Pages: 1 2 [3] 4   Go Down

Author Topic: Lightroom performance improvements...  (Read 18717 times)

hogloff

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1187
Re: Lightroom performance improvements...
« Reply #40 on: July 26, 2017, 08:35:43 am »

Also in the real world, a vote for one improvement need not automatically translate into a vote against another aspect. While I don't think any business has infinite resources, I also think a multinational corporation that measures their revenues ... and profits in factors of billions of dollars have the means to address most concerns all users have in the area of performance improvements.

After all, if you have time to sleep while importing your images, I doubt you'd be impacted negatively, regardless of which area Adobe should focus their development dollars.

Butch...don't get so excited. I never said anything about not wanting improvements during the import phase...it's just if we are voting I choose to focus on the develop module as I am in that module a lot more than import.

As far as Adobe being a big company...yes they are, but they also have an awful lot of products to manage which require an awful lot of resources. And typically product groups are smallish teams and adding a bunch of resouces to a team usually leads to a disaster.
Logged

rdonson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3263
Re: Lightroom performance improvements...
« Reply #41 on: July 26, 2017, 08:35:53 am »

The place for "voting" was the Adobe Lr survey. It does little good to "vote" here. 

Why don't we just unearth and discuss ALL performance Lr improvements needed here in this thread.

Logged
Regards,
Ron

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Lightroom performance improvements...
« Reply #42 on: July 26, 2017, 09:07:59 am »

The place for "voting" was the Adobe Lr survey. It does little good to "vote" here. 

Why don't we just unearth and discuss ALL performance Lr improvements needed here in this thread.

OK Ron, you started this thread and maybe it would be good if you could clarify what in your mind were the "Terms of Reference" for this discussion. By "performance", did you mean to limit the discussion to matters of process efficiency (for example speed), or did you also mean the effectiveness on outcomes of existing features and the desirability of new features to improve application effectiveness?

Of course Adobe has its own forums for these discussions, but having a discussion here is useful for all those who don't partake of those resources and thereby to broaden the audience for those interested in Lr's future development.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

rdonson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3263
Re: Lightroom performance improvements...
« Reply #43 on: July 26, 2017, 02:50:38 pm »

Fair enough, Mark.  Thanks for asking.

Background:  I recently upgraded to a 27" iMac - 4.2 GHz quad core i7, 512GB SSD, 40GB RAM and Radeon Pro 580 8GB VRAM.  It replaced a 6 year old 21.5" iMac with quad core i5, 16GB RAM, 512GB HD and a modest graphics card.  I was surprised that many things in Lightroom weren't all that quicker on the new machine.  Then Hogarty sent out his missive that Adobe wanted to know our concerns about Lr performance and a survey was available.  I provided my feedback in the survey.  I wondered what others were thinking on this topic.

My concerns were limited to speed of existing features.  I really didn't think about workflows not prescribed by the Lr features.  I was simply interested in what other people were experiencing with regards to performance.  I thought that sharing experiences might be useful and we might gain insight and perhaps some tuning tricks in configuration or preferences.

Responses detailing alternate workflows that alleviate some performance problems were/are interesting as long as we all understand the underlying performance issue being worked around.

I'm open to expanding things to include effectiveness of outcomes and probably to new features to improve application effectiveness as long as it's reasonable and not fantasy stuff or digs on Lr because product XXX does something better.  The later seems to lead down a rat hole.

If someone wants to indicate that they're experiencing the same performance concerns that's valuable and confirmation of the problem.  I don't find value in "voting" on performance problems to be addressed.  Perhaps that's bias or misunderstanding on my part though but I don't think it's of much use for people in this forum to try to prioritize Adobe's tasks by "voting".

This is where I'm at and I'm open to all thoughts on this. 


   




Logged
Regards,
Ron

ButchM

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 749
Re: Lightroom performance improvements...
« Reply #44 on: July 26, 2017, 03:04:43 pm »

... I don't think it's of much use for people in this forum to try to prioritize Adobe's tasks by "voting".

I agree ... a cursory 'vote' for improvements to Lr performance in a particular aspect should not be equated as  a vote against performance improvements elsewhere in the app. We aren't electing a politician, we are pointing out the problem areas and performance bottlenecks that users experience. Each user contributes equally for their licensing fees, so one 'vote; should never be of more value than another.
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Lightroom performance improvements...
« Reply #45 on: July 26, 2017, 03:08:02 pm »

OK Ron thanks - that's helpful context for orienting the discussion. And I agree, none of this should be an issue of "voting" - it's useful for sharing experience and suggesting options. Some of these ideas should likely end-up in the Adobe Lr Forum from where there would be greater probability of uptake.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

adias

  • Guest
Re: Lightroom performance improvements...
« Reply #46 on: July 28, 2017, 02:25:19 am »

...  I recently upgraded to a 27" iMac - 4.2 GHz quad core i7, 512GB SSD, 40GB RAM and Radeon Pro 580 8GB VRAM.  It replaced a 6 year old 21.5" iMac with quad core i5, 16GB RAM, 512GB HD and a modest graphics card.  I was surprised that many things in Lightroom weren't all that quicker on the new machine.

