Image:IBM16 Palette Map.png
From Wikipedia, the free encyclopedia
Size of this preview: 100 × 600 pixels
Full resolution (515 × 3,089 pixels, file size: 38 KB, MIME type: image/png)
| This is a file from the Wikimedia Commons. The description on its description page there is shown below.
|
Contents |
[edit] Summary
The IBM16 (CGA text modes, EGA smaller graphics modes) palette shown in relation both to its own 4-bit mapping and to the full 2 bits per sample of the CGA and EGA monitor circuitry (the 4-level, 64-colour palette from which EGA can pick any colour in its large graphics mode). To the left are the colour swatches with the 6-bit binary number representing the intensities of the Red, Green and Blue samples with regard to the full palette; to the right are 4-bit arrays of Intensity, Red, Green and Blue flags signalling the particular colour in the IBM16 subset.
[edit] Image information
Portable Network Graphics image, 515 pixels wide, 3089 pixels high, 8 bits per sample, RGB with alpha channel, 3780 pixels per metre (96 pixels per inch) resolution, 1/2.2 (0.45455) gamma, sRGB with relative colorimetric rendering intent, background colour #000000. Created with The GIMP 2.2.8, Netpbm 10.27.0 and pngcrush 1.5.10 on Tuesday the 23rd of May 2006.
[edit] Technical notes
With a range of 0–3, the values for Red, Green and Blue (on the left side) can be off, one third, two thirds or full (intensity of 0, 1/3, 2/3, 1). Obviously, having recurring decimals, it would not be practical to draw those numbers in the image itself; and vulgar fractions would be too similar to the binary values to be worth it (the binary number converted to decimal and divided by 3 gives the intensity). The stepping should be upscaled to the target bit-depth, which in this image is 8 bits per sample, giving 0 (hexadecimal 00), 85 (55), 170 (AA) and 255 (FF). The values 84 (54), 168 (A8) and 252 (FC) you see ever so often are the result of doing, from a 6-bit value, left-shift + zero-fill instead of linear scaling or left bit replication:
101010 (2A) → 10101010 (AA) : correct value 101010 (2A) → 10101000 (A8) : rounding error
See also the section Sample depth scaling of ISO/IEC 15948:2003 (section 12.5 on the W3C edition).
[edit] Orthopraxy
For any PNG screenshot from MS-DOS that uses the IBM16 palette or a subset of it, ie any CGA or EGA image (but not VGA), it is recommended to add an sBIT (significant bits) chunk with a value of 2 for each of Red, Green and Blue. In addition, the colour values should be checked for accuracy and fixed if erroneous (see above for the most common inaccuracy).
[edit] Other versions
- Image:IBM16_Palette_Map-reduced.png is a reduced-colour version (no antialising, palette with only one transparency entry) for decoders that ignore the alpha channel or generally for where resources are short.
[edit] Licensing
File history
Click on a date/time to view the file as it appeared at that time.
| Date/Time | Dimensions | User | Comment | |
|---|---|---|---|---|
| current | 20:28, 23 May 2006 | 515×3,089 (38 KB) | Shlomital | (new version: added #C0C0C0 borders) |
| 22:39, 22 May 2006 | 512×3,072 (38 KB) | Shlomital | ||
| 19:14, 22 May 2006 | 512×3,264 (39 KB) | Shlomital | (IBM16 palette shown in two relations) |
