From painting to photography, from stills to moving images, from analog to digital, technological advances continue to change how we make images. There is no doubt that CGI and AI technology will shape visual communication, so photographers need to understand how it will affect their industry and how they can make the most of it, rather than being left behind by the new technologies.
The first use of computer-generated images (CGI) goes as far back as the 1970s. With increasing computer power and better software, the use of CGI is now firmly established in both still photography and video. Today, it’s AI-generated imagery that is making headlines. So, what are its applications, and how will they change photography?
This is a huge field which is used extensively across all visual media. Both AI and CGI are applied across image editing, still photography, as well as moving images. The terms are often used interchangeably, but they are very different. Wonderful Machine defines AI/CGI photography as images that are (at least partially) created with a computer rather than a camera, and are usually intended to look like a photograph. But, within these two technologies, there are many differences.
Here is a list of terms commonly used to describe different types of software and technologies to create, synthesize, or alter images.
Software, such as Adobe Photoshop, which combines and alters different elements, creates backdrops and props, and is often used to create new and fantastical settings, or in drafting 2D and 3D models.
Software, such as Luminar or Photoshop, which uses AI to enhance images optimally. They use machine learning and algorithms to optimize color, contrast, luminosity, and other parameters.
CGI is a term usually reserved for images created using 3D models, software, and rendering techniques. Models are often ‘built’ from 3D shapes, which are then filled in with textures and colors to represent objects. Traditionally, CGI uses 3D software and/or algorithms that create pixel-based models, for example, of a face or shape, which can then be moved. Now AI is used to help with tasks as well as the animation itself.
Deep Fake is a term usually used for AI software that utilizes images and sometimes sound recording as the basis for an animated video clip of the person in the photo moving and/or speaking. It is often characterized by subversion or the intention of spreading false information.
Software that uses a “bank” of millions of images, often scraped from the internet, as reference material to create new images. Therefore, some argue the output is more synthography than photography, as images are derived (whether merged or synthesized) from millions of existing images.
Midjourney and DALL·E 2 are popular generative AI programs, but many others exist. Adobe has joined the race as well with Firefly, adding AI-generated imagery and prompts to its offering. For example, users can leverage word commands to change images by asking the program to expand an object, change its color, or create an entirely new item in a space by typing “add a vase with flowers on the table.”
CGI is most common in film — one of the first examples is the 1977 Star Wars film A New Hope, which no doubt would have had top industry talent working night and day to realize entirely new (and now iconic) worlds. In film, CGI is often used to animate people and animals as well as to add dramatic special effects. It’s also used in still photography to create backgrounds or fantastical worlds or just to create a prop that’s difficult to come by.
Perhaps the area where CGI has made the biggest inroads so far is product photography. For this, a 3D model of the object is created and then placed in the desired background. The software can add lighting effects, change camera angles, or create entire environments around that object. It also allows for very creative images, as the photographer is not limited by the laws of physics, or which props the production team can get ahold of. It can also be a creative and cost-effective way of producing product photography.
Car photography (especially the interiors) is now also largely produced with CGI technology, which lends itself to producing very sleek, luxurious, and immaculate textures.
Sometimes these images are produced by business owners, designers, or photographers. As technology advances, CGI will likely make further inroads into other photography specialties. In particular, when combined with generative AI image software.
I use CGI to create elements of an environment or key props that I don’t have access to or won’t work with the budget.
On a separate assignment for Viasat, John used CGI to experiment with lighting and POV to ensure all the elements came together correctly before the shoot.
Pamplona-based photographer Mikel Muruzabal also specializes in CGI photography. He said,
CGI was primarily used in car and jewelry imagery. I was involved in the fashion industry at the time, which led me to specialize in combining real-life shots of people taken in my studio with meticulously crafted digital environments. I extended this method to product and advertising photography, particularly for flooring companies launching campaigns for items not yet in production. The process involved creating a 3D representation of the manufacturer’s designs, designing a suitable digital environment, conducting studio photography sessions similar to a fashion shoot, and compiling these elements into a high-resolution composite image for advertising.
While CGI rendering requires users to craft the objects themselves, generative AI image software does it for you. It harnesses vast image banks to train a deep learning model built to recognize underlying patterns, structures, and textures of objects. It then uses the learned features to create new images by synthesizing images from the training set.
The right prompts are key to realizing a vision, a point noted by Cincinnati-based food/drink photographer Teri Campbell.
It’s all about getting the prompts right, which can take some time. Things are also changing very fast. What worked last month may not work anymore this month, because things have moved on already. They move so fast! I have also been playing with Photoshop’s generative fill. You can use word prompts to instruct Photoshop to expand an area, for example, or fill it with an object, like a pot on a stove. I learn something new every day.
