Lightroom Export and 1:1 Previews - Time & Temp Tests

Discussion in 'MacBook' started by newellj, Apr 18, 2015.

  1. newellj, Apr 18, 2015
    Last edited: Apr 29, 2015

    newellj macrumors 601

    Joined:
    Oct 15, 2014
    Location:
    Boston, MA, US
    #1
    Updated to include time to build 1:1 previews. Operating temps and CPU usage was comparable to previously observed results for the export operation.

    I had some time today and ran a comparison of the rMB with some other computers performing an export with Lightroom 5.7. I used a test library containing 215 20mp raw files aggregating ~4.5 gb. The CPU figures are CPU temps and % utilization shown by iStat, averaged by means of the Mk1 eyeball. The CPU temps and %s were roughly equivalent for both the exports and the 1:1 preview builds.

    TLDR: the rMB did quite well; for what it is, there is a large degree of awesomeness in this little machine.

    rMB (2015) - Core M 1.2 gHz/8 gb
    1:1 preview time: 30:04 minutes
    Export time: 31.15 minutes
    CPU: ~91*C at ~55%
    note on exterior temps measured with IR thermometer (note that these are F, not C):
    palm rest: 86*F speaker grill: 100*F bottom/base: 92-97*F - the rMB never got more than a tad over body temperature, in spite of running maxed out for almost half an hour.

    rMBP 13.3" (late 2013) - i5 2.4 gHz/8 gb
    1:1 preview time: 25:17 minutes
    Export time: 25:15 minutes
    CPU: ~78*C at ~58%

    Dell XPS 13 (2013) - i7 2.0 gHz/8 gb
    1:1 preview time: not tested
    Export time: 25.15 minutes
    CPU: not observed

    Mac Mini (2012) - i7 (quad) 2.4 gHz/8 gb
    1:1 preview time: 16:54 minutes
    Export time: 16:55 minutes
    CPU: not observed

    BYO Z87 - Core i7 4770K @ 3.5 gHz / 32gb
    1:1 preview time: 10:08 minutes
    Export time: 8:20 minutes
    CPU: not observed
     
  2. thedeske, Apr 18, 2015
    Last edited: Apr 18, 2015

    thedeske macrumors 6502a

    Joined:
    Feb 17, 2013
    #2
    Thank You!

    This light machine is something many photo tour travelers will consider and
    Lightroom performance is important. Ver6 will introduce GPU access and speed it up further, so it's looking very interesting. Exporting is a benchmark. Most people Cull daily images and make backups and wait to return home for any editing on a desktop system.

    This and an SSD loaded Delite enclosure from Frimtek would make for a very
    stealthy rig on the road with almost no weight.
     
  3. iRun26.2 macrumors 68000

    Joined:
    Aug 15, 2010
    #3
    Very impressive!

    Thank you for the test data.
     
  4. Meister Suspended

    Meister

    Joined:
    Oct 10, 2013
  5. Eric-PTEK macrumors 6502

    Joined:
    Mar 3, 2009
    #5
    I cannot imagine doing any real work with Lightroom considering its barely usable on a MB Air if your looking at doing any heavy duty stuff.

    We'll go out and shoot(as a hobby) and may come back with 200 shots between the two of us. Trying to do that on Lightroom on a MB Air was way too slow to get work done in a reasonable period of time.

    I picked up a rMB Pro because of it. Performance was one part of it, and the 1 or 1.5 hour of battery life you got out of a MB Air using Lightroom heavily was the other reason.
     
  6. newellj, Apr 18, 2015
    Last edited: Apr 18, 2015

    newellj thread starter macrumors 601

    Joined:
    Oct 15, 2014
    Location:
    Boston, MA, US
    #6
    I try to avoid doing photo editing on laptops for several reasons, but I don't find the rMB noticeably different for editing than my rMBP. Interestingly, one big plus with the ForceTouch trackpad is that very small adjustments are easily made with a slight roll of your fingertip, which doesn't work as well for me with the "old" trackpad. Having said that, I'd rather use a mouse, and I'd far rather use a desktop with a large display than use any laptop of any size.

    Even for people who have no interest in Lightroom, I thought that the export comparison would be interesting because it is a CPU-intensive test that can be done on a basis that allows a consistent comparison.

