Skip to content
FoodPhoto.ai
Step-by-step guide

How to Photograph Food for Delivery Apps

A realistic workflow for restaurants shooting their own delivery photos — or preparing inputs for an AI tool — with exact settings that pass DoorDash, Uber Eats, Grubhub, Deliveroo, Rappi and iFood approval.

Minimum resolution
1000 × 1000 px
Aspect ratio
1:1 square
Max file size
5 MB
Color space
sRGB

Delivery apps do not want magazine-style food photography. They want clear, appetizing, consistent thumbnails that read at 200 pixels wide. Most phone photos fail not because of the camera but because of four fixable mistakes: bad lighting, wrong angle, cluttered background, and incorrect export.

This guide walks through the exact setup restaurant operators can use in their own kitchens — or what to provide to an AI tool so the output is platform-compliant on the first try.

1. Lighting: the single biggest factor

Good delivery app photos start with soft, even, front-or-side lighting. Overhead fluorescent kitchen light is the worst case; it flattens colors and produces unflattering shadows. The best no-cost setup is natural daylight from a window, diffused through a white sheet or translucent curtain.

  • Shoot near a large window between 9 am and 3 pm for the cleanest neutral light
  • Diffuse direct sun with a white bedsheet, tracing paper or a softbox
  • Avoid kitchen fluorescents — they cast a green tint that delivery apps flag
  • Use a second white surface (foam board, cutting board) opposite the light source to fill shadows
  • If you must shoot indoors at night, a single LED panel at 5600K with a diffuser beats any phone flash

2. Angle: match the dish, not the trend

The angle rule is simple: shoot the dish from the angle a customer would see it if it were in front of them. That usually means 45° for burgers, stacked sandwiches, pasta, and most proteins; 90° overhead for bowls, pizza, noodle dishes, charcuterie and drinks with latte art.

  • 45-degree angle: burgers, sandwiches, steaks, plates with height
  • 90-degree overhead: pizzas, bowls, flat-lay tacos, drinks with top detail
  • 15-degree eye-level: tall drinks, desserts in glassware, layered parfaits
  • Never shoot at 0° (straight on) — dishes look flat and unappetizing on thumbnails

3. Framing: fill the square

Delivery apps crop to 1:1 automatically. Shoot in square mode on your phone if possible, or frame so the dish occupies 70 to 80 percent of a square crop. Leave a thin, consistent border of background — never let the dish touch the edge of the frame.

  • Enable 1:1 or square mode in your camera app before shooting
  • Center the main dish; supporting items (fries, sauce) sit in the lower-right or lower-left
  • Leave roughly 10 percent of the square as background on each side
  • Shoot every item at the same zoom distance for menu consistency

4. Background and styling

Delivery app research consistently shows that clean, neutral backgrounds outperform busy ones. Wood, slate, matte white, and unpainted concrete photograph well. Avoid branded tablecloths, competing dishes, phones, hands, or napkins with logos.

  • Approved backgrounds: matte white, dark wood, slate, concrete, linen
  • Avoid: glossy surfaces, patterned tablecloths, branded props, hands in frame
  • Add 1-2 simple props only if they clarify the dish (parsley sprig, lime wedge, dipping sauce)
  • Keep the same background across your whole menu — consistency doubles as a brand signal

5. Exports that pass every platform

The final step is where most shoots break. Even great photos get rejected for wrong dimensions, color profile, or file size. Use a single universal export spec that meets or exceeds every major delivery app.

  • Export 1600 × 1600 pixels, 1:1 aspect ratio, JPEG at 85 percent quality
  • Convert to sRGB color space (not Adobe RGB or P3)
  • Keep file size under 5 MB — at 1600 px the typical result is 600 KB to 1.5 MB
  • Strip EXIF data that contains GPS or camera metadata before uploading
  • Name files descriptively: classic-cheeseburger.jpg, not IMG_4382.jpg

6. The shortcut: AI with a phone photo as input

If shooting a full menu in-house is not realistic, the practical alternative is taking one decent phone photo per dish and letting an AI tool handle lighting, background, color grading, and the per-platform export. The input quality that works best follows the same rules above — good light, 45° or overhead, clean frame — but is far more forgiving than a traditional shoot.

  • Take 3 to 5 phone photos per dish so the AI can pick the strongest frame
  • Avoid shadows across the dish itself; the AI can relight but not reconstruct details
  • Include the entire dish in the frame — AI can re-crop but cannot invent what is off-frame
  • For consistency, shoot every dish the same distance from the camera

Frequently asked questions

Do I need a professional camera to shoot delivery app photos?

No. Modern phones (iPhone 13+, Pixel 7+, Galaxy S22+) shoot at resolutions well above the 1000 × 1000 px minimum required by every major delivery app. Lighting, angle, and export settings matter more than camera.

What time of day is best for shooting food?

For natural light: between 9 am and 3 pm near a north-facing window (Northern Hemisphere) or south-facing window (Southern Hemisphere). Avoid direct harsh sunlight — diffuse it through a white sheet. If shooting after service at night, use a single 5600K LED panel with a diffuser.

Should food photos look exactly like what the customer will receive?

Yes. DoorDash, Uber Eats, Grubhub and Deliveroo all have authenticity policies — photos must represent the actual dish. Styling is allowed (fresh garnishes, clean plate, good lighting) but not misrepresentation of portion, ingredients, or presentation.

Can I use one photo across DoorDash, Uber Eats and Grubhub?

Yes, if you export at 1600 × 1600 pixels, 1:1 square, JPEG, sRGB, under 5 MB. That single file meets the spec on every major delivery platform. Country-specific platforms like Rappi, iFood, Swiggy and Zomato also accept this export.

How many photos do I need per menu item?

One strong primary photo per item is enough for most platforms. DoorDash allows multiple angles per item — for signature or high-margin dishes, a second photo (detail shot, cross-section) can lift conversion. For LTOs and promotions, a single hero image is standard.

Why do my photos look amateur even with good light?

The usual culprits are: white balance tint (fix: manually set white balance before shooting), post-shoot over-saturation (fix: export at natural saturation), and inconsistent framing between dishes (fix: use a tripod or tape marks for repeatable distance).

Ready to produce menu-ready photos?

Start with 20 credits for $3 — no contract, no photoshoot, no scheduling.