Jacket photography is the image set that sells an outer garment: a front on a body, a back, a hardware detail, a lining or interior frame, and a laydown. Amazon’s clothing imaging guide recommends two different crops for jackets and outerwear, and the hem decides which one applies. It is also the category where the picture itself carries a claim about fill, fur or insulation that a regulator can read.
Most jacket guides open with lighting. Diffuse the key, rake the leather, stop down, keep the background clean. That advice is correct and it is not what costs a listing. Brands follow it, shoot a beautiful bomber at waist crop, and then discover the parka in the same drop needed a full-length frame and a different setup.
This guide takes the other route. Which Amazon rule your jacket falls under and why it changes per SKU, the two frames outerwear listings skip most, what the Federal Trade Commission says about a picture of a down garment, how the four fabric families behave under a camera, and which frames survive a generative pass.
Key Takeaways
- A jacket sits in two Amazon crop classes at once. Amazon’s 2021 clothing guide recommends a full-length shot for sweaters, jackets and outerwear that fall at or below the knee, and a top-body shot for those above the knee. The hem sets the frame (Amazon Seller Imaging Guide, Clothing and Accessories, Spring 2021, informational, accessed September 2026).
- Sort the drop by hem length before anyone books a day. A mixed outerwear capsule needs more setups than a mixed tee capsule, and that is a budget line, not a styling preference.
- Amazon does not allow any part of a mannequin in the main image. A ghost mannequin still works in the alternate slots, and for a structured jacket it is the shot that shows shoulder and collar roll without a model.
- The picture is part of the claim. FTC guidance says it examines an ad in context, including “words, phrases and pictures”, and asks for proof of express and implied claims alike (ftc.gov, accessed September 2026).
- Unqualified “down” carries a number. FTC guidance says production should yield down content of more than 70% for a product labelled down, and a stated percentage may deviate by no more than 5% (ftc.gov, accessed September 2026).
- The back frame is the fit frame. Seat length, yoke and centre-back seam decide whether a jacket reads tailored or boxy, and a front frame cannot answer that question.
- Generated backgrounds have a measured upside on apparel. Live tests of retargeting ads on mostly-apparel catalogues lifted click-through about 15% over baseline, in a range of about 4% to 40% across later tests (RecSys ‘24, arXiv:2408.12392, accessed September 2026).
- Hardware is where generation fails first. Zip teeth, two-way pulls, snap plackets and woven labels come back as approximations. Count the pockets in every result before it ships.
What is jacket photography?
Jacket photography is product photography for a structured outer garment: blazers, bombers, bikers, overcoats, parkas, shells, denim and puffers. It differs from the rest of apparel in three ways that change the shoot.
The garment has internal structure. A shoulder pad, a canvas, a collar roll and a button stance exist whether or not anything is wearing the jacket, so an empty jacket still holds a shape that a tee never does. That is why off-figure outerwear shots read better than off-figure knitwear.
The garment has function. A jacket claims warmth, weather resistance, pocket count and closure type. Those are the things a shopper is buying and they are all visible decisions in a frame.
The garment layers. It sits over something, which means the frame has to show the fit with the intended layer underneath, or the sleeve and chest read tighter than the size chart says.
Why Amazon files a jacket in two crop classes
Amazon’s 2021 clothing guide recommends a crop by where the garment falls on the body. Amazon marks that guide as informational. Read the outerwear rows and the logic is immediate.
| Crop | Garment falls | Outerwear this covers |
|---|---|---|
| Full length | At or below the knee | Overcoats, long parkas, trench coats, longline puffers |
| Top body | Below fingertips, above the knee | Bombers, bikers, blazers, shackets, cropped puffers, shells |
Both rows name sweaters, jackets and outerwear explicitly. The same brand, in the same drop, sends two different frames depending on where the hem lands.
Two consequences follow, and both are budget items rather than styling notes.
You cannot shoot the whole capsule at one focal length and crop later. Framing a full-length coat and then cropping in for the bomber throws away resolution on the tighter shot. That matters because Amazon needs at least 500 pixels on the longest side, and 1,000 pixels or more to turn on zoom (Amazon product image guide, accessed September 2026).
A mixed outerwear drop needs its shot list sorted by hem before anyone books a studio. Eight SKUs split five short and three long is two setups, not one. The parent rule set for every other garment category sits in the clothing photography shot list Amazon already wrote.
The jacket shot list, sorted by hem
Amazon recommends a main image plus at least six more images and one video. Here are the six outerwear frames that carry the most information. Amazon’s variant codes run from PT01 to PT99, and the 2021 clothing guide’s example set puts the back in PT01 and the laydown in FL01. The other codes below are a suggested order.
| Shot | Slot | On or off figure | What it proves |
|---|---|---|---|
| Front, closed | MAIN | On-figure | Silhouette, button stance, length |
| Back | PT01 | On-figure | Seat length, yoke, centre-back seam |
| Front, open | PT02 | On-figure | Lining, inner pockets, how it wears layered |
| Hardware detail | PT03 | Either | Zip type, snaps, buckles, storm flap |
| Ghost mannequin | PT04 | Off-figure | Shoulder and collar roll with nobody in it |
| Laydown | FL01 | Off-figure | Full colourway and true proportion |
Five of those six carry only the rules that apply to every product image. One of them, the main, carries its own rule sheet. That imbalance is the most useful thing to know about an apparel shoot, because it tells you where to spend.
Amazon’s product image guide asks for a pure white background (RGB 255, 255, 255), the product at 85% of the image, and no part of a mannequin. Adult clothing is shown on a standing model. Amazon’s clothing image guidelines add no borders, logos, watermarks, text or other graphics on the main image. The 2021 clothing guide adds no props and a model shown from the nose down. The alternate images are far more open. They still have to show the product accurately, but the styled frame is where brand voice belongs.
The picture is a claim about the fill
This is the part outerwear guides leave out, and it applies to no other apparel category as hard.
The FTC does not treat imagery as decoration. Its guidance on feather and down says the FTC examines the ad in context, including “words, phrases and pictures”, to decide what it conveys to consumers. It also says advertisers “must have proof to back up all express and implied claims that a reasonable consumer would take from an ad”. The worked example in that document is a picture: the claim “nature’s best insulation” beside a picture of a goose suggests goose down (ftc.gov, accessed September 2026).
Three numbers from the same guidance are worth writing on the brief:
- A product may not be called “pure down”, “all down” or “100% down” unless that is true.
- Production techniques should yield down content of more than 70% for a product labelled “down”.
- A stated percentage may deviate by no more than 5%, and a producer “may not shoot for the tolerance”.
Now apply that to a generated image. A puffer rendered with baffles the real garment does not have, a hood rendered with fur trim that is not on the SKU, or a shell rendered with a laminated sheen the fabric does not carry, is an implied claim about the product. The label is not the fix. If the image flatters the garment beyond what arrives in the box, the answer is a different image.
The disclosure question is separate. Article 50 of the EU AI Act has applied since 2 August 2026 (European Commission FAQ, accessed September 2026). The Commission published guidelines on it on 20 July 2026, and the guidelines are not binding. They treat realistic AI-generated human avatars or personas as persons, so a photorealistic AI model is likely in scope of the deep fake rule. They also say colour correction, background changes for clearly aesthetic reasons and rescaling usually have only a minor effect. This is general information, not legal advice. The full breakdown, including which assets in a drop trigger a label, is in labelling AI-generated fashion images.
How the four fabric families behave
Outerwear is four capture problems wearing the same word. Brief them separately.
Leather and coated finishes. Grain is a specular property, so it only appears where light rakes across it. Overhead soft light flattens a biker jacket into a shape with no surface. Use a directional key with visible falloff and let one side of the sleeve go darker than the other. The failure mode to watch is the opposite of underexposure: a blown highlight on a shoulder deletes the grain entirely and the jacket reads as vinyl.
Wool and melton. Nap has direction. A wool overcoat photographed with the nap running away from the light looks flat and slightly dusty, and the same coat turned around looks rich. Shoot a test frame both ways before committing the setup. Wool also holds a crease differently from twill, which is a drape claim, and drape is the thing that carries a coat. The gauge and stitch problems that apply to sweaters under the coat are worked through in knitwear photography.
Technical shell. Ripstop, laminates and DWR finishes reflect. A shell will pick up a colour cast from anything in the room, including the model’s skin and any coloured surface behind camera. Flag it. A shell that photographs slightly green is a colour accuracy problem, and colour accuracy is a returns problem.
Down and quilting. Baffle geometry is a product identifier. Box-wall versus sewn-through changes both the look and the warmth claim, and a shopper comparing two puffers is reading exactly that. Light it so the baffle edges hold a shadow line, because flat light collapses a quilted panel into a printed pattern.
The two frames outerwear listings skip most
The back. Seat length, yoke width and centre-back seam decide whether a jacket reads tailored or boxy, and a shopper cannot infer any of it from a front frame. Amazon’s 2021 clothing guide puts the back in PT01 in its example set. A back frame answers a fit question before the order, and that is returns work.
