Sharpen AI v4.1.0

Thank you. I’m afraid I can’t identify a blue callout on the first page of the official website. Would you be sooooo kind as to send me a screen shot of the callout you refer to?

Make sure popups are not blocked by your browser … you can see it on every page of the website …


This isn’t ON1, he is talking about Topaz Sharpen AI.

Stupid of me. I have deleted the post

1 Like

Here’s what the button looks like in Safari. Note that I have pop-ups blocked for Topaz but it is still showing.

I think I got it right this time!

So many thanks: indeed my browser was blocking these popups. You made my day…

Have I found a new bug in version 4.1.0 (Windows)? When I change the brush size in “Refine” of the selection, the opacity of the selection changes to the same value!

Thank you for your answer. That shows us, how intrested you are in your customers.

Just raise a support request at the main website… this is a user to user forum.

1 Like

Hi… downloaded the trial version of Sharpen AI with a view to buying… I have been testing all day various .dng files in the stand alone by just dropping the image into the program… does a decent job of removing noise and sharpening.

however on saving (as .dng) I get a much brighter image than the original with lots of artifacts in the shadow areas that look like nasty colored noise similar to when brightening an underexposed image… I am willing to supply the specific image for your testing… tnx /p

I suggest trial Photo AI, it has updated Sharpening, as Sharpen AI is not going to be updated according to Topaz Labs,


I tried Photo AI also but I need to buy the licence before I can export to see the results in Lightroom,… after my experience with Sharpen AI exporting a brighter image than the original I am not prepared to buy a program just to see if it will do the same then have to request refund…

hence I am leaning towards DXO photolab with is about $60 more than Photo AI… and I can export without any watermarks for the next 30 days… Topaz should have enabled exporting with watermark insted of forcing me to buy the program before I can export.

1 Like

You do have a 30 money back guarantee, no questions asked.

BTW DxO Photolab is a good option also, I just finished a trial but I would wait to see what v7 brings very soon.


Is this for certain? No more updates to Sharpen AI? If so, disappointing!!!


Looks like Gigapixel, Sharpening and DeNoise have been discontinued and won’t be updated. Topaz says Photo AI does a better job, but I don’t agree.

Just had Sharpen AI for about 10 days now.

Generally it dies a good job but is very slow, for example a 50mb file, takes around 8 minutes, whilst same file on Nik Efex takes about 4 seconds. The results are very similar, unless there is a problem with the image, then Sharpening AI, is better.

There are a few issues with TSAI, it does not sharpen everything, say on a wooden post, it won’t sharpen near the top, it also shifts or adds colours, sometimes.

If this won’t be updated, then I will probably get a refund, I don’t like Photo AI, I’m sure it will suit some users but I’m leaning towards DxO at the moment.


I’m getting colours now appearing in my images, when I sharpen B&W images. This was reported back in 2020, but is still happening now. Maybe that is the issue on Gigapixel, where I have a similar issue.

If I edit a photo in Lightroom, then Photoshop, then Topaz Sharpen and save, then save in Photoshop so it ends back in Lightroom, how do I then bring back that photo back into Topaz Sharpen to see what Topaz settings I originally used? I am comparing several sharpening methods and I need the info to make the comparisons.

I don’t use Lr. So someone else may have a different answer b/c you’re meandering through a very circuitous route via multiple hosts and plugins. But here’s my suggestion just for the Photoshop (Ps) portion of your post-processing travels…

Once you have you image opened into a Ps layer. Convert that layer to a Smart Object (personally, I’d duplicate the layer that came from Lr & use the duplicate as the smart object launch layer…) before launching the Sharpen AI (SAI) plugin. Make your SAI plugin settings adjustments and save back to the Ps layer. Once your SAI adjusted layer is back in the Ps layer stack, save the layers stack (don’t flatten it!) as a .PSD doc. Then go back to Lr.

If you save the entire Ps layer stack (that includes the Smart Object layer), you should be able to double click to re-launch SAI (from inside Ps) & it should (that’s how smart objects should work in principle) maintain the settings you used for you to see what they were. Once you flatten the stack, you lose the ability to relaunch your layer back into SAI plugin with your prior settings maintained.

.PSD files are large. So unless you have tons of storage, you may not want to do that for every image and keep those .PSD files for a long time. I suppose it depends on the quantity of images you process and may want to revisit the settings for.

Anyway, that’s my take and a suggestion to try - if I’m understanding your question. Ciao!

It crashes on me a lot. I use it as a plug-in to Lightroom Classic and the crashes seem to occur more frequently when I use one of the STANDARD sharpen models. I haven’t had any crashing problems with Photo AI, so sometimes, I just use the sharpen feature there.

Does anyone know what the advantages/disadvantages of using sharpening in Sharpen AI vs. Photo AI? Is Topaz Labs continuing development and support of Sharpen AI or are most of the development resources going to Photo AI?

UPDATE: after searching through the community messages, I came across this comment from a Topaz Lab developer:

“Photo AI is the only one receiving regular updates. Our plan is to improve the Photo AI performance to be better than the individual apps combined, plus some additional features as well. We appreciate your understanding while we build this app for you.”

So, it looks like the individual apps are end-of-life and will no longer get any updates (or at least regular updates, maybe they’ll still do updates for bugs).