AI-generated images are increasingly popping up on social media. Everybody from charities to influencers uses the software to illustrate a point. And more professional photographers are experimenting with it, too. Miami-based producer and director Buster Cox sent AI-generated images to a real estate client, who described them as “useable.” With clients expressing openness to these AI-generated photos, an emerging trend will likely catch on with greater rates of adoption.
Adobe Stock now accepts AI-generated images, but they must be labeled as “Generative AI” and an “Illustration” rather than a photograph. Users also need to have the right to sell it, which depends on the agreement with the software company.
For an interview with Fstoppers, renowned commercial photographer Tim Tadder called generative AI “the most powerful creation tool ever made.” It’s no surprise then, that countless media professionals have acted as harbingers, predicting the end of photography. However, John urged more caution.
“Authenticity” has been the emphasis in much of advertising for years. So I’m curious to see if “simulated authenticity” (no matter how well done) of people and moments will be accepted by advertisers and consumers to save money and for ease of producing content.
Whether someone stumbles upon or actively engages with it, using generative AI platforms can be an eye-opening experience. Teri leaned on AI when trying to create the perfect kitchen background for a shoot. He was mesmerized.
From the day I started this, it has been so exciting. It’s unleashing creativity for me. Hiring a model and doing a big shoot, including hair and make-up, is a great expense. Now I can do shoots with people, without the costs of models or cars. It’s opened me up creatively in a way that hasn’t happened since I got my first camera.
Teri now routinely uses AI software to generate ideas, mock-up images, backgrounds, and set designs for his commercial clients.
John also experimented with AI-generated images.
I’m intrigued by how I might similarly use generative AI as I do CGI. I plan to develop in this area after I finish client projects this summer. Also, I’ll be using AI for first-tier ideation, mock-ups, and treatments.
Mikel initially saw the unauthorized use of online images and art for AI training as an affront to artists but has recently begun experimenting himself.
This tool fuels the ideation process, resulting in concepts that I present to my clients. I also have projects lined up for my revamped website. However, achieving notable results requires dedication, a clear vision, and an extensive understanding of visual culture and references.
Cincinnati-based Creative Director Dale Doyle believes AI will become an important tool for commercial photography. He recently worked with Teri on a project using AI-generated images. He can’t share any photos yet, but said,
As a creative director, I’m always looking for the bleeding edge and AI is the hot topic of the moment. I truly believe it will be a great partner moving forward. I think it would be foolish if photographers think they don’t need AI in the future. It’s going to be a vital tool.
AI-generated images of still life are looking pretty convincing but still have quite a way to go. AI-generated images of people are really rough right now, but they will be much better in the years to come.
It’s early days, but I do like the flexibility that AI offers. You’re only as good as your prompts, but Teri has been working on this for quite some time, and is doing a great job managing the real and unreal in all of this.
When monetizing AI-generated imagery, the photographers we spoke to recommended embracing the software entirely. First, become literate in using it and observe its potential. For Teri, this potential has him going farther along the digital path without looking back.
I don’t ever want to shoot film again — digital made me a better photographer, and now I can create so much more with AI. Let’s go further with AI! As long as you’re open and willing to adapt, the possibilities far outweigh the negatives.
As it’s a new field, it can be difficult to convince potential clients. Still, there also tend to be emerging opportunities for photographers who do use AI, since many photographers are not yet offering AI imagery as a solution. The photographers and media professionals we spoke to provided the tips below.
With any new technology, there come new challenges. Currently, AI-generated images can’t be copyrighted in the US, so companies worry about photos being stolen and used in other contexts.
Separately, many argue that AI images violate the copyright of the photographers whose images have been scraped from the internet and used in machine learning algorithms to create new images. Theoretically, every AI-generated image contains tiny elements from thousands of training images. Adobe avoided this legal hurdle by using Adobe Stock to train its algorithms.
In addition, AI-generated images are currently generated at low resolutions, which can be increased with software such as Photoshop. But if recent trends in processing power continue to accelerate, it’s only a matter of time before this issue ceases to exist.
It’s clear that the potential for AI-generated images is vast — both for good and bad. It can supercharge creativity or be used to manipulate and spread lies, as this MIT article illustrates. Some people see opportunity, while others see threats to humankind.
But if the past is anything to go by, this new technology is here to stay. It will change visual communications. Experience also shows that photographers are incredibly adaptable, as they have seamlessly moved from analog to digital, and from stills to moving images. The tools may change, but the hands behind them will keep doing what they do best: using their talent to create striking images and get their message across. In an increasingly visual world, the skills of photographers to communicate through images are as valuable as ever, even if they use a machine as their photo assistant.
Wonderful Machine: Revolutionizing Photography – Teri Campbell Experiments with AI
Wonderful Machine: AI and Avante-Garde — Clay Cook’s Endless Possibilities for KMAC Couture