2016 MBP Real World Lightroom Test vs '12rMBP and '15 5KiMac

[arbitrage]

So my main power usage for my MBP is Lightroom...I have used a 2012 rMBP while travelling and my 5KiMac at home for most of the heavy lifting. The 2016 MBP is to replace the 2012 for on the road if it shows enough improvement.

So today I ran some tests of my normal workflow:

The players:
1) 2012 15" rMBP 2.7GHz i7, 16GB, 512GB SSD and 750M
2) 2015 27" 5KiMac 4.0GHz i7, 24GB, 512GB SSD and Radeon R9 M395X 4GB
3) 2016 15" rMBP 2.9GHz i7, 16GB, 1TB SSD and 460

My test was to import 104 Canon 5DSR Raw files and have LR generate 1:1 Previews. Import was from a CF card. I ran the test twice on each machine. One test I imported the files to the internal SSD on each machine. The second time I imported the files to an external Seagate Backup Plus 1TB portable drive for the two laptops and my Thunderbolt RAID for the iMac.

The results:
Importing to SSD:
1) 2016: 10:07.24
2) iMac: 10:36.18
3) 2012: 12:34.19

Importing to the external drives:
1) 2016: 10:54.11
2) iMac: 11:08.23
3) 2012: 14:09.93

So basically this new machine is getting me performance on par with the best iMac you can buy currently and actually did beat it by a few seconds in each test. And saves me a lot of time over my 2012. This was with 104 files. I typically ingest 1000-2000 at a time from a days shoot (bird photography mostly).

I'm thoroughly impressed with this new machine....I didn't think it would come close to the iMac but it beat it!!

Hope this helps some other photographers out there...

 

[deadworlds]

are there more computationally intensive tests you could run? Maybe something like applying a filter to all 104 images or something along those lines? Im not a professional photographer by any means, but I feel that your measurement of importing photos only tells us how fast the I/O transfer is, and does not really tell us about the computation power.

 

[arbitrage]

are there more computationally intensive tests you could run? Maybe something like applying a filter to all 104 images or something along those lines? Im not a professional photographer by any means, but I feel that your measurement of importing photos only tells us how fast the I/O transfer is, and does not really tell us about the computation power.

It is generating 1:1 previews on 50MP 5DSR files...that is the computation power....I have no issue even on my 12"rMB using sliders and doing adjustments in LR. Export would be another test but I export in really small numbers.

The actual ingest times to write to the drives was around the 1min mark for the 2016 and iMac. About 2mins for the 2012.

One other note is that the 2012's fan sounded like a jet engine....the 2016 was very quiet.

 

[Bryan Bowler]

arbitrage,

Wow, I can't thank you enough for measuring and posting these results! I'm always interested in real-world stats from fellow photographers, but yours really intrigued me because I have nearly the exact same set-up as you in terms of mid-2012 MBP, iMac, and I have the exact same late-2016 MBP on order. Heck, I even have a Thunderbolt RAID system and a 5DsR.

About the only difference in our workflow is I use PhotoMechanic and then after culling and metadata, I bring the images into Capture One Pro 9. BTW, I don't know if you've given any consideration to PhotoMechanic, but it is absolutely superb for culling and metadata. I shunned it for years because I didn't want to pay $150 for tasks that my existing DAM could do, but now that I'm using it, I can honestly say that I would happily pay double for it due to the time I've saved. And especially when it comes to birds and sports. You know as well as I do how taxing that can be to try and quickly import and cull through 1000 images. I can get it done so quickly in PhotoMechanic it's unreal.

At any rate, back to the original topic at hand and that's your real-world observations on the new MBP. I love it and thanks again for sharing! My new MBP is not due until the end of December, but I will run some side-by-side tests too and see what I come up with. Between now and then, I look forward to hearing anything else you might have to report.

Cheers...

 

[arbitrage]

It is generating 1:1 previews on 50MP 5DSR files...that is the computation power....I have no issue even on my 12"rMP using sliders and doing adjustments in LR. Export would be another test but I export in really small numbers.

arbitrage,

Wow, I can't thank you enough for measuring and posting these results! I'm always interested in real-world stats from fellow photographers, but yours really intrigued me because I have nearly the exact same set-up as you in terms of mid-2012 MBP, iMac, and I have the exact same late-2016 MBP on order. Heck, I even have a Thunderbolt RAID system and a 5DsR.

About the only difference in our workflow is I use PhotoMechanic and then after culling and metadata, I bring the images into Capture One Pro 9. BTW, I don't know if you've given any consideration to PhotoMechanic, but it is absolutely superb for culling and metadata. I shunned it for years because I didn't want to pay $150 for tasks that my existing DAM could do, but now that I'm using it, I can honestly say that I would happily pay double for it due to the time I've saved. And especially when it comes to birds and sports. You know as well as I do how taxing that can be to try and quickly import and cull through 1000 images. I can get it done so quickly in PhotoMechanic it's unreal.

