iGary said:
I added a few things like brightness/contrast - levels adjustment and highlight/shadow tool. But i didn't actually take anything out...
eerie
I wish i had done more testing with my other drive (any testing*)
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
'Real World' Photoshop Benchmark - Speed Test
- Thread starter BakedBeans
- Start date
- Sort by reaction score
I added a few things like brightness/contrast - levels adjustment and highlight/shadow tool. But i didn't actually take anything out...
eerie
I wish i had done more testing with my other drive (any testing*)
CMYK: 115.0%
LAB: 122.0%
RGB: 107.8%
Invert: 98.8%
Gaussian: 198.0%
Guess we know which actions are multi-aware, eh?
LAB: 122.0%
RGB: 107.8%
Invert: 98.8%
Gaussian: 198.0%
Guess we know which actions are multi-aware, eh?
iGary said:
I think i know why your action has not gone slower. you are not using the updated action
this is the direct link
http://www.liverpoolvideos.com/downloads/speedtest.zip????
I think it must have been in your cache or something - it doesnt have invert in it anymore
the action (when in the folder) should be called retouch artists.atn not retouch artists action.atn
EDIT - also. I knew that 2 processors where supported in some and not in some others (that was planned) i was meaning about the quad (i guess if 2 is supported then 4 will be?)
This bench is very disk dependent due to the history buildup. It filled up 3GB of disk space on my PowerBook in less than half the run-through, thus I can't give you a result but I can give you the warning that the vast majority of notebooks will score hugely lower than their desktop counterparts thanks to their typically faster drives and also that if your scratch disk is nearly full, the bench will be much slower, as HFS-based filesystems become much slower as they get fuller (much moreso than the natural decline in disk speed at the end of the drive).
From what I observed, you'd probably get much better results if you set your scratch volume to be an empty (or nearly empty) external firewire disk as my menumeter was rarely reading over 60-70% CPU usage throughout much of the test.
From what I observed, you'd probably get much better results if you set your scratch volume to be an empty (or nearly empty) external firewire disk as my menumeter was rarely reading over 60-70% CPU usage throughout much of the test.
that in itself doesnt make it an unfair test.
it just means that the faster the drive the better - just like in the real world. and a separate scratch disk will be better - just like in the real world.
I think that what really counts is that the tasks are often used in the creative industry. what effect they have on disks or which is is cpu intensive isnt really important - no point in optimizing it for anything - it need to just be natural.
it is demanding and if your scratch disk doesnt have a lot of space then it will hurt... just like in the real world.
it just means that the faster the drive the better - just like in the real world. and a separate scratch disk will be better - just like in the real world.
I think that what really counts is that the tasks are often used in the creative industry. what effect they have on disks or which is is cpu intensive isnt really important - no point in optimizing it for anything - it need to just be natural.
it is demanding and if your scratch disk doesnt have a lot of space then it will hurt... just like in the real world.
4God said:
Yeah, it is just as noisy as the maxtor - you can hear it read and write. not worse than the maxtor though.
never heard the disk in any other computer though.
4God said:
Mine ran faster the second and third times as well: Cache.
I wish Apple would go back to Seagates - my 2.7 has one Maxtor (came with) and one Seagate - which you cannot hear if you try.
My DOA Quad that came last week had a Western Fracking Digital.
If the replacement has that, I'm taking it out immediately.
4God said:
Clear the cache manually and restart - open nothing else but Photoshop when everythign is booted - that would be interesting to see.
BakedBeans said:
Hmm... I don't even find myself running out of space whilst working with multiple 12mpixel-resolution images from my camera with 2GB of RAM, so it seems a little bit unfair to me.
Besides, in a nearly full-disk situation, it gives a huge advantage to Windows machines (NTFS isn't nearly as inefficient with nearly full disks). I bet my MBP running WinXP or my MSI Turion laptop would flatten my 15" PowerBook in these benchmarks when in reality, the PowerBook is the best machine to use.
Azurael said:
Hence the reason why Photoshop pros have empty scratch disks and dont use a laptop packed with data (and runs the OS)
this action shows the downside of doing what you are doing.
