I’ve been using image editing and restoration tools professionally for decades as a filmmaker, photographer, and editor, and I have a suggestion that I think could make Gigapixel even more powerful. Instead of relying only on AI to guess what’s happening in an image, why not let the user briefly describe the image in plain English before processing? For example: “This is a 1988 concert performance under purple stage lighting with smoke and intentional motion blur. Preserve the atmosphere, don’t overcorrect the lighting or haze, and don’t aggressively reconstruct faces.” Large language models have shown how effective natural language instructions can be, and I think image restoration could benefit from the same approach.
It would also be incredibly useful to include intelligent presets such as Concert Photography, Film Scan, VHS Capture, Television Frame Grab, Portrait, Wildlife, Architecture, and Documentary Footage. Each preset could tell the AI what characteristics are intentional and should be preserved versus what should actually be corrected. As someone who has spent years shooting live performances and restoring archival material, I often see the AI mistake stage lighting, smoke, film grain, and motion blur for defects. Those elements are often part of the image, not problems to be “fixed.” Giving the software context before it begins processing could produce much more accurate and artist-friendly results. I think this would be a significant step forward for professionals who depend on Gigapixel to preserve the intent and character of the original image rather than simply making everything look sharper.
Patrick R. Epstein
Filmmaker • Photographer • Editor
Former collaborator with Prince (11 years)
