ACDSee Photo Studio Ultimate 2023 at a glance:
- Facial mapping for face editing
- Auto selection tools for sky replacement
- Photomerge now includes panorama stitching
- $8.90 per month, or $150 a year
Monopolies aren’t healthy for anyone, so it is always good to see imaging software applications evolving to counter the dominance of Adobe. I don’t have anything at all against Adobe, and the company is powerful in this area for a number of very good reasons – I just like healthy competition. So many other applications have a range of strong points, but so many weak spots and missing features, that I find myself going back to Photoshop for my everyday needs.
ACDSee’s Photo Studio Ultimate 2023 seems to cover most bases that are important to most of us though, and with this latest update brings a number of promising extras that could be enough to lock new users in.
The more striking new features include an artificial intelligence face detection and editing tool that allows us to make extensive adjustments to a subject’s face shape, AI options for the selection and replacement of objects, skies and background, and a new merging tool that allows photo stitching, stacking and the creation of high dynamic range images.
For those who aren’t familiar with ACDSee’s Photo Studio, here’s a short explanation of the package and what it does. This is an image editing application that allows users to process RAW and JPEG files, and which features an organisational element for photographers to catalogue their files. It doesn’t force us to use the cataloguing feature, so gladly we don’t have to import images to work on them. We can just open, adjust and close it we want to.
Much like Adobe’s Lightroom, and other programs, the path of editing is broken into different stages with a series of operating windows in which we browse our catalogue/computer folders, and then ‘develop’ and edit our images. Newcomers will find it takes a while to comfortably navigate and to understand the process, but it is mostly pretty straightforward.
There is no separate workflow for RAW and JPEG files, so they both go through the same set of controls and tools, which is quite nice – we just have to be careful to notice when the JPEGs begin breaking up under extensive manipulations. I like the tools and their flexibility, and many offer the facility to moderate their impact after their application, so if you realise you’ve overdone it you are able to rein the effect back in a bit.
There is a whole pile of pre-set looks and effects that can be applied to create instant results, but most of these are so extreme and ghastly they should be treated with a good deal of caution or used in moderation.
While the range of controls is extensive and varied, I found it took me a lot longer than usual to process my images. This isn’t due to the number of choices – though it was at first while I explored the application – but due so many standard tools needing their own window and their own ‘apply’ process that even some simple things become quite a task.
The program seems quite resource-heavy too and loading the image into a new interface for the next task involves all the previous filters and effects being applied once again in sequence before the image is ready for further work. There is a bit of a lag too in places between a slider being moved and the effect being shown, so it’s easy to push sliders too far without realising the strength of the effect.
I also found the program hangs and needs restarting when faced with folders that contain both still images and videos, which is really frustrating.
It seems unnecessary that so many tools are duplicated between the Develop and Edit panes, as they seem to do the same things, and moving from the Edit pane back to the Develop pane requires we save the image to ensure changes applied in the Edit pane carry forward. It all seems a bit convoluted and long-winded.
In the Repair window the Smart Erase tool is exceptionally good, but each individual repair takes so long to effect that dusting a product shot would be completely impractical.
However, despite all that, the results the application produces are very good and there isn’t too much we can’t do to our images using this program. I just wish it was a bit more streamlined.
I also object to the fact that the application has rather invaded my Windows operating system and has put itself as the default for ‘Open With…’ options and shows all image files in Explorer with the ACDSee icon. As disorganised as I am, I would prefer to choose how image files are displayed and in which application they get opened in. It’s like having a cuckoo in my PC.
It is also worth checking that the application is compatible with the RAW files from your camera – we got off on the wrong foot when I discovered it couldn’t read the images I’d taken for this test with the Lumix GH6.
The Face AI is the most fascinating element of this new application and I looked forward to seeing what it can do. This part of the program lives in the Edit pane and clicking AI Face Edit loads your image into a new interface specifically for this task. The AI part of the application is that it detects a face, or a number of faces, in the frame and identifies the key features of each.
Eyes, eyebrows, nose, chin, jaw lines, and cheeks are detected and mapped, so the characteristics of each can be adjusted. Eyes can be opened a bit more, widened, narrowed, lifted, brought together or spaced wider and whitened. We can even change the direction in which they are looking – though unfortunately not individually. Wrinkles around the eyes can be removed and bags reduced.
The rest of the face can be slimmed or bulked, made longer or shorter, foreheads extended, chins reduced and jawlines made more or less rugged. You can, in fact, alter someone’s appearance so much it hardly looks like them at all, but at the same time with minor tweaks and twitches we can moderate an unfortunate expression or make someone having a bad day look like they slept well, that they’ve been keeping hydrated and as though they don’t eat fast food on a regular basis.
If adjustments are looking a bit lop-sided it’s possible to inspect and correct the program’s mapping of the face, should hair, a hat or some other obstruction have confused the detection process.
