Photo Roadmap Update (March 2023)

Product Updates

Since last month’s roadmap update, we’ve focused on improving the usability and customization options in Topaz Photo AI:

Basic Brush to edit subject selection

You can now access a new basic brush to manually refine auto-detected subjects:

horse-mask

You can use the Basic Brush alongside the default subject selection and the AI Brush. Basic Brush is enabled when the AI Brush toggle is disabled in the right panel:

basic-brush@2x

You’ll also notice that softness / feathering is applied all at once instead of per stroke. Because the brush is for subject refinement rather than precise selection, it’s intended to be fast, simple, and easy instead of a comprehensive painting tool.

Apple Photos extension

You can now select Topaz Photo AI in the Apple Photos EditExtensions (…) menu:

Edit In

The processed image will replace the original in Apple Photos, but you can undo the edit by clicking the Revert to Original button. You can also first duplicate your photo before running it through Topaz Photo AI if you prefer.

Note: don’t use the right-click → Edit In menu because TPAI will open in standalone mode, which will prevent the finished image from automatically appearing in Apple Photos.

Better cropping

The Crop tool now always stays centered and works more intuitively than before:

CleanShot 2023-03-02 at 18.04.15

We’ve also fixed a ton of edge cases and potential crashes, so the whole cropping experience will now be a lot more reliable.

Enhance Resolution parameters

You can now manually adjust Enhance Resolution parameters for more control over upscaling results:

CleanShot 2023-03-02 at 08.53.18@2x

These settings describe the input image rather than the strength of the effect. In other words, for each image there’s a perfect combination of optimal values that Autopilot tries to detect. If you feel like Autopilot’s not doing a good job, you can now experiment with them yourself:

  • Suppress Noise: the level of color / luma noise in the input image
  • Remove Blur: the degree of softness or motion/lens blur in the input image
  • Fix Compression: the level of compression artifacts like jpeg blockiness and aliasing in the input image

Please let us know if you see Autopilot consistently misapplying settings for certain images.

Smaller faces are now sharper

v1.1.9 introduced a Face Recovery issue that caused unintentionally blurry output with smaller faces. This is now resolved:

Also, Face Recovery strength now defaults to 80 instead of 100, which leads to more natural results without visible compromise in recovery strength. We’ll continue investigating other known Face Recovery issues in the Roadmap.

Other improvements and fixes

  • Added support for more DNG types (HDR, Monochrome)
  • Add compression option for PNG exports
  • Support for opening jif, jfi, jfif, and jpe files
  • Lightroom Classic Plug-ins Extras integration now copies Rating and Keyword data, stacks images, and preserves metadata from sidecar files
  • Fix issues with the command-line interface (CLI) for Windows and Mac

For the full list of changes, please read the individual release notes for v1.2.4, v1.2.3, v1.2.2, and v1.2.1.


Feature Roadmap

We’re now actively working on:

  • New Remove Noise model for non-raw images that improves slider responsiveness, better preserve textures, and offer adjustable sharpening
  • New Sharpen model that produces fewer artifacts
  • Add image straightening to crop tool
  • Improve Autopilot Configuration flow and add Preferences panel
  • Improve raw handling, particularly with rendering colors and metadata issues
  • Resolve occasional blurry patches in Enhance Resolution

Improving the performance of our AI models takes a while, but we think the new denoising and sharpening models will be worth the wait. We’re first evaluating the performance of these models before considering subject-only denoising and more Autopilot Configuration options.

Next up:

  • Improve performance for preview and processing
  • Improve Autopilot decision-making
  • Improve Face Recovery edge cases: “pop-up” faces in the background, eye colors, better blending
  • Explore approaches for more natural-looking text when upscaling

As always, please let us know what you think. If you have images that don’t work well for a certain feature, send us the image and comment below.

We look forward to working with you to make the product even better in the future. Thanks for using Topaz Photo AI!

16 Likes

Hi Eric,
Very happy with new brush. I will be eagerly waiting for Denoise and sharpen models.

2 Likes

Love the new masking brush - thank you!!

1 Like

Thank you. Your software, purchased some days ago, really changed my wildlife photographic experience to the better, since I am so much less afraid of higher ISOs. I already had an allegedly good demonising software, but Topaz beats, in particular its subject detection capabilities.
Speaking of: I still don’t really understand whether and how subject detection is used for differentiating demonising. I
It would be outstanding, if one could control the demonising parameters for subject vs. other areas of the image separately.

2 Likes

This is helpful to know. Thank you.

1 Like

Dumb question, perhaps, but what does the new (March) extension return to Apple Photos? Previously, it was a JPEG. New the image is labeled as “RAW.”

New brush is good addition , we also need edge aware added to the brush,

5 Likes

In Video AI you have option to process more than one thing at a time, can the same be done for Photo AI.

Batch processing is supported in PAI. Do you have something else in mind specifically that you can’t do now?

In Video AI you can have up to 4 processes at once, I usually run 2 as I have 2 gpus. I would like to process my Photo AI batch with both GPUs, 2X faster. Sometimes even 2 process on one GPU can be faster.

1 Like

I’ve been a Topaz Denoise user for a year or so and just started using Photo AI for some less than stellar raw files from a Canon 5D MIV.

What would be great to see is Photo AI saving settings for each file that is processed with the tool. So if files need to be reprocessed in the future you start where you left off.

I processed an entire folder of images and while editing the DNGs in Capture One I realized that I missed the fact that there were artifacts in some images and I had to go back reprocess them.

It would be nice to load that folder of files and have all the settings that were originally used loaded at the same time. By being able to see what corrections were used it would help narrow down the source of the artifacts. Then I could go through and quickly change what need to be changed without having to start from scratch.

3 Likes

If the settings are for the input image, then perhaps it’s more intuitive to use wording that emphasizes the input? Based on the descriptions, something like this:

  • Level of Noise in Original Image
  • Degree of Blur in Original Image
  • Amount of Artifacts in Original Image

Is there any visibility when the new sharpen model will be available?

The new “Standard - Beta” Sharpen model was tested in the V1.2.7 Beta version. However, many users feedback that it is worst then the original sharpen model, that is why developer decided not to include the new model in V1.2.8

If you want to test it out, you can message Topaz staff / Moderator, to let you join the beta group.

Thanks. That’s a pity its is worse…

I would suggest to Topaz to prioritize Sharpen Model improvements above all else.

2 Likes

For me, this is one of the most delicate aspects that needs to be improved!

I am looking forward to reading the April roadmap.

2 Likes