1.2 Tracking Trees in PFTrack


This tutorial is part of Level 1. An Introduction to Matchmoving in PFTrack of The Pixel Farm’s Training Course for PFTrack. Find out more and register for the next available live class.

The Introduction to Matchmoving in PFTrack class covered three ways of creating tracking tree nodes in PFTrack and during the course of the class, nodes have been created to form a tree for a simple tracking task. This guide shows you some of the more advanced features in PFTrack’s tracking trees, which you can use to simplify and speed up your workflows.

01. The Custom Category: Where to Save Your Favourite (or Most Used) Nodes

Nodes in PFTrack are grouped into categories to make them easier to find. For example, the Auto Track and User Track nodes are in the Tracking category, whereas nodes like Camera Solver or Survey Solver are in the Solving category. There is one category, however, which, by default, is empty: The Custom  category. It is for you to decide which nodes should be in this category; Custom can act as a collection of your most used nodes, leaving them all in one handy place. But don’t worry, the nodes will still be part of their original groups as well.

Adding Nodes to the Custom Category

To add a node to your collection, simply drag it with the mouse over the group icon and drop it into the group.

02. Branching Out and Exploring New Avenues

One of the biggest advantages of a tree based workflow is its flexibility, which gives you the freedom to explore new approaches to a task, without risking to lose the work you have already completed. Branches provide a natural way of doing just that in a tree based workflow. Every output connector can be connected to multiple nodes, allowing you to branch out from anywhere in your tree.

03. Open to New Connections

As explained above, every output connector can be connected to multiple nodes, but is this also true for input connectors? The answer is no. Let’s take the Auto Track node as an example. The Auto Track node takes a clip as input and outputs the same clip with tracked features  (There is a bit more than this going on behind the scenes but we keep it simple for the sake of this example). But what if we could connect more than one clip into the Auto Track node? The Auto Track node could then track all these clips and output them all. However, that would be the same as having one Auto Track node for each clip; and having an Auto Track node per clip would be the preferred of these two options. It is easier to understand what’s going on when looking at your tree, and you couldn’t accidentally change the trackers for one clip while working on the other.

But does that mean that every node only has one input? There are some nodes that can work on two or more clips, but each clip will have its own connector. Unlike output connectors, input connectors can only ever have a single connection.

There are nodes that require multiple clips to work properly. One example would be Stereo nodes such as Stereo Auto Track and Stereo Camera Solver. A stereo clip consists of a left eye clip and a right eye clip and the stereo nodes need both to work. These nodes will have two input connectors, one for each clip.

The Spherical Orient and Spherical Render nodes, even have 6 input connectors, one for each face of a 360º cube map.

Then there are also some nodes that can have a varying number of inputs. As they do not require a fixed number of clips, these will normally start with a single input connector. The little ⊕ at the top right corner indicates that additional input connectors can be created as needed.

To add an additional input, simply drag another connection from another node’s output connector into the node. If all inputs are already connected, a new one will be created.

In any node with multiple inputs, the Current clip menu allows you to select on which clip to work on.

The selected clip will also be visible in the Cinema, and its path highlighted in the Tree View.

Most nodes (but not all) that have multiple inputs also have the same number of outputs. If a node has the same number of input and output connectors, they usually correspond, and the first connector outputs the data for the clip in the first input, the second connector the second clip and so on..

04. Improving Clarity Through Information

Once you have multiple nodes of the same type in your trees, it may be harder to remember what the purpose of each node was. To help you keep track (no pun intended) of your trees, you can give your nodes meaningful names to make it easier to understand your tracking trees. You can change a node’s name in the Notes panel.

If a node needs an even longer explanation than is possible with changing its name, you can easily store additional notes with each node.

05. Belonging Together

As your project and trees grow, finding your way through the tracking tree can become harder and harder. You can simplify the layout of your trees by logically grouping nodes. Groups can be collapsed to free space in your Tree View, and expanded again if you need a closer look at the group’s members.

To create a group, select the nodes you want to group and click the nodePageGroup button.

To remove a group again, simply select its nodes and click the nodePageGroup button again

You can change a group’s name in the Notes panel of any of the group’s nodes.

06. A New Blank Page

When, despite having all your nodes neatly organised in groups, your Tree View still looks too cluttered, but you don’t want to start a new project just yet, you can create a new page for your new trees.

If you want to close a page again, make sure it is empty (by deleting all its nodes), then click the nodePageMinus button at the top

07. Reuse Your Work

Copying and Pasting Nodes

You can copy and paste nodes and even whole trees within a project when you want to reuse (parts of) your work.

Pasted nodes will have the same parameters that worked in their source node, and where possible, some pasted nodes even keep their content (e.g. trackers in an Auto Track node), providing a shortcut to trying slightly different approaches.

Saving Presets

Often used tree layouts, but not individual node parameters, can also be stored as presets and reused. Stored presets will be available across projects and are stored as XML files, which can be copied and shared.

To create a preset, select any node in a tree and click the storeTreePresetButton button. This will create a preset of the complete tree, except Clip Input nodes. The presets’ icons show a miniature representation of the saved tree.

Double click on an icon to rename the preset.

Click the treePresetCurrent  button to recall the selected preset in the current Tree View page or click the treePresetNew button to create a new page with the preset.

08. Conclusion

Tracking Trees are the backbone of any work completed in PFTrack, and this document introduced the many ways of how the Tree View can help you to master  your trees. From key concepts such as branching and multiple inputs, to duplicating and grouping nodes, these features are just a mouse click away.

Related Posts


Submit a Comment

Your email address will not be published. Required fields are marked *

Share This

Share This Post On: