Is Lightroom Getting Slower

[Ronald30553178lwuv]

There are a number of threads where community members have complained that Lightroom 13 is a lot slower than version 12. The complaints usually involve severe slowdowns. I have not experienced these severe slowdowns. However, runs of the Puget Systems benchmark suggest to me that Lightroom is indeed getting slower.

With Version 12 typically the benchmark ran around 820. On version 13.0x this dropped to around 740. After moving to version 13.1 scores run around 660. There are possible causes other than the program being slower. My computer is an LG Gram with an I7-1260P CPU, 16 gigs of memory and Intel Iris graphics. The LG Gram is said to have a minimal cooling system. It's possible newer versions of LRC utilize more CPU cores thereby overwhelming the computer's cooling to the point of thermal shutdown. Maybe there is something about the design of the benchmark program which invalidates comparisons across different versions of LRC. There is the slight possibility that my computer is suffering a cumulative problem in hardware or software. However, I suspect the benchmark results are indicative of a problem. Note that I have tried to no avail the fixes suggested around here like resetting preferences or making a clean reinstall.

Does anyone have thoughts about this?

[Keith Reeder]

We've been seeing (apparently) increasing complaints about Lr slowing up since around version 5 - it's hard to put much store in what seems to be happening nowadays. Personally I just think that more people are complaining because it's easier to do so, and the additional online "chatter" about it motivates people to share their experience too.

"Intel Iris graphics."

I suspect that's your problem right there - the Real World minimum GPU recommendation is for a dedicated card with at least 8gb, and as many tensor cores as you can afford to pay for. You singularly don't have that.

"Note that I have tried to no avail the fixes suggested around here like resetting preferences or making a clean reinstall."

But not the fix of using a computer with the power to run Lr properly. Shared-RAM iGPUs are really not Lr's friend.

[Ronald30553178lwuv]

If it "seems to be happening" then it isn't real, so why should I go out and buy a new computer for something which isn't real. There's a database of Puget benchmark results and results have been dropping as versions progress for users of 8 Gig dedicated graphics cards. I suppoose all the other people having problems just seem to....

[jrsforums]

As software adds functionality, the resources to drive it need to increase.  The original PC had 4K ram and a processor that could light a match today....but it worked great with the software it had...but you couldn't even start to load today's SW..  It is just a fact of life....the train moves on.  Unfortunately, if you want to continue to upgrade the software, with new functionality, at some point you will need to upgrade your hardware.  I, for one, am happy with the improvements LR is making/adding....and some of them a speed improvements, such as the read ahead they added some time back.  But everyone needs to make their own choice.  You could look for lower functioning software that you existing hardware can support with good responsiveness.

[Ronald30553178lwuv]

The program is reasonably responsive. I can understand that adding functionality increases hardware requirements. What I don't understand is how moving from 13.02 to 13.1 entails a 10% performance hit. The problem is gradual degregation rather than an extreme hit to responsiveness. Actually, I am looking around for faster hardware. The problem is I don't want to lug around a notebook which weights 7 or 8 pounds when the power supply is included.  The new 14th gen notebook CPU's from Intel offer some promise as the power requirements for desktop replacement processors have dropped to thin and light levels.

[dj_paige]

The program is reasonably responsive. 

By @Ronald30553178lwuv

Then you have a local problem that is very likely causing "reasonably unresponsive", and that is something that you (we) should try to fix. There are plenty of people whose LrC is "reasonably unresponsive" and this forum has figured out what the solution is (but the success rate of solving these problems is not 100%).

"Reasonably unresponsive" is different than a 10% decrease in speed. The two are unrelated.

By the way, many of us are not seeing either "reasonably unresponsive", or a 10% drop in speed, or a slow steady decline in speed.

[Ronald30553178lwuv]

You have misquoted me. I wrote "reasonably responsive" not "reasonably unresponsive". The later doesn't make any sense. While increased functionality may be a cause of slowdowns, there is little or no increased functionality when going from 13.02 to 13.1 yet the benchmark shows a 10% slowdown. Please explain that to me.

It seems the prevailing opinion is an 8 Gig GPU is the solution. My view is that is that is increasingly out of step in a mobile computing world. My view is the developers are ignoring the potential of recent improvements in CPU's where the mumber of threads and cores have greatly increased. 

[Keith Reeder]

"If it "seems to be happening" then it isn't real"

That's one interpretation of the phrase, but not the only one: you suggested that the number of  threads complaining about deteriorating performance indicated something. I'm simply saying that this has been going on for a long time, and that - maybe - what you're seeing isn't indicative of the worsening (thread title: "Is LightRoom getting slower?") problem your comment implies, after all. 

Maybe it is, maybe it isn't - we have no convincing evidence either way. But there is ample empirical evidence that integrated GPUs can be a significant (often show-stopping) problem.

