Skip to main content
Known Participant
April 20, 2023
解決済み

Denoise takes several minutes to complete

  • April 20, 2023
  • 返信数 10.
  • 49648 ビュー

I posted this on a Youtube message board and apparently I'm not alone...so posting here.

 

I have a relatively fast Win 11 desktop, with 32Gb of RAM, and plenty of free disk space on both my cache and program SSD drives. Lightroom's new denoise takes 360 seconds (6 minutes) to process a single 50Mb file from my Sony a1. Topaz deNoise AI takes 80 seconds for the same file. Another user with a similar Win11, SSD setup with 64Gb RAM reported 5 minutes to process a 52MB Pentax file.

 

The Youtube creator 'showed' a processing time of 7 seconds. While it's unclear whether he used a technical trick to speed that view on his video, he did suggest reporting our long processing times to Adobe/Adobe Community to see if there's something other than Lightroom causing these long times?

解決に役立った回答 Keith Reeder
quote

It's not lack of ram or hard drive space and my graphics card was updated so I'm stumped. 


By @frankt98015343
Correct - it's an inadequate GPU. Same as everyone else with the same problem.
 

 

返信数 10

Participant
April 25, 2023

Adobe denoise is unuseable for those of us with "older hardware". I refuse to spend hundreds or (alot) more dollars on a new system when Topaz does the same thing in a fraction of the time. 

Community Expert
April 25, 2023

You do realize that absolutely nobody is forcing you to upgrade your hardware? Lightroom will work just fine if you don't use its denoise. I think it is a wonderful new tool but you really don't have to use it. Nobody forces you.

 

I use Topaz' tools all the time especially the motion blur recovery in Sharpen AI for wildlife images that I used too slow a shutter speed on and it is great (but finicky) for that. I also like gigapixel quite a bit better than Lightroom's super resolution tool. It is very clear however there is a big difference in quality in the denoise results with the new tool in Lightroom being quite a bit better than Topaz. Topaz' denoise creates lots of artifacts that are very hard to get rid of even when working directly from the raw files. I often have to do lots of manual recovery from the bad job the Topaz denoise tool does so I am elated with the new AI denoise in Lightroom.

hofendisasd作成者
Known Participant
April 25, 2023

Totally agree with your second paragraph re Topaz Sharpen AI and have been happy until now with Topaz deNoise. After quite a bit of testing -and spending almost $400 on a new(er) graphics card -I will likely transition to LRC denoise.

 

That being said, your first paragraph (which echoes an Adobe rep on this thread) is hard to accept. Adobe is rolling out -with some fanfare - a new tool that users will generally have to spend hundreds of dollars to effectively use. Already paying for a subscription service, most users that I've heard from are saying the same thing that Chris states: no way are they going to open their pocketbooks to upgrade to use this and will stick with Topaz or other denoise software. This was a topic of discussion locally here today. Virtually everyone said they were disappointed to have a new tool that they can't use (although they said they'd very much appreciate being able to use a single piece of software, IF they could.) Aside from me, no one was planning to upgrade their computers/cards and a couple talked about switching to Luminar (ugh) or On1, etc.

 

So, rather than have happy users that are excited about this new tool, Adobe overwhelmingly is going to have disappointed users looking to see what other options are out there. From a business point of view, you really don't want your users even thinking their are other options. And there are several.

 

But hey, not my call on either side. The market will speak for itself.

Participant
April 23, 2023

I have a similar problem on my computer.  One picture estimates 10 seconds, but usually takes 15 to 20.  I can do a similar image next and it says 10 seconds, but takes 5 or more minutes.    I will continue to use Topaz.

hofendisasd作成者
Known Participant
April 23, 2023

Update:

  • One of the other people with the 'slow processing' problem upgraded his graphics card from a 4G to 12G and reports that processing times are now at least, tolerable. Topaz deNoise is still 9x faster, but the 6-minute processing for LRC is now under a minute. I'm going to do the same upgrade and will report back. (Of course, we shouldn't have to spend hundreds of dollars on new hardware to make this new feature work - especially if the competition can do it faster on the same cards we have now.)
  • Another user corrected me to say that LRC DOES allow batch processing of denoise files. It's a different process (not the "sync settings" that we're used to, but instead one selects the group of files, moves the slider as desired, then processes. Still 6+ minutes per file, but hopefully the new graphics card will make that tolerable.
  • After processing another 20+ files in Topaz and LRC, I've concluded that LRC does a very good job and is comparable to Topaz. One apparent advantage that LRC has (over just the Topaz RAW model) is that you can make some adjustments before denoising, which you can't do with the Topaz RAW model (you CAN do it with the other models). Note, we always want to do denoise early in the workflow - certainly before texture, curves, dehaze, clarity or sharpening.
  • The LRC DNG output files remain about 1/2 the size of the Topaz DNG output files my original 50Mb RAW file becomes 160Mb after processing in LRC and 291Mb with Topaz deNoise. That IS significant, although after giving to my clients and waiting some time, I'll probably want to save in a format that takes less space.
TheDigitalDog
Inspiring
April 23, 2023
quote

Update:

  • One of the other people with the 'slow processing' problem upgraded his graphics card from a 4G to 12G and reports that processing times are now at least, tolerable. Topaz deNoise is still 9x faster, but the 6-minute processing for LRC is now under a minute.

