We’re excited to release the next version of Photo AI. This version includes a new text recovery model which can improve results of text in images, a re-design of the right panel controls, and long awaited support for Z9 High Efficiency compressed files. For details about various larger features, please visit the Roadmap post (also linked above).Īs always the full change log can be found below. Please give us any feedback or report issues with this release. We’ll be updating TPAI regularly to address those pieces of feedback and issue reports. If there’s a specific image you’d like us to see, you can send it us at this dropbox link. It also fixes problems that many photographers will be working quite hard to avoid in the first place.Nikon Z9 High Efficiency compression support.Added support for Nikon’s High Efficiency compression.You may need to login again and claim seats from pre-1.4.0 versions.Auth system internals changed to allow Alpha, Beta, and Release builds to use the same seat.Note for MacOS users: The new minimum supported OS version is now 11.0 (Big Sur). When it doesn’t, it’s a bit disappointing, especially in view of the cost and the time it takes. When it works, Topaz Photo AI is very good. Topaz Photo AI is expensive and quite slow to use, and while it can fix some photo problems remarkably well, they have to fall into what I’ll call its ‘fixability window’, and you have to have enough of these problem photos in the first place to make it worth the cost. With phone images I found it tended to upscale the phone processing artefacts rather than finding or adding new detail. The upscaling works really well on images with good intrinsic detail and not too much processing – such as those from a DSLR or mirrorless camera. The Upscale and Enhance Resolution tools will often be used together. There is a ‘processed’ look about the results, but it’s still an effective tool for rescuing or enhancing unrepeatable people shots. It works surprisingly well on people who are just out of focus or not quite sharp. The Recover Faces tool kicks in when the software recognizes faces in the frame and thinks they need fixing. Shots that were just slightly soft underwent a pretty dramatic transformation, and shots with poor focusing had variable outcomes – often with obviously processed edge detail and ‘filling in’ of the sort you see with over-processed phone images. I found that shots on the wrong side of its ‘fixability’ threshold were made worse – including any kind of double-image blur from camera shake. The Sharpen process can be spectacular or bad, depending on the image. I wouldn’t put this in the same league as DxO’s DeepPRIME XD processing. I found myself pushing the Detail slider up to maximum and the Strength slider down to zero to get results that looked smooth and crisp. The noise removal is very effective but quite aggressive by default. The results vary, depending on the quality of the image you’re starting from and its particular issues. The recovered detail in the rocks and the leaves, top right, is pretty remarkable. This is the most spectacular outcome, though the Autopilot didn't think this needed sharpening it all, so it was done manually.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |