How to Restore Photos for Obituaries: A Practical Guide for Tight Deadlines

When a family member passes away and a newspaper obituary deadline is twenty-four hours out, finding and preparing a good photo is one of dozens of tasks falling on people who are already overwhelmed. The photo matters — it is often the first thing people see when they read the obituary, and it becomes part of the family's permanent record. But the best available photo is often old, small, or damaged.

AI photo restoration has made it possible to take a wallet-sized print from the 1960s or a faded snapshot from a shoebox and produce a newspaper-quality image in under an hour, without professional help. This guide covers exactly what you need to know: newspaper requirements, funeral slideshow specs, how to select the best photo under time pressure, and what results you can realistically expect from AI tools.

What Resolution Does a Newspaper Actually Require for an Obituary Photo?

Newspaper obituary desks are more specific about technical requirements than most families realize. The standard is expressed in DPI — dots per inch — at the printed size.

Most U.S. daily newspapers require a minimum of 200 DPI at their print dimensions. For a standard obituary portrait printed at 2x3 inches, that means your image file needs to be at least 400x600 pixels. Papers that prefer 300 DPI at the same print size require 600x900 pixels. Some papers, particularly those with large circulation and high print quality, request 600 DPI, which would require 1200x1800 pixels for a 2x3 inch print.

When submitting digitally — which most papers now require — ask the obituary desk for their specific pixel dimension requirement rather than the DPI figure. Digital submission bypasses some of the DPI complexity since the paper's systems handle the final rasterization.

What AI upscaling delivers: Real-ESRGAN, the upscaling model used in ArtImageHub, can produce a 4x upscale with meaningful detail recovery — not just interpolation. A wallet-sized print scanned at 600 DPI produces an image of roughly 850x1100 pixels. After Real-ESRGAN upscaling, that becomes approximately 3400x4400 pixels — well above what any newspaper requires, with sharp detail rather than the blurry softness that conventional bicubic upscaling produces.

How Do Funeral Home Slideshow Specifications Differ?

Funeral home memorial slideshows are displayed on HD screens, and the technical target is different from a printed newspaper photo. The standard display resolution is 1920x1080 pixels — standard HD — and photos that are smaller than this are stretched or letterboxed by the slideshow software, which visibly softens the image.

For a portrait photo that occupies roughly half the screen in a typical slideshow layout, a source image of 960x1080 pixels or larger is adequate. For full-screen display, aim for 1920x1080 or above.

AI restoration makes this achievable even from old source material. A 3x5 print from the 1980s scanned at 600 DPI gives you about 1800x3000 pixels of raw scan data — before any AI enhancement. After NAFNet denoising (which removes the film grain and scanning artifacts that accumulate in old prints) and Real-ESRGAN upscaling, the final file is sharp enough for full HD display.

Most funeral homes accept JPEG files at high quality settings (85 or above) or PNG. If the home uses a cloud portal for photo uploads, confirm whether there is a file size cap — a full-resolution restored image can be 10 to 30 megabytes as a PNG, and some portals have limits.

How Do You Select the Best Photo Quickly When Time Is Limited?

When a family has a box of photos and a 48-hour window, the selection process should be systematic. Evaluate candidates in this order:

Face size first. The subject's face should fill at least 20-25% of the frame. AI face restoration with GFPGAN works by analyzing the facial region — a face that is a tiny element in a wide group shot gives the model very little to work with. A head-and-shoulders portrait from any era will restore better than a distant group photo from the same period.

Original sharpness second. A faded photo that was originally sharp will restore far better than a crisp-looking photo that was originally soft or motion-blurred. Hold the photo at arm's length and look at the eyes — if the catchlights (small reflections of the light source) are visible, even as small dots, the original negative was in focus. If the entire face looks equally soft, the original was likely blurry.

Physical condition third. Condition matters less than families usually expect. Heavy fading, yellowing, surface foxing, and even significant creases are all addressable with AI restoration. What cannot be fully recovered is information that was never in the original capture — deep physical tears through face areas, or water damage that has caused the emulsion to separate and distort.

For a formal obituary portrait, a studio photograph from any decade will typically outperform an informal snapshot from the same period because studio lighting and professional focus produce better original capture quality.

What Can AI Restoration Actually Recover from a Wallet-Sized Print?

A wallet-sized print is typically 2.5x3.5 inches. Scanned at 600 DPI, it produces a file of approximately 1500x2100 pixels. This is the starting point — and it is a workable starting point for AI restoration.

GFPGAN specializes in face reconstruction. In a wallet-sized portrait where the face occupies most of the frame, GFPGAN can restore fine facial detail — the clarity of the eyes, the definition of the lips, the texture of skin — even from a heavily faded or grainy source. The model uses a reference set of high-resolution facial images to reconstruct plausible fine detail from degraded pixels. The result is not guesswork; it is a statistically informed reconstruction from the degraded data that exists.

Real-ESRGAN handles the overall upscaling. After GFPGAN restores the face region, Real-ESRGAN upscales the full image with detail-aware processing that produces sharp edges and recovered texture rather than the blurry output of conventional upscaling methods. A 1500x2100 pixel wallet scan can be taken to 6000x8400 pixels with Real-ESRGAN — far beyond newspaper requirements.

