Apples to Apples II: Dualies and Droolies
Page 4 of 4
Web graphics and video
The Web graphics and video category is a new one for our benchmark series, and hence we have no numbers for the old 933 MHz G4. Nevertheless, we were able to run tests on all of the new systems and the old dual 1 GHz G4 for comparisons. [an error occurred while processing this directive]
For vector and SWF encoding, we used Wildform Flix Pro 3.1, which takes excellent advantage of multiple processors. (In fact, its speed is nearly twicew that of the previous release of Flix.) And for video encoding, we used the brand, spankin' new Cleaner 6 from Discreet, which is only available for the Macintosh right now. Cleaner 6 now also makes excellent use of multiple processors on the Mac and does particularly well in Mac OS X 10.2. (Look for my separate, formal review of Cleaner 6 soon.)
For Flix Pro, I ran three tests. The first one was a two-pass, variable bit rate encode of a video file into the Flash MX format (using the "512K CD/Broadband High Video MX" preset). The second was a Flash 3-6 encode of the same source file (using the "512K CD/Broadband High Video 3-6" preset). And the third used Flix Pro's unique conversion of raster video into animated vectors.
Flix 3 Tests | ||||||
(All results listed in seconds.) | ||||||
Dual 1.25 GHz | Dual 1 GHz | Dual 867 MHz | Dual 1 GHz | Dual 1 GHz Old Model Running OS X 10.1.4 | Single 933 MHz | |
Test 1 | 219 | 268 | 306 | 270 | NA | NA |
Test 2 | 39 | 48 | 55 | 49 | NA | NA |
Test 3 | 101 | 126 | 146 | 126 | NA | NA |
Totals | 359 | 442 | 507 | 445 | NA | NA |
Our Cleaner 6 benchmarks also involved three tests. There are so many options in the new Cleaner 6 that I won't list every single parameter used in the tests. The first one was a simple QuickTime transcode from Photo JPEG on a 13-second D1 source file to the Apple Animation codec. The second used the same source file on a transcode to MPEG-2 with a numeric crop. And the third was a transcode to the Windows Media 7 format with a high data rate for audio and video.
Cleaner 6 Tests | ||||||
(All results listed in seconds.) | ||||||
Dual 1.25 GHz | Dual 1 GHz | Dual 867 MHz | Dual 1 GHz | Dual 1 GHz Old Model Running OS X 10.1.4 | Single 933 MHz | |
Test 1 | 65 | 76 | 89 | 97 | NA | NA |
Test 2 | 99 | 123 | 145 | 127 | NA | NA |
Test 3 | 35 | 44 | 52 | 48 | NA | NA |
Totals | 199 | 243 | 286 | 272 | NA | NA |
Conclusions
Benchmark tests can be a bit tricky. So much depends on how well the software is written for the hardware. This is particularly sticky when comparing software on different platforms. Fortunately, we don't have that problem with these particular benchmarks, since we're comparing five Macintosh G4s. Still, when it comes to the value of multiple processors, the particular piece of software you use can have a huge impact on the responsiveness you see from the system. Adobe's software is not particularly friendly to multiple processors, though this depends largely on the filters and functions being tested.
It's in this area alone that the low-end dual 867 MHz G4 failed to keep pace with the previous generation of G4s. But in all other tests, even this low-end machine churned out some impressive results. Apple, of course, likes to position its lower-end machines more toward consumers than professionals, but for those of you working with Web graphics or video compositing, editing and effects, the dual 867 might very well find a place in your production workflow, particularly with its relatively low price point.
The mid-range and high-end models, however, blow away the dual 867 and even the old dual 1 GHz model. Their faster system busses, more sophisticated system controllers, faster memory and faster disk throughput give them some distinct advantages over the previous generations of G4s so that even with half the level-3 cache, the new dual 1 GHz model manages to outperform the old, sometimes significantly. Both are well suited as workstations for graphics, video and encoding in any professional environment.
At any rate, our benchmarks don't touch on some of the other significant advantages of the new systems. Dual processors are, of course, ideal for multitasking, which is an issue rarely covered but which is quite important for workflow. And the new systems offer other architectural advantages, like the PCI architecture, which performance benchmarks like ours can't touch. Nevertheless, our tests should give you a good idea of how the machines will perform for you on everyday creative work. If possible, we will add more benchmark results in the coming weeks, so look for those. In the meantime, if you have further questions about these tests or would like to see benchmarks using other applications, be sure to drop me a line or post a message in one of the forums I host, which are listed below.
Contact the author: Dave Nagel is the producer of Creative Mac and Digital Media Designer; host of several World Wide User Groups, including Synthetik Studio Artist, Adobe Photoshop, Adobe InDesign, Adobe LiveMotion, Creative Mac and Digital Media Designer; and executive producer of the Digital Media Net family of publications. You can reach him at dnagel@digitalmedianet.com.
Prev 1 2 3 4
[an error occurred while processing this directive]
