Several Further Notes on images.

The resultant final quality and file size of any particular image depends on many things. In truth, there are no concrete rules, only guidelines. Ultimately it will depend on each individual image and what you wish to achieve. Photographic images tend to reproduced better using the JPEG format while illustrative or drawing/clip art images tend to be better as GIFs. If the image can be acceptably reduced to 256 colours or less, GIF is often the best option, particularly in cases where there are large areas of the same colour within the image. Also, JPEGs cannot contain any transparency and thus can be more difficult to merge with page backgrounds.

Binary Bingo
There is at least one more important point to mention regarding GIF images.

The computer stores all information in bits and uses the binary numbering system. Binary number systems use a base of 2 and therefore double at each increase in notation (bit) size (e.g. 1, 10, 100, 1000 for the decimal equivalents of 1, 2, 4, 8, etc.).

The decimal system uses a base of 10 and increases by a factor of 10 at each notation size increase (0, 10, 100, 1000 etc.). The number of bits defines each exponential increase in a binary system. A bit size of 0, 1, 2, 3, 4, 5, 6, 7, 8 relates to the following maximum values: 0, 2, 4, 8, 16, 32, 64, 128, 256. Any value between any of these points will still contain the same number of bits as the largest value at each point.

Thus a GIF image with 96 colours, contains the same number of bits per pixel as a 128 colour image. A GIF image with 32 colours requires 5 bits per pixel but a 33 colour GIF image requires 6 bits per pixel. However, increasing the number of colours to 64 still requires only 6 bits per pixel.

When choosing the number of colours for an image you should choose the maximum number of colours at each bit size interval. For instance, if your image looks good at 48 colours, then choose 64 as the number of colours. This then gives the image more colours to use but will not increase the file size as each pixel is still described in 6 bits.

The Optimised palette in Corel PhotoPaint will tend auto reduce to the minimum number of colours required to that equal to or less than that specified in the number of colours roller. Often, when one checks the actual number of colours used in an image it may in fact be less than that which you specified, especially when increasing the numbers just after decreasing them. Note that once you actually press OK to reduce image colours, there is no way to then increase them other than returning to full RGB mode.

Caveat carrot
The only caveat with the above relates back to the GIF image file system. The more colours there are in an image the less likely it is that there are large sections of the same colour index.

Therefore, sometimes, choosing a number of colours between any of the binary number segments (say 41 rather than 64) may actually reduce the file size by forcing more adjacent pixels to be the same colour during the colour reduction process.

Farewell till next time
I hope you have enjoyed this tutorial on the preparation of images for the WWW. Many of these issues can be used of images for other purposes also such as the production of 8 bit animations and for the production of images with reduced numbers of colours for print and screen print purposes. Thus until next time - Keep on discovering the seemly endless potential within the best professional graphics package about - Corel PhotoPaint.

Please don't forget to view the Glossary section of this tutorial if you have not already done so.

Also, there is a little bit of fun hidden here - if you're game, why not read my version of the Book of Genesis, ' Pictorius Neogenesis'.