MB RAM running Windows 2000 and iTunes 6.0.2 converts aiff to 128 Kbps AAC at 10 x speed. Hopefully Apple will fix this quickly.
Originally posted by webmail
Yup, agreed iTunes under OS X intel is either not native or has another problem it's not the same.
I thought that converting to universal binaries was very easy? Shouldn't all the apple applications in 10.4.4 already be in universal binary format?
Originally posted by e1618978
I thought that converting to universal binaries was very easy?
Converting them is one thing, make them work properly is another.
Originally posted by e1618978
I thought that converting to universal binaries was very easy? Shouldn't all the apple applications in 10.4.4 already be in universal binary format?
Probbly parts of the apps. The converting mechanism probably not. And besides sjobs doesn´t always tell the truth
Originally posted by e1618978
I thought that converting to universal binaries was very easy? Shouldn't all the apple applications in 10.4.4 already be in universal binary format?
Easy for some apps that were written natively on OS X right from the start. But remember that iTunes was written on OS 9 and comes from the SoundJam codebase. There's probably some very ugly code in there.
I've been hoping that Apple would rewrite iTunes from scratch for over 3 years now. It hasn't happened so far...and look at what trouble iTunes gets into now.
Still doesn't explain why the Windows version of iTunes is faster but oh well.
Originally posted by keikojaa
seems to me that itunes 6.0.2 is "universal" .. i did chek with right click on itunes and "read the informations" (in french macosx LIRE LES INFORMATIONS..) and it mentions "Universal".. so i guess iTunes is ready for intel. So i can't explain the so slow import speed...
Perhaps you did not understand what kim kap sol said above. iTunes is an old Mac application with code from the OS 9 days, something that would make it difficult to convert it properly into a Universal Binary. Mind you, it has been converted but it is not yet optimal. So it seems at least.
Originally posted by PB
Perhaps you did not understand what kim kap sol said above. iTunes is an old Mac application with code from the OS 9 days, something that would make it difficult to convert it properly into a Universal Binary. Mind you, it has been converted but it is not yet optimal. So it seems at least.
Not yet optimal????
Let's not kid ourselves, 4.5x encoding speed is total, utter and complete crap. My 500 MHz PowerBook G4 (originally introduced exactly 5 years ago) is faster than that.
is that itunes, at my job, on a pentium 4, is encoding fast at 20x minimum. so pentium is intel x86, itunes is apple and it is possible to make itunes scream on a pc.. why not on a intel mac?
This is very unscientific, and we know that the computers are fast.
If a proper benchmark is shite, you can panic *then*.