    Note that even driven up to 91* C for almost half an hour, the fanless rMB was only 20% slower than the base rMBP. You could open that difference up by going up to the top CPU on the rMBP (currently 2.9 gHz), obviously.
     
  7. iRun26.2 macrumors 68000

    Joined:
    Aug 15, 2010
    #7
    Your test results are quite telling: The rMB is no slouch!

    (especially considering that the last few years of improvements from Intel have brought very minor processor speed bumps)
     
  8. Eric-PTEK macrumors 6502

    Joined:
    Mar 3, 2009
    #8
    I should have been a bit more clear. I wasn't even talking editing. A lot of times we are someplace all day or for a few days and we like to review what we just shot. So loading in a couple of hundred pics and waiting for full res previews at 5-6 seconds per pic is a long wait when we want to do a quick review without editing at all.

    That is dealing with 20mp RAW images, its not as bad on the rMBP but the big advantage is battery life which on the light laptops is terrible with LR especially if its building previews.
     
  9. Miltz macrumors 6502

    Joined:
    Sep 6, 2013
    Location:
    New York
    #9
    Great Work... The 1.2GHz holds it's own.... I'm excited about the 1.3GHz model with a higher turbo. It should get close to 25min. Can you take a screen shot of Lightroom and or photoshot and upload it full resolution so I can see the working space for photo editing? Thank You.
     
  10. newellj thread starter macrumors 601

    Joined:
    Oct 15, 2014
    Location:
    Boston, MA, US
    #10
    The delay in getting 100% previews could definitely be painful. I didn't try to test getting the full res previews because it would be hard to time. Actually, I guess I could delete all preview for this catalog then have LR build 100% previews and time that. (I am thinking out loud and haven't had any coffee... ;) ).
     
  11. sdugoten macrumors regular

    Joined:
    Jul 23, 2010
    #11
    wonder how fast the battery drain when you are doing the test?
     
  12. newellj thread starter macrumors 601

    Joined:
    Oct 15, 2014
    Location:
    Boston, MA, US
    #12
    I'm doing this on mains power - one less variable and hard to test impact on the battery.
     
  13. newellj thread starter macrumors 601

    Joined:
    Oct 15, 2014
    Location:
    Boston, MA, US
    #13
    Added times for building 1:1 previews. Interestingly, with the exception of the Z87 desktop, the times were very close to the export times.
     
  14. Geert76 macrumors 65816

    Joined:
    Feb 28, 2014
    Location:
    the Netherlands
    #14
    I am a complete noob reading the OP's results. Is there anyone that can tell this noob from The Netherlands ;) what these results mean particular about performance for me as a future retina Macbook owner?

    Thanks
     
  15. newellj thread starter macrumors 601

    Joined:
    Oct 15, 2014
    Location:
    Boston, MA, US
    #15
    Here's a screen shot. I have the left side panel set for hidden/hover. Because of the vertical height limitations, doing the same to the right side panel doesn't increase the size of the image.

    I'm not certain (either way) about whether the 1.3 will do any better. Both of these operations pegged the CPU at its thermal limits. I posted 91*F but there were short spikes up to 94*F, at which point it would throttle, fall back to 88-89*F, then climb to 91-92*F for a while before repeating that cycle. If the 1.3 gHz throttles similarly, the results might be only slightly different? Not sure.

    [​IMG]

    ----------

    Weeellllll.... ;) A few thoughts still without adequate caffeination...

    1. Assuming you are not directly interested in using the rMB to edit photos with Lightroom, it gives you a comparison between the 1.2 gHz rMB, the 2.4 gHz rMBP and the 2.4 gHz quad Mini for a CPU-intensive operation. The fan-cooled rMBP with a 2.4 gHz i5 CPU did better than the fanless rMB 1.2 gHz Core M, but notice that the difference is only about 20% for an operation that redlined the CPU performing an operation that required roughly half an hour. I think you can extrapolate that an operation that doesn't run as long and/or does not redline the Core M CPU may be for practical purposes be indistinguishable in performance terms if you're comparing the rMB to an rMBP or MBA.

    2. It also gives you an idea about operating temps (CPU and external) with the rMB. It will run up to the low 90s C and surprisingly maintains quite moderate temperatures on the exterior (note that I measured external temps after almost half an hour at ~91*C).
     
