ArtCraft: Free Adobe Alternative, But Can You Trust It?

You've no doubt seen ArtCraft all over the internet over the last week or so; it's a set of free apps that look an awful lot like Photoshop, Lightroom, Premiere and the rest of Adobe's apps, and they've all been built using AI. There's been video after video, article after article and thousands of comments cheering it on, but having read through a lot of it, there are a few things I haven't seen anybody mention.

Before I get into it, I want to be upfront about where I'm coming from. I do occasional work for Adobe, recording tutorials for Adobe Learn, presenting and so on, and I use Creative Cloud every single day. I also use Claude, which is the AI that was used to build these apps, and I love it. I'm editing more and more of my videos in DaVinci Resolve rather than Premiere Pro, and I love that too, so I'm all for competition. I just want you to know where I stand so you can weigh up what I've got to say.

What's actually happened

A software developer called Brandon Thomas has released seven free apps under the ArtCraft name. PhotoCraft is his version of Photoshop, LightCraft is Lightroom and FilmCraft is Premiere, and there are versions of Illustrator, InDesign, After Effects and Acrobat as well. He says he built them using Anthropic's AI, Claude, and they're open source, which means the code is published for anyone to look at.

The reaction has been huge. GitHub is where software developers share their code, and people can give a project a star, a bit like a like on YouTube. Eleven days after it all came out, PhotoCraft alone had 44,000 of them. He also tried posting about it on the Adobe Reddit forum, but understandably that post didn't stay up for very long.

Early, very early

What's had a lot less attention is how finished it is. By PhotoCraft's own roadmap, it's at about 25% to 35% of what a professional needs for daily work. Only two of the seven apps are even labelled as early alpha; the rest are still in development. When it launched he promised it would match Adobe completely within a month, and when people started pointing out what was missing, that became 99% within months.

It's also not trying to hide where it comes from. PhotoCraft's own website describes it as "the open-source image editor you already know how to use", and one look at the screenshot tells you why: the menus, the tools down the side, the Curves panel, the Layers panel. You already know how to use it because it's been made to look like Photoshop, and that's the whole pitch.

What can Adobe actually do about it?

I'm not a lawyer, so take this as my understanding rather than legal advice. Copyright protects the actual code, the written instructions that make a program work, and things like the artwork and the icons, but it doesn't protect an idea or what a program does. If he really hasn't copied any of Adobe's code, which is what he says, a copyright case isn't straightforward.

An intellectual property attorney called Steve Hanson made a really good point about this in [his own video]([FILL IN: link to Steve Hanson's video]). He reckons copyright probably isn't where Adobe would go at all, and that it's more likely to be the contract, the terms you agree to when you sign up. Section 17 of Adobe's terms says, in Adobe's own words, that they don't allow you to recreate, or attempt to recreate, their products.

Then there's the look. There's something called trade dress in the US, and passing off here in the UK, which is about whether a product looks so much like another that people could be confused about who made it. The menus, the panels and even the keyboard shortcuts match Photoshop, and that's exactly the area people are pointing at.

So far Adobe hasn't said a word, and there's good reason it might stay that way. Taking a free, open-source project to court would hand it more publicity than it could ever buy, and the code is already out there for anyone to copy; people already have, and there's another free editor built on PhotoCraft's code. Even if ArtCraft stopped tomorrow, the apps wouldn't disappear.

I remember when Affinity Photo first came out and everyone was saying how close it was to Photoshop, in how it looked and what it could do. We all kind of thought something was going to happen, but nothing did, and today Affinity is owned by Canva and it's free. It's going to be really interesting to see what happens with this one.

Why he built it

This is the part I find really interesting. He's said Adobe charged him its cancellation fee more than once, and in his own words, "They charged me the 'cancellation fee' on multiple occasions and I've hated them ever since." Elsewhere he said he'd been "bit by the dark pattern cancellation fee one too many times."

So this didn't start as a better way to edit photographs; it started with someone who hates Adobe.

I do understand the frustration. In March this year Adobe agreed to a $150 million settlement in the US over how it sold and cancelled Creative Cloud subscriptions. Half of that was a penalty and half was free services for customers, and Adobe didn't admit any wrongdoing. Plenty of people have felt exactly what he felt, and Adobe brought a lot of this on itself, but I'd want to know a lot more about someone who built software out of a grudge before I put it on my computer, let alone gave them my card details.

Who's behind it

Brandon Thomas is a software developer with around 15 years' experience, and he's worked at Square, the payments company. Before ArtCraft he was CEO of a company called Storyteller.ai, and ArtCraft's contact email still goes to Storyteller. Storyteller's best-known product is FakeYou, a site for generating the voices of celebrities and characters using AI, and in 2022 Italy's data protection authority opened an inquiry into it because it could reproduce the voices of public figures, including Italy's Prime Minister, from whatever you typed in.

ArtCraft itself started out as a tool for making and editing AI-generated images and video, and that's still what it's selling.

How it's paid for

The apps are free. What ArtCraft sells is credits for AI image and video generation on a monthly subscription, at $10, $35 or $60 a month, using AI models from Google, OpenAI and others, including Google's Nano Banana, the same model Adobe has put into Lightroom's Prompt to Edit.

So a project that launched with the message of no more subscriptions is paid for by a subscription. To be fair, the site does say you can use the apps without paying, and it's actually a very similar idea to Affinity now, where the app is free and you pay Canva for the AI features. The difference is who's behind it. With Affinity it's Canva, a company you can look up and hold to account; here, it's someone who's told us he hates the company he's copying.

A question if you're worried about AI

There's been a lot of anger about AI in Adobe's apps, and about AI being trained on work photographers and artists never agreed to. I totally understand that; I've shared videos about my own concern over too much AI in Lightroom and Photoshop, and plenty of you have told me how you feel about it.

If that's you, what do you make of this? A set of apps built by AI, copying Adobe's software and paid for by selling access to AI image generators. I'm not saying it's wrong, I'm just asking whether it sits differently with you because it's aimed at Adobe.

A link versus an install

There's one more thing about all of this I just can't get past. We're all really careful about clicking a link in an email we're not sure about, yet installing software gives it far more access to your computer than clicking a link ever could, and thousands of people have installed this without a second thought. That's the bit I can't get my head around. Why would you be more careful with a link than with a program you're putting on your computer?

When you install software from someone you know nothing about, can you be 100% sure what's in it? I'm not saying there's anything in it that shouldn't be there, and there's no evidence of that at all, but can you be completely sure? I couldn't.

