How to load a facial scan and work with the 3D camera in Cympha Studio

Registering a facial scan to the dental scans and CBCT in Cympha Studio gives repeatable 3D camera views, which you save to the project gallery for the dental case.

Audience: Clinicians and case reviewers who want to see the patient's face alongside the dental and CBCT data — for smile-design review, communication, and consistent screenshots across visits.

You'll come away knowing: how to bring a facial scan into Cympha Studio, register it to the dental scans (and CBCT when one is loaded), use the 3D camera to take repeatable views, and ship the result into the project gallery so it travels with the case.

Time required: Around ten minutes for the first scan you load. Two or three minutes per case after that.

Before you start

You need:

A facial scan of the patient as a single mesh file — .stl, .ply, or .obj. If the scan has a texture, you also need the matching .png, .jpg, or .jpeg image (the mesh must carry UV coordinates for the texture to map; .ply with UVs is the most reliable, .stl typically cannot be textured this way).
A reference layer to register against. This is usually the patient's upper arch (an intraoral scan you've already loaded) and, when available, a CBCT volume in the same patient frame.
Cympha Studio open at studio.cympha.com.

If you also plan to align the facial scan to bone (not just to the teeth), open the CBCT first so it's in the scene before you start registration — Studio takes a visibility snapshot when registration begins, and a hidden CT cannot be used as a registration target.

Step 1 — Switch to the Facial Scan workspace

In the top bar, open the Workspace nav and pick Facial Scan. Studio narrows the hamburger menu to just the file actions you need for a face workflow — DICOM, optical scan, facial scan, pictures, project actions, reset camera, clear scene — so the menu stays short.

Switching workspaces does not change anything in the viewer or in any layers already loaded. It only filters the menu. Your choice persists across sessions in this browser.

If you open a .stl / .ply / .obj file from a link (a ?model= URL) while in a narrower workspace, Studio switches you automatically to a workspace that supports optical scans. For files you load by hand, switch the workspace yourself first.

Step 2 — Load the facial scan

Three ways, all equivalent:

Hamburger (☰) → Open optical scan file… and pick the mesh. To bring in the texture in the same step, Ctrl-click the matching image inside the same file picker so both files are selected together. The mesh must have UVs for the image to map.
Top toolbar → Add model — same rules, same Ctrl-click trick for adding a texture image.
Drag-and-drop onto the viewer — drop the mesh by itself, or drop exactly one mesh plus one image. Studio refuses to mix multiple meshes with a texture in a single drop and explains why in the status line at the bottom of the viewer.

The scan appears in 3D in the coordinate frame it was exported in. Most facial scanners write the scan in their own scanner space, so the face will not yet line up with the dental scans — that's the next step.

Step 3 — Register the facial scan to your reference

You align the face to the dental scan (and to CBCT bone when available) with Register to… on the mesh that should move:

Right-click the facial scan in the 3D view. The context menu opens.
Choose Register to…. The Register model panel opens. The facial scan is the source (the layer that will move when you press Register).
Pick at least three landmark pairs in strict alternation: a point on the source (facial scan), then the matching point on the target (the upper arch mesh, or CT bone if the volume is visible). Source landmarks show as reddish spheres, target landmarks as greenish spheres.
Useful pairs on a face-to-arch align: the labial midline at the incisal edge of 11/21, the cusp tip of 13 and 23, and a posterior landmark such as the cusp tip of 16 and 26 if visible through the lips. Spread the points around the arch — clustering them all in one spot makes the rigid solve slide.
Need at least three complete pairs (six clicks). Until you have three pairs, Delete removes the last pick. Once you reach three pairs, deletion is disabled — use Cancel to restart picks.
Click Register. Studio computes a rigid transform (rotation + translation, no scaling) and snaps the facial scan into place. If you registered against CT, turn on Show in 2D for the facial scan (right-click menu) so it appears in MPR overlays.

If you started picks by mistake, Cancel restores the facial scan to the exact pose it had when you opened registration; Close dismisses the panel without applying anything (the mesh stays wherever it currently is).

Step 4 — Use the 3D camera to set repeatable views

Once the face is aligned with the rest of the scene, the main 3D view is your "3D camera." Mouse controls are the same across Cympha Studio:

Left-drag rotates the camera around the scene.
Middle-drag pans.
Right-drag zooms (rubber-band style), or use the mouse wheel.

For consistent framing visit-to-visit, use the arch toolbar at the top-center of the viewer. When a GuideMia OrthoPlus case is loaded, it offers Both / Upper / Lower jaw modes plus front and lateral orthographic presets. With a CBCT loaded but no case, you get front / patient right / patient left orthographic buttons instead.

The orthographic presets are the closest thing Studio has to a "canonical camera position" — they put the camera at a known angle every time, which is what makes screenshots comparable across visits.

Step 5 — Capture and store the view

When you have the camera where you want it, click the Screen capture icon (camera) on the top toolbar. Studio writes one PNG of the viewer column. With CBCT loaded and MPR active, the capture stitches the MPR slice row and the main 3D row into one image as they appear on screen; with CT off, it captures only the main 3D view.

A browser prompt asks for an optional comment — type a note such as "post-op week 6, full smile" so the image is self-describing later. Press OK to save, or Cancel to discard.

The image lands in the Media gallery (open it from the pictures icon on the top toolbar). Each project picture shows the filename plus the comment you typed beneath the thumbnail. Right-click a project picture to remove it; case media that came from a .gvw archive is read-only in the gallery.

To save the entire workspace — facial scan, registered transforms, captures, and notes — use ☰ → Save project or Save project as… to write a Cympha .cyp project. The captures are stored under <Name>_CypFile/Pictures/ alongside the manifest, so the gallery rehydrates exactly as you left it when you reopen the project.

What happens next

The next time you open this case in Studio (☰ → Open project… and pick the folder containing the .cyp), the facial scan returns at its registered pose and the gallery is repopulated with your captures and their comments. (Opening a project doesn't change your workspace — the workspace choice persists per browser, so you'll still be on Facial Scan if that's where you left it.) You can take a new round of captures at the same orthographic presets and compare them side by side with the earlier ones in the gallery.

Troubleshooting

The facial scan loads but isn't textured. Studio textures only when the mesh carries UVs and you select exactly one mesh plus one image in the same picker (Ctrl-click). .stl typically lacks UVs in this pipeline. Re-export the scan as .ply with UVs and try again.

Register to… won't accept clicks on CT. Studio takes a visibility snapshot when registration starts. If the CT was hidden when you opened Register to…, it stays hidden through the target stage. Cancel registration, turn on the CT in the model browser, and start registration again.

The facial scan rotates the wrong way after Register. That means the three landmark pairs were clustered too tightly — the rigid solve can slide along a flat region. Cancel, place pairs that span the arch (left, right, anterior, and ideally a vertical landmark such as the nasion or chin point if visible on both layers), and re-run.

The capture shows MPR slices I don't want. Toggle off the CT or switch out of the MPR layout before capturing — Studio captures whatever is on screen in the viewer column.