Refocus on core features (Gigapixel/DeNoise/Sharpen) - No more fancy features

Is Topaz completely in Holidays? :thinking:


Feels like all of their resources are in marketing/sales…
Black Week / Black Friday / Cyber Monday / Tuesday whatever…
Infusions of fresh blood :drop_of_blood: :wine_glass:


Thanks for the 100th vote :+1:

1 Like

The latest update seems to be causing some problems.

Not done it yet, as not been doing any editing recently.

Seems each update fixes a few things but causes other issues.

I find the quality of Sharpen AI and DeNoise AI results is superior in every way to the mish-mash, hit-and-miss results from PhotoAI. This all-in-one product is a disappointment in use on my OM Systems and Leica images.

On my 5k iMac screen I see modest benefits at best from PhotoAI no matter whether I run it automatically or under my control nor whether I process the RAW or Lightroom-modified TIFF. Too often the software over-sharpens or fails to find the (admittedly minor) noise I can confidently expect DeNoise AI to deal with (well). I can only assume it’s meant for phone-photographers.

I’m holding onto the earlier products while they still work and have no intention at this stage of renewing a subscription to Photo AI. I’m also tired of the continual minor updates that seem to me to bring no discernible improvement in quality.


That is because of the unreliable “AutoPilot”.
The AutoPilot often pick wrong model or setting, make the image worst than it should be.

1 Like

I also miss the multi-preview feature of the old products. Where you were able to see how the image looked in 4 different models at one glance… Now you have to click each model, keep the result “in the back of your mind” and click the next model…


Quite right, this is a big step backwards: the autopilot can be remedied by deactivating it but there is no remedy for the functions no longer present

Finally in version 2.2.0 a new feature :open_mouth::
“This week we’re introducing a new masking tool - “Object Selection Brush”. This brush is intended to be a point-and-click masking tool that provides quick and accurate results. Currently there is no on-hover mask preview for the tool, however this feature is planned for a future release.”
As always, half implemented with the promise that it will be completed in future releases! :face_with_symbols_over_mouth:

1 Like

PAI is just continuous Beta testing. I have never known a software with so many problems.

AutoPilot does not work. It does not select the correct model for sharpening, or noise, unless it happens, like a broken clock, to be correct twice a day.

Why does AutoPilot take so long to run and produce poor results?
I think it has got worse.

I agree. Even old PAI 1.3.2 had better results, people looking like people not Barbie and retaining much more detail and less strange artifacts. I was hoping that they include some way that the user can fine tune the PAI Autopilot (like the sliders they added in the preferences, which sadly don’t seem to do much). If I was able to get results similar to 1.3.2 in 2.x I would be happy, since I could use the CLI and so on.

Simple comparisons with the (in)famous 2001 picture of photo_jmr *g

I think that a serious company should make a plan for the next versions, decide which contents to develop, plan the necessary tests and decide the release date. If the release date is not respected, it is a failure for the company.

I agree with most of what you say, especially the need to plan and require tests. However, a fixed release date seems like a holdover from old development approaches which were driven by marketing plans rather than product quality. As a customer I would rather have unpredictable releases of great software than reliable releases which may have been rushed out the door by managers “under the gun” to meet a deadline.

Some of the best open source software has a “release when ready” approach, and I don’t see why commercial software can’t be the same.

I completely agree that a product must be released only when “it is ready”, i.e. after a careful design, development and functional and non-regression testing phase. However, I wanted to point out that for a company that needs to make profits, completely different from those who make open source, it is absolutely essential to have a product roadmap in which a release date is set, even with a margin of tolerance: an excessive delay can mean a significant increase in development costs which may lead to a redefinition of the sales price or, in some cases, the crisis of the company itself. For those who develop open source these problems do not exist.
Serious companies do not communicate the availability date of a product, they keep the roadmap reserved for internal staff only where it is often updated, until they are more than confident that they will be able to meet the deadline exported to the public.

I fully agree, I have always used the Topaz products as a plugin in Photoshop and Lightroom and I will continue to use these tools in this way. I do not need a replacement for better Lightroom/ Photoshop functions. Focusing on your core features and trying to offer the best solution for sharpening, denoising and upscaling would be better

1 Like

Absolutely. We use 99% Gigapixel AI here. It’s like included Sharpen AI most of the time and brings wonderful results - timesaver, sometimes livesaver.

Since Adobe has the new AI noise remover in LrC, I use that very often with my RAW filesdirectly in LrC which is great, too. But besides that workflow and for upscaling right now TopazLabs Gigapixel AI is our one and only choice here. So PLEASE focus on that step ahead you have, TopazLabs.

1 Like

If I understand correctly, version 7 of Gigapixel is about to be released: if it is satisfactory (so not a reduced version of Photo AI) and if I have the feeling that Topaz will give attention to this product again, I think I will abandon Photo AI to use Gigapixel exclusively. Let’s wait…

I’m concerned that GPAI 7 is little more than a port of the current PAI tech back into the GPAI shell just to catch it back up to date.

I also suspect that this is the case: I don’t understand the reason other than having a by-product at a lower cost than PhotoAI; at the same time for Topaz it would be another software to maintain: is it worth it?