When someone's prepared to go this far over a grudge with a company, building copies of seven of its apps and giving them away, I'd also want to know how they'd treat me, someone they've never met, if I ever fell out with them or if they ever needed something from me.

Would I install it?

No, not as it stands. It's early alpha software, it's not clear what it'll look like in a year's time, and the person behind it has told us he hates the company he's copying. That doesn't mean he's done anything wrong, but think about what's on the computer you'd be installing it on. For me it's every photograph I've ever taken, but it's also my emails, my passwords and the same computer I log into my bank on. That's my whole day-to-day life, not just my pictures, and I'd want to know a lot more history behind something before I gave it that kind of access, and well before I handed over a card number.

If you're curious, there's a version that runs in your web browser, so you can have a look without installing anything at all.

Competition is good

I'm all for competition. Look at what's happened with cameras over the last few years, Canon, Nikon and Sony pushing each other, or Apple and Android. That's when we win, with better products and better prices.

You're free to use whatever you want; it's your choice, not mine. I only use what's fit for purpose, whoever makes it, which is exactly why I've moved more and more of my video editing over to DaVinci Resolve. Affinity, DaVinci Resolve and Capture One are all really good alternatives to Adobe and I'd encourage anyone to look at them. The difference is that there's a company behind each one with a history that you can look up, and that you can hold to account if something goes wrong.

Five questions before you install anything

Whatever you're thinking of installing, ArtCraft or anything else, these are the five things I'd ask first.

• Who's behind it, and what have they made before?

• How is it paid for? If it's free, how is the person making it earning a living?

• Is the installer signed by a named publisher? Your computer will usually warn you if it isn't, and that warning is worth reading rather than clicking straight past.

• Where do your files and your card details go?

• If it disappears tomorrow, can you get your work back out of it?

If you can answer all five and you're happy with the answers, go for it. If you can't, I'd wait.

Updated Reflection Removal, But is it BETTER?

On the 29th September Adobe released Lightroom 9.6 and Lightroom Classic 15.6, and one of the things in the update was a set of improvements to reflection removal. Adobe's wording talks about cleaner results on complex surfaces, with fewer smudges and ghosting artefacts when you're removing reflections from windows, glass and water, which sounds great, but I think it's fair to say my experience of it up until now has been a little bit hit and miss. When it works it works incredibly well, but there are pictures where you'd think it should do the job and it just doesn't.

So I wanted to see how the update copes, and then show you something else that's now in Lightroom which could be a workaround for the times it doesn't. I haven't tried it on water yet; I wanted to see how it got on with ordinary reflections in windows first.

Where it works

The first picture I tried is one where the window is the dominant part of the scene, with me clearly reflected in the glass. In the Remove section on the right-hand side you've got Reflections, so I ticked Apply, using Preview so I wasn't waiting around too long, and after a few seconds it had done a really good job. Drag the Amount slider all the way to the left and you can see the reflection, drag it all the way to the right and it's gone completely.

That's the version of this tool people get excited about, and rightly so.

Where it doesn't

The next picture is me standing outside a gallery photographing it, and again I'm very clearly visible in the window along with plenty of other things being reflected. Same steps, same few seconds' wait, and nothing really happened, which is odd, because the window is still the main part of the picture.

The third was outside my friend Nicola Catherine's fashion knitwear and art studio, with me reflected in the window again. This time it looked really good at first, until you start looking properly; around the picture frame and up towards the top you can still see the reflection, and with the Amount slider dragged across I could still make out my teeth, my beard and my glasses.

That's where the tool is right now, and I'd fully expect it to keep getting better. The question is what you do in the meantime when it won't play ball.

Prompt to Edit

I've previously done this in Photoshop with a selection and Generative Fill, but there's another new addition in Lightroom which I think works as a workaround for now. Down at the bottom of the right-hand panel there's a small icon for Prompt to Edit. My first thought when I saw it was that I didn't really want that kind of thing in Lightroom, but Lightroom has a lot of users with different wants and needs, and it'll suit plenty of them. It's early access at the moment, so this isn't the finished version.

It works the way prompting does in any generative AI tool. There are presets for things like fixing dust and scratches, colour restoration and black and white colourisation, plus effects such as adding dense fog or replacing the sky, and you can type your own prompt and pick earlier ones from a history list underneath. The important part is the note at the top saying the result is saved as a new DNG, because what it's actually doing is regenerating the entire image to make whatever change you've asked for.

Right now the only models on offer are Nano Banana 2 at 2K, which uses 20 credits, or Nano Banana 2 at 4K, which uses 30, and there's no preview, so you don't get to see whether it's going to work before the credits are spent. I'd really hope more models appear once it's out of early access, because there are others that are better at working on just part of an image while holding on to the original.

Trying it on three windows

The picture I used has three windows with some pretty severe reflections, and the big white patch in the right-hand one is my softbox from when I lit the portrait. I typed "remove the reflections from each of the three windows", hit Generate, and after a short wait a second file appeared next to the original Fuji raw file, with "-prompt" added to the file name.

The reflections were gone, but flicking between the two with the arrow keys the saturation had clearly gone up, and in the right-hand window, where you couldn't see anything at all through the softbox, it had simply made something up. For your own pictures that might be fine, but if you're trying to keep any integrity in the image I don't imagine a client would be too keen.

So I deleted that one and tried again, adding "do not change the colour, do not increase the contrast, keep the original contents exactly the same" to the prompt and choosing 4K this time. The windows came out really well, but the saturation was still up, and zooming in showed the bigger issue; even at 4K it's a much lower resolution than the original photograph, and a lot of the lettering had changed. The menu was nonsense and the Costa sign now said "burned", which isn't the best advert if you're selling coffee.

Bringing it into Photoshop

However, all is not lost, because you only need the generated file for the glass. I selected the original and the new DNG in Lightroom, holding Shift to get both, then went to File, Edit In, Open as Layers in Photoshop.

In the Layers panel the difference in size between the two thumbnails is obvious straight away, which is a very visual way of seeing how much resolution you lose. With the generated layer on top, I selected both layers and went to Edit, Auto-Align Layers, left it on Auto and clicked OK, and it lined the two up really well. A quick crop took care of the edges, along with a little bit of the softbox showing in the top right corner.

Then I held Option (Alt on Windows) and clicked the layer mask icon, which adds a black mask and hides the generated layer completely, and painted the windows back in with a white brush. I nudged the brush hardness up so it stayed tidy against the window frames, because a very soft brush spills over them.