The open front. A jacket is bought partly on its interior: lining, inner pocket, storm flap, drawcord. A closed front hides every one of them. This is also the frame that shows how the jacket wears over a layer, which resolves the sizing question a structured garment always raises.
The off-figure equivalent of both is the ghost mannequin, and it is the outerwear technique that gains the most from a structured garment. Amazon does not allow it in the main image, and it works in the other slots. The channels that take the opposite position are covered in where a ghost mannequin shot is allowed to run. In DesignerBox, a ghost mannequin template makes the same effect from a photo of the garment on a form.
Where AI holds up on jackets, and where it breaks
The split follows the constraint map above, and it is sharper for outerwear than for soft apparel.
Where it holds up. Scene and background swaps for the styled frame. Additional colourways of a jacket already photographed, which for outerwear is often the same shell in four colours. Crops and aspect ratios for a paid social calendar. Angle coverage from one source photo, which a multi-angle template makes as several views. There is measured upside here: live tests of retargeting ads on mostly-apparel catalogues lifted click-through about 15% over baseline with generated backgrounds, in a range of about 4% to 40% across later tests (RecSys ‘24, arXiv:2408.12392).
Where it breaks. Hardware first, and it fails quietly. Zip teeth render as a plausible texture rather than the actual tooth pitch. A two-way zip comes back one-way. Snap plackets lose a snap. Woven labels and embossed logos garble into near-text. On an expensive biker the hardware is the price signal, so a wrong pull tab matters.
Second, quilting. Baffle count and stitch spacing drift between generations, and on a puffer that pattern is the product.
Third, pile direction on shearling and fur trim, which reverses or feathers in ways that read as a different material.
The working rule is the same one that applies across apparel: generation moves the alternates first and the main image last. Photograph, or verify frame by frame, the shot that carries the 85% rule. And check the output against the real garment with a literal count. Pockets, zips, snaps, baffles.
Two DesignerBox starting points fit this rule. Both start from your real product photo, which carries the shape of your jacket into the run. You still check the hardware. A colourways app changes the colour of the jacket in that photo, which is the safer version of a colourway variant. When a jacket has to appear worn, a virtual try-on template puts the real garment on a model. Virtual try-on starts on the Premium plan. Keeping that model identical across a whole capsule is its own problem, worked through in consistent AI fashion images.
Model choice is one step in the workflow. DesignerBox runs 8 image models, listed on the model page. For outerwear, fabric fidelity is usually the deciding axis, so run the same jacket through two models and compare the fabric before you save the workflow.
What an outerwear drop costs to shoot
Run the arithmetic on the shot list, not on a day rate, because the shot list is what scales.
An 8-SKU outerwear capsule at six shots per SKU is 48 images, across at least two setups because of the hem split. The budget carries a photographer’s day rate, studio rental, a model, styling and retouching. Each extra setup adds time to the day, and the second colourway of the same shell costs close to the first.
Generated images change the unit. In DesignerBox, the cost of each run depends on the model you pick, and it is shown before the run. Pro is $35 a month billed monthly for 1,000 credits, and the commercial licence starts on Pro. The plans are on the pricing page.
The number that compounds is the second colourway, not the first shoot. Build the outerwear pass once on one jacket, as a workflow. A saved workflow runs the same way on the next jacket, in this drop and in the next season’s. Additional colourways, crops and scene variants all come from the file you already have. For a category that ships the same shell in six colours every season, that is where the numbers change.
Two limits stated plainly. AI video starts on Premium, and an 8-second clip costs 40 to 560 credits, depending on the model. You see the cost before you run it, so budget video separately from stills. Team features and shared brand kits are on the Ultra plan at $200 a month billed monthly, and every plan below Ultra is one seat.
A jacket set in DesignerBox
- Photograph the main image properly. One clean front frame of the real jacket, correct crop for its hem, white background. This is the file everything else derives from.
- Sort the drop by hem. Long coats to full length, short jackets to top body, before you run anything or book a studio.
- Run the angle coverage. An AI product photography workflow makes the back and three-quarter frames from the single source photo.
- Add the off-figure frames. Ghost mannequin for structure, laydown for colourway.
- Put it on a model where the fit story needs one. The Dress my model app works from the garment photo. Compare the result with the real jacket.
- Run the count check. Pockets, zips, snaps, baffles and labels, against the real garment, on every result.
- Save it for the next drop. The whole sequence saves as a workflow, and it runs the same way on the next jacket. Batch is coming, and it will run one workflow over a whole sheet of products.