At any rate, back to the original topic at hand and that's your real-world observations on the new MBP. I love it and thanks again for sharing! My new MBP is not due until the end of December, but I will run some side-by-side tests too and see what I come up with. Between now and then, I look forward to hearing anything else you might have to report.

Cheers...

Thanks for the thanks Bryan.

I have tried the trial of PhotoMechanic. I brought it along on my slow 2015 12"rMB on a recent vacation where photography wasn't my main goal. It works very well. I mostly wanted it to do the initial culling of rejects and then import to LR allowing LR to have a lot less 1:1 previews to build. My normal workflow is import to LR and build 1:1 previews so I can go through them at 1:1 checking sharpness as I reject them. PM should save me time with this but I just haven't made the effort to make it a permanent thing in my workflow.

I think you will really like the new laptop. I am still surprised it matched the iMac building 1:1 previews. I had assumed that would be an all CPU job and the iMac can run faster and would finish quicker. But what I really wanted was a noticeable improvement in speed over the 2012 and I got that. 2mins on a 100 files will be 20mins on 1000 and 40mins on 2000 so that is some real time savings. The fans were barely noticeable on the 2016 which is also a big improvement. Battery life looks to be much better but I need a few more days to see how it is.

I will try to think of some more tests to run and report back if I have any more results or other thoughts.

Cheers.

 

[Macintosh IIcx]

So my main power usage for my MBP is Lightroom...I have used a 2012 rMBP while travelling and my 5KiMac at home for most of the heavy lifting. The 2016 MBP is to replace the 2012 for on the road if it shows enough improvement.

So today I ran some tests of my normal workflow:

The players:
1) 2012 15" rMBP 2.7GHz i7, 16GB, 512GB SSD and 750M
2) 2015 27" 5KiMac 4.0GHz i7, 24GB, 512GB SSD and Radeon R9 M395X 4GB
3) 2016 15" rMBP 2.9GHz i7, 16GB, 1TB SSD and 460

My test was to import 104 Canon 5DSR Raw files and have LR generate 1:1 Previews. Import was from a CF card. I ran the test twice on each machine. One test I imported the files to the internal SSD on each machine. The second time I imported the files to an external Seagate Backup Plus 1TB portable drive for the two laptops and my Thunderbolt RAID for the iMac.

The results:
Importing to SSD:
1) 2016: 10:07.24
2) iMac: 10:36.18
3) 2012: 12:34.19

Importing to the external drives:
1) 2016: 10:54.11
2) iMac: 11:08.23
3) 2012: 14:09.93

So basically this new machine is getting me performance on par with the best iMac you can buy currently and actually did beat it by a few seconds in each test. And saves me a lot of time over my 2012. This was with 104 files. I typically ingest 1000-2000 at a time from a days shoot (bird photography mostly).

I'm thoroughly impressed with this new machine....I didn't think it would come close to the iMac but it beat it!!

Hope this helps some other photographers out there...

Thanks a ton for doing this, arbitrage. Much appreciated!

I'm going to use Lightroom on my Sony A7r2 raw files, so your tests are really useful for my needs. Now you got me really excited about getting mine. I needed that with the long waiting we are experiencing in Europe. Anybody has one yet.

One other note is that the 2012's fan sounded like a jet engine....the 2016 was very quiet.

Excellent. This is really interesting and important information too.
[doublepost=1479651003][/doublepost]

I will try to think of some more tests to run and report back if I have any more results or other thoughts.

Cheers.

arbitrage, is the dGPU enabled by default on the new MBP/Lightroom combo? And does it make any difference when it comes to doing work in the Develop module and zooming and out?

 

[groundcontrol]

My thanks as well @arbitrage! I'm a photographer who shoots raw and am so ready to leap from my USB 2.0 2011 MBP (upgraded with an ssd but still bottlenecked by USB 2.0) to even just basic USB 3.0, nevermind all the additional improvements with the 2016 model. Your findings are very encouraging.

I also appreciate the discussion about Photo Mechanic. I bought it last year on the strong recommendation of a few fellow photographers but, out of ingrained habits, haven't yet incorporated it into my workflow. I need to give it a proper try. Voices in this thread certainly encourage me to do so.

 

[Bryan Bowler]

In regards to PhotoMechanic, I put it off for a couple of months after buying it. But then I finally dived in. Once you know what you're doing with it and you get your own workflow ironed out with it, it is a massive timesaver.

The user interface may not seem as polished, but it is fast once you get it sorted out.

 

[groundcontrol]

Thanks @Bryan Bowler. I'll give myself a push to try it after one of my upcoming shoots. So many can't be wrong!

 

[arbitrage]

arbitrage, is the dGPU enabled by default on the new MBP/Lightroom combo? And does it make any difference when it comes to doing work in the Develop module and zooming and out?

I believe the dGPU is enabled when LR is open. I have LR set to use the 460 and I have the gFxCardStatus program and it won't let me run the integrated intel chip when LR is open. Even if I tell LR not to use the 460 it still won't let me use the Intel 530 when LR is open.

I will need to do some more developing before coming to a conclusion on if it helps at all.