This is where you get to choose what you keep. One sign had changed colour, so I left the original there. The wording at the top of the TripAdvisor sign had gone a bit odd, so I painted that back with black. The menu had turned into a strange made-up language, so I kept the original there too; there's still a touch of reflection on it, but it's not dominant. In the left-hand window a chalkboard had vanished while a hanging decoration had been kept, which is pretty clever, but it's artistic licence, and the right-hand window is completely invented. That's usable for your own work, but for a client I very much doubt they'd want new things appearing behind the glass.

Fixing the Costa sign

The Costa sign was a different problem, because masking couldn't sort it without the colour shifting, so the wording itself needed changing. I added a new blank layer, drew a selection around the word with the Rectangular Marquee tool and used Generative Fill from the contextual taskbar with Firefly Image 5, which is the one to pick here; it's designed to edit while preserving detail. I typed "change to served", and it swapped the word over in what looks like the same font, which is really useful.

The finished picture is the full resolution original everywhere except the glass, which comes from the regenerated file, and flicking between before and after it's not bad at all. Ideally you'd do all of this with Reflection Removal inside Lightroom, and I'm sure that's where it's heading, but as things stand in October 2026, with Prompt to Edit in early access and reflection removal still a bit hit and miss, it's another good option to have for the pictures the built-in tool can't manage yet.

Are we thinking WRONG about AI Editing?

My last two videos had ChatGPT editing my photographs, first in Lightroom, where it went through a folder of my own edited pictures, worked out my style and then edited a raw file the way I would, and then in Photoshop, where it used Photoshop's own tools to remove a colour cast, add a long exposure effect to a sky and do a cutout, hair included.

In both of them it didn't generate a new picture. It moved the same sliders and used the same tools that you and I would, and every one of those adjustments was still there to change afterwards.

The comments on those videos have been brilliant and they've really made me think, but one kept coming up in different words, and the gist of it was this: so now every picture is going to look like yours.

When you watch what ChatGPT is doing in those videos, it's an understandable thing to think. I just don't think it's right, so rather than argue about it, I decided to test it, and the video above is the result.

Putting it to the test

I gave ChatGPT a folder of my portraits, with my edits on them, and asked it to work out how I edit. Then I gave it the raw file of a portrait I'd already edited myself and asked it to edit that in my style.

I'm not going to go through what I did to mine; the two results are side by side in the video, so have a look and see if you can tell which is which before I give it away.

Here's the thing, though. Whichever one you prefer, they both look like one of my portraits, and that's not because of the edit, it's because I took the picture. Everything that makes it look like one of mine happened before the shutter was pressed, and ChatGPT had none of that to work with; it just had my raw file.

The edit is the icing on the cake

The edit is the part of the process we can see on screen, so it's easy to think that's what makes a picture yours, but it's only one part of a much bigger process. I really enjoy editing, and it gives a picture its final look, but for me it's the icing on the cake.

The big part is what goes on between my ears: how I think about a picture before I take it and how I compose it. With a portrait, it's how the person looks in front of the camera and whether they're relaxed or not. With a seascape, it's what goes in the frame and what doesn't.

None of that is in the edit, so even if somebody had the exact colouring, contrast and tonality I use, their pictures still wouldn't look like mine. They'd look like theirs, with my colouring on them.

We can all buy the same kit, spend the same money on the same camera and lens, stand in exactly the same spot and point it at the same thing, and the pictures still won't look the same. There'll be something in each one that's uniquely yours, because of what you like and what you don't, how you were brought up and everything you've looked at over the years. All of that goes into a photograph, and none of it comes out of an edit.

An opinion, or a misunderstanding?

People have their own opinions about AI, and that's a good thing; having an opinion is part of being human, not something artificial. Sometimes an opinion just isn't right, though, and I know that sounds a strange thing to say. The idea that everybody's pictures are going to end up looking the same isn't really an opinion at all. It's a misunderstanding of what photography is.

Using the software, not replacing the picture

That's also why the way ChatGPT did those edits interests me so much. There's a difference between asking an AI to generate a new version of your picture and having it use the same software you and I would. When a picture's generated, you hand it over and get a new one back. When the AI works the sliders and the masks itself, you get your own raw file back at full quality, with every adjustment still there to change.

It's a bit like the difference between someone redrawing your picture and someone sitting next to you showing you what they did with it. Both have their place, but they're very different things.

Two very different photographers

Where it gets interesting is who all of this is for, because there are definitely two different audiences. Take Adobe as an example. On one side they've got working photographers, people with clients and deadlines, editing a shoot with hundreds of pictures that all need the same look. On the other they've got enthusiasts, people who take pictures because they just love doing it.

For the working photographer, imagine being able to hand an AI an actual part of your process. I don't mean something that makes a new picture; I mean something that applies the same look to a whole set of images inside the software, moving the sliders and making the masks without touching the originals, so you can go in and change whatever you want afterwards. That could be a massive time saver.

The hobbyist is a different story, because for them editing isn't a job to get through. They go out and take their pictures because they enjoy it, then come home and edit them with no rush at all, sitting in a room with a bit of music on and maybe a nice glass of something. For them the editing is part of the thing they love, and a one-click fix isn't for them.

However, some of what AI brings is. The Remove Tool in Photoshop has been a massive benefit; jobs that used to take an age now take moments. Anything that helps with selections, hair and fine cutouts especially, is going to be a huge help as well. None of that takes the enjoyment away. It takes away the tedious bits and leaves more time for the parts you actually enjoy.

How I use it

I use AI every single day, for the admin, the research and the writing that goes with running everything on my own. I'm a one-man band, so that's a lot of work it takes off my desk, and I'm not worried in the slightest that other people's pictures are going to look like mine.

What it gives me is time, and that matters to me more than anything, because I absolutely love doing what I do. I got into this creative world because I love being creative, and I love the freedom of it and the escapism it gives me. I don't want anything to replace that. If anything I want more of it, which is exactly how I'm using AI; it takes care of the jobs I'd rather not be doing and gives me back more time with a camera in my hand.

The mood of the moment

When I gave ChatGPT my Lightroom and Photoshop to use, it didn't edit my pictures exactly the way I would. It can pick up the things I tend to do, but when I'm editing I'm very much in the mood of the moment, chopping and changing as I go, and that freedom of just going with it is a big part of why I enjoy editing at all. AI doesn't know the mood or the feeling I want to give a particular picture, and creating that is what I've always loved.

