How to Enhance Low-Resolution Old Photos
Old photos are often tiny by modern standards — a wallet print, a low-dpi scan from the early 2000s, a small crop off a group shot. Enlarge one and it turns soft and pixelated. Enhancing it means adding resolution and detail so it stays crisp when printed or shown on a big screen.
This is where AI upscaling shines, but it's also widely oversold. Understanding what upscaling genuinely adds — versus what it plausibly invents — is the difference between a clean enlargement and an artificial-looking one.
Enhance your low-resolution photo
Upload a small or pixelated photo and Jobim enlarges and sharpens it with real, believable detail — free on iPhone.
Why old photos look low-resolution
- Small originals. Wallet prints and photo-booth strips simply don't hold much detail to begin with.
- Low-quality scans. A photo scanned years ago at 72–150 dpi captured a fraction of the print's real detail. Re-scanning the print at 600 dpi is often the single biggest upgrade — it recovers real information rather than inventing it.
- Heavy cropping. Pulling one face out of a group photo leaves you with only a handful of pixels to work with.
How AI upscaling works
A super-resolution model has been trained on huge numbers of image pairs — a high-resolution photo and its shrunken version — until it has learned what fine detail typically corresponds to a given blurry patch. Given a low-resolution image, it predicts and reconstructs that detail: sharper edges, believable skin and hair texture, cleaner lines.
Crucially, it's adding statistically plausible detail, not recovering the exact original. On real-world textures the guess is usually excellent and indistinguishable from a true high-resolution photo. On unique detail — the exact features of a barely-visible face — it's an informed reconstruction, so identity can drift when the input is extremely small.
Enhance a low-resolution photo with Jobim
- If you have the physical print, re-scan it at 600 dpi first — real detail beats any upscale.
- Upload the photo to Jobim.
- Run restore first to fix any fading or damage, then upscale to enlarge and sharpen it.
- For old family photos, the face-reconstruction step recovers convincing facial detail alongside the resolution boost.
- Save the enhanced copy at its new, larger size.
How big can you go?
A modest enlargement — 2× to 4× — is where upscaling looks completely natural. Pushing a tiny, heavily-cropped face to poster size asks the AI to invent most of what you see, and the result becomes more illustration than photograph. Enhance for a clean, believable improvement, not a miracle.
Frequently asked questions
Can you increase the resolution of an old photo?
Yes. AI upscaling reconstructs plausible fine detail to enlarge a low-resolution photo cleanly. A 2–4× enlargement looks natural; pushing far beyond that means the AI is inventing most of the detail.
Does upscaling add real detail or fake it?
It adds statistically plausible detail learned from millions of photos. On ordinary textures the result is essentially indistinguishable from real high resolution; on tiny unique features like a distant face it's an informed reconstruction.
What's better — upscaling or re-scanning?
Re-scanning the physical print at high dpi, every time you have the option. It recovers detail that genuinely exists in the print. Upscaling is for when the small digital file is all you have left.
Keep going
Related restoration guides:
Restore your photo in seconds with Jobim
Upload the old photo, tap once, and Jobim rebuilds faces, colors and detail with AI — right on your iPhone.
AI restoration reconstructs plausible detail — it can revive faded color, sharpen faces and repair damage, but it cannot recover information the photo never captured. Always keep your original scan. Results vary with the condition of the source image.