[Ronald30553178lwuv]

I find your comments to be contradictory on their face. You say the problem has been going on for a long time but it doesn't exist. I agree that a dedicated GPU will speed things up, but that doesn't change the fact that the program continues to get slower for everyone and that includes users of dedicated GPU's. The evidence is out there, just look at the database for the Puget benchmark.  I'm looking for ideas outside the buy a new computer box.

[D Fosse]

The GPU is a very large part of this, and the obvious place to start.

A few years ago the GPU was basically irrelevant. Now a lot of functionality relies entirely on it.

For me, LrC is a lot snappier now than it was 5-6-7 years ago. But there's no denying that applications increase in complexity all the time. As does the operating system, and all the hardware. So unless all the ducks are very neatly lined up, I would assume that the general risk of problems all over tends to increase.

[Ronald30553178lwuv]

Unforttunately, when one has a notebook a new GPU  means a new notebook. I can understand the measured drop in speed from V12 to V13 but the drop in measured speed from 13.02 to 13.1 is inexplicable. The only new functionality is expanded camera/lense support. There is a bugfix for a display problem. 

[jrsforums]

More than one bug fix

https://helpx.adobe.com/lightroom-classic/kb/fixed-issues.html

[Ronald30553178lwuv]

Thank you for that information. Still, there is no new functionality yet the program runs measurably slower. In particular, the fix regarding image export using the GPU when it shouldn't ought to speed things up for non dedicated GPU users. 

[jrsforums]

I have fairly powerful system.  I have not noticed any speed charge, but does mean there hasn't been...just not noticed.  Complex software can be funny....particularly with the huge variety and mix of hardware, accessories, driver level, operating systems (and current level), other apps installed, etc.  what affects one system may not be seen on another.  Companies like Adobe do their best to have their release test bucket and beta users wring out the bugs, but it is impossible to even come close to testing every thing out there.  My method over the years has been to 'over provision' my system...to cover speed glitches (didn't help with 2nd monitor problem thought 😀)

[Keith Reeder]

"I find your comments to be contradictory on their face. You say the problem has been going on for a long time but it doesn't exist. "

You're misrepresenting me - I said no such thing. Not even close: of course it exists. I'm just pointing out (quite unambiguously) that you're making unsupportable assumptions about it getting worse.

" I'm looking for ideas outside the buy a new computer box."

Then I imagine you'll be looking for a long while yet. What you want, and what there is to be had - two very different things. 

Feel free to ignore the depth of evidence and experience in favour of your wish-list, though...

[Ronald30553178lwuv]

I found that by disabling GPU acceleration I could get the same measured speed with 13.1 as I had with 13.02 with limited GPU acceleration on. It would be greatly appreciated if participants around here could think outside the box and instead of just telling me to get a new notebook they would make some suggestions which would work with my existing hardware. 

[jrsforums]

If that 'patch' is acceptable...OK.  It is not a long term solution as more and more functions, old & new, we probably need GPU support....and faster cpu, bus, memory, and more memory.

[Ronald30553178lwuv]

Maybe the developers will try to get more umph out of the newer CPU's with more threads. 14th gen Intel notebook chips are said to have improved GPU's. 

I tested this on version 12 and GPU support made no difference. 

[Keith Reeder]

I highly doubt that - even if it was possible - Adobe is going to fundamentally rewrite its entire processing pipeline to make life easier for users who aren't prepare to abide by the well-publicised minimum spec requirements for LightRoom.

[Ronald30553178lwuv]

Thank you for your pure speculation.

[jrsforums]

Ronald, first CPU's & GPU's provide different capabilities for processing, which is why for this type of work, they are using GPUks more often.  Either way, it is going to require hardware upgrades on your part....I suspect you don't have 13 gen, much less new 14..

[Ronald30553178lwuv]

I have a 12th gen. There difference between an i7-1260P and the 1360P is minimal at the CPU  level and none at the GPU  level. So far there are no 14th gen notebooks for sale with discrete graphics.  However, I am quite certain that rather than taking advantage of the increased number of cores in 12th gen CPU's Adobe is going down the route of increased reliance on the GPU which I find disappointing. What does it take to get anyone around here to realize that telling someone to go out and buy a new notebook isn't being helpful.

[jrsforums]

I'm sorry, but the facts of life are that lack of GPU, slower CPU (mobile are slower than desktop to use less power), only 16GB memory....all these combine are going to be more constrained and show more effects of increased software demand as function is added.  I suspect your performance is not 'bad'...just has slowed from what you had before.  Maybe it's the add functionality...maybe it's bad code somewhere....it is what it is....it may get fixed....it may get worse....who knows.  As I said earlier, over provisioning to keep up with these changes is necessary if you really want optimal performance.

[Ronald30553178lwuv]

Maybe I should spring for an M3 MacBook pro.  I wonder if this will do the job. The latest Puget benchmark for Photoshop shows MacBook pro's running at high end desktop speeds.

[jrsforums]

Apparently Apple's 'M' series have good on chip GPUs.