What it cannot do: If a face is in deep shadow with no recoverable detail — a region that is essentially solid black — no AI model can reconstruct what was not captured. Severely damaged areas where the photographic emulsion has physically separated or where the image is printed on top of a torn region are the practical limits.

What Is the Workflow, and How Long Does It Take?

The complete workflow from physical photo to newspaper-ready file:

1. Scan the original print at 600 DPI on a flatbed scanner. If you do not have a scanner, many pharmacies and office supply stores offer scanning services. A phone camera photograph of a print is a distant second option but workable in an emergency — use even lighting and hold the phone directly above the print to avoid distortion.

2. Upload to ArtImageHub. The preview processes in seconds. You will see the restored version before committing to the download.

3. Review and download. The one-time $4.99 unlock gives you the full-resolution restored file. For a newspaper, export as a high-quality JPEG. For a funeral slideshow, JPEG at 95% quality or PNG both work.

4. Submit to the newspaper or funeral home. Most obituary desks accept email submission or have online portals.

Total elapsed time from starting to scan to having a ready file: thirty to sixty minutes for most families. The AI processing itself takes seconds.

Does This Work for Grayscale (Black and White) Photos?

Yes — and grayscale photos are often the cases where AI restoration shows the most dramatic improvement. Many obituary photos come from the 1940s, 1950s, or 1960s when black and white film was standard. These photos suffer primarily from fading and grain rather than color degradation, and both are directly addressed by NAFNet (denoising) and Real-ESRGAN (upscaling).

If you want to colorize a black and white photo for the slideshow while keeping the original restored version for the newspaper, DDColor can produce a natural-looking colorized version that makes the image feel more contemporary and emotionally immediate. Both versions — restored black and white and colorized — are available in the workflow, and you can use each for different purposes.

Is $4.99 Worth It Compared to Professional Restoration?

Professional photo restoration at a portrait studio or specialty service runs $50 to $300 per image with a turnaround of one to two weeks. In an obituary situation, that timeline is completely incompatible with the deadline.

ArtImageHub's $4.99 one-time unlock delivers a full-resolution restored file immediately, with no subscription or ongoing charge. In the context of median U.S. funeral costs of $7,000 to $12,000, it is a negligible line item that produces a permanent, high-quality record — something the family will keep long after the obituary has been published.

For families managing funeral logistics under emotional and time pressure, the combination of low cost, instant results, and no technical expertise required makes ArtImageHub a practical tool for exactly this use case.

Frequently Asked Questions

What resolution does a newspaper obituary photo actually need?

Most U.S. daily newspapers require a minimum of 200 DPI at their print dimensions. For a standard 2x3 inch obituary portrait that means at least 400x600 pixels, while papers preferring 300 DPI need 600x900 pixels. Ask the obituary desk for the exact pixel dimension rather than a DPI figure, since digital submission bypasses some of the DPI complexity. AI upscaling with Real-ESRGAN can take a wallet-sized print scanned at 600 DPI and produce a file well above 3000x4000 pixels — far beyond any newspaper requirement — with sharp recovered detail rather than blurry interpolation.

Can AI restoration fix a blurry or damaged photo in time for a newspaper deadline?

Yes. The entire AI restoration workflow on ArtImageHub takes under five minutes per photo once you have uploaded the file, with the AI processing itself completing in seconds. A realistic timeline from deciding on a specific photo to having a newspaper-ready file is thirty to sixty minutes, including scanning. Newspaper obituary desks typically accept submissions up to twenty-four hours before publication, and the 24-72 hour window most families have is more than sufficient. GFPGAN handles face clarity specifically, which is the most important quality factor for a portrait published in an obituary.

What are the technical requirements for a funeral home slideshow?

The most common standard is 1920x1080 pixels — standard HD — for full-screen display. Photos smaller than this are stretched or letterboxed by slideshow software, which softens the image visibly. For a portrait occupying half the screen, a 960x1080 pixel source image is adequate. AI upscaling with Real-ESRGAN can bring a 400x300 pixel old snapshot up to 1600x1200 or higher with recovered detail rather than just stretched pixels. Contact the funeral home to confirm their preferred file format — JPEG at 85 quality or above and PNG are both widely accepted.

How do I quickly identify the best photo to restore when time is limited?

Prioritize by three factors in order: face size, original sharpness, and physical condition. A photo where the subject's face fills at least a quarter of the frame gives AI models the most detail to restore. A faded photo that was originally sharp restores far better than one that was originally blurry — check for catchlights (tiny reflections in the eyes) to confirm original focus. Physical condition matters least: fading, yellowing, foxing, and creases are all addressable, while original blur from camera motion or a soft lens cannot be fully recovered. Studio photographs from any decade will typically outperform informal snapshots because studio lighting and focus produce better original capture quality.

Is $4.99 a reasonable cost in the context of funeral expenses?

The median U.S. funeral costs between $7,000 and $12,000. In that context, $4.99 for a one-time AI restoration that produces a lasting high-quality image for the obituary, memorial cards, and family records is negligible. Professional restoration services at portrait studios or specialty shops typically charge $50 to $300 per image with one-to-two-week turnaround — far outside the window an obituary deadline allows. ArtImageHub's $4.99 one-time unlock gives you the full-resolution restored file immediately, with no subscription or recurring charge.