In this guide, I’ll take a closer look at the AI technologies driving the future of photography, and give you the top AI tools that every photographer should know about.
2023 will be remembered as the year that popularized AI tools for the masses. Overnight the world was taken by surprise by ChatGPT that in just 2 months earned over 100 million users.
But a few months before this chatbot became popular, OpenAI (the same company behind ChatGPT) developed Dall-e 3 (they just launched version 3), an image generator that again took everyone by surprise due to its incredible quality. I talked extensively about this and other similar tools and how you can use them to actually make money (if you haven’t yet done so, check out my guide).
The idea behind a text-to-image AI tool is fairly simple: you write what you need, and the AI generates a series of images. The difficult part is to actually get exactly what you had in mind. But with a bit of practice, you can get great results.
AI is here to stay so it’s a good idea to get some familiarity with it. After all, early adopters of new technologies are those who reap the greatest benefits.
In this article, I want to expand a bit on other tools that every photographer should consider using or at least know that they exist. Whether you want to use AI to generate art from scratch or you want to integrate it with your post-production skills, I’m sure you’ll find something useful.
This is by no means a comprehensive list since new tools pop up daily, but I’ll try to keep it updated.
Key AI Technologies in Photography
I organized the tools by category so that it’s easier to browse them. These include: Image Generation, Image Editing and Other Tools.
Image Generation
This is a type of technology that generates new images mainly using a text prompt input (text-to-image). Special algorithms analyze existing images and then create new ones. These tools can be used to create portraits, landscapes, animations, and even abstract art.
They are far from being perfect though and some of the main limitations are:
- Quality of generated images: In some cases, the images generated are blurry, poorly lit, or not very visually appealing. It can take some time to learn which prompts to use to produce high-quality content.
- Limited control over the output: It can be difficult to specify exactly what kind of details the image you want to generate should or should not include. At the moment, many text-to-image tools have serious trouble generating realistic hands.
- Understanding the model’s limitations: The model may not be able to generate images of certain objects or scenes based on the data it has been trained on.
- Bias in the training data: If the training data contains biased images, the model may also generate biased output.
- Time and computational resources: Generating images can take a while especially if you use free tools that don’t allow you to speed up the process.
Image Editing
These tools will help you to refine your images. They often overlap with text-to-image tools, since they can help you to edit images that were previously AI-generated (inpainting).
The main disadvantage of these tools is that the best ones can be expensive and they require some level of technical skills and knowledge to use them effectively. Some knowledge of classic photo editing software like Photoshop is also required.
Top Image Generation Tools
Midjourney
It’s one of the most popular text-to-image tools and also one of the best. For some reason, it was decided to use Discord as a graphic interface which I personally don’t like. Regardless, the images are quite good. Actually, all the images in this article, excluding the screenshots for the websites, have been generated by Midjourney.
Pros
- One of the best in terms of image quality
- Reasonable pricing
- You can see other people’s work
Cons
- Uses Discord
- Hard to use
- All images are public (unless you pay more)
DALL-E 2
Dall-E 2, developed by Open AI (the same company behind ChatGPT), is the website that has popularized text-to-image AI. Like Midjourney, it can create incredible images. The advantage is that it doesn’t require Discord and can also edit part of the images generated (inpainting).
Pros
- First mover advantage
- Easy to use
- Can enhance existing images
Cons
- Images tend to be dull if not provided with specific parameters
- The free plan is very limited
Dreamstudio
DreamStudio by Stability AI is a text-to-image AI system powered by Stable Diffusion. Stable Diffusion is Open Source so technically anyone can use it for free, but not everyone has the skills and/or computational power to run the AI. This is where DreamStudio comes in handy.
Pros
- Clean and easy-to-use interface
- API access for developers
- Cheaper than the competition
Cons
- Small images dimension
- On a local installation, you need very high-end hardware
Nighcafe
NightCafe Creator is an AI app that lets you create and share art using new methods of AI image processing. You can use a few different styles and methods of creating, like style transfer to transform photos or text-to-image prompts.
Pros
- Allows you to upload your own images and edit them
- Easy to use
Cons
- Very limited free plan
- Much slower to generate images than the competition
- Image quality is not as good as the competition
Rundiffusion
Rundiffusion is a website that provides an account on a virtual server where you can run Stable Diffusion. Instead of spending hours installing all the libraries, not to mention the fact that most computers would not be able to run them, you can “rent” a computer remotely and use it for as long as you need. There are lots of libraries, and creating a working environment is really very easy.
