Learn Phoduit

9.2 Geometry

9.2.1 Crop

The Crop node cuts a rectangular region out of an image. It’s usually used for fine tuning composition.

See Cropping in Chapter 4. Essentials for an example on how to use the Crop node.

Input

Image
Image to be cropped.
Left
Offset of the remaining region after crop from the left edge in pixels.
Top
Offset of the remaining region after crop from the top edge in pixels.
Right
Offset of the remaining region after crop from the right edge in pixels.
Bottom
Offset of the remaining region after crop from the bottom edge in pixels.

Output

Image
The remaining rectangular region of the input image as specified by the left, top, right, and bottom input fields.

Mode

The Canvas View combo box is locked to the input image and the Canvas becomes interactive so you can move or resize the cropped region. An exact region can be specified in the Crop dock window. You can also constrain the size of the cropped region to a specified aspect ratio in the Crop dock window.

9.2.2 Lens Correction

The Lens Correction nodes corrects or distorts images based on various camera lenses. It’s usually placed right after the Demosaic node to correct lens distortion in raw images.

Use of this node prevents transform aware editing on any nodes on the input side of this one while any nodes on the output side are displayed by the canvas.

Input

Image
Image to be corrected or distorted.

Output

Image
The input image corrected or distorted.

Mode

You can select a camera model and lens model as a basis to correct or distort your image.

9.2.3 Mirror

The Mirror node flips the input image either horizontally, vertically, or both horizontally and vertically.

Input

Image
Image to be mirrored.
Axis

The direction the image should be mirrored.

  • Horizontal: Mirror from left to right. For example, the top left corner will now be the top right corner.
  • Vertical: Mirror from top to bottom. For example, the top left corner will now be the bottom left corner.
  • Both: Mirror left to right and top to bottom. For example, the top left corner will now be the bottom right corner.

Output

Image
The input image mirrored along the specified axis.

9.2.4 Perspective

The Perspective node transforms the input image so that it looks like it was created from a different perspective. The area not covered by the transformed image is made completely translucent. It’s useful for straightening the edges of architecture so they look parallel.

Input

Image
Image to be transformed.
Size

Size of output image.

  • Original: Crop to the input image’s size, even if perspective transform would make the image smaller or larger.
  • Fit All: Fit the entire perspective transformed image, even if smaller or larger than the input image.
Method

Interpolation algorithm.

  • Nearest Neighbor: Produces blocky images but very fast. Recommended when interacting with this node.
  • Bilinear: Produces smooth images that can sometimes appear blurry.
  • Bicubic: Produces smooth images and retains some sharpness but very slow. Recommended when exporting.

Output

Image
The input image with its perspective changed.

Mode

The Canvas View combo box is locked to the output image and the Canvas shows interactive handles that you can left click and drag to change the perspective. This can be very slow on older computers so it’s recommended that you set the Method field to “Nearest Neighbor” whenever you are changing the perspective.

9.2.5 Rotate

The Rotate node spins the input image around its center. The area not covered by the rotated image is made completely translucent. It’s often used to straighten photographs.

Input

Image
Image to be rotated.
Angle
Angle to rotate the input image in degrees.
Size

Size of output image relative to input image and angle rotated.

  • Original: Use the input image’s size. Angles that are not a multiple of 180 will cause the image’s corners or sides to be cut off.
  • Fit All: Fit the entire rotated image without cutting off any corners or sides. Angles that are not a multiple of 90 will produce images with more pixels.
  • Crop: Use as much of the rotated image as possible without including the translucent background.
Method

Interpolation algorithm used.

  • Nearest Neighbor: Produces blocky images but very fast.
  • Bilinear: Produces smooth images that can sometimes appear blurry.
  • Bicubic: Produces smooth images and retains some sharpness but very slow.

Output

Image
The input image rotated by the specified angle.

90 Degree Rotations When an image is geometrically transformed, there is almost always a loss in quality. Rotating an image by an angle that is a multiple of 90 degrees is one exception – there is no loss in quality.

9.2.6 Scale

The Scale node changes the size of the input image.

See Resizing in Chapter 8. Output for an example on how to use the Scale node.

Input

Image
Image to be scaled.
Unit

Unit that width and height are specified in.

  • Percentage
  • Pixels
Width
Width of output image. Specified in Unit set above. Click the Link button on the right to lock or unlock the aspect ratio.
Height
Height of output image. Specified in Unit set above. Click the Link button on the right to lock or unlock the aspect ratio.
Aspect Ratio

If and how the ratio between the input image’s width and height should be preserved when scaling.

  • Ignore: Output image will be scaled to the width and height specified.
  • Keep: Output image will be scaled so that its original aspect ratio is preserved while being no bigger than the width and height specified.
  • Expand: Output image will be scaled so that its original aspect ratio is preserved while being at least as big as the width and height specified.
Method

Interpolation algorithm used.

  • Nearest Neighbor: Produces rough images but very fast.
  • Bilinear: Produces smooth images that can sometimes appear blurry.
  • Bicubic: Produces smooth images and retains some sharpness but very slow.

Output

Image
The input image scaled to the specified width and height.

9.2.7 Translate

The Translate node moves the input image along the X and Y axes. A positive X or Y value makes the output image larger along the specified axis. A negative X or Y value clips the input image along the specified axis but does not change the output image’s size.

Input

Image
Image to be translated.
Unit

Unit X and Y are specified in.

  • Percentage: X is in percentage of the input image’s width. Y is in percentage of the input image’s height.
  • Pixels: X and Y are in pixels.
X
How much to move the image horizontally. Use negative values to move the image left and positive values to move the image right.
Y
How much to move the image vertically. Use negative values to move the image up and positive values to move the image down.

Output

Image
The input image translated by the specified X and Y.