The Photoshop side was different, because those were tasks. The colour cast, the sky and the hair cutout were jobs where I knew what the finished result needed to be, and what I found really interesting was the thought of handing those kinds of tasks to an AI to do by hand, using the software the same way I would, so I've got more time for the creative side. It was an experiment, but it's certainly one that's got me thinking.

ChatGPT Can Now ACTUALLY Use Photoshop, and It's Impressive

ChatGPT can now take control of your computer, moving the mouse, clicking and typing its way through a job the way you would, and the question I've been asked the most about it is whether it can use Photoshop. I've been testing it, and yes, it can.

What makes it interesting is that it isn't generating a new picture. It's working inside Photoshop with the same tools you and I use, which means everything it does is left as layers and masks you can go back into and adjust yourself, at full quality.

To see how far it could go, I gave it three jobs on three different pictures: fixing a colour cast, creating a long exposure sky, and cutting out hair, which is the thing I get asked about more than anything else in Photoshop.

How it works

I'm using ChatGPT 6 Astra on the high setting, in the Work section of the app. I type out what I want in plain language, send it, and it replies with what it intends to do. The first time it needs to use the computer it asks for permission, I click allow, and from then on it's hands off; I just sit back and watch it work.

With each job I asked it to keep everything non-destructive and editable, and not to save the file, just leave it as layers. You can see everything it does as it does it, and you can stop it at any point.

Fixing a colour cast

The first picture was a portrait of a biker with a definite colour cast, a slightly green and magenta look to it. I asked ChatGPT to look at the image, decide for itself which tools to use, and give me a neutral result.

It added a Curves adjustment layer, worked on the individual red, green and blue channels, and set the layer's blend mode to Color so that it only changed the colour and left the brightness and contrast alone, which is a genuinely sensible way to do it. The original background layer was untouched.

Turning the layer off and on, the cast was gone. It's not too bad at all, and I was quite impressed with that.

A long exposure sky

For the second one I wanted to step it up and see whether it could create an effect rather than make a correction. The picture was of my car on a very wet, stormy day, and I asked for the look of a ten second long exposure, with fast moving clouds coming over the top of us.

What really caught my attention was that it used the same technique I use for this myself. It selected the sky, put it onto its own layer, then converted that layer to a Smart Object before applying a Radial Blur, so the filter stays editable afterwards, which I thought was really clever.

Better still, it moved the centre of the blur. By default a Radial Blur spreads out from the middle of the picture, but ChatGPT pulled that point right down to the bottom, so the blur looks as if it's coming from the horizon and the clouds are streaming towards you. That's exactly the thinking you'd want.

I did have to stop it at the very end, because it was trying to save the file, but the effect itself was done.

Cutting out hair

The last job was the hardest, and the one I was most intrigued by. It's a picture of my friend Simon from a Viking shoot, on a grey background, but with lots of fine, loose flyaway hairs and fur cuffs on what he's wearing. I asked for the best cutout it could manage, body, hair and everything as one, output as a layer mask so it stays editable.

Watching this one was fascinating. It went into Select and Mask and used the Refine Edge Brush around the hair, and it kept switching the view mode to check its own work. At one point it reported that the fine strands were coming through, but that the black preview was showing grey left between the looser locks, so it went back to work on those and the fur before checking the final mask. Right at the end it even renamed the layer, which is maybe something all of us could learn from.

The result is a pretty good job. It picked up a lot of the fine hairs, it's a little smudgy in places, and there's some transparency through the fur. To judge it properly I put a solid colour fill layer behind Simon, a sky blue rather than anything too different from the grey he was photographed against.

The transparency in the fur is easily fixed. Duplicate the layer with its mask (Command or Ctrl + J) and the density builds up, then brush out any areas where the background starts to show through.

Doing the heavy lifting

It isn't perfect, and I don't expect this to be the end point, but as a way of doing the heavy lifting so you can go in and do the fine tuning yourself, it's pretty impressive. The colour correction was very good, the long exposure was really well done, and the hair cutout, the hardest of the three by some distance, was far better than I expected.

This is also the worst and the slowest it's ever going to be, given how quickly all of this is moving. I'll be doing a lot more testing, and one of the things I want to try is whether it can write down the steps it takes and grab screenshots as it goes, so you could almost create your own tutorials from it.

I Let ChatGPT Edit My Photos in Lightroom

The photograph in the video above was edited from start to finish in Lightroom, and I didn't make a single one of the adjustments. ChatGPT did. It used my Lightroom, moved the sliders, made the masks and took the picture from a slightly underexposed raw file to a finished edit, in my style.

ChatGPT has been able to use your computer for a while now. You give it permission and it takes over the mouse and keyboard and works through a job the way you would, opening apps, clicking and typing, and you can see everything it does as it does it and stop it at any point. With the recent update to Astra it's taken a real step up, which got me wondering whether it could actually edit a photograph. I don't mean generate a new picture; I mean sit inside Lightroom and edit a raw file with the same tools I use, and learn to do it the way I would.

Teaching it my style

The first job was getting it to understand how I edit. In Lightroom I set up an album called My Seascape Style with 11 of my images in it. Some still have their adjustments visible, masks included, so it could see exactly what I'd done. Others have the edit baked in, because I'd sent them over to Photoshop for some heavy lifting and brought them back, so there's nothing in the panels to look at, only the finished picture.

I used GPT-6 Astra with the effort level set to high. Day to day I find medium is plenty, but for studying my edits I wanted it to do an extra good job, even though it takes longer.

Then I told it what I wanted, dictated rather than typed because it's quicker. Look at every image in the album, study the adjustments, the masking, the detail and the colour grading, and try to understand not only what I'd done but why. For the images with nothing visible in the panels, study the mood and the atmosphere and the tonality instead, and work out what I'd have done to get there.

As soon as I sent it, ChatGPT asked for permission to use Adobe Lightroom, and I allowed it for that conversation. From then on you can watch it working away on its own, opening images and panels and looking around.

Watching it work

What really interested me was how far it went. It dived into the masking and looked at how I'd masked the rocks in the foreground and the adjustments I'd used to bring them out, which is exactly the kind of fiddly detail I'd half expected it to skip.

It posts updates as it goes, and part way through it had noticed that I recover the highlights and open up the shadows, then bring the depth back with control over the black point and selective detail. It had even picked up that I use one film-style profile for the warmer, pastel pictures and a different one for the cooler surf shots.

The whole thing took 11 minutes and 30 seconds. At the end it said it had reviewed the adjustments across all 11 images, including 27 local masks, saved an image-by-image reference of my style, and, thankfully, left my images unchanged.