Try Rundiffusion now (use code fabionodari15 to get a discount)
Pro
- Dozens of open-source models are available
- Intuitive interface
- It takes little time to start a new instance
Cons
- There are not many cons other than the fact that it is a Pro tool and therefore difficult to use for non-experts
Astria AI
Special mention for Astria AI. It’s an interesting tool because it’s an artificial intelligence model trained using a set of images provided by the user himself. You need to pay $5 to use it, and the service allows you to generate images using text requests (with a limit of 500 requests).
Pro
- You can use your own images
- Very cheap
Cons
- The results are not always the best
I tried many other text-to-image apps and despite having nice features like the ability to add or remove details from your pictures and so on, honestly, they are not very good at generating realistic images from a text compared to Midjourney or Dall·e 2. Personally, my go-to AI generator is Midjourney despite its ugly user interface. It’s simply great most of the time and quite cheap.
The best AI image generator for semiprofessional use is Midjourney, despite its ugly Discord-based user interface. On the other hand, if you want to take it a step further and become a pro, I recommend using Stable Diffusion. If you don’t have a powerful PC or don’t have time to put in the time to study how to run Stable Diffusion, you can use Rundiffusion which will make your life much easier.
Top Image Editing Tools
Topaz Photo AI
Topaz Labs offer some of the best post-processing and video editing tools. They are famous for the great noise reduction algorithm and this particular tool does a bit of everything: sharpen, remove noise, and increase the resolution of the photo using AI. All these enhancements don’t come cheap and you’ll have to spend 199$. You can read my detailed review here.
Luminar Neo
I’ve been using Luminar for many years and I can say that it’s an amazing product. Luminar Neo is a software that uses AI to greatly speed up post-production and is especially good for replacing the sky, removing unwanted elements from the image and soon it will allow to inpaint new elements using the AI.
Try Luminar (use the code Fabio10 to get a 10$ discount) – You can have a free trial using this link
AI Image Upscaler
AI Image Upscaler offers AI-powered tools for upscaling and enhancing images. It claims to increase the resolution of the pictures up to 4x while maintaining great quality. I tried it and I was honestly impressed. Despite a slight loss of image quality, the result was actually good. The website also offers other useful AI-powered image tools.
Image Cleaner
Image Cleaner is a free image cleanup tool that uses AI to instantly remove unwanted objects from images and pictures in a few clicks. It can clean up pretty much anything including people, scratches, graffiti, watermarks, and text, and it can also retouch pictures for commercial use. The tool is 100% free. Give it a try.
Cutout Pro
Cutout Pro is one of those tools that are not really necessary but speed up your post-processing. Its AI tools can help you with image background removal, photo enhancing, upscaling and other things. To be clear you can manually to the same things in Photoshop but this is faster and cheaper. Additionally, it offers API integration.
Leia Prix
This Ai convert converts 2D images into 3D Lightfield images. You can then export to Leia Image Format, Side-By-Side 3D, Depth Map, or Lightfield Animation. Light field images contain information about the direction and intensity of light rays in a scene, which can be used to produce a range of imaging effects, such as adjustable focus and perspective.
Photoshop
Although lagging behind the competition for a while, Adobe has recently added powerful new AI features to Photoshop. One of the latest is the ability to generate and/or remove elements in the photo (inpainting and outpainting).
Photoshop has a huge user base on its side, and there is no denying that this is an advantage since there is now the ability to use these tools within one familiar interface.
Other Useful AI Tools for Photographers
Camira is a good all-in-one suite of apps to assist photographers. It can help to plan a shooting, give you ideas for the composition, or even write a script for a video.
Another cool tool I found is Replicate. It allows you to upload an image and it suggests a text prompt to create similar images. It’s a good tool to get ideas from text-to-image prompts. Give it a try.
Lexica is a search engine for images and hints generated by artificial intelligence. Since every image has a prompt and seed, you can use Lexica to copy and remix existing images.
Final thoughts
Is AI taking over the world? Will photographers be out of jobs soon? Is it really over for artists? Well, I personally think that none of these things are gonna happen any time soon. There is no need to panic but one thing is for sure: we are right in the middle of a technological revolution similar to the creation of the Internet. The more AI becomes more efficient, the more it will become part of our lives.
AI is just a tool (a very advanced one indeed), but like any other tool, it can help us do a better job only IF it’s used correctly. Rather than being afraid of it, I think we should give it a try. It’s not going anywhere and the sooner you start to use it, the better it is since it will give you an advantage over later adopters.
It’s also important to set the right expectations. All these AI tools are far from perfect, in fact, many of them are still in beta testing.
Regardless I’m quite excited about all the new opportunities and new tools that AI will create for us photographers. I really looking forward to seeing what’s coming next.
If you enjoyed reading this article don’t forget to check out my guide to selling NFT and microstock images.
good