Channel specs beyond Amazon, including Shopify, Etsy and TikTok Shop, are collected in the channel-by-channel image spec.
To build the first jacket set, start from a template, add your brand and one jacket photo, and run it. The cost is shown before the run.
FAQ
How do you photograph a jacket for ecommerce?
Shoot six frames: a front on a body for the main image, a back, an open front showing the lining, a hardware detail, a ghost mannequin for structure, and a flat laydown for the colourway. Frame the main image to Amazon’s crop class for that garment’s hem, on pure white, with the product filling at least 85% of the frame.
What crop does Amazon require for a jacket?
It depends on the hem. Amazon’s 2021 Clothing and Accessories imaging guide recommends a full-length crop for outerwear at or below the knee, and a top-body crop for outerwear above the knee. Amazon marks that guide as informational. A long parka and a bomber from the same brand take different frames.
Should a jacket be photographed on a model or a mannequin?
On a model for the main image if you are selling adult apparel on Amazon, because Amazon’s product image guide asks for adult clothing on a standing model. Use a ghost mannequin in the alternate slots, where it is allowed and where a structured jacket benefits most from it.
Can you use a ghost mannequin for a jacket main image on Amazon?
No. Amazon’s product image guide says not to show any part of a mannequin in the main image. A ghost mannequin frame still works in the variant slots, which is where most outerwear brands should use it.
How many photos does a jacket listing need?
Amazon recommends a main image plus at least six more images and one video. For outerwear, six frames carry most of the information, because the open-front frame carries the lining and the layering story.
How do you photograph a leather jacket without losing the grain?
Light it directionally rather than flat. Grain is specular, so it only shows where light rakes across the surface. Watch the shoulder highlight, because a blown highlight deletes grain and makes leather read as vinyl.
Do you need a back shot for a jacket?
Yes, and it is a frame many listings skip. Seat length, yoke and centre-back seam decide whether the jacket reads tailored or boxy. Amazon’s 2021 clothing guide puts the back in the first variant slot in its example set.
Can AI generate jacket photos that keep the real hardware?
Not reliably, and hardware is the first thing to check. Zip tooth pitch, two-way pulls, snap counts and woven labels are where generation drifts. Generate the styled and colourway frames, verify the hardware detail frame against the garment, and count pockets, zips and baffles on every result before it ships.
Does a picture of a down jacket count as a claim?
Yes. FTC guidance says it examines an ad in context, including pictures, and advertisers need proof for implied claims as well as express ones. An image showing baffles, fur trim or an insulation finish the garment does not have is a claim you would need to substantiate.
Sources
- Amazon Seller Imaging Guide, Clothing and Accessories, Spring 2021 edition (m.media-amazon.com), accessed September 2026. Recommended crop classes for full length and top body, the example variant codes, no props and the nose-down main image. The guide carries a Spring 2021 stamp and is marked informational, so confirm the current style guide for your subcategory before a launch.
- Amazon Seller Central, product image guide (sellercentral.amazon.com), accessed September 2026. Pure white background, 85% of the image, no mannequin, adult clothing on a standing model, 500 and 1,000 pixel thresholds, and the recommended image count.
- Amazon Seller Central, image guidelines for clothing (sellercentral.amazon.com), accessed September 2026. No borders, logos, watermarks, text or other graphics on a clothing main image.
- Federal Trade Commission, “Down…But Not Out: Advertising and Labeling of Feather Down” (ftc.gov), accessed September 2026. Express and implied claims, the reading of ads in context including pictures, the 70% threshold for unqualified down, the 5% percentage tolerance, and the pure, all and 100% down restriction.
- European Commission, guidelines on the transparency obligations under Article 50, published 20 July 2026 and not binding (ec.europa.eu), and the Article 50 date of 2 August 2026 (European Commission FAQ), accessed September 2026.
- Czapp et al., “Dynamic Product Image Generation and Recommendation at Scale for Personalized E-commerce”, RecSys ‘24 (arXiv:2408.12392), accessed September 2026. Click-through lift from generated backgrounds in retargeting ads on mostly-apparel catalogues.
- DesignerBox plans, credits and feature gating (DesignerBox pricing, designerbox.ai/pricing, September 2026).
Amazon requirements verified from Amazon’s product image guide and the 2021 clothing guide, and FTC guidance from ftc.gov, as of September 2026. Marketplace specs and disclosure rules change, so check the current guide for every channel you list on before briefing a drop. This is general information, not legal advice. Individual results vary.