Handing it a raw file

For the edit itself I picked an unedited raw file from the Local tab in Lightroom, a seascape that was slightly underexposed, which felt like a fair test. I told it to edit the image in my style using everything available in Lightroom, including the masking, the colour grading and denoise, then sat back with my arms folded to prove I wasn't touching anything.

When it finished, it described what it had done as my pastel seascape treatment: pink and lavender grading, cool water, deep textured rocks and luminous foam. The edit included four separate masks, denoise, selective sharpening, a custom tone curve, subtle grain and a vignette, plus a horizon correction.

What it actually did

Going through the Edit panel, it had chosen a profile and made adjustments right through the Light section. In the curve it had done one of the things I love to do, which is lift the blacks so they aren't a deep zero black, and it did that on a point curve. It used the colour mixer, and in the colour grading it worked on the shadows and midtones as well as the blending and balance. It used denoise too, and it built all those masks.

Hand on heart, that's pretty much how I would have edited it. I wouldn't leave it exactly as it is, I'd go in and tweak a few things, but it got me about 90% of the way there in roughly 10 minutes.

Try it with your own photos

If you'd like to give this a go, the detailed prompt I used is free to download. Head over to the classroom in my free Photography Community on Skool and you'll find it under YouTube download files:

https://www.skool.com/photography-community/about

Where this leaves us

There's been a lot of resistance to letting AI edit our photographs, and for good reason. When it generates the picture for you the quality drops, at the moment anyway, and you can't go back in and adjust what it's done. This is very different; it's using my software to make the same manual adjustments I'd make to my own raw files, at full quality, and every one of them can still be changed.

It's not something I'd do myself, because editing is a big part of how I make a picture mine and I just love the process. However, if you're new to editing and don't know where to start, or you know what you want but not how to get there, this is a bit like having someone who's really good with Lightroom sit next to you and show you what they'd do. You can watch every move it makes and undo any of them, and I think there's real potential in that as a way of learning.

Capture to Print: One-Light Complete Portrait Workflow

For the past three months I've been touring with Wex Photo Video around the UK on their Summer Roadshow, showing how I take portraits with minimal kit in limited space. I came away with portraits I'm really happy with from every date, but there was one look in particular I was asked about more than any other, so I asked my friend Michael if he'd sit for me and we went outside with one small flash to make it again.

The video above goes through the whole thing, from the shoot, through every step of the edit in Lightroom and Photoshop, to a finished print. Here's an overview of what's in it.

One small flash, outdoors

The idea is simple. Before the flash comes into it at all, I set the camera so that it doesn't see any ambient light; take a picture and you get a black screen. That way the only light in the final shot is the light I put there.

Outdoors in bright sun that takes a bit more doing than it does in a store. I ended up at f/16, which was as far as the lens would go, around ISO 125 and 1/250th of a second, keeping within the flash sync speed rather than reaching for high speed sync or ND filters. I did wonder whether a 100 watt second light could cope with sun that bright, but it wasn't a problem.

The kit was:

• the Westcott FJ100 on power 9, so pretty much full power

• a 15 inch octa with the grid fitted

• a pistol grip handle to hold it steady

• the camera on a tripod

• a reflector just below chest level

The camera goes on a tripod so Michael has one fixed point to look at, rather than following me around as I move, and I trust the eye autofocus to nail the focus so I don't need to be right behind it.

How the light falls

I raise the light nice and high and get Michael to lean forward a little towards the camera. Because of the grid, the light doesn't go everywhere; it comes down and skips across the front of his face, and it doesn't spill onto the background, which I wanted to keep dark.

What I really like about it is that the light is so directional you get shadow either side of it, so it's almost like having a strip of light running down the front of the face. It's a look I definitely want to work more on.

The one thing that made the biggest difference was the reflector. Flat and horizontal, it did very little for the eyes. Lift the far end so it's angled up towards him and the light bounces back into his face, and suddenly there's plenty of detail in the eyes to work with.

The edit in Lightroom

Only one job happens in Lightroom, and that's the eyes. I use masking to select the iris and pupil, finesse the mask with the feather and edge sliders so it sits just inside the eyes, then lift the exposure slightly and add a little texture, clarity and a touch of dehaze. At 100% the difference with the mask on and off is night and day.

After that the image goes over to Photoshop, because that's where the heavy lifting gets done.

The heavy lifting in Photoshop

My rule of thumb here is to look at the image and fix whatever jumps out first. On this one it was the shine down the middle of Michael's nose, which I took out with the Remove Tool on a new layer, then lowered the opacity to around 50% to bring some of the natural shine back. After that I used the Spot Healing Brush on a blank layer to clean up the more obvious marks in the concrete wall behind him, which was sharper than I'd like because of shooting at f/16.

Most of the work is dodging and burning on a 50% grey layer set to Soft Light, with the Dodge Tool at 10% exposure and the Option key (Alt on Windows) held down to switch to the Burn Tool with the same settings. I brighten the highlights and darken the shadows either side of them, around the eyes, down the nose, the eyebrows and the mouth, and then the folds in his top. It sneaks up on you; you often don't see it until you turn the layer off. Working on a grey layer rather than the image itself also means that if you go too far, you can paint over it with 50% grey and it's gone.

From there it's a handful of smaller things:

• a loose lasso round the teeth and a Hue/Saturation adjustment on the yellows, lifting the lightness slightly so they're cleaner without looking like he's had them done

• my 2010 technique, two passes of Unsharp Mask on merged layers, amount and radius at 20 then 10 with the threshold at 0, each painted in through a black mask, the first over the middle of the face and the second over the eyes, nose and mouth

• a skin texture effect using Reduce Noise at full strength with everything else at zero

• a High Pass layer at a radius of 2, set to Hard Light and placed on top, to bring back the sharpness in the eyes, hair and skin detail that the texture effect softens

The 2010 technique is superb on close-up portraits; it makes the face look as if it's coming forward off the screen, and even more so in print.

Camera Raw and a finishing touch

In the Camera Raw filter I used Lens Blur to soften the concrete wall and give the picture a bit of fake depth of field, brushing blur in around the top of the head, the ears and the shoulders so it feels more like f/2.8 or f/4 than f/16. I still wanted to be able to just about see the wall, so it isn't blacked out completely.

A linear gradient with the exposure brought down takes the edge off the bright top he's wearing, which was pulling my eye, and a radial gradient in the middle of the face at +0.2 exposure adds a little light where it matters. I do that last one on pretty much every portrait.

The finishing touch is a warm colour grade from Optics by Boris FX, using the Titanic look in the Portrait colour section, brought down to 50% opacity.