By @hofendisasd

Those of us with modern hardware have reported speeds, so yes, older GPUs are the issue if 'time is important**

You can't compare this with another product like Topaz; apples, and oranges. "We" don't know how that product uses CPU or GPU (or both), but it is known that Adobe pushes GPUs hard for many new features like DeNoise etc. 

 

You can wait for the processing on older equipment or update your older equipment. You can use another product. But there isn't anything here for Adobe 'to fix'; it is what it is. 

 

** Middle of 1990 or so, Photoshop 1.0.7, Mac IIci with 8MB of ram. Opened a 15MB document that needed to be rotated 1 degree clockwise. Took 15 minutes! And yes, MB is the correct size above. We've come a long way in terms of hardware and software functionality. What took 15 minutes to accomplish back then (and why at the time I asked Adobe for the 'Beep when done" option), couldn't be accomplished at all a year earlier. 

Author “Color Management for Photographers" & "Photoshop CC Color Management/pluralsight"
hofendisasd作成者
Known Participant
April 23, 2023

Yes, things have come a long way. My first computer was a mainframe we were coding with Fortran 4 and my first home computer had 4Mb of RAM and a $250 hard drive, also with 4MB of storage (that's Mb not Gb). I now have 8T of storage on a RAID drive, with SSDs and tons of RAM. My 4G graphics card, while not reaching current gaming standards, is not a dinosaur either. All together, not really legacy "older" equipment.

 

I get your point that Adobe doesn't do things the way that competitors do, so "nothing to fix." Yet I'm the past president of a photography club with 150 members, probably more than 3/4 of whom use Lightroom and almost all the other 1/4 use ACR. My computer is probably in the fastest/newest 10% of theirs. So for our members to use this new feature - the way that Adobe engineers CHOSE to build it - our members would likely spend $60k+ on new graphics cards, if not new computers. Can a case be made that the Adobe engineers could have chosen to make the feature backwards compatible to comparably run in the same time zone as the competition?

Participant
April 22, 2023

Six minutes?  I wish!  Just tried my first file, and what took 16 seconds with Topaz took ~35 minutes with Adobe.  Yeah, you can point to my weak GPU, but I think it's the software. While running, my CPU is 13%, memory is 75%, and GPU is 1% utilized.  Ugh.

frankt98015343
Participant
April 22, 2023

I'm with you guys on the problem.  My files are listing as more than 15 minutes, yet I can get the denoised in Topaz in less than a minute.  This has happened on both my systems.  It's not lack of ram or hard drive space and my graphics card was updated so I'm stumped. 
I was on line with Adobe Help but they weren't any help. 
I'm hoping Adobe reads these posts and puts out a solution to the problem.  At first I thought it was just my computer, but judging by the traffic on this site, my system is not the problem. 

Come on Adobe - Fix this thing.

Participant
April 25, 2023
quote

It's not lack of ram or hard drive space and my graphics card was updated so I'm stumped. 


By @frankt98015343
Correct - it's an inadequate GPU. Same as everyone else with the same problem.
 

 


It does work on my computer, but for the same raw image file times vary
from seconds to minutes. Sounds like a software bug to me. This is more
like a beta release of denoise than a finished product. Also, no support
of tiff, hdr, or panoramic images. This is a half release. They should
have waited to get it right.
Participant
April 21, 2023

I'd be happy if denoise took 6 minutes. My sony a99ii files take 20 minutes each, compared to just about 4 minutes for Photolab DeepPrime, and I'm not sure that the result is any better. My machine is:

Intel(R) Core(TM) i7-8700T CPU @ 2.40GH

64 bit windows 10

32G RAM

Intel UHD Graphics 630

 

The GPU is on 100% for that 20mins.

Software problem I reckon - hopefully Adobe will fix it.

D Fosse
Community Expert
Community Expert
April 21, 2023

Integrated GPUs, such as the UHD 630, have never worked well for Adobe applications. They certainly don't measure up in Photoshop, and now Lightroom.

 

That's the weak link in your hardware.

 

For those trying to stay on a reasonable budget, I can recommend the Nvidia RTX 3060, finishing up 60 megapixels in 33 seconds here. That seems comparable to any of the other best results posted so far. It would be very interesting to see if the higher RTX numbers do significantly better (3070/80/90).

Community Expert
April 20, 2023

Another factor that should not be forgotten is that whatever Adobe did, it is leading to FAR better quality than Denoise AI/Photo AI give. Far fewer weird artefacts, no hot pixels that get amplified, no weird grid artefacts, etc. I am frankly blown away by the quality of these denoise results and it is completely reviving old images that I thought were not really scalable to the sizes I like to print. I will happily give up a bit of speed for this. You do need a very beefy GPU or a recent Apple Silicon mac to really use this comfortably though.

hofendisasd作成者
Known Participant
April 21, 2023

Well, "far" better is somewhat subjective. I'm certainly not finding that using the deNoise RAW model. A "bit" of speed is somewhat less subjective, LR taking 6 minutes vs 40 seconds for Topaz. And if you happen to need batch processing as I need for my clients' multiple images, well, LR is simply not an option. The 100 4000-6400 ISO images I provided to my clients this week would have taken over 10 hours for LR (and me having to stop and run each one individually, so who knows how long it really would have taken) vs me batching them in Topaz and coming back after dinner to find them all finished.