Because Lr/PS are not CPU limited (and not RAM limited either if you already had 16GB). More 'horsepower' does not equate here to higher performance... Higher performance will come from parallelism being taken advantage of in the core code.
Logged

rdonson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3263
Re: Lightroom performance improvements...
« Reply #47 on: July 28, 2017, 10:38:47 am »

Because Lr/PS are not CPU limited (and not RAM limited either if you already had 16GB). More 'horsepower' does not equate here to higher performance... Higher performance will come from parallelism being taken advantage of in the core code.

Please help me understand.  So doubling clock speed and hyper threading, faster RAM and a dramatic improvement in GPU speed and VRAM are of little value in Lr/PS?
Logged
Regards,
Ron

adias

  • Guest
Re: Lightroom performance improvements...
« Reply #48 on: July 28, 2017, 11:54:47 am »

I restate... Lr/PS performance has not been CPU limited by recent (10 years or newer) hardware. Sure, if you use performance metrics, specific processing is faster with faster CPUS, but it does not make any perceived difference to the end user, especially in the editing stage. Where you will notice an improvement is in batch processing - exporting a bunch of images, for example.

I bet that Adobe is working on core performance improvements, such as using multiple CPU cores, and parallelism. That will make the app snappier.
Logged

rdonson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3263
Re: Lightroom performance improvements...
« Reply #49 on: July 28, 2017, 12:11:12 pm »

Thanks for your response. 

There are many things where I think CPU performance should show dramatic improvements such as building previews but I don't see a marked difference.

Adobe has said for a long time that PS and Lr supports multiple cores.  Perhaps their use of the cores and hyperthreading isn't where it should be.   

This article caught my attention a while back with regards to performance. 

https://www.pugetsystems.com/labs/articles/Adobe-Lightroom-CC-6-Multi-Core-Performance-649/
Logged
Regards,
Ron

rdonson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3263
Re: Lightroom performance improvements...
« Reply #51 on: July 28, 2017, 12:14:50 pm »

Thanks, John, for the more recent test results!
Logged
Regards,
Ron

Ethan Hansen

  • Full Member
  • ***
  • Offline Offline
  • Posts: 148
    • Dry Creek Photo
Re: Lightroom performance improvements...
« Reply #52 on: July 28, 2017, 01:56:38 pm »

John - Thanks from me as well. One hopes Adobe will start better using the multiple cores coming to workstations from Intel and AMD. For grins I tried running LR on one of our Broadwell-EP servers used for ICC profile calculations. Despite having eight times as many cores and sixteen times more memory than an i7 Kaby Lake system, the server crawled through both imports and conversions. A base clock only half the desktop's cratered performance.

Your comparison test shows Intel systems handily besting AMD's latest on Windows. I wonder how much of this is a result of Intel being more responsive at pushing compiler optimizations into Visual Studio and how much is purely processor.

Alan Goldhammer

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4344
    • A Goldhammer Photography
Re: Lightroom performance improvements...
« Reply #53 on: July 28, 2017, 06:01:03 pm »

https://www.pugetsystems.com/labs/articles/Lightroom-CC-2015-10-1-CPU-Comparison-Skylake-X-Kaby-Lake-X-Broadwell-E-Kaby-Lake-Ryzen-7-973/ is more recent
LOL, the LR machine that they have configured is almost identical to the one I built for my new workstation 18 months ago for less than 1/2 of what this one is priced at.
Logged

Damon Lynch

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 330
    • http://www.damonlynch.net
Re: Lightroom performance improvements...
« Reply #54 on: July 29, 2017, 03:37:12 pm »

LOL, the LR machine that they have configured is almost identical to the one I built for my new workstation 18 months ago for less than 1/2 of what this one is priced at.

RAM has doubled in price in less than 12 months. Plus those guys have a business to run. Some people find their services add value they truly need, such as the system validation, after sales service, etc. I'm not one of them, because like you I prefer to build my own machine and save the $$.  Meanwhile, I'm happy to read their benchmark results  :D
Logged

rdonson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3263
Re: Lightroom performance improvements...
« Reply #55 on: July 31, 2017, 10:57:45 pm »

https://www.pugetsystems.com/labs/articles/Lightroom-CC-2015-10-1-CPU-Comparison-Skylake-X-Kaby-Lake-X-Broadwell-E-Kaby-Lake-Ryzen-7-973/ is more recent

Looking at the charts it's seems that modest gains (10-17%) in performance are seen between generations of Core I7, 4 core processors.  I guess that's to be expected.  It seems looking at this and the previous tests that Lightroom really doesn't take advantage of more than 4 cores, not that many of us would go that route anyway.

