DIY Photoshop Torture Test

DIY Photoshop Torture Test

Summary: Back in February I posted some Photoshop benchmarks pitting the MacBook Pro 2.0GHz against the PowerBook G4 1.5GHz. Now it's time for you to test your machine.

TOPICS: Hardware

Photoshop-torture-test.jpgBack in February I posted some Photoshop benchmarks pitting the MacBook Pro 2.0GHz against the PowerBook G4 1.5GHz. Not surprisingly, the old-school PowerBook G4 won the pixel-pushing death-match in a unanimous decision due to the fact the Photoshop is written to natively address the PowerPC chip whereas the MacBook Pro is sorely hobbled by the constraints of Rosetta emulation.

If you've ever wondered how your Apple iron stacks up to the Photoshop benchmarks I posted, I am making available the Photoshop action and test files that I used in the test. Feel free to download them and run them on your own machine. After you run the Photoshop action, please post your results in the TalkBack below!

Click through for the downloadable files...

Download the Photoshop Torture Action and Test files (2.27 MB, .zip)

In the download you'll find two test images and a Photoshop action set that will torture your new MBP or any computer for that matter (it even runs on Windows).

Put the Photoshop action "PS CS Test Actions.atn" in:
    Applications > Photoshop CS > Presets > Photoshop Actions

TEST 1 creates a Kaleidoscope from and existing image. The action creates a Kaleidoscope image by copying layers and rotating them. Will wrote this action because it has scaling, rotation, blurring etc.
1) Open the Action palette (Window > Action)
2) Open the Action palette's flyout menu (the triangle in the upper right corner of palette)
3) Select "PS CS Test Actions" which will put the actions in your actions palettes.
4) Open the image "Kaleidoscope Test" and go to the Paths palette. Click on the "Work Path" action to make the path active.
5) In the Action palette click on the action "Make Kaleidoscope"
3) Click the small "play" button in the bottom of the palette, and START your stopwatch.
4) STOP when the completed image (pictured) appears on screen.

Will's TiBook 1GHz (1GB RAM, 100GB HDD) completed the action in 01:08 (one minute, eight seconds) in Photoshop CS2.
My MacBook Pro 2GHz (2GB RAM, 120GB HDD) completed the action in 01:30 (one minute,30 seconds) in Photoshop CS2.

TEST 2 takes a 10MB image and scales it losslessly to about 550MB. It does so in 110 percent steps with sharpening in between.  It creates a 550MB file and it tortures the processor, RAM and HDD. Will's TiBook 1GHz completed the action 06:45 (minutes:seconds).

1) Open the image called "Resize Test" and run the action called Resize CS
2) The final file is actually scaled 540 percent plus with little or no loss of quality
3) There is a "Purge" steps in between each scaling step otherwise you would be using well into the 4TB of RAM and scratch space.
4) You will have to click on a "continue" button twice, so pay attention.
5) STOP time when the final dialog displays a stop button.

Will's TiBook 1GHz (1GB RAM, 100GB HDD) completed the action in 06:45 (six minutes, 45 seconds) in Photoshop CS2.
My MacBook Pro 2GHz (2GB RAM, 120GB HDD) completed the action in 04:40 (four minutes, 40 seconds) in Photoshop CS2.

Note: Both PowerBook and the MacBook Pro were running with the "Better Performance" option selected in the Energy Saver control panel. I will be running the new benchmark files on a PowerBook G4 1.5GHz (aluminum) over the weekend and will update the story with the results. 

Download the test files and post your results in the TalkBack section below.
(Thanks to Will Hammond for the test files!)

Topic: Hardware

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Hmmmm... A Dual G5 doesn't fare too well...

    Time: 4:30 with a Dual 2.5 G5 and 1.5 GB RAM... considering this
    machine has a 7200 rpm drive and dual chips, i thought it would
    fare well. Guess I was wrong...
    It looks like the MacBook more than holds its own. Now i'm
    thinking twice about replacing my 550mhz G4 tibook...
  • My screamin' machine got . .

    . . .3:52 (3 minutes and 52 seconds

    Powermac G4 / 400. 832MB RAM. 20 and 30GB 7200 RPM
    ATA-66 drives. Maybe it's time for an upgrade??
  • G5's Do Fare Well

    System: Dual G5 2.0 GHz, 1 GB ram, 250 GB HDD, GeForce 6600
    256MB Video Card

    Results: 17.9 second average on Kaleidoscope (15.2 second fastest
    time); 2:06 minute:second average time for re-size test.

    • Confirmed on dual G5

      dual G5 2.0 with similar specs (except half the VRAM & twice the
      regular RAM); [b]EXACT[/b] same result.
  • MacBook Pro 2.16 GHZ, 2 GB RAM, 7200 HD

    Test 1: 41 secs
    Test 2: 3:21 secs
  • G5 Quad

    Kaleidoscope took 10.0 s on a Quad G5 with 3 Gbyte of ECC
    memory; Image Resize took 1 m 58 s. 10.4.5, Adobe Photoshop CS.

    Yes, the Quads are optimized for image processing, among other
  • Yikes!

    Kaleidoscope : 15 sec
    Resize 1:23

    I considering trading my G5 (Dual 2,3 - 2 GB RAM) for a MacBook
    Pro. As a multimediadeveloper I use Photoshop a lot, mostly at
    screen size but with a lot of layers.

    This upgrade is gonna hurt until Adobe comes with universals!!
    • CS2

      PS, how come my dual 2,3 beats JGurman's Quad at the resize test,
      maybe because I use CS2?
  • G5 Quad, 4.5 Gb RAM

    Test 1: 7 secs
    Test 2: 34 secs

    Why is this so much faster than the other G5 Quad?
  • The story, is true! D, d, d, drag(net)

    My brand spankin' new MacBook (2`Ghz, 2GB RAM, CS2) just
    scored the following:

    Kaleidoscope 1:15
    Resize: 5:03

    PS, Installing CS2 was also horribly slow, over 20 minutes!
    • times, G5 2.3 dual, PS 7, 2.5 gb ram

      O'Grady's actions work with older PS...

      test 1: 15 seconds
      test 2: 1:50

      I wouldn't trade in any G5 for graphics work until Adobe makes
      their sw universal and faster on the Mactels.
  • Suggestion for testing

    When I wrote these actions I didn't give much thought for VM files that may already be on you hard drive. If your maching got a particularly low score, as some G5 systems did, do a cold restart (shutdown completely and then start up again). Launch only Photoshop and try the test again. I had my airport disabled as well. I am curious to see how much other apps/processor functions affect Rosetta performance.

  • Dual Xeon Test Results

    Dual 3.0 Ghz Xeon on Supermicro X5DAE Motherboard
    2GB Registered ECC Ram
    256 MB GeForce 6800 GT

    Kaleidoscope 1:17
    Resisze Test: 2:53
  • Times on my G-5 Duo

    I heard about your test on Scott Kelby's blog. Ran it on my 2.5 Dual
    G-5 tower (3 gigs RAM)

    Test 1 took 31 seconds
    Test 2 took 48 seconds

    Ran it on my G-4 iBook. Still waiting for results. Can't wait to get
    my MacBook Pro. It is on order. Thanks for the test.
  • G5 1.8 SP 1.5gig Ram

    I was surprised by the times posted here. I only have a Single
    Processor 1.8 G5 running 1.5gigs of Ram and my times were
    pretty darn good compared to what is listed here

    Kaleidescope: 24 sec
    Upsize: 1:51

    I thought that was great, especially since I thought I noticed the
    computer starting to slow down a bit, must have just been my

    Thanks for test!
  • Dual Opteron

    The first test takes 7.5 seconds and the second test takes 27 seconds.
    My system is a dual opteron 854 (2.9 GHz single core) with 4 meg ECC ram.
    On this system, everything is memory bound, there is no disk IO which most lkikely accounts for the very good scores.
  • iMac G4 - 800mhz - 767Mb

    Several tests, clean system, each time double verified
    Test #1 - 1:11
    Test #2 - 10:24

    Hopefully someone will do a score on the latest G5 & Intel iMacs -
    I'd like to buy a new iMac but would love to see the comparision.
  • PC Benchmark

    TEST 1: 0:10.052
    TEST 2: 0:41.052

    Athlon X2 64 4400+ (dual core 2.2GHz)
    2GB PC 3200 DDR RAM
    1 x 300GB SATA-3
    2 x 300GB SATA-II
    1 x 2 Mirror Raid ATA-133 300GB
    1 x 120 GB ATA-133
  • imac g5 rev a 1.8ghz, 2 gig ram cs1 vs cs2

    Interesting that on my iMac G5 that the Kaleidoscope test gets the
    same results on both CS1 and CS2 : 25 seconds

    but for the resize test:

    CS1 : 1:10
    CS2: 4:20 (!?)
  • Quad G5 with 4GB RAM

    Kaleidiscope: 18 Sec
    Resize: 25 secs (did not get the "purge" prompts mentioned