Conrad_C
Community Expert
Community Expert
April 20, 2023

I think the Ian Lyons post that Victoria linked to may hold part of the answer to this. Hardware acceleration specifically for machine learning/AI might be a missing link that helps explain why some computers can process AI features so much faster than others, and might help explain why older computers take much more time.

 

On my mid-range but recent laptop, I have so far never seen a Denoise time longer than a minute. I just did some tests and below is what I got. My hardware list is at the end…really nothing special from a CPU/GPU point of view, so I strongly suspect it’s the Neural Engine machine learning hardware acceleration that cuts the times. And that machines with older AI acceleration hardware, or none at all, will have to take longer. The Neural Engine is only in Apple Silicon Mac processors, which might help explain why owners of older Intel CPU based Macs are similarly reporting much longer AI Denoise times in the minutes. On the Windows side, I wonder if times are consistently faster using an NVIDIA GPU new enough to have their more recent AI acceleration hardware. (I notice that D Fosse above only took 33 seconds using an RTX 3060 with 12GB VRAM. That GPU model was released the same year as my laptop.)

 

In my tests, while the GPU is very busy, the CPU is very quiet during AI Denoise, with other background processes using more CPU than Lightroom Classic. This looks very similar to video editing, where newer computers render video very quickly because they have both a powerful GPU and hardware acceleration for popular video codecs. An older computer lacks both, forcing rendering to the CPU, which takes many times longer, and costs a lot more power and heat.

 

Of course, this only looks at performance within Adobe AI Denoise alone. It does not explain why AI Denoise might be slower than Topaz or others on the same images. I expect Adobe will optimize competitive performance further after this first release.

 

- - -

 

Test results

 

24 megapixels (Sony ARW raw)

ISO 3200

Default Develop state, no edits

AI Denoise set to 40
Estimated time: 40 seconds
Actual time: 36 seconds

 

16 Megapixels (Panasonic RW2 raw)

ISO 800
Many edits and masks
AI Denoise 25
Estimated time: 30 seconds
Actual time: 28 seconds

 

16 Megapixels (Panasonic RW2 raw)
ISO 800

Reset to default Develop settings
AI Denoise 25
Estimated time: 30 seconds
Actual time: 27 seconds

 

Equipment:

MacBook Pro M1 Pro laptop (1.5 years old), 8 CPU cores and 14 GPU cores (yes, it’s the base model), with 32GB unified memory.

(Unified memory means the GPU integrated with the Apple Silicon SoC can use any amount of unused system memory.)

Lightroom Classic uses about 5.5GB memory through most of AI Denoise processing, but that more than doubles to 12–13GB near the end of AI Denoise. That still may have left large amounts of system memory available to the graphics/AI acceleration hardware.

No special preparation: 6 days uptime, lots of apps open behind Lightroom Classic.

Legend
April 20, 2023

"Remarkably slow" is a reasonable result if you have a low-powered GPU. So reporting it to Adobe will just result in having Adobe point to your GPU. AI Denoise is an extremely computationally intensive process, more than any other process in LrC.

Known Participant
April 20, 2023

My LRX Denoise is taking 12 minutes to process. I guess it's my graphics card, but should it be THAT slow? Here's what I have:

NVIDIA GeForce GTX 1050 2GB Ram

Processor Intel(R) Core(TM) i7-7700 CPU @ 3.60GHz, 3601 Mhz, 4 Core(s), 8 Logical Processor(s)

Windows 10 64  32GB RAM

So, if it's my card... what do you recommend I replace it with??  Thanks

Michael

Known Participant
April 20, 2023

Typo.. I meant LRC

JohanElzenga
Community Expert
Community Expert
April 20, 2023

It's your GPU that determines how fast Denoise will be.

 

-- Johan W. Elzenga
Victoria Bampton LR Queen
Community Expert
Community Expert
April 20, 2023

My 16MP images are running about 9 seconds on one machine and 3 minutes on an old one. What's your graphics card? That's the main difference in performance.

Victoria - The Lightroom Queen
hofendisasd作成者
Known Participant
April 20, 2023

I have an NVIDIA GeForce GTX 1650 Super. 4Gb of GDDR6 RAM

TheDigitalDog
Inspiring
April 22, 2023
quote

I have an NVIDIA GeForce GTX 1650 Super. 4Gb of GDDR6 RAM


By @hofendisasd

 

Underpowered for this functionality. My 2022 MacBook Pro with 64GB of unified memory processes a Canon R6 Mark II raw in about 12 seconds. 

Author “Color Management for Photographers" & "Photoshop CC Color Management/pluralsight"