Midjourney users have discovered an innovative technique to enhance the photorealism of AI-generated images: incorporating camera filenames into their prompts. By adding terms like “IMG_1234.CR2″—a format typical of Canon RAW image files—before their descriptive prompts, users can achieve significantly more lifelike results.
This method gained attention when Min Choi, a user on X (formerly Twitter), shared a series of AI-generated images employing this approach. For instance, Choi used the prompt “IMG_7249.CR2” followed by a description such as “casual selfie at X,” resulting in images with enhanced realism.
The rationale behind this technique is that including a camera filename cues Midjourney to emulate the characteristics of actual photographs, thereby producing images with more authentic lighting, textures, and depth. This approach aligns with other strategies aimed at achieving photorealistic outcomes, such as specifying camera models, lens types, and detailed photographic settings within prompts
In addition to using camera filenames, users have found success by avoiding terms like “hyper-realistic,” which can inadvertently lead Midjourney to generate images resembling paintings rather than photographs. Instead, descriptors like “high quality” and “ultra detailed,” combined with negative prompts such as “-no painting, drawing, unreal,” help steer the AI toward more realistic outputs.
Midjourney’s recent updates, particularly the release of version 5, have further enhanced its ability to produce photorealistic images. This version introduced improvements in language understanding, accuracy, and stylistic flexibility, allowing for more nuanced and lifelike image generation.
For those looking to explore this technique, here are some practical steps:
- Start with a Camera Filename: Begin your prompt with a typical camera filename, such as “IMG_1234.CR2” or “DSC_5678.JPG.”
- Add a Detailed Description: Follow the filename with a comprehensive description of the desired image, including subject, setting, lighting, and mood.
- Specify Camera Settings: Incorporate details like camera model, lens type, and photographic techniques to guide the AI toward specific visual styles.
- Use Appropriate Modifiers: Include terms like “high quality,” “ultra detailed,” and avoid “hyper-realistic” to prevent stylized outputs.
- Apply Negative Prompts: Use “-no painting, drawing, unreal” to exclude non-photographic elements.
By experimenting with these elements, users can leverage Midjourney’s capabilities to produce images that closely resemble real photographs. This technique opens new avenues for creators seeking high-quality, realistic AI-generated visuals.
For a visual demonstration of creating photorealistic images in Midjourney, you can refer to the following video:
