PXLtools — Tool guide

Camera Matchmaker — single-image camera match

Load a still photo onto the canvas, draw axis lines along real-world parallel edges, set the origin, and Camera Matchmaker solves the camera — Horizontal FOV, Pan, Tilt, Roll — from the vanishing points. One click applies the result to a Maya camera with an image plane attached. A three-step gated flow walks you from canvas setup through scene data to output.

Maya v0.4.0 Prototype

The interface

Overview

Early prototype — under active development. Camera Matchmaker v0.3.0-alpha is an early prototype. The UI shown here and the solver behaviour may change between versions. Test on non-critical shots before integrating into a production pipeline.

Camera Matchmaker is an interactive single-image camera matching tool for Maya, inspired by fSpy. Load a reference photograph onto the canvas, draw axis lines along edges you know to be parallel in the real world, set the scene origin, and the tool solves the camera using a weighted least-squares vanishing point algorithm (Guillou et al., 2000).

The solve reads out Horizontal FOV, Pan (Ry), Tilt (Rx), and Roll (Rz) live as you adjust the lines. When you are happy with the result, one click creates a new Maya camera (or applies to an existing one) with the solved values and attaches the reference image as an image plane. It shares the same UI kit as the rest of PXLtools.

Requirements

Autodesk Maya 2025
Python 3.11 (included with Maya 2025)

Maya 2025 already includes PySide6. You do not install it, you do not pip anything, you do not touch your Python environment.

Availability

Camera Matchmaker is part of Cristian's PXLtools production toolkit and isn't publicly downloadable yet. The TurnTable tool is the first public PXLtools release; the rest of the suite is rolling out gradually. Want early access, or to be notified when it ships? Get in touch.

Using Camera Matchmaker

The tool has three collapsible sections: 01 MATCH CANVAS, 02 SCENE DATA, and 03 OUTPUT. Work through them in order.

The step flow

Each numbered section is independent but designed to be completed in sequence. The solve updates live as you draw and adjust lines, so you can see the FOV and rotation values before committing to a camera.

Canvas shortcuts

Scroll — zoom

Scroll the mouse wheel to zoom in and out on the canvas.

Middle mouse — pan

Hold and drag the middle mouse button to pan around the canvas.

F — fit / drag handles — adjust lines

Press F to fit the image back to the viewport. Drag the coloured endpoint handles to reposition any line.

01 MATCH CANVAS

Load your reference image and draw the axis lines that define the vanishing points.

Load an image. Click the canvas drop zone (or drag and drop an image file onto it) to load your reference photograph. Supported formats: JPG, PNG, TIF, BMP, EXR, HDR.
Draw X Axis lines (red). Select X in the axis toolbar and click + Add Line, then click and drag on the canvas to draw a line along a horizontal parallel edge — e.g. a floor edge, window sill, or ceiling cornice going sideways. Add at least 2 lines. The vanishing point is computed live and shown as a red marker.
Draw Z Axis lines (blue). Select Z and add at least 2 lines along depth-parallel edges — edges that go away into the scene. The solver needs both X and Z vanishing points to compute the camera.
Optional: Y Axis lines (green). Select Y and add lines along vertical reference edges (columns, door frames) to constrain the vertical vanishing point.
Optional: Ref line (yellow). Select Ref and draw across a known real-world object to set scene scale for the position solve. Configure the real-world size in 02 SCENE DATA.
Set the Origin. Click Set Origin and click on the canvas to place the scene world origin (0, 0, 0) marker. Used for the camera position estimation.
Read the solve. The readout below the canvas shows Horiz. FOV, Pan (Ry), Tilt (Rx), and Roll (Rz) as soon as the solver has enough lines.

02 SCENE DATA

Optional ground-truth constraints for the position solve.

Camera Height. Real-world height of the camera lens above the ground plane. Enabled by default at a sensible default for your Maya scene unit. Required for position estimation — uncheck if unknown.
Camera Lens. Override the computed focal length with a known lens value (assumes 36 mm full-frame sensor). Leave off to derive from vanishing points.
Measurement Ref. If you drew a Ref line in the canvas, enable this and enter the real-world size of the object the line crosses. This scales the position solve.

03 OUTPUT

Apply the solved camera to Maya.

CREATE NEW CAMERA. Creates a new Maya camera named MatchCam_01 (editable) with the solved FOV and rotation, and optionally attaches the reference image as an image plane.
APPLY TO CAMERA. Applies the solve to an existing camera you select from the dropdown. Choose which of FOV, Rotation, and Position to set.
ADD IMAGE PLANE. Attaches the loaded reference image to the selected camera without changing its transform.

Tip: After applying, use Place Test Object at Origin in the Verify tab of the OUTPUT section to place a reference cube at (0, 0, 0) and look through the solved camera to check alignment against the photo.

Troubleshooting

The + Add Line button is disabled.

Load an image onto the canvas first. The axis toolbar is inactive until a reference image is loaded.

The solve readout shows “—” for FOV and rotation.

The solver requires at least 2 X Axis lines and 2 Z Axis lines to compute a result. Draw both sets of lines and the readout updates automatically.

The Ref line sets scale but the position looks wrong.

Enable Measurement Ref in section 02 SCENE DATA and enter the correct real-world size of the object the yellow Ref line crosses. The Ref line drawn on canvas only registers as scale data when this checkbox is ticked and a value is set.

CREATE NEW CAMERA and APPLY TO CAMERA are greyed out.

The output buttons activate only after a valid solve is available. Draw the minimum required X and Z axis lines until the solve readout shows real numbers, then the buttons become clickable.