Adding and Removing Noise
In images, “noise” refers to randomly-colored pixels. Noise can be good or bad; e.g., you can apply noise to computer-generated graphics to make them appear more photographic. You can also use a filter that removes noise to minimize the appearance of tiny scratches or other artifacts present in the source material or introduced during digitizing. Noise commands work with all image modes except Black & White and Indexed.
To Add Noise to Selections:
- Select one or more paint objects to adjust. Select an area in one image to adjust the selected area only.
- Choose Image | Filter | Noise | Add Noise.
- Enter 1 to 999 for Amount, or use the slider, to specify how far the color of the noise can vary from the original color.
- Choose the Uniform or Gaussian distribution option:
- Uniform: To apply colors randomly picked within the Amount specified. Canvas X Draw evenly distributes the color of the noise across a range of colors. This option gives the smoothest effect.
- Gaussian: If you want the noise to favor lighter and darker colors within the specified range. This option creates a more pronounced effect than Uniform.
- Select Monochromatic if you want to add noise of different brightness levels of the original color.
- Click OK to apply the noise settings.
Removing Noise from Selections
You can remove noise from an image or selection using the Median, Despeckle, or Dust & Scratches filters. The Median filter removes noise by averaging the color of pixels. The Despeckle and Dust & Scratches filters remove noise by selectively blurring regions of the selection.
To Use the Median Filter:
On a pixel-by-pixel basis, the Median filter applies the median color value of all pixels within the specified radius. Although the filter ignores extreme values in its computations, higher radius values can still wash out an image.
- Select one or more paint objects to adjust. You can select an area in one image to adjust the selected area only.
- Choose Image | Filter | Noise | Median.
- Type a value from 1 to 16 in the Radius text box, or drag the slider. Smaller radius values produce subtler effects.
- Click OK.
To Use the Despeckle Filter:
The Despeckle filter can remove defects such as dust and other speckling in images. The filter blends pixels with the lightness values of neighboring pixels. It’s a good idea to select areas that need correction before applying the filter.
- Select an area in an image to adjust. If you don’t make a selection, the entire image is affected.
- Choose Image | Filter | Noise | Despeckle.
- Drag the slider or type a number in the text box to set the Threshold value. Higher values produce greater blending of pixels.
To Reduce Dust and Scratch Marks:
The Dust and Scratches filter can remove dust specks by replacing a pixel’s value with a median value. The filter does not change a pixel’s value unless the absolute value of the difference of its gray value and the median gray value of its neighborhood is greater than the Threshold. Larger numbers of pixels are replaced by the median value when the Threshold is low.The practical effect of this is that larger pixels regions (larger “specks”) are reduced or removed with a lower Threshold. Therefore, you can use the Threshold to control the size of the artifacts you want to remove.
- Select an area to adjust in an image. If you don’t make a selection, the entire object is affected.
- Choose Image | Filter | Noise | Dust & Scratches.
- Enter a value from 1 to 16 in the Radius text box. Smaller radius values produce a subtler effect than larger ones.
- Type a value from 0 to 255 in the Threshold text box. After entering the settings you want, click OK.
Smoothing Video Images
Because video images contain two interlaced pictures, you can sometimes see a slight banding effect in images acquired from video-recording devices. You can correct this by using the De-Interlace filter and then applying the Unsharp Mask filter. De-Interlace works with all image modes except Black & White.
To Smooth Video Images:
- Select one or more paint objects to adjust. You can select an area in one image to de-interlace the selected area only.
- Choose Image | Filter | Video | De-Interlace.
- Click Odd fields or Even fields to select bands to eliminate.
- Choose a replacement method for the eliminated pixels:
- Duplication: To fill the area by inserting a copy of an adjacent band.
- Interpolation: To fill the area by inserting intermediate color values based on the color values of neighboring pixels. This option creates a smoother, more accurate fill than Duplication.
- After entering the settings you want, click OK.