BakedBeans said:
Well unfortunately, my LaCie Big Disk isn't very portable. Thus, I don't use it as a scratch disk. If this is a real world benchmark, it should probably apply to real world situations. I consider myself a Photoshop pro, and while I don't have any qualifications to prove that, I'm quite certain my clients (including several UK music magazines) would agree. If I, as a semi-professional photographer am not using the tools of the trade in a typical fashion, I do apologise.
Now, does anybody know of a real, real world benchmark for those of us who don't have a pocket RAID array, or shall I just record some typical actions myself?
Azurael said:
With an attitude like yours i am surprised you have any clients at all.
you don't need a raid array - you just need more than a completely full hardisk. Strange how ibook users have managed this test without complete gridlock.
Mods - please can you tidy this thread up a little it is completely destructive to the thread. Maybe he would like to consult the amount of photoshop users and photographers that i have over the last two weeks and also contact the NAPP and create something low end (I look forward to high end users coming in with a score of 1.6 seconds and complaining its not real to them).
If indeed you do work for uk magazines then you will have to use certain aspects of this thread. please accept my apologies for not catering to your exact workflow. I just hope you realise that people have bigger things to deal with than a few fuji point and shoot images - this test can't alienate high end users just to cater for people with no space on the HD
Photoshop is a PRO application.
I'm not complaining about my inability to run this test. What I am complaining about is the fact that I believe that the results on laptops are likely to bare little resemblance to real world performance, so to suggest that this test has any relevance to 'real world' performance is a little misleading.
I'm sorry you find it hard to accept that the vast majority of photographers do not find the need to convert an image at 300% of it's original size to three different colour formats in sequence and numerous other unlikelihoods. In reality, one would only apply changes to the image in such a way as to lose as little data as possible from the image, thus the most direct path to acheive a task is the preferred one. Unless working with multiple images of this size simultaneously, one should rarely even run into scratch space with 2GB+ of RAM in reality. What I am saying, in essence, is that this benchmark bears little, if any relevance to most people's real world use of the application.
It's entirely possible that I am just a little irate over the lack of any useful Mac benchmarks (yours is better than XBench, I suppose) but questioning my attitude certainly says something about yours.
I'm sorry you find it hard to accept that the vast majority of photographers do not find the need to convert an image at 300% of it's original size to three different colour formats in sequence and numerous other unlikelihoods. In reality, one would only apply changes to the image in such a way as to lose as little data as possible from the image, thus the most direct path to acheive a task is the preferred one. Unless working with multiple images of this size simultaneously, one should rarely even run into scratch space with 2GB+ of RAM in reality. What I am saying, in essence, is that this benchmark bears little, if any relevance to most people's real world use of the application.
It's entirely possible that I am just a little irate over the lack of any useful Mac benchmarks (yours is better than XBench, I suppose) but questioning my attitude certainly says something about yours.
You are missing the point. This test is purely testing the computer's power; nothing more. It just so happens that this test uses Photoshop, other than that, Photoshop has nothing to do with this test.Azurael said:
And I agree, the Mods should delete all posts that do not have test results in them, or are otherwise vitally important to the thread.
EricNau said:
No, primarily it tests the speed of the available space on the scratch disks on machines with less than about 4GB of RAM. That's the only reason G5s are turning in such good results. As an example, my Athlon64 desktop with 4GB of RAM and two 7200RPM HDDs in RAID 0 runs through this suite twice as fast as my PowerBook, but I know which feels faster to me in my real world work. Maybe I am the only one who wouldn't dream of using anything like the sequence of actions in this script in reality, but if so, I don't want to see how much resolution you lose from your photos when editing them.
Why are you so focused on the photos? The end photo after running the test is not important.Azurael said:
Then surely it's not a real world test? That's my point. A real world test would invole making modifications to the image which a real photographer might make to produce a real result which a real photographer might. Anyway, my opinion is clearly not wanted here, and admittedly, I did hijack the thread just a little so I'll leave before I'm pushed out.EricNau said:
lol, your point about not enlarging photos 300% is crazy.
Very very funny and makes your "argument" even more weak.
People do print photos more than 6x8 you know. That part of the action was recommended by a fine art photographer and is often used when printing large images.
this test is not misleading.
Very very funny and makes your "argument" even more weak.
People do print photos more than 6x8 you know. That part of the action was recommended by a fine art photographer and is often used when printing large images.
this test is not misleading.