Poses

Describing a pose in words rarely gets you that exact pose. "Hands in pockets, weight on the left leg, looking over the shoulder" comes out close, but not the same. A pose lets you show the body position instead: you give Brandmachine a photo of the pose, and the model in your look takes it.

Poses are a team library, like fashion models and set locations. Make a pose once and use it on any of your team's campaign looks.

How a Pose Works

When you make a pose, Brandmachine redraws your photo as a plain mannequin in the same position: a figure drawn as a fine grid, with no face, hair or clothes, and empty hands. On the pose's page it is called the wireframe. The render receives the mannequin, not your photo.

That is on purpose: whatever the person in your photo holds or wears (a bag, a scarf, a coat) would otherwise come into your shot. The mannequin carries only the body position. Your look keeps its own fashion model, products and set.

Preparing a pose takes about 30 seconds. While it prepares it shows Preparing; then Ready, or Failed if it didn't work.

What makes a good pose photo

These are recommendations, not hard rules:

One person, with the whole body in the photo if you want a full-body pose.
Limbs clearly visible, not hidden behind other people, furniture or a long coat.
A camera at roughly eye height. Body position carries over well; the camera angle less so (see Q&A).

Any photo you have the right to use works: a test shoot, a mood image, or one of your own earlier Brandmachine images (download it, then upload it as a pose).

Making a Pose

On the Poses page

Open Poses in the sidebar, under Shared Resources.
Click New pose and pick a photo: JPG, PNG, WEBP or AVIF, up to 15 MB.
You land on the pose's page, which shows Preparing while the mannequin is made.
Give it a name: click the title (new poses are called "Pose") and type a new one.

From your moodboard

You can also make poses from the images you already collect on a campaign's moodboard:

On the moodboard, each image has a Used for row of labels (Model, Location, Pose and so on). Click Pose on the images whose pose you want. That only tags them; nothing is made or charged yet.
On the campaign's Plan tab, the Ready to plan? box lists your tagged images. Click Plan with the agent. The agent goes through your Pose images: if a pose in your library clearly has the same body position, it uses that one; otherwise it proposes a new pose, together with any castings and sets it wants to make, and asks you to confirm once.
The agent assigns the poses to looks while it plans: each look gets at most one pose, the one whose image suits that look. It doesn't wait for the poses to finish preparing; the looks show the pose as Preparing until it is ready. You can change any look's pose on the plan board afterwards (see below).

Before it makes anything, the agent names each pose it wants to create, so you see what you confirm. Each new pose costs 4 cents; reusing one from your library costs nothing.

To make one pose without the agent, click Create pose next to that image in the Ready to plan? box.

The moodboard's Poses card shows each Pose image and whether its pose is preparing, ready or failed.

Check the Mannequin

Open the pose: its page shows the reference photo and the wireframe side by side. Compare them before you use the pose:

Is the body position the same: weight, lean, arms, legs, head turn?
Is anything missing or added: a limb, an object the mannequin seems to hold?

Renders follow the mannequin, so what it lost (a lean, the height of the camera) the shot loses too. A pose can't be edited or redone once it is ready, so looks that use it keep rendering the same way. If the mannequin is wrong, archive the pose and make a new one, if possible from a clearer photo.

Using a Pose on a Look

Open the look and find the Pose card.
Click Select, pick a pose from your library and click Select again. Use Clear to remove it.
In the look's Instructions, remove the sentences that describe a body position or camera angle ("standing with arms crossed", "shot from below"). Keep everything else. Conflicting text works against the pose.

On the campaign's plan board, the Pose chip in each look's header shows and changes the look's pose, until the look is generated. The board is also the quickest way to see which looks use which pose.

A pose needs a fashion model on the look. A pose that is still preparing can be picked straight away; generating waits until it is ready.

A look can have a pose and a Look Reference at the same time. The Look Reference then sets only mood, lighting and colour; the pose sets the body position and framing.

You can also ask the agent to put a pose on looks. It tells you which pose goes on which looks and asks you to confirm first.

Changing a pose later: pick another pose (or clear it) on the look's Pose card and generate again. Images already generated don't change.

Archiving and Deleting

Archive when you don't want a pose offered any more. It disappears from the library for new looks, but looks that already use it keep rendering with it. Restore it any time from the Archived tab. This is the usual choice.
Delete only when the pose must be gone, for example because you may not use the photo. It removes the pose and its reference photo for good, and looks and moodboard images that used it lose their pose (images already generated stay as they are). You can only delete an archived pose.

If Preparing Fails

The pose shows Failed with the reason. Click Retry to convert the same photo again. A failed attempt is not charged; only a successful one is. If it fails again, try a clearer photo with New pose.

Cost and Roles

Making a pose costs 4 cents, charged once, only when it succeeds. Using a pose doesn't change what a render costs; renders are charged as usual, also when you generate again with a different pose. Roles, from fewest to most rights: Viewer, Member, Creator, Approver, Admin, Owner.
Member and above can put poses on looks, and archive or restore them.
Creator and above can also make poses (including Create pose on the Plan tab), retry them and rename them.
Approver and above can also delete poses.

See Team Members & Roles for what each role can do.

Q&A

Does the pose also set the camera angle? Not reliably. The model aims for the camera height in the mannequin, but a view from straight above or below often doesn't carry over. If the angle matters, check the first render.

Does the pose adapt to my fashion model's body? Yes. The look keeps its fashion model; the pose only sets the position. A model with very different proportions from the person in the photo may hold the pose slightly differently.

The photo shows only part of the body. What happens? The mannequin shows the same part. For a full-body look, use a full-body photo; otherwise the model fills in the rest freely.

The photo has several people. Which one is used? Use photos with one person. Poses are made for one model; a photo with several people may produce an unexpected mannequin.

Can I mirror a pose? Not in Brandmachine. Flip the photo before you upload it and make a new pose.

Will the model hold my product, for example a bag? The mannequin has empty hands, so the pose itself never adds an object. How your products are worn or carried comes from the look's products and instructions, as before.

What about long coats, wide skirts or dresses? The pose sets the body position; the garments come from your products. Where a garment hides the legs, the leg position matters less and may not show.

My look has a different shape than the pose photo. Is that a problem? The look renders at its own shape. If the shapes differ a lot (for example a wide photo of a lying pose on a portrait look), the figure may be placed or cropped differently than in the mannequin.

Can I use a photo of a seated pose? You can, but it was not part of our testing. A chair or bench in the photo isn't part of the mannequin, so check the result.

The render doesn't match the pose. What do I check?

Compare the photo and the mannequin on the pose's page; if the mannequin is off, make a new pose.
Remove body-position and camera text from the look's Instructions.
Check that the Look Reference (if any) doesn't show a very different pose.
Generate again: like any render, results vary.

Can I use poses in Product Studio? Not yet. Poses work on campaign looks.

Can several looks use the same pose? Yes, as many of your team's looks as you like.

Can other teams see my poses or photos? No. Poses and their reference photos belong to your team.

Why does generating say "The pose is still being prepared"? The pose isn't ready yet. Wait a minute and generate again.

Why does generating say "A pose needs a fashion model"? Poses position a model, so the look needs one. Pick a fashion model for the look.

For further help, contact support@brandmachine.shop.