Image - Convert To

The Convert To command converts images into different methods of handling color: RGBa, RGB, Grayscale and Palette images. Convert To changes the image data permanently. The Convert To dialog will also report the estimated size of the resulting image. The Convert To command may also be used to convert linked images into local images within the project.

 

Target

Desired type of image: Compressed, Grayscale, RGB, RGBa or Palette.

Palette

Choose one of the preset palettes for a specific usage, when converting to a Palette image.

Metric

Method for computing proximity of colors to each other when judging if colors are close to each other in the color cube. Displayed when converting to palette images. Use Euclidean: the other settings are rarely used metrics provided for expert use in special situations.

Colors

The number of colors to use to create the image. Displayed when converting to palette images.

Dither

Check to use dithering. Displayed when converting to palette images.

Ratio

Compression ratio. Displayed when converting to compressed images. Larger Ratio numbers will result in greater compression at the cost of lesser image quality. A ratio of 10 will create a compressed image that is one-tenth the size of the original image.

Save result as new component

Instead of converting the same image component, places the converted image into a new component.

 

The Target box will contain only those choices other than the method currently employed for the image. RGB images, for example, will only have Palette and Grayscale appear as choices.

 

See the Dither topic for information on dithering.

 

Palettes

 

Adaptive

Sample colors appearing most frequently in the image and use these to choose a best fit to the given number of colors from the entire RGB space.

Linear

Choose colors that are evenly spaced in the RGB cube space for the given number of colors.

Photoshop

Use the 90-color Photoshop swatches palette. Handy for many restricted-color graphics arts images.

System

Use the 256-color Microsoft Windows system palette.

System (Mac)

Use the 256-color Apple Mac system palette.

VisiBone 2

Use the VisiBone web designers palette.

Web

Use standard 216-color palette included by default in all web browsers.

 

Palettes used by other applications and systems are provides so that images destined for those systems can be created using an exact match of palettes.

 

Grayscale and palette images may be converted into RGB images without any loss of color values. Thus, one may take a grayscale image, convert it into RGB, and then convert it back into grayscale without any loss of color data.

 

The reverse is not true: when converting RGB images into grayscale or palette images there will be a permanent loss of color information as the RGB color information is interpolated down into a much smaller set of tones that are available in grayscale or palette images.

 

RGBa are RGB images with an extra channel that allows setting a percentage transparency for each pixel. This is called an "alpha" channel. RGBa images are used when many smaller images are stacked together in a map to achieve photocomposition effects as is done with professional image editors such as Adobe PhotoShop. See the Layers topic for an example.

 

Compressed Images

 

Compressed images are decompressed when converted into other image types. Since compressed images are often very, very large be careful to consider the resulting size of image when converting into other image types. Keep in mind that compressed images also use a sophisticated display algorithm to more rapidly display the compressed data. When decompressed and converted into other image types the speed of redisplay available with a compressed image will no longer be available.

 

When converting images into Compressed form, although the visual appearance of the image will be almost exactly the same (or even may appear to be enhanced) the actual information content of the image will be reduced. The high-speed compression algorithms used are normally lossy algorithms based upon wavelet compression or similar compression strategies. Although they preserve or enhance visual appearance, the specific pixel values will change.

 

images\img_compressed_before.gif

 

Consider the RGB image seen at high zoom level above. We have zoomed in so that individual pixels are visible. The scene illustrated is the Northern part of Baja California in Mexico and the Sea of Cortez.

 

images\img_compressed_after.gif

 

After conversion to a compressed image, the scene at the same zoom level appears above. Note that individual pixels are no longer discernable because the compression technology used synthesizes a scene using full screen resolution to utilize however many pixels are available in the computer monitor's display.

 

Although the scene looks very realistic (considering the high zoom into what was originally a low-resolution image), it is very important for scientific purposes to understand that the information content in the compressed image is less than that of the original. The compressed image only looks good visually because a clever decompression-on-the-fly algorithm was used to synthesize a visually appealing image from a smaller number of bytes.

 

When converting to other image types a compressed image can expand into a dramatically larger image in terms of storage space required. Pay careful attention to the estimated size reported by the Convert To dialog for the resulting image to avoid creating unexpectedly large images.

 

Tech Tip

 

Read only images (for example, from an Enterprise server if Enterprise Edition is being used) can be converted to other types of images if the Save result as new component box is checked.

 

Attempting to convert an image to a compressed format using Convert To will display any error messages returned by the ECW / JPEG2000 compression code should the conversion fail.