  16. Geert76 macrumors 65816

    Joined:
    Feb 28, 2014
    Location:
    the Netherlands
    #16
    thanks for the feedback...I dont use LR, but indeed, it shows how the temps are working in heavy loads.

    thanks from The Noob! :cool:
     
  17. newellj thread starter macrumors 601

    Joined:
    Oct 15, 2014
    Location:
    Boston, MA, US
    #17
    But again, it also shows a worst case comparison between the Core M and a fan-cooled i5, and even under these worst case circumstances the Core M was only 20% off the i5. If you were using Word, Excel or similar programs for normal file operations, you would never see the difference.

    ----------

    Sent you a PM.
     
  18. iRun26.2 macrumors 68000

    Joined:
    Aug 15, 2010
    #18
    (I'd like to see it too)
     
  19. driftless macrumors demi-god

    driftless

    Joined:
    Sep 2, 2011
    Location:
    Chicago-area
    #19
    Hi newellj,

    Thank you for spending the time to run these tests. 200 RAW pics is a good measure of a reasonable day of shooting. Nice puppy! For me, the rMB will be a replacement for an iPad with Photos/Snapseed to cull pics and do a few minor edits. All of my heavy lifting will be done on my iMac.

    David
     
  20. newellj thread starter macrumors 601

    Joined:
    Oct 15, 2014
    Location:
    Boston, MA, US
    #20
    Yup. I can often shoot 200+ photos in a day, but to put the tests above into context, if I am traveling and forced to do some preliminary editing on a laptop, the first thing I'm going to do is cull the keepers from the rest. That can be done with standard previews (much faster rendering).

    Evaluating the results of the first round culling may involve some 1:1 previews to check critical focus, but that's a relatively small percentage of the overall number of photos. That will dispose of a lot of photos.

    The final set that results from those two rounds of culling might be only 10%, possibly but infrequently 20%, of the total, so the times (and time differences) shown above would be proportionately reduced. In other words, the 20% advantage of the rMBP looks a lot less significant if we're talking about 20% of an operation that takes 2-5 minutes rather than 25 minutes.
     
  21. hypno macrumors regular

    Joined:
    Oct 30, 2012
    #21
    Exactly. This laptop fits in my bag easily with my lenses too. Nearly as good as the rMBP but easier for portability. Thanks for doing this test.
     
  22. newellj, Apr 19, 2015
    Last edited: Apr 19, 2015

    newellj thread starter macrumors 601

    Joined:
    Oct 15, 2014
    Location:
    Boston, MA, US
    #22
    It makes my hair stand on end a bit, but the rMB could be stashed in the back of a camera bag, the way I sometimes do with my iPad. I guess it shouldn't bother me, because the *camera* gear in the bag is worth many times what the iPad or even the rMB are worth!
     
  23. bcaslis macrumors 68020

    Joined:
    Mar 11, 2008
    #23
    Very interesting tests. Lightroom is one of my primary apps for my 13" rMBP. So would you get rid of the rMBP in favor of the rMB? I saw one for the first time on Friday and it was love at first sight. If the Apple store had any I would have bought one but I'm still concerned about replacing the rMBP.
     
  24. newellj thread starter macrumors 601

    Joined:
    Oct 15, 2014
    Location:
    Boston, MA, US
    #24
    I had not planned to.

    Logically, I should.

    Emotionally, it sounds funny, but I'm extremely bonded to my rMBP, so I'm undecided. But I did put an ad on the local Craigslist... :confused:

    If you are seriously interested, perhaps, when they become more readily available, buy one and give it a try for 12 or 13 days? There is a lot that is new - the keyboard or trackpad, for example, might not agree with you.
     
  25. bcaslis macrumors 68020

    Joined:
    Mar 11, 2008
    #25
    I tried the keyboard and trackpad briefly in the store and they seemed fine.

    This has got to the weirdest laptop launch in Apple history. I've never seen one like this where none of the stores has any and have no clue when they are getting any. And the Apple website lead time is 4 - 6 weeks! And the Best Buy website has it in stock? Simply bizarre.

    I'm pretty bonded with my rMBP also. But the tasks I use it for are general tasks and Lightroom. Losing the SD card slot would be a pain but it could be worked around. The rMB in space gray was just stunning though. I can't remember an Apple laptop that looked so good.
     

Share This Page