Shane Hulbert Artist & Academic | info@shanehulbert.com

Compositing with Masks and Stencils in Live Picture

Compositing with Masks and Stencils in Live Picture

Hulbert, S. 2002, Compositing with Masks and Stencils in Live Picture. EFX Art & Design [Xperimental], issue 35, Summer 2002. Studio Matchbox, Stockholm, Sweden.

As a non-pixel editing program, Live Picture makes extensive use of layer masks to edit and manipulate images.  In fact, an image insertion layer in Live Picture really is just an image with a mask inserted at 100% – if the mask was inserted at 0%, the image would still be there, but would not be visible due to the mask.  If you want to blend one layer with another layer, then the mask is inserted at less than 100%.  If you only want part of the mask to show (regardless of the percentage), then you have to create a stencil for the mask to show through.  This principle is the underlying theory behind the way masks and stencils work in Live Picture.  You don’t alter the image, you edit the mask or stencil.

It therefore stands to reason that Live Picture would offer extremely sophisticated ways to create, edit and manipulate masks and stencils.

Creating Screen Masks / Image Transfers:

The screen mask / image transfer technique uses the mask of one image to define the shape and appearance of another image. The technique is visually similar to the traditional method of chemically removing images, then transferring them, either through rubbing or pressing, onto another image or medium.  The process is quite simple, and is a good way of understanding the functional differences between masks and stencils in Live Picture.

Live Picture layers: In Live Picture a layer can either be selected or active.  Selected layers are for positioning and using the Layer Menu commands, whist active layers are for editing.  A selected layer is any layer in the stack bordered in blue, whilst a layer is active when it is the uppermost visible layer surrounded by a grey border in the layers palette.  Single click a layer to select it, double click a layer to activate it.

Step 1: Define the area of the mask:  Insert the image that will create the mask (the transfer image) as an Image Silhouette layer (Create > Image Silhouette…). For this example I used a pencil rubbing produced by hand on watercolour paper then scanned.  The paper needed to be coarse so small areas within the pencil were left unfilled (white). The white space within the pencil rubbing emphasises the effect, and provides the illusion of dimensionality and texture.  Use the Auto Brush Tool, set the inside control area, adjust the tolerance to between 10% and 20%, then click the pencil areas of the image.  Ensure that only the pencil areas are highlighted as the inside.  Set the outside control area and fill the rest of the image.  Calculate the mask using the Color Compensation computing mode.  Set the Edge Precision to High and the maximum transition edge to the default 20 pixels.

Step 2: Insert the source image: This is the image whose shape will be defined by the mask created in Step 1 (Create > Image Insertion…).  Place this layer behind the transfer layer.  Make sure the source image layer is selected, then switch to positioning mode (single click on the position mode icon above the toolbar) and move the layer until the source image is placed in the required location under the transfer image.  When lining it up make sure the top mask is covering all of the source image in the areas you want it to.

Step 3: Copy the mask of the transfer layer to the stencil of the source layer.  Select the source image, click and <option> drag the mask from the transfer image onto the stencil of the source image.  To copy the mask hold down the option key as you drag the icon from the mask to the stencil.  Delete or turn off the transfer image layer (select the layer – then command D to remove it from the FITS file).

Creating a Reverse Stencil:

Unlike most forms of Live Picture masking, reverse stencilling is not based on defining specific inside or outside areas of an image silhouette.  Reverse stencil uses the Luminance computing mode to create an inverted stencil that only reveals lines.  These lines are the thickness of the gradient between the inside and outside areas set in the luminance mask setting histogram.

Step 1: Choose an image with lots of random detail, or a specific form whose outline you wish to isolate.  In Random Inventory I used old damaged photographs. The effect is more suited to shadow values, so be sure to choose an image with plenty of dark areas.  Shadow areas create strong, defined lines, whereas highlight areas create lines that are often difficult to see.

Step 2: Insert the image as a Silhouette Layer (Create > Image Silhouette…).  Click on the compute mask icon.

Tip: You can use the eyedropper tool to work out where certain values are in the image, for example, if you are wanting to define a specific shape, run the eyedropper over the dark areas of that shape to determine what the levels are, and use those levels in the next step.

Step 3: Set the Computing Mode to Luminance and click Preview. Luminance is used to define global areas based on numerical values (levels).  In the Mask Setting Histogram define the values of the mask using shadow values, and keep the gradient between the black and white sliders to a minimum (between 5-10 levels), as it is the values between the black and white sliders that are actually being defined in a reverse stencil.  To do this place the black slider on any value below 75 (for example 45) and the white slider on a higher value within 10 levels (for example 55).

Tip: Remember that when using the Luminance computing mode and the mask setting histogram, any value above the white slider is included in the mask, any value below the black slider is excluded from the mask, and the distance between the black and white slider is the gradient between what is included and what is excluded.  The sliders may be placed anywhere on the histogram.

Step 4: Copy the mask to the stencil.  Note that if the layer is active it must first be de-activated in order to access the mask and stencil of that layer.  To do this click once on the background layer. To copy the mask onto the stencil hold down the option key as you drag the mask over the stencil.

Step 5: Invert the stencil.  The areas in the image that are now visible will only be a defining line, the softness and width being determined by the difference between the inside and outside areas set in the luminance mask setting histogram.  If these values are close together, ie. less than ten pixels, then the reversed Stencil will appear as random solid lines.

Step 6: Insert the mask at 0% in areas that need to be removed from the image.  This is especially effective if you only want a specific shape that you were unable to define when making the mask.

Step 7: Optional: Duplicate the layer and rotate it using the positioning controls.  This will increase the number of lines as well as create repeating patterns and forms in the composite.  To duplicate a layer hold down the option key as you drag it above or below itself in the layers stack.  To rotate (or flip, move, skew etc) a layer click on the Mode icon (above the toolbar) to enter the Positioning Mode.  Manually re-position the image or enter in numerical values.

Step 8: Add a Colorize Layer (Create > Colorize…).  Using the Colorize layer will intensify the effect of the outlines.  Either fill or manually brush the colorize layer in using the darken mode at anywhere between 60% and 100%.  If you use the brush to manually insert the colorize layer you can control the placement of the colorizing, which helps to create depth in the image and enhance subtle shifts in contrast and tone.

Variants:

In Step 3 set the brush tool on auto and select the shadow values as the inside area and the remaining values as the outside area.  Once this is done click on the Compute Mask icon and set the Computing Mode to Color Compensation.  This method generates lined areas of various opacity and gradient, creating a softer effect.  The Edge Precision setting determines the transparency of the fill areas in relation to the lines.

The default setting Maximum transition edge of 20 pixels, with an Edge Precision setting of High, creates a subtle fill constrained within soft lines.

Setting the edge Precision to Low  gives much softer edges, and less defined shapes, as well as very little tone within the shapes.

Turning Hard Edge turned on produces only hard lines, similar to using the Luminance method, although the lines are much harder.

Leave a Reply

Your email address will not be published.