Browse Prior Art Database

Conversion Process for IBM Graphics Printer Bit Image Data

IP.com Disclosure Number: IPCOM000061264D
Original Publication Date: 1986-Jul-01
Included in the Prior Art Database: 2005-Mar-09
Document File: 3 page(s) / 44K

Publishing Venue

IBM

Related People

Daniels, P: AUTHOR [+4]

Abstract

A conversion process is described which permits bit image data originally created for the IBM 5152 Graphics Printer to be accurately printed on other printers with (1/240 inch) vertical print resolution. There are two parts to the conversion process. First is the treatment of the general conversion algorithm. Finally is the treatment of special cