A source product photo is the single frame everything else gets built from. It has to be sharp on the product edge, evenly lit, color-true, and large enough for the biggest place that product will ever appear. Get those four right on one shot per item and the rest of the set is built from that one frame: angles, backgrounds, styled scenes and on-model looks.
New stock lands every month. Every item needs pictures before it can go on the site, and the pictures are usually what is late. So the shoot gets compressed into an evening, the light is whatever the room had, and three weeks later someone is asking why the generated angles look like a different product.
This guide covers what that one frame has to carry, how big the file needs to be, and what a good source photo can and cannot produce. It also gives a product photography setup at home that you can repeat on every drop without a studio.
Key Takeaways
-
One frame per item is enough, if that frame is right. Sharp edge, flat even light, true color, and resolution headroom. Everything downstream inherits all four.
-
Shoot for the largest display size, not the thumbnail. Shopify recommends 2048 x 2048 px for square product images and accepts up to 5000 x 5000 px or 25 megapixels (help.shopify.com, September 2026).
-
A phone clears the resolution bar easily. A 12-megapixel photo is roughly 4,000 pixels on the long edge. Resolution is rarely the failure.
-
Photograph the things a generated angle would have to invent. Printed backs, logos, clasps, texture, anything asymmetric. One extra frame each saves a wrong product image later.
-
Bad light cannot be recovered. Blown highlights and a soft edge are missing information. Color and framing can be corrected. Those two cannot.
What is a source product photo?
A source product photo is the reference frame an AI creative tool builds every other asset from. It is not a finished listing image. Its job is to record the product truthfully: real shape, real color, real surface detail, with nothing important hidden in shadow or cropped away. Every packshot, scene, angle and video made afterwards inherits whatever that frame captured, and nothing it did not.
That is the difference between this and ordinary product photography. A listing photo has to look good. A source photo has to be complete.
The four things the source shot has to get right
A sharp, unbroken product edge. The outline is what separates product from background on every later pass. A soft edge or a shadow that merges the product into the surface produces halos, clipped corners and a slightly wrong silhouette in everything derived from it. Focus on the nearest edge, not the center of the object. Every angle and scene in a generated product shoot inherits that edge.
Flat, boring light. Dramatic light is information loss. A hard highlight erases surface texture and a deep shadow erases the shape underneath it. Window light at roughly 90 degrees with a piece of white card bouncing back into the shadow side is enough for almost any product, and it repeats identically next month.
True color. Color is the failure mode most likely to pass a glance and fail a customer, and it drives returns. Shoot under one light source rather than mixing daylight with a warm bulb, and put a plain white card in one frame so you have something to correct against.
Resolution headroom. You are not shooting for the size the image appears at. You are shooting for the crop someone takes out of it later, which is how a detail shot gets made without a second setup.
Standardizing these four across every item matters more than which model you generate with. Mixed inputs produce mixed output whatever the tool, which is the point behind what moves first-pass acceptance rate.
How big the file needs to be
Shopify’s own guidance is a useful floor because it is the largest thing most small shops publish to. It says 2048 x 2048 px usually displays best for square product images, accepts anything up to 5000 x 5000 px or 25 megapixels, and requires the file to be under 20 MB (help.shopify.com, September 2026).
A 12-megapixel phone photo is about 4,000 x 3,000 pixels. That clears the recommendation with room to crop. So resolution is rarely the reason a source photo fails.
What does fail: a screenshot of a supplier PDF, an image saved from a wholesaler’s site at 600 px, a photo cropped hard on a phone before it was sent. Those arrive already missing the detail the set needs. Which legacy photos are worth restoring and which need a reshoot covers where that line sits.
Every channel then has its own spec on top. The five-channel version of that, and the one master file that clears all of them, is in ecommerce product photography, five channels and five specs.
What one source photo can produce, and what it cannot
Some outputs come reliably from a single frame. Some need a second reference, and pretending otherwise is how a wrong product image reaches a listing. Once the frame is ready, the guide to creating product images with AI covers the method and the brief. The five source checks in our product photography checklist confirm the frame before you start.
| Output | What the source has to carry | From one frame? |
|---|---|---|
| Background swap, clean packshot | Sharp product edge | Yes |
| Styled scene, lifestyle setting | Correct perspective and scale cues | Yes |
| Flat lay | A flat, top-down source | Yes, if shot flat |
| Relight, color variant of the scene | Even base light, no blown highlights | Yes |
| Detail crop | Resolution headroom | Yes, if the file is big |
| New angles | Consistent form, plus any asymmetric side | Partly. Add a frame for printed backs and logos |
| On-model, garment on a body | Garment shape and drape | Depends. A flat garment gives the model less to work with than one on a form |
| Video from the still | A still that holds at motion | Yes, and it inherits every flaw at speed |
Anything that changes the world around the product works from one frame. Anything that has to show a part of the product the camera never saw needs that part photographed.
That is why the extra frames matter more than the extra angles. A generated three-quarter view of a plain ceramic mug is safe. A generated back view of a printed tee is a guess, and shoppers who receive the wrong graphic send it back. The seven views a listing needs, and which ones are worth capturing yourself, are set out in the complete product photography shot list.
Product photography setup at home
A product photography setup at home needs a table by a window and a white backdrop. Light it with the window or one daylight lamp. Add a white card as a reflector, and hold the camera on a tripod or phone mount. Fix them in one place and mark their positions. The same positions then give you the same framing and light direction on every drop.
The value here is repeatability. A setup that lives in someone’s head gets rebuilt slightly differently every month, and the whole catalog drifts.
This DIY product photography setup costs little and fits on one table:
- A table next to a window.