It also seems that any real performance gains that we might like to see will depend on Adobe programming expertise and efforts.

Now.... back to your favorites for Lr performance improvement candidates.  Where would you like to see improvements? 
Logged
Regards,
Ron

Hoggy

  • Full Member
  • ***
  • Offline Offline
  • Posts: 207
  • Never take life, or anything in it, too seriously.
Re: Lightroom performance improvements...
« Reply #56 on: August 02, 2017, 09:06:52 pm »

Now.... back to your favorites for Lr performance improvement candidates.  Where would you like to see improvements?
For me, I would have to say brushings..  That and stopping the lag when using the Wacom with Windows Ink enabled (to allow for pressure sensitivity) - PS fixed it, so why is it taking years to fix it in LR.
As for the brushings, that includes any brushing, be it adj brush, radial brushing, grad brushings.  I always have to stop any pointer movement for anything to show up at all.  Sure, there are workarounds - disabling this, disabling that - but they are still just that: workarounds.  Also the speed of doing anything else after there are a lot of local adjustments, is dreadful.
I keep wondering how in the hell could professionals be using LR with all it's performance deficits.  I'm just an enthusiast and I can't stand the speed.  How professionals can be using this, just boggles my mind.  My guess is workarounds galore.
« Last Edit: August 02, 2017, 09:12:55 pm by Hoggy »
Logged
Cams: Pentax K-3, K-30 & Canon G7X, S100
Firm supporter of DNG, throwing away originals.
It's the hash, man..  That good hash!

rdonson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3263
Re: Lightroom performance improvements...
« Reply #57 on: August 03, 2017, 08:11:20 am »

Well said.  I experience the brushing lag you site as well.  The only workaround I've found is to the exit and restart Lightroom. That's more than annoying for me.
Logged
Regards,
Ron

john beardsworth

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4755
    • My photography site
Re: Lightroom performance improvements...
« Reply #58 on: August 03, 2017, 12:55:36 pm »

I keep wondering how in the hell could professionals be using LR with all it's performance deficits.  I'm just an enthusiast and I can't stand the speed.  How professionals can be using this, just boggles my mind.  My guess is workarounds galore.

Because not everyone has problems?

Despite two or three reminders, where's the flood of replies suggesting favourite Lr performance improvement candidates? Maybe that's because the experience is varied. Your specific issue sounds driver-related and I see nothing like it on my system (I did have crashes caused by the latest AMD drivers which AMD fixed with an update), but I do get slowdowns in Develop with lots of spotting corrections on a single image.

Other problems are more in Adobe's realm. For instance:
  • Deleting all the History recently solved a severe slowdown on a client's catalogue. So one favourite candidate would be that Lr should automatically cull History in some way.
  • Another seems specific to one brand of camera, Fuji. Exporting from Fuji raw files takes 40% longer than from DNGs made from those files. So there are opportunities in caching.
  • Another was introduced by third parties - accessing presets. I recently found that some users of my X-LR plugin found it started very slowly on systems with lots of third party Develop presets, especially the ludicrously-expensive ones like VSCO. I believe preset encryption caused the issue, so maybe Adobe could greatly reduce the number of calls to the list of presets (every time you select an image, Lr tries to see if it matches an existing preset - does it need to do this at all?).
  • Raw data is accessed unnecessarily and multiple times, so could Develop automatically create smart previews, standard and 1:1's, and could Export use previews (eg if you export full size it should use a 1:1 preview if it's up-to-date and of enough quality).

These are four of mine, as well as using embedded previews, but notice how diverse they are.

John
Logged

rdonson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3263
Re: Lightroom performance improvements...
« Reply #59 on: August 03, 2017, 04:04:15 pm »

John, thanks for responding to Hoggy.

I have experienced lagging when doing lots of brushwork on my new 27" iMac with an AMD Radeon Pro 580 with 8GB of VRAM.  That's not specifically listed by Adobe as supported/tested although it far exceeds their minimal requirements.  https://helpx.adobe.com/lightroom/kb/lightroom-gpu-faq.html 

Anyway, the next time it happens I'll try your suggestion to delete some of the history states as a workaround until Adobe fixes the underlying problem.



Logged
Regards,
Ron
Pages: 1 2 [3] 4   Go Up