It Seems All a Bit Upside Down at The Moment - AI and Topaz

As much as I love Topaz…

Photo editing tools traditionally are judged by their end result, how easy they are to use, what amazing differences they make to the image, that sort of thing. Who really cares about the coding process if it works? Topaz have always been good at this, making things work - until recently. Puzzlingly, the focus has been turned on its head and has shifted away from wooing customers with amazing end-result features to worshiping the actual process of how it is put together, putting it under the epithet of the much overrated AI moniker. Which is misdescribed, as artificial intelligence does not exist. Not yet, anyway. What Topaz and a dozen other photo software developers are actually using is Machine Learning and calling it AI.

AI apps are huge, they take longer to download and install, they gobble up resources and are slow to use. But are they they really any better? I’m not seeing any hard evidence that they are. Studio 2 is a constipated stripped-down version of the original and, despite all the updates, those amazing end-result developments have practically stalled. Mask AI is not faring much better, and AI Sharpen produces reasonably good results if you have all day to wait for it to process, very slowly, one image at a time.

But, looking on the bright side, I bet the coding is a joy to behold!


I was in the IT realm for the last 22 years of my career and I noted that if a company is not pushing the envelope with software development they will not retain the best and brightest developers. Profit and cash flow, however, is also key to a companies success. Management pushes for quicker release cycles and testing often takes a back seat. Here where there are almost limitless combinations of hardware and installed software Beta testing takes on a very key role. It is a balancing game in a highly competitive environment.

1 Like

I think most software companies release their product before it is finished, look at Microsoft for an example. I have seen this with many other software companies as well, probably because of competition, as well as recovering R&D costs.
I miss the layout of TS 1 being able to bring in multiple images to blend or just compare each result. Mask AI needs to add tools to mask glass and sheer fabric like ReMask and add Mask AI to TS 2. I can understand why Topaz has changed their format. Free doesn’t pay the bills. I still love Topaz and will continue to use their products.


I was a senior developer In the publishing and printing industry. If we didn’t deliver what customers wanted we didn’t get customers. They rarely gave a toss of how we got there, only that we did, and losing sight of the end product was a fast track to disaster. Just saying.


Maybe not so much… And, I’d love to know how much of the installed (& prospective) base of Topaz users are high-end gaming machine users. I’m sure Marshall McLuhan (sp?) would have something to say about how things have evolved (devolved?).

I wonder when the remaining features of TS1 will be added to Studio 2?


Couple of things here for clarity…

  • Machine learning is a component of AI
  • Studio 2 is actually ‘Artistic’ application and not AI based

And for those not appreciating AI: Mask AI, GigaPixel AI, DeNoise AI and Sharpen AI ARE best in class. And the best comparisons you can make are with the ‘supposed’ best out there such as DXO Prime Noise Reduction and the recently released PS Sensi object selection.

AI is here to stay and, with machine learning, will only get better … the challenge will always be getting AI to apply that which is no more than necessary. I guess the best example of ‘more than necessary’ is the luma NR with the AI Clear model.


I see you have changed some of the wording in my post, Don. As much as I respect your opinions I wish you wouldn’t do that.

1 Like

Didn’t change any wording in your post. Changed the title to something meaningful to all and changed the category.

In changing the title you altered the perceived purpose of the post to an attack on AI, which it isn’t. I wanted the post to contrast the emphasis on AI against the emphasis, or seeming lack of emphasis, on development.


I have added your words back in the header, but what I changed gave the actual discussion meaning. As for the rest I will let others judge what the meaning of the post is.

1 Like

Thank you. In your role as moderator, you have a duty to remain neutral.


I am still using Studio ! because of my reliance on the multiple images and Image Layer too.
We seem to be i a brave new world with software and the rush to release when something isn’t ready. I doubt we would buy a new car when the engine wasn’t yet working properly.


Agree. If a new customer came to Mask AI now they are likely to pass it by and look elsewhere - and perhaps never return to see where it finally got to. They may also not think much of the company to check Topaz products again either. Strange business logic.
If Topaz’ cash flow has been suffering from their promise not to charge for upgrades (which all of us seem to be seeing this as), perhaps they should not offer that deal with anything completely new they may come up with in future rather than damage consumer confidence with what has happened over Mask AI


The fact is that that was the deal they did make with the customer. And it worked well to build a customer base that was not only impressed with their product - they were a very loyal legion that passed the word around to the photography community which brought more photographers to the fold. Topaz’s reputation for greatness may be short lived.


I am a long time Topaz user, and I own all of their original software up to Topaz Studio. I never bothered with Topaz Studio because I did not see any need to put another program between Photoshop and my original Topaz programs. All my original software is still running fine with Photoshop (now up to version 2020) and to quote Buzz Lightyear, I can still take my photos “To infinity and beyond” with those original Topaz programs.

Of course I understand that Topaz needs to expand their business and customer base, but I’ve had my doubts about the company direction since they adopted the Studio concept. I visit this website almost every day and at least half of the posts are people struggling with various problems ranging from installation to upgrades.

So, I wish Topaz much success but it’s clear that they need to improve their quality control to minimize all these technical problems. And I can’t help but feel they have cornered themselves into a box with their Studio concept.


I’ve just noticed this post on here and I have to say I’ve come to much the same conclusion as both the original poster and other subsequent posts with the change to AI based software. Like others who have posted on here I’m also a long time user of Topaz software and have tried the AI versions as they’ve become available but currently I’m underwhelmed with their performance.

I’ve found AI is slow and it rarely seems to deliver the same quality of result previous versions of the non AI software did, particularly AI Noise reduction which I’ve stopped using for the time being until it improves both in terms of speed and output quality. My other gripe with AI driven photographic software is your limited ability to adjust the output at which point its easier to do it using software giving you full control over the output.