Making the print

I print pretty much every image I'm happy with onto A4, and this one went onto Fotospeed Natural Soft Textured Bright White 315, a fine art matte paper and one I use most of the time, on the Canon PRO-300.

The steps are:

• duplicate and flatten the image

• resize to 10 inches on the long edge at 300ppi

• soft proof with the paper profile, comparing relative and perceptual rendering intents

• a slight brightening just for the print

• save as a TIFF in Adobe RGB

• print from Canon Professional Print & Layout, with Heavyweight Fine Art Paper as the media type, the paper's ICC profile, perceptual rendering intent and black point compensation on

For this picture perceptual was the clear choice. Relative blocked the blacks up quite a lot, whereas perceptual gave it more contrast, which suited it.

Straight out of the printer it was looking really good, and on a matte paper like this it only gets better once the ink has had time to dry and settle. You can just about see detail in the back wall, which is what I was after, and the 2010 technique does exactly what I hoped, bringing the face forward off the page.

Why I always print

Our screens are fantastic, but it's not until you have something physical in your hands that you can really study it and start noticing the things you missed while editing. I truly believe printing can help you get better as a photographer, and certainly better at retouching.

If you'd like to follow along with the video step by step, the raw file I used is free to download from the classroom in my Photography Community on Skool:

Adobe Now Owns Topaz: Is This GOOD News or BAD?

On the 23rd of September 2026, Adobe completed its purchase of Topaz Labs. If you use Photoshop, Lightroom or any of the standalone Topaz applications, the obvious question is: what does that mean for us?

Will we get better tools? Will more of the technology be included in what we already pay? And what happens if you prefer using Topaz on its own?

There are some answers, but there’s still plenty we don’t know. In the video above, I share my thoughts on the acquisition, where I can see some real benefits, and what I’d personally like Adobe to do next.

What Has Adobe Confirmed?

Adobe announced the proposed acquisition in June. The September announcement confirms that the deal has now gone through.

Adobe says the Topaz brand will remain, and its applications and models will continue to be available as standalone products. If you use Topaz separately, perhaps alongside a different editor, that’s encouraging.

It’s also worth clearing up the subscription question. Topaz introduced its newer subscription offering in September 2025, before this acquisition. Whatever your feelings about subscriptions, that change wasn’t introduced by Adobe buying the company.

Has My Opinion Changed?

Earlier this year, I made a video about the Topaz integration in Lightroom, and I was frustrated.

At the time, it felt as though Adobe could be taking the easier route by bringing in somebody else’s technology. Given Adobe’s experience and resources, I wanted to see what it could develop itself. I still would.

But I can also see the reasoning behind this purchase. Topaz has specialist knowledge and people who have spent years working on image and video enhancement. Bringing that experience into Adobe could be a sensible way to improve the tools we use.

That’s something I probably could have given more weight in my earlier video.

There is, however, one point I still stand by: when I’m working in Photoshop or Lightroom, I want the tools to feel like they belong there. I want them to work together, follow a consistent approach and give me enough control to get the result I’m after.

Changing who owns the technology doesn’t automatically achieve that. It takes development.

Where Could This Benefit Photographers?

The most obvious opportunity is a better editing experience.

Say you’ve cropped a photograph quite heavily and want to make a print. Having a suitable enlargement tool available inside the application you’re already using could be really useful—especially if you can judge the result properly and adjust the processing without a complicated round trip between applications.

That’s the kind of integration I’d welcome.

Adobe doesn’t have to invent every piece of technology from scratch for it to be valuable. What matters is whether bringing these teams together produces something better when we sit down to work on a photograph.

There’s also the possibility of more processing happening on our own computers. Adobe has highlighted Topaz’s NeuroStream technology, which helps demanding AI models run locally.

Depending on how Adobe uses it, that could reduce the need to upload files for processing. But it isn’t a promise that every feature will work offline, or that every computer will suddenly process everything quickly. We’ll have to see how it develops.

Give Us More Control

For me, one of the most useful places to start would be the controls.

The integrated Topaz tools haven’t offered the same experience as the standalone software. The controls have been limited, with what’s available differing between applications and models.

Having the technology available is useful, but I also want some say in the result.

We’ve all had that moment when an adjustment gets close, but isn’t quite right. Perhaps the effect is a little too strong, or some of the detail doesn’t look as natural as we’d like. That’s where being able to refine the processing makes a difference.

It doesn’t have to mean putting every possible slider on screen. A simple option is useful when it gives you what you need. When it doesn’t, I’d like the option to go further.

Adobe hasn’t announced that the acquisition will bring the extra controls I’m asking for. It’s something I’d like to see.

How Does It Fit Into the Lightroom Workflow?

In Lightroom, Generative Upscale creates a separate DNG file, leaving your original photograph intact. Adobe recommends applying it at the end of your editing workflow.

Understanding that you’re creating a new file matters when deciding how to fit the tool into your editing process. Over time, I’d like that process to become as flexible and clear as possible.

There’s also an important distinction between Lightroom and Lightroom Classic. A feature being available in one doesn’t automatically mean it’s available in the other.

At the time of writing, Adobe’s published FAQ says there are no plans to bring this particular Gigapixel integration to Lightroom Classic. That refers to the built-in integration; it doesn’t mean you can’t use the separate Topaz application with Classic.

What About Generative Credits?

This is where the practical questions become particularly important.

At the time of writing, Adobe’s published charges for the integrated tools are:

ToolFinished output sizeCredits per generationTopaz Gigapixel in Photoshop or LightroomUp to 25 megapixels10Topaz Gigapixel in Photoshop or LightroomAbove 25, up to 56 megapixels20Topaz Bloom in Photoshop1–9 megapixels35

These are Adobe generative credits for the integrated features. You can check the current rates in Adobe’s credit table.

If you have several thousand credits each month and use these tools occasionally, that might be perfectly manageable.

But imagine you have 200 credits remaining. At 20 credits a go, that gives you ten generations, assuming you don’t spend any elsewhere.

That’s ten attempts, rather than necessarily ten finished photographs you’re happy with. Generate another version and you use more of your allowance.

The 200-credit figure is just an example. Your allowance and access depend on your plan, so check your own account.

For me, this connects directly with the question of control. When each new generation uses part of your allowance, being able to get a result you’re happy with becomes particularly important.

Could Ownership Mean Better Value?

When one software company uses another company’s technology, there’s usually a commercial agreement behind it. That might involve licensing fees or usage charges.

