19 April 2020

The AI Battleground in Photographic Editing

Industry

by Arwyn Pennington-Bailey

Whenever someone has said to me, “You can just photoshop that, can’t you?” I have to restrain myself from warmly shaking them by their throat. The implication of this statement is that photographic editing is a breeze and therefore my time is inexpensive.

The idea that it’s all so easy is fuelled by the latest smartphone tech being so readily available. With a single click, the image is captured. At a second click, an effect is applied. Within less than a minute or two, experiences are captured and shared with a global audience.

How can a professional photographer produce anything different to that which can be captured on a device or tablet? More and more the answer lies within AI.

When I’m back at my studio, after a shoot, I open up my copy of Luminar designed by the software house Skylum. I have access to a range of features driven by AI. Luminar recognises where the eyes, pupils, eyebrows, teeth, lips and mouth of my subject are. Adjustments such as enlarging eyes, whitening pupils or the reddening of the lips can now be carried out within seconds. Add in the smoothing of skin blemishes, changing the skin tone, or creating a B&W equivalent, I create a hi-res, high quality image ready for any client without complexity.

Adobe Photoshop still offers a hugely multi-faceted editing piece of software but it’s complicated to learn and use. Having an eidetic memory is a plus when working with this software, otherwise referencing tutorials, videos and online help forums is a continual necessity.

An example of this is the addition of an Orton filter. Many photographers will need a refresher tutorial if using Photoshop, at the very least. Whereas, within Luminar, a user can go straight into the Orton effect filter. It’s easy and intuitive to use, even for a first-time user.

The availability of AI doesn’t stop there. When using Sharpen from the software house Topaz, the sharpening of an image, the removal of camera shake, or the improvement of the focus point creates amazing results. Removing noise from darkly lit subjects, or cutting out unwanted objects, is accessible through yet more AI based software.

But how does this all match up to the well-known software house Adobe, in particular Photoshop?

The advantage of Photoshop is that extensive editing tools lie within this one piece of software. To gain a similar functionality from other software houses, a number of different programmes are required.

However, one advantage of the new boys on the block, such as Skylum and Topaz, is that they have created software from the ground up and maintain uniformity across their respective software suites. The layout and GUI for these new tools is easy to learn.

This new software has also broken new ground in visualisation. A vertical bar divides the image on the screen between a before and after during the editing stage. Adobe offers nothing like this. A small preview box is about the best it gets in Photoshop. This feels and looks outdated and it is not intuitive.

Let’s take another piece of software from Topaz: Gigapixel. The claims by the makers are bold, but do they stack up? Yes, they do! It is now possible to take an image shot on a smartphone and enlarge that same image.

To perform a similar function within Photoshop, you have to read and learn about Bi-Cubic options and then record and run your own scripts. This is cumbersome. It is true that Adobe is using AI within its products via the Sensei engine, but the delivery of the technology to the end-user remains complex, clunky, non-intuitive and it requires lengthy episodes of learning.

Is there a downside to the new AI software? There is one. A high-end processor, graphics card and lots of memory is required. Be prepared to make lots of coffee whilst your machine whirs away on each image as AI does its work.

Software outside of the established workhorse of Photoshop offers a range of affordable software that is filled to the brim with AI, which is only going to be improved upon.

Will the likes of Skylum and Topaz software steal a march on the giant that is Photoshop? It is true that the market share owned by these companies is a comparatively minuscule part of the market. Will this always be the case? Who remembers ICI, Nokia, RIM and the Blackberry? You could hardly move in London without seeing investment bankers typing onto these devices or holding them firmly to their ear! Never say never, perhaps.

It is also important to understand that any image still requires skill when the button on the camera or device is pressed. AI will not, yet, do away with the skills of the person behind the lens.

Adobe Photoshop can be seen as a rather ageing Ferrari. It needs to be brought out from time to time as it can still out-perform the new generation of hybrid cars in some things. However, the Ferrari is becoming ever harder to justify, due to the ongoing expense. 

The simplicity of affordable tech-based cars is becoming ever more attractive and they are far less costly to run. They’re also more likely to be used for those everyday trips about town by Ferrari owners. For new drivers, without question, these new cars will feature more prominently and probably become the first choice for new drivers.