What “pure white” means in numbers
Pure white in a product photo is RGB 255, 255, 255. Not off-white, not “looks white on my monitor”: 255 across red, green and blue, on every pixel of the background, and nowhere near the product itself unless the product genuinely is white.
I check this with the eyedropper tool in whatever editor I’m using that week. Click the backdrop, read the numbers off the panel. If I see 250, 248, 251, that is not pure white. It might look fine on a laptop screen at midnight in a flat in Bethnal Green. Sat next to a marketplace background coded to 255, it reads as a faint grey box around the listing.
Eight points off white is invisible until it is not
Amazon is explicit about this for main images; a scan that comes back at 250 or below can get a listing flagged. Etsy and eBay are looser in policy, but the eye is not.
Clients have handed me their own attempts, shot on a white sheet pinned to a wall in a Hackney studio space, convinced the background was white. It was not within eight or nine points of it. That gap is invisible on its own screen and obvious the moment it sits beside a genuine 255. Get the numbers right and nobody notices the background at all, which is the point.
Where a pure white background is required, and where it is not
This is where most guides go wrong. They teach one white background technique and imply it answers every use case, when a marketplace, a website and a printed catalogue often want opposite things from the same photograph.
| Where the image is used | Background required | File spec | Cut-out needed |
|---|---|---|---|
| Amazon main image | Pure white, 255/255/255, no exceptions | JPEG, sRGB, minimum 1000px longest side | Yes, a clean masked cut-out |
| Shopify product page | White or very pale grey, the brand’s choice | JPEG or WebP, sRGB, 2000px or more | Usually, unless a soft shadow is kept |
| Etsy listing | White is the norm, not mandated | JPEG, sRGB, 2000px square recommended | Optional, many sellers skip it |
| eBay | White strongly preferred, not enforced everywhere | JPEG, sRGB, 500px minimum | Rarely worth it at this price point |
| Google Shopping | White or plain, no props, no watermarks | JPEG or PNG, sRGB, 800px minimum | Yes, the feed can reject the image otherwise |
| Printed catalogue, CMYK | Whatever the layout calls for, rarely flat white | TIFF or high-res PDF, CMYK, 300dpi | Yes, but built for print bleed, not web |
| Wholesale line sheet | Consistent across the whole range, white common | JPEG or PDF, print resolution | Yes, buyers expect a genuinely clean edge |
Notice how much of that column labelled “background required” is a preference rather than a rule. Amazon and Google Shopping will actually reject a bad image. Everyone else is asking politely. I’d still shoot to the strictest spec across the board, because it is much easier to sell a 255-white image into a looser platform than to reshoot a looser image once a client suddenly needs an Amazon listing.
Getting it right in camera
The setup is a sweep of white paper or a perspex riser, two lights aimed at the background rather than the product, run at roughly two stops brighter than the key light on the subject. Flags on either side stop that background light spilling forward and washing out the edges of the product, which matters more than people expect.
Lighting a bottle of skincare on white and lighting a plate of food for a menu shoot use the same principle at different scale. I’ve written more on that approach in a piece on food photography in London.
Blowing the background out in camera eats the edge
Here’s where I disagree with a lot of standard advice. The usual instruction is to clip the background to white in camera, so it needs no work afterwards. I do not do that, and I would argue you should not either.
Push the background that far and the light bleeds into any pale or slightly reflective edge: a cream jumper, a frosted glass, a matte white ceramic mug all lose their outline to the glow. Under-light it by a stop, keep the edge intact, and finish the last few points to 255 in editing. That is two extra minutes an image and it saves reshoots.
The cleanup, and the edge problem
For a simple, solid, non-reflective shape, a levels adjustment on the background channel will clean a near-white sweep to a true 255 in seconds. That covers most boxed goods, most tools, most straightforward homeware. Where levels falls apart is a soft or complicated edge, so check a range for these before assuming the quick method holds:
- Hair or pile: a wig, a soft toy, a sheepskin insole
- Fur or feather trim on a coat
- Fine transparent detail such as a wine glass stem
- Loose threads, fringing, unfinished fabric edges
A wine glass stem is twenty minutes of path work
On any of those, levels leaves a grey halo or eats the detail, and neither survives a zoom. A pen-tool clipping path earns its keep instead, usually £8 to £25 an image depending on the edge. A clean path around a thin stem and a rounded foot takes fifteen to twenty minutes; a solid plastic case takes ninety seconds.
White or transparent products against white are the hardest version of the job, because you are separating white from white by tone and shadow alone. It’s usually faster to shoot those on pale grey paper and swap the background digitally than to chase a clean path from a true white shoot.
The shadow question
For an Amazon main image the rule is simple: no shadow, product floating on pure white. For a Shopify page or a brand’s own site, I nearly always add a soft contact shadow underneath, even when the client has asked for the “clean Amazon look” by default.
A product with no shadow at all, sat on a page that is not Amazon, reads as edited to most buyers. It looks cut out and dropped in, because it has been. A faint shadow costs nothing to add and does more to make the product feel present on a shelf than another few points of background brightness. You can see the difference across a range in my packshots and product work.
Doing it yourself with a window and a light tent
Honest answer: for small stock, cheap stock, do it yourself. A £40 light tent and a north-facing window in a spare room will get you a usable white background for twenty items worth under £15 each. North-facing matters in London because it gives flat, consistent daylight without direct sun swinging the colour temperature every twenty minutes, which is exactly what happens with a south-facing window in a Peckham flat around midday.
It stops working the moment the product fights the method. Jewellery throws hard reflections a tent cannot control without proper diffusion and several angles. Glassware picks up every seam in the tent fabric as a visible line. Anything glossy, a phone case or a varnished box or patent leather, turns the tent itself into a mirror.
That’s the point where paying for a shoot starts making sense, and it is often the same photographer doing your food and product shoots from £400 for a menu or a range launch, rather than a separate specialist per category.
Pure white is not always the right white
This is the section most guides skip, because it complicates the pitch. A true 255 white background is correct for a marketplace listing and wrong for a lot of brand photography that is trying to say something other than “buy this now.”
A warm grey or a linen-textured backdrop sells ceramics and candles better than clinical white does, because it implies a home rather than a warehouse. A styled set with props photographs better for social feeds and email campaigns than a floating product ever will. That work belongs with social content packages from £500 rather than a packshot brief.
The category I’d flatly refuse to shoot on plain white, unless a client insisted and paid for the reshoot risk, is anything food-adjacent not going straight onto Amazon: a jar of jam, a bar of chocolate, a bottle of oil. White drains the warmth that makes food photography sell, and clients are happier with a wooden board or a linen cloth once they see both side by side. Pure white is a tool for a specific job. It isn’t a house style.
Send one product and I’ll shoot a test frame before you commit the whole batch. You’ll have the file back within two working days, ready to drop into your listing template and measure.