We don’t know the details of Adobe’s previous agreement with Topaz, so I can’t say how much of the credit charge reflected that arrangement.

But now Adobe owns Topaz, it’s reasonable to ask whether that changes the calculation. Could fewer credits be required each time? Could more usage be included in our existing plans?

There are still costs involved in developing and running the technology. Ownership doesn’t make those disappear.

The acquisition announcement doesn’t promise lower credit charges, either. Better value is something I’d like Adobe to address, rather than something we’ve been told to expect.

Clear pricing, a practical allowance and more control over the result would make this integration more useful.

What I’d Like to See Next

I can see why Adobe has bought Topaz, and I can see opportunities for photographers to benefit.

Closer development between the teams could improve the tools. Local processing is interesting. Better integration could make editing easier. And continuing to offer standalone Topaz products preserves a choice that matters to plenty of photographers.

For me, the test will be what Adobe does next.

I want to see useful improvements, more flexibility and tools that feel like a complete part of Photoshop and Lightroom. If Adobe’s resources and Topaz’s expertise produce that, I’ll be very happy to give them credit.

What do you think? Would you like more of the standalone Topaz controls inside Photoshop and Lightroom, or does the simpler approach already give you what you need? And if you use Topaz separately, what would you most like Adobe to preserve or improve?

Let me know in the comments.

Details reflect the published information checked on 24 September 2026. Features, credit charges and plan allowances may change.

How to Fix Photoshop Scratch Disk Full (once and for all)

This warning has popped up a few times recently when I’ve been using Photoshop, and if you’ve seen it too, you might be wondering what a scratch disk actually is… and what you’re supposed to clear.

So, to accompany the video, I thought I’d put the steps here so you’ve got something to refer back to. There’s also a way to change your scratch disk if Photoshop won’t even open.

First, check which drive Photoshop is using

Before doing anything else, save your work if Photoshop will let you.

Then head to:

• Mac: Photoshop > Settings > Scratch Disks

• Windows: Edit > Preferences > Scratch Disks

Look at which drive is selected and how much space is available. That’s the drive we need to check.

You might have another drive connected with loads of room, but unless Photoshop is configured to use it, that space isn’t helping.

What exactly is a scratch disk?

Think of it as extra working room.

Imagine working on a print at a table. The print itself might be quite small, but you still need space around it for everything you’re using.

Photoshop needs that working room too. It uses space on a drive to store temporary information whilst you edit, so the size of your saved photograph doesn’t tell you how much space Photoshop might need to work on it.

There’s also a difference between a warning that space is getting low and an error that stops you opening Photoshop, editing or saving. A warning doesn’t necessarily mean the drive is completely full, but dismissing it doesn’t give Photoshop any more room.

Free up some space

Adobe recommends having at least 100 GB free on your primary scratch drive when troubleshooting this error. That’s guidance rather than a guarantee that every job will fit. Larger projects may need more.

Start with files you recognise: old video exports, downloads or projects you’ve already finished.

If you move them to another drive, check that the copies are there and open properly before removing the originals. And remember, putting files in the Trash or Recycle Bin doesn’t immediately free up the space. Review what’s in there before emptying it.

Once you’ve made room, save your work if you can, quit Photoshop completely and reopen it. Then try whatever you were doing when the warning appeared.

Give Photoshop another drive to use

If storage on your computer is always tight, another option is to give Photoshop a suitable drive with plenty of available space.

I use an SSD as a dedicated scratch disk.

Go back into Scratch Disks, select the drive and, if you’ve selected more than one, use the arrows to set their order. Adobe recommends putting the fastest drive first.

Click OK, then restart Photoshop.

Oh, and changing this setting doesn’t move your photographs. You’re choosing where Photoshop can put its temporary working data. If you’re using an external SSD, keep it connected whilst Photoshop is using it.

What if Photoshop won’t open?

You can’t get into Settings if Photoshop refuses to launch, but there’s a shortcut for that.

Hold these keys whilst launching Photoshop:

• Mac: Command + Option

• Windows: Ctrl + Alt

This lets you choose a scratch disk during startup. Select a suitable drive with space available, then continue opening Photoshop.

Don’t add the Shift key. That’s the shortcut for resetting preferences, which we’ll come to shortly.

Of course, if your only drive is full, the shortcut won’t create any extra space. You’ll need to make room outside Photoshop first.

Still getting the warning?

First, double-check that the free space you’re looking at belongs to the drive selected in Photoshop.

Then try quitting Photoshop completely and reopening it. Its temporary working data can build up during a session, and closing your photographs doesn’t necessarily release everything.

A restart can help, but it won’t solve a drive that’s consistently running out of room.

It’s also worth checking History States, under Settings / Preferences > Performance > History & Cache.

This controls how many previous editing steps Photoshop keeps available for you to return to. Keeping more can require more scratch space.

Adobe’s documented default is 50. If you’ve increased yours to several hundred, consider whether you really need that many. Reducing the number may help, but you’ll have fewer steps you can go back to.

Be careful with Purge

You might have seen advice to choose Edit > Purge > All.

This can release memory and scratch space, but understand what it does before clicking it: purging clears temporary editing information, including your history.

It doesn’t return your photograph to its original appearance, but you lose the ability to go back through the history you’ve cleared. You can’t undo the purge.

Save your work first if possible, and treat this as an optional troubleshooting step.

Resetting preferences

If you’ve checked the correct drive, made room and restarted, but the error keeps returning, resetting Photoshop’s preferences is another troubleshooting option.

This can help if damaged preference files are contributing to the problem. It doesn’t create storage, and it can remove your custom settings, so back those up first.

With Photoshop open, go to Settings / Preferences > General > Reset Preferences on Quit. Confirm, quit and reopen Photoshop, then check your scratch disk selection again.

If Photoshop won’t open, hold these keys whilst launching:

• Mac: Command + Option + Shift

• Windows: Ctrl + Alt + Shift

That’s the shortcut with the extra Shift key. You’ll be asked whether you want to delete the Photoshop settings file.

Start with the Scratch Disks panel

The main thing is to check which drive Photoshop is actually using and how much room it has. Then free up space or give it another suitable drive to work with.

Make one change at a time and repeat whatever caused the warning. That way, you’ll know what helped.

The video above takes you through where to find everything.

Right, I hope that’s useful.

Glyn

Who is Photoshop for now? 🤷‍♂️ (update 27.10.0)

Here's a question I've been carrying around for a while now, and I still haven't got a tidy answer for it …

Who is Photoshop for now?

