Browse Prior Art Database

Emulating a 6-Bit System on an 8-Bit System

IP.com Disclosure Number: IPCOM000036175D
Original Publication Date: 1989-Sep-01
Included in the Prior Art Database: 2005-Jan-28
Document File: 3 page(s) / 32K

Publishing Venue

IBM

Related People

Haigh, DC: AUTHOR

Abstract

This disclosure describes a way of making a video system with an 8-bit color palette and digital-to-analog converter (DAC) emulate a 6-bit system. As a reasonable approximation it is assumed that the missing two bits are the low order bits of the 8-bit DAC. Several approxima- tions are discussed and a flexible solution to implement all approximations is described. The logic required to generate the two missing bits requires three threshold comparators.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 51% of the total text.

Page 1 of 3

Emulating a 6-Bit System on an 8-Bit System

This disclosure describes a way of making a video system with an 8-bit color palette and digital-to-analog converter (DAC) emulate a 6-bit system. As a reasonable approximation it is assumed that the missing two bits are the low order bits of the 8-bit DAC. Several approxima- tions are discussed and a flexible solution to implement all approximations is described. The logic required to generate the two missing bits requires three threshold comparators.

A 6-bit system divides the range from zero to full-scale output (FSO) into 63 steps, and an 8-bit system divides it into 255 steps. Because the only common factor of 63 and 255 is three, there will be only 2 places between zero and FSO where an 8-bit DAC could generate the same output as a 6-bit DAC. These matching places are at one-third and at two-thirds of the FSO. At all other places, the 8-bit DAC cannot generate the same output as the 6 bit DAC. The worst error occurs with an input of 10 to the 6-bit DAC; the closest approximation to this is with an input of 40 to the 8-bit DAC, giving an error of nearly 1.2%.

The approximation is controlled by how the missing 2 bits are generated when driving an 8-bit DAC with 6 bits. Different applications are likely to have different criteria for what is the most suitable approximation, but for all reasonable approximations the missing 2 bits are the low order bits of the 8-bit DAC.

Some applications will require the closest possible numerical approximation. This will require the input to increase from 0 in increments of 4, except that in three places increments of 5 will be necessary in order to arrive at the full-scale input value of 255. This means that some steps in the DAC output will be greater than others, which may be a disadvantage.

Another disadvantage of the closest numerical approximation method is that one of the three guns may be driven with a voltage that is 1.2% too low, while the voltage drive to another gun may be 1.2% too high. This will cause the color to differ slightly from the desired colour, and this effect is made worse by the non- linear (gamma) characteristic of the CRT. The approximation method which does not give rise to this defect is to supply zeroes to the missing DAC bits. This allows the gun drive voltages to remain in step, but it has the disadvantage that the FSO is not achieved, because the maximum possible input to the DAC is 252, and hence the maximum output will be 1.2% less than that obtainable by other approximations and by full 8-bit use of the DAC.

Other applications may prefer to perform some compensation for the non- linear characteristic of the CRT, which makes low-level steps less distinct than high-level steps. One way to perform such compensation is to make the initial step from zero larger than the others; to do this, the missing bits are both turned on for all values of input other than zero, so the values presented to the DAC as the input sweeps...