Selecting through the Alpha channel

When stacking layers, Autograph builds a selection map based on both the position of each element in the Stack from bottom to top, and each layer’s Alpha channel. This lets you select elements you actually see in the Viewer, without worrying about layer bounding boxes and transparent pixels that might prevent you from selecting a visible element. Clicking within the Viewer goes through each layer's Alpha channel to eventually reach a pixel with an opacity higher than 0.5 and selects the corresponding layer.

This selection takes transparency into account after applying Modifiers such as a Color Difference Key, used to remove a green screen.

When you move your mouse over the Viewer and the cursor enters an area of this selection map, a flash appears on the layer that will be selected if you click it. A yellow outline defines the selected layer(s). It’s possible to add new layers to the current selection by pressing the SHIFT key and clicking on them.

Blending modes and selection map

The Alpha channel is not the only factor used when creating the selection map. The layer blending mode used also defines areas through which the click can pass through.

In the following example, a dust image without an Alpha channel is added to the top of the Stack in Screen mode. The dark parts let you see the rest of the Stack while the brighter parts hide it.

Screen mode considered the black parts as transparent and Autograph takes this into account when creating the selection map. So it’s possible to click through these disappearing black parts to select lower layers in the Stack. This selection is accurate to the pixel. By moving your mouse over the parts of the dust where the relative transparency is higher than 50%, you can select the dust layer by clicking it.

Selecting based on bounding boxes

For layers containing very thin elements such as text, the transparency-based selection may not be very convenient. Autograph provides a selection mode based on layer bounding boxes instead, which is available in the top bar of the Viewer.

This selector contains two modes:

• Select with Alpha channel (default)
• Select with bounding box

In this second mode, the selection map is made from the full bounding boxes, without checking transparency.

In order to select layers more easily, especially those hidden behind others, you can also rely on the Visual stack.

Connecting to the Timeline

When hovering over a layer, it lights up in the Timeline at the same time so you can see its position in the Stack.

Moving layers in the Viewer

Once one or more layers are selected, placing the mouse over the Viewer brings up a transformation widget.

By default, it displays:

• A blue frame to resize the layer
• Controllers to change the size and position of the layer

Pressing the Alt/Option key replaces the size and position controllers with alternate ones for modifying rotation and skew. Pressing this key also switches the position icon at the top left of the Viewer with a rotation icon. Alternately, you can click directly on this icon to switch between controllers.

Moving options

At the top left of the Viewer, there are icons that represent layer movement options.

From left to right:

• Individual origins/ Median anchor point: when a rotation is performed, it is defined around the respective anchor points of each layer or according to a median point.
• Local/Global mode: defines if the widget's moving arrows should be rotated and skewed.
• Uniform Scaling ON/OFF: determines whether or not to keep the original aspect ratio of the layer when scaling.
• Center Scaling ON/OFF: defines whether or not scaling should be done through the center of the layer.
• Selection mode: switches between Alpha channel and Bounding boxes.
• Snapping angle: when clicked, activates snapping and defines the steps in degrees during a rotation.
• Tool Spacing: allows you to increase the distance between the widget tools.
• Ripple mode: allows you to move an anchor point while keyframes have been added to the position. These will then all be recomputed.
• Pixel snapping: when moving a layer in the Viewer, the position values will be rounded

Transform

After linking a graphic element to a layer’s Source parameter, you can use the Transform section to define the Positioning, Rotation, Scale and Skew of the layer using it.

You can access this section either by unfolding a layer in the Stack or by clicking on it to display its parameters in the Properties panel:

Transform sub-section

In this section, you will find the following parameters that use Autograph's specific coordinate system:

• Position: 2D param, in unified mode by default
• Scale: 2D param, in single mode by default
• Anchor Point: 2D param, in unified mode by default
• Rotation: 1D param, in single mode only
• Skew: 2D param, in unified mode by default
• Skew order: non-numerical param; defines if the X Skew will be applied before or after the Y Skew
• Amount: weight of the global transformation. Can also take values outside the range 0-1, even negative ones
• Inverse Transform: allows you to invert the transformation matrix

Anchor Point

The Anchor Point is the reference point used when determining a layer's Position values. Since an image is made up of a large number of pixels, saying that it is currently at Position (160,80) for example, does not make it possible to determine what it is in relation to - is it its upper right corner, lower left corner, or its center that is at these coordinates?

The values displayed in Position are therefore those of the Anchor Point positioned at the center of the image and with values (0,0) by default. This is also used as a pivot point during a rotation or as a suction or expansion point during a scale change.

In order to view the position of the Anchor Point, simply position the cursor over the Viewer with the Move layers tool activated. It is represented by a small plus sign in the middle of the Transformation Widget.

Changing the Anchor Point values will not change the position of this reference point; but, it will cause all image pixels to shift accordingly.

Let's use the example of an image with a size of 600×300 pixels. By dragging it into a composition, it will be placed in the center of the composition, with an Anchor Point in the center of the layer source.

If you move this layer, the Position values will be modified, but not those of the Anchor Point which will remain positioned at the center of the Source.

If your image is 600 pixels wide, that means it extends 300 pixels to the left and 300 pixels to the right. By entering the value 300 in the X component of the Anchor Point parameter, we place it on the right edge of the image; but, since the Position values will remain the same, all Source pixels will move to the left, until reaching the right edge.

Anchor Point tool in the Viewer

The Viewer has a tool that allows you to easily move the Anchor Point to a corner, an edge, or the center of the image, without moving the image pixels. New values will be calculated for both the Position and the Anchor Point.

You can choose to move it according to:

• The Source of the currently selected layer
• The composition coordinate system

Transformation amount

Amount is the last parameter of the Transform section and it allows you to determine the influence of this transformation. By setting this value to 0, the transformation has no effect and the layer returns to its original place when it was added to the stack - in the center of the composition.

Note that you don't have to limit yourself to the 0-1 interval. If a layer in the center of the composition is moved 200 pixels to the right and the Amount is set to 2.0, then the movement will be twice as much, bringing the layer 400 pixels to the right.

Similarly, setting this value to -2.0 will not only move the layer twice as far, but it will also move it in the other direction, to the left at -400 pixels.

Transformation Matrix and Modifiers

All transformation parameters that make up the Transform section, including Amount, end up applying what is called a Transformation Matrix on the layer. It is a mathematical object that allows you to store changes in Position, Rotation, Scale and Skew in a single element.

When you apply Transform Modifiers, such as Card3D, Corner Pin or Follow Path, they will act directly on this matrix, and concatenate between them. Once this resulting matrix is obtained, Autograph applies the corresponding transformation in one go, which allows you to have only one rasterization, thus avoiding the appearance of pixelation effects.

But a very important point to keep in mind when adding these transformation Modifiers and ONLY in the case of Transforms:

Concatenating matrices is always done starting from the last one, going back gradually to the first one. Autograph chooses to represent these Modifiers in the same order and not take any liberties with this mathematical tool and respect this norm.

As a result, there is a major difference with all the other Modifiers in the software. The list must be read from bottom to top in order to know the order in which these Transform modifiers are executed.

Inverse Transform

This option allows you to invert the transformation matrix. It is particularly useful when one transformation is linked to another, allowing for compensation when stabilizing a tracked plate, for example.

3D Transformation

When a layer has its 3D option activated, the current Transform section is replaced by another one with 3D parameters. 2D transformations still exist and can be combined with 3D transformations. To learn more about this, please refer to the 3D Mode section.