Click for ThunderBolt Upgrades!
Click for ThunderBolt Upgrades!


One Click on the Banner shows support to my Sponsors

Accelerate Your Mac!  - the source for performance news and reviews
The Source for Mac Performance News and Reviews
Don't forget to check out all the other site features!

iMovie 1 vs iMovie 2, QT 4 vs QT 5 Performance Tests:
Using G4/500 & Dual G4/500 Systems
First Published: 10/5/2000 (iMovie 1 vs 2)
and 10/18/2000 (QT 4 vs QT 5)



QT 5 Preview vs QT 4.1 - Dual vs Single G4 Systems iMovie2 Tests:

I tested Quicktime 4.1.2 against Quicktime 5 public preview on a Single G4/500 and Dual G4/500 system. It's clear from the results that dual processors are not a benefit when exporting to QT in the current version (2.01) of iMovie2. (Preview tests here included below showed iMovie 2 didn't support dual processors.) The upside to Quicktime 5 however is both systems were much faster than when using Quicktime v4. It's not often a software update delivers a 50% gain in any task.

Vincent Lugnier had sent a note that the Quicktime 5 developer PDF noted it supports dual processors: "Altivec and multiprocessor support fo H.263 and DV" so I decided to run a few tests in iMovie 2 on my dual G4/500 system and my single G4/500 system. (The dual had an OEM Radeon card, the single CPU system had the retail Radeon card. Both had VM off, OS 9.04.)

QT 5 Preview was appx. 55% faster than QT 4.1.2. My original thoughts were dual processors were supported, but that's clearly not the case. Due to slight extension set differences or perhaps overhead from the dual G4 system, the single G4 system was slightly faster in all tests. (FYI - the default codec used in these tests is H.263.)

iMovie 2 Export Movie Tests
(default CDrom Medium Size Movie settings)
QT
v4.1.2
QT 5
Preview 2
Dual CPU G4/500 Tests
4 Min, 12 Sec. 2 Min, 43 Sec.
Single CPU G4/500 Tests
4 Min, 3 Sec. 2 Min, 40 Sec.

The gain seen over QT 4.x I assume may be due to better Altivec support in QT 5. It's nice to see such a boost from just a software update.

iMovie 2 vs iMovie 1 Performance on Dual G4/500 System:
Awhile back a reader had commented that it appeared that iMovie 2 not only has new features but also support for dual processors. I didn't have a chance to test this before my G4/500 DP arrived, but just did - with some interesting discoveries.

I have iMovie 1 and 2 on the G4/500 DP and used the Tutorial file from iMovie 1 as a test. I found that iMovie 2 did render the same movie in less than 1/2 the time of iMovie 1 on the same system using the Export movie "Medium Sized CDROM Movie" setting. One iMovie 1 performance drag is that I've found no way to disable the preview of the movie while it's exporting. In iMovie 2 there seems to be no way to enable this, so I can't test what effect this has on performance.

Despite the fact iMovie 2 doesn't preview during export, the fact it took less than half the time to export the movie made me dig a little deeper into the differences. I compared the output file sizes and found the iMovie 1 file was 11.1MB and the iMovie 2 file was 6.6MB. The size, FPS and audio settings were identical, so it had to be the compression codec. I should have noticed in the export dialog that iMovie 1 uses Sorensen by default for the Medium Movie export and iMovie 2 uses H263. Looking at the two movies side by side you can see the iMovie 1 Sorensen codec video is much sharper with better color fidelity.

Although it's still not a fair test since iMovie 1 previews the movie during export (but I did select lower quality previews to help a bit), I then ran a test with the same codec and settings using expert mode. (Sorensen codec, medium quality, 15FPS, keyframe every 5 frames, 600kb/sec data rate limit, 44KHz stereo audio, IMA 4:1 compression)

    iMovie 1 vs 2 on G4/500 Dual CPU
    (same codec and settings used)
  • iMovie 1: 9 Min, 47 Sec.
  • iMovie 2: 3 Min, 0 Sec.

Despite identical settings, the iMovie 1 file was 500KB larger (21.2MB vs 20.7MB). Using QT Pro's movie info option I saw that this was due to a slightly higher data rate on the iMovie 1 file.

I wish I could have disabled the preview during export in iMovie 1, or enabled it in iMovie 2, so that I'd have identical test conditions. Even if the 2nd CPU was 100% efficient, that can't explain the 3x+ faster performance in that test.

To determine if the 2nd G4 was really supported in iMovie 2, I ran the same export file test on my G4/500 single CPU system. This test clearly showed that iMovie 2 does not use the 2nd processor.

    iMovie 2 Export - Dual vs Single G4/500
    (same codec and settings used)
  • G4/500 Single CPU: 2 Min, 53 Sec.
  • G4/500 Dual CPUs: 3 Min, 0 Sec.

The G4/500 single CPU system has a slightly faster drive (7200 RPM WD vs 5400 RPM Maxtor) which may be the reason for the slightly faster times, and there may be some overhead with a 2nd CPU for apps that don't use it. The results however make it clear that iMovie 2 doesn't use the 2nd processor, although the much faster performance (and additional features) over iMovie 1 is a big plus.

For those that missed it, Hiram Wilson wrote an interesting iMovie article several months back that compared making movies with Film to iMovie titled 'iMovie - A Revolution Takes Hold


Test System Configurations:

(Linked items go to reviews here of that card/device)

  • Apple G4/AGP Single CPU
  • 500MHz G4 CPU (CPU rev 2.8)
  • 1MB of L2 Cache at 250MHz
  • 100MHz system bus
  • OS 9.04 with Firmware update
  • Virtual Memory Off
  • 256MB PC100 RAM (222)
  • 45GB IBM 75GXP IDE Hard Drive
  • 6X DVD ROM (IDE)
  • ATI Rage128 PRO AGP w/16MB SDRAM (iMovie1 vs iMovie2 tests)
  • ATI Radeon AGP w/32MB DDR RAM (QT 4.1 vs QT 5 tests)

  • Apple G4/AGP Dual CPU
  • Dual 500MHz G4 CPUs
  • 1MB of L2 Cache at 225MHz (x2)
  • 100MHz system bus
  • OS 9.04 with Firmware update
  • Virtual Memory Off, 128K Disk Cache
  • 1GB PC100 RAM (two 512MB 222 dimms from OWC)
  • OEM 20GB (Maxtor) IDE Hard Drive
  • ATI Radeon OEM AGP w/32MB DDR RAM


= Other Site Topic Areas =
Systems | CPU Cards | SCSI | IDE | Firewire | Cache | Video | Forums | FAQ

Back to XLR8YOURMAC.COM
Your Source for the best in CPU/SCSI/VIDEO card reviews, daily news, and more!

Copyright © Michael Breeden, 2000.
No part of this site's content is to be reproduced in any form without permission.
All brand or product names mentioned here are properties of their respective companies.
Disclaimer: Users must read and are bound by the Site Terms & Conditions of Use.