- A backdrop. A sheet of white paper or board that curves from the wall down to the table surface.
- One light. The window, at roughly 90 degrees to the product. On a dark day, use one lamp with a daylight bulb in its place. Do not mix the two.
- A reflector. A piece of white card or foam board on the side opposite the light, to bounce light into the shadow side.
- A camera mount. A tripod for the camera, or a phone clamp on a tripod, so the height and the distance stay the same.
- Tape, to mark where each piece stands.
Then run the same six steps on every item:
- Fix the kit in place once. Put the table, the backdrop, the reflector and the camera mount in the same positions every time. Mark the table with tape so each piece returns to the same place next month.
- Fill the frame, keep a margin. The product should take most of the frame with a little air around it. Cropped-to-the-edge shots remove the room every later pass needs.
- Shoot one clear angle. Three-quarter for anything with volume, straight-on for anything flat. One frame, product square to the camera, no tilt.
- Take one frame with a white card in it. Shoot it, then remove the card and shoot again. The first frame is your color reference and never gets published.
- Shoot the exceptions. Printed back, logo placement, clasp, closure, base, fabric texture. One frame each. These are the parts nothing downstream can invent.
- Check at full size before the item is packed away. Zoom to 100% on the edge and on any fine detail. A product you already returned to stock means a second setup.
The setup is the slow part, and you only build it once. How to take product photos without a studio adds a pass or fail check to each step.
What to do with the photos you already have
Most shops are not starting clean. There are supplier images of varying quality, phone shots from last season, and a folder nobody has opened since the last site migration.
Sort them into three piles. Usable as source: sharp, evenly lit, big enough. Fixable: right content, wrong size or color, worth a restoration pass. Reshoot: anything soft on the edge, anything with a blown highlight across the product, anything where you cannot tell the real color.
Supplier imagery is its own category, because it usually arrives inconsistent across a hundred items at once. Run it through a single standardizing pass before it becomes source material. Dropshipping product photos covers how to bring supplier images to one look. The product imagery templates are one place to start that pass.
Before any of it feeds a catalog run, it is worth putting the output through the five accuracy checks that catch a misleading product image.
Where this fits in the month
The reason to care about one frame is the calendar. Stock arrives, and until it has pictures it cannot be sold. That gap is the only part of the process most shops can compress, and where a launch waits works through where the weeks go.
A monthly cadence that holds looks like this. Stock lands. Everything gets shot in one session against the same fixed setup. The source frames go into one place. The listing sets, ad frames and scenes get generated from them over the following days, while the products are already back on the shelf. Build the job once on the first item, with the light, the crop and the brand rules fixed. Then run the same job on every item behind it, so each product meets the standard you set on the first.
The source photo in DesignerBox
In DesignerBox a workflow starts from your actual product photo, and every result is built from that frame. Check each result against the real product before it goes live. The AI product photography page covers the product-photo side of this, and DesignerBox for ecommerce brands covers shops with stock that turns over monthly.
For the angle problem, a multi-angle template makes extra views from a single shot. This is where the rule about photographing asymmetric sides yourself pays off. A scene placement template handles the setting swaps. For a full drop, save the setup as a workflow. Batch runs one workflow over a whole sheet of products.
There is a free plan, and you see the cost of a run before you start it. Two plan gates matter before a catalog run. Uploading your own photos and the commercial license start on the Pro plan. AI video, virtual try-on, upscaling, the image editor and the video editor start on the Premium plan. Plans and credits are on the pricing page.
The other half is the whole job in one place: the image editor, the video editor, your brand rules, your Assets and the AI models. The full workflow from the first product photo to the finished ad, in one subscription.
To build the job on your first item, start from the templates, add your source photo, and run it.
FAQ
Can I use a phone for the source product photo?
Yes. A 12-megapixel phone photo is roughly 4,000 pixels on the long edge, which clears Shopify’s 2048 x 2048 px recommendation with room to crop. What matters more than the camera is fixed light, a sharp product edge, and shooting the same way every time.
How many photos do I need per product?
One good source frame, plus one extra frame for anything a generated view would have to invent: a printed back, a logo, a clasp, a distinctive texture. For most plain products that is a single shot. For printed or asymmetric items it is two or three.
Can AI fix a badly lit product photo?
Color casts and framing can be corrected. A blown highlight and a soft edge cannot, because the detail was never captured. Those are the two failures worth reshooting for, and both are avoidable with flat, even light.
Do I need a white background for the source photo?
No, but a plain, evenly lit surface makes the product edge easier to separate cleanly. Backgrounds get replaced downstream anyway. What matters is that the product does not merge into a shadow at the base.
What resolution should I shoot at?
Shoot as large as the camera allows and keep the original. Shopify recommends 2048 x 2048 px for square product images and accepts up to 5000 x 5000 px or 25 megapixels, with files under 20 MB (help.shopify.com, September 2026). Extra pixels are the crop budget for detail shots.
Should I shoot new stock before or after it goes on the shelf?
Before. The expensive version of this process is retrieving a product from stock, restaging the setup and reshooting because nobody checked the frame at full size while the item was still in front of the camera.
Can I make a video ad from the same source photo?
Yes, and the video inherits everything the still carried, including its flaws, at speed. The same sharp edge and even light rule applies, only stricter.
Sources
- Shopify Help Center, Product media types: the 2048 x 2048 px square size, the 5000 x 5000 px and 25 megapixel maximum, and the 20 MB file size limit, accessed September 2026
- DesignerBox pricing page (designerbox.ai/pricing), plans, credit allocations and feature gating, September 2026
Image specification figures verified against Shopify’s own documentation as of September 2026. Platform requirements change, so re-check the channel you publish to before a large run. Individual results vary.