I found, as with most image manipulations, that less is more and, if you intend your human to remain looking human, moderate changes that might not be noticed even by their best friends are the most effective. Bags and lines can be subdued rather than eliminated, and faces slimmed or plumped to healthier days. So long as you maintain this attitude your subject will look good and be still recognisable, but also the software will be able to work effectively.
As soon as you try to make dramatic changes the artificial intelligence becomes more artificial than intelligent, and your changes will have a destructive impact. With this in mind, the measures of the adjustment strength could be tempered a little to make smaller changes easier.
You might also use caution when altering the eyebrows of glasses wearers as the software will sometimes adjust the shape of the frames rather than the brows. It seems a little crazy too that the Skin Smoothing tool hasn’t been integrated with the Face AI so it can actually detect what to smooth rather than just smoothing the whole image.
A new AI selection section now exists in the Edit pane. These tools allow the user to make selections with a single click, as the software itself makes judgements on what is and is not required. Automated options are listed for removing a background, selecting a subject, selecting a sky, blurring a background and for making the background black and white. These tasks rely on the AI understanding what the subject is and where it sits in the frame, and then determining which bits we’ll want selected.
I am really very impressed with how well the AI can identify what the subject is in the frame. With images where the depth-of-field is shallow the selective focus gives the software a massive clue, but I thought I’d be clever and offer it a rare shot taken at f/22 that has everything in focus – and it still managed to detect what the subject was. It will sometimes include other areas too, which are easy to eliminate from our enquiries, but in the main it did really rather well.
It is rather less good though at actually selecting the subjects, skies and backgrounds than detecting where they are, and elements such as hair, out-of-focus features and glasses/headphone cables/accessories are sometimes left behind.
The same principals and problems apply to actions to remove the background and to turn it black and white. Blurring the background can be made better by performing it on a duplicate layer and then reducing the opacity to make it less obvious when not everything has been selected properly, and of course there are tools to clean up the AI selections.
I found the Select Sky surprisingly good too, and was quite amazed at how well it works. I offered it a simple sky-and-mountains shot at first which it understood pretty well. Then I gave it a much more complicated sky-visible-through-trees picture which it managed even better.
Using this I was able to trade patches of white Cretan sky showing between leaves with a much bluer and more summery sky from Portsmouth – who would believe? Dropping the opacity of the post-Brexit imported sky layer allowed me to create something quite credible. I do not have the patience to make manual selections like this, so was pleased the AI action has produced something I can use.
Selecting a background to make it black and white should be illegal, but performing it on a duplicate layer allows us to lightly and more tastefully desaturate a background using the opacity slider.
We’ve had photo merging in ACDSee Photo Studio Ultimate before for HDR and Focus Stacking, but this new version also offers us the chance to stitch panoramas together from multiple images. To get my hand in I also tried out the HDR feature and was left with a slightly flat and dull image that was not as good as the result of processing a single RAW file of the same shot.
The automatic panorama stitching gave me mixed results. Well, I say mixed, but I suppose I mean unsatisfactory. I shot eight different panoramas and joiners, and asked the program to put each set together, and only one of those was useable. The application seems incapable of creating a panorama without leaving some pretty obvious clues about where the stitches are.
I allowed masses of overlap but the duplicated details weren’t used to create smooth joins and aligned sections of the scene. I worried that I’ve given the program too much overlap to work with, so tried merging fewer images with less overlap. But while that created a different result in some cases, feeding the processes less information isn’t a failsafe answer.
I found repeating a merge gave slightly different results each time, so the actual answer might be to keep trying until you get a result that either works or which can be easily fixed.
For comparison I passed the same sets of images through the Photomerge function of Adobe Photoshop and, while the results weren’t perfect in every case and each merge took longer, they were consistently very much better. I found the ACDSee algorithm added coloured fringing in places it didn’t exist in the individual images, double exposures, missing edges and areas of fluffy detail where there may have been at least two images providing the information needed.
ACDSee Photo Studio Ultimate 2023 Verdict
If I were to come back to ACDSee Photo Studio Ultimate 2023 it would be for the Face AI adjustments if I needed to ‘help’ a portrait sitter look their best. While it’s easy to over-use this feature, moderate applications can be very beneficial to the sitter and the photographer at the same time.
I have been impressed with the facial mapping, the customisation available and the range and effectiveness of the adjustments. I doubt I would need them all in any one sitting, but each has a place and together they present an excellent set of tools to ensure sitters will be pleased with their images.
The general package is very full and the application offers a massive range of image editing options that will keep everyone happy. I’m not sure the workflow is as smooth and efficient as it could be, but many individual tools – such as the Smart Erase tool – stand out as really great performers.
My main concern, and the reason I’d find it hard to come back to the program, is that things just take too long to do. An image I rotated had to un-rotate and re-rotate itself every time I adjusted its size on-screen, and the number of tools that need their own window and their own processing time would drive me mad.
The new selection and merging functions deliver mixed results and could do with some automatic tuning or more manual refining options for the user. I often found a ragged selection or a fluffed merge meant that the image either needed fixing or just couldn’t be used. The program isn’t bad for the money, but I’m not especially impressed with all the new features.