For most of the years I've been using it, Photoshop has been the place I go to disappear. I'll open an image with a rough idea of where I want to take it, start working, and the next time I look up properly two hours have gone. That isn't just productivity, it's creative escapism, and I think that deep flow state is why so many of us fell in love with the application in the first place.

Lately though, opening Photoshop feels a bit different. It feels like it wants something from me. It wants me to write a text prompt, it wants me to keep an eye on my generative credit balance, and it seems remarkably keen to show me what it can do without me.

I want to be clear about where I'm coming from. I love Photoshop; it's what I do, it's what I built my career on, and I'm not going anywhere. But when you genuinely care about a tool you pay close attention to the direction it's pointing in, and looking at what's arrived in recent releases like version 27.10.0, I think it's worth having an honest conversation about where things are heading.

Two interfaces and a bit of an identity crisis

There are a few significant shifts in the interface in the newer updates.

The contextual taskbar now has a dedicated Prompt to Edit button, so instead of making a careful selection first, you can prompt changes right across an entire image using various AI models.

Then there's the AI Assisted Editor workspace, which you can enable from the menus. Switch to it and Photoshop strips away your layers, your familiar tools and your adjustments, and replaces the lot with a prompt driven layout that looks and behaves like Adobe Firefly sitting inside Photoshop.

There's also the Pro Editor label. The workspace you and I have used for decades, with its layer stacks, blend modes, masks and smart objects, is now officially labelled the Pro Editor to tell it apart from the AI Assisted mode. Photoshop never used to need a qualifier; it was just Photoshop.

I do understand the logic from Adobe's side. The full Photoshop interface is intimidating for beginners and a simplified prompt box lowers the barrier to entry, but turning the whole canvas into a prompt box disables the very foundation that makes Photoshop worth learning in the first place, which is non destructive layers, precise pixel control, and being able to unpick your creative decisions months down the line.

If someone purely wants to generate an image from text, they've already got dedicated web tools and mobile apps for that. What they don't need is desktop Photoshop with its core strengths stripped out.

Resolution and control

Beyond the workflow philosophy there are some practical hiccups when you lean on whole image generative edits.

Run a full prompt, asking the software to add weather effects say, and it generates an entirely new flattened layer. Even with the recent improvements to generation resolution, the output is often softer and lower in detail than your original camera file.

If you then try to mask through to recover the original sharpness on your subject's face, you run into problems fairly quickly. Perspective shifts, subtle facial alterations and mismatched lighting mean the generated background no longer lines up cleanly with the original capture, so you end up spending extra time fixing blend modes, manually aligning shifted anatomy and correcting colour mismatches, just to make the composite work.

The credit counter

There's another subtle shift that changes how the software feels to use, and that's the generative credit counter.

Adobe have made the right decision by being transparent and showing you exactly what a generation will cost before you press the button; hiding the cost would be far worse. But seeing that meter tick down does introduce a psychological hesitation.

You pause before generating. You wonder whether this variation is worth spending a credit on or whether you ought to save it, and if the result comes back slightly off, you've then got to decide whether a second roll of the dice is worth the cost.

In twenty years of editing I've never opened a Curves adjustment and wondered whether I could afford it. I've never dodged and burned a portrait with a bill running in the corner of the screen. Creativity thrives when you're free to experiment, fail and try again without an active meter running.

The quiet lane and the loud lane

I don't want any of this to sound like a complaint about modern features. In recent updates to Camera Raw, Adobe introduced a new Glow adjustment covering diffusion, bloom and halation. It doesn't need a prompt. It's practical, it's well engineered, and it's built directly for photographers and retouchers, which is Adobe at their absolute best.

Historically Photoshop has moved forward in two lanes at the same time. There's the loud lane, the flashy headline grabbing features that generate all the excitement, and there's the quiet lane, the foundational unglamorous refinements to core tools, performance, selections and menus that quietly save you ten minutes on every job for the next five years.

My concern isn't that generative AI exists in Photoshop. My concern is that the quiet lane slows down because all the engineering focus goes into the loud one. We need generative tools when they genuinely speed up complex retouching work, but we also need the core engine to keep advancing, and we shouldn't have to choose between modern AI assistance and the deep precise craft of manual image making.

I'd be interested to know where you've landed on it, whether the generative tools have changed the way you work, or whether you're staying firmly in the layer stack.

Reflection Removal NOT WORKING in Lightroom? Try this

The Reflection Removal tool in Lightroom and Camera Raw is a brilliant addition, but it does have its limits at the moment, and that comes down to the way it's been designed. As things stand it's built for windows that are the dominant part of the picture, so if the glass you're working on is only a small part of the frame, you can hit apply, watch the bar do its thing, and end up with an image that looks exactly the same as it did before.

Someone in my photography community got in touch about this. He had a raw file of a cabinet indoors, he'd followed all the usual steps, and the reflections just wouldn’t shift.

However, there's a good reason for that, and there's a simple way round it.

Why it ignores some reflections

Reflection Removal sits under Distraction Removal, and right now it's looking for a big window; a properly dominant one, filling a good chunk of the frame.

Put it on a glass display case, or a pair of glasses, or a window that only takes up a modest part of the scene, and the algorithm often does nothing at all. It isn't broken; it just doesn't think there's anything there worth working on.

I'd expect that to change. These tools tend to get better quickly, and smaller areas will almost certainly be handled in time. For now though, we take a slightly different route.

The Photoshop route

Worth saying first, doing it in Camera Raw or Lightroom doesn't use any generative credits, which is one of the nicer things about it, but taking the file into Photoshop and using AI will use a credit, but you do get total control over what happens, and the result is clean.

It works like this.

1. open the image in Photoshop, which from Lightroom is a right click and Edit In > Adobe Photoshop

2. grab the Polygonal Lasso Tool, or whichever selection tool you get on with, and draw around the glass with the reflections in it

3. click Generative Fill on the Contextual Taskbar

4. in the model picker, choose Firefly Image 5

5. type in the prompt: remove the reflection from off the glass do not change the original content

6. hit Generate

That's it.

Firefly Image 5 is the part that matters here; it's designed to edit while holding on to the detail that's already in the file, so Photoshop isn't off inventing new objects and quietly swapping your subject for something it made up.

What comes back

Because it's a detail-preserving model, you get one result rather than the usual three variations to choose between.

The reflections go, and the things that were sitting behind the glass all along, book spines, shelves, whatever happens to be in there, come through properly. It looks very much like you'd screwed a polariser onto the front of the lens and dealt with it at the time of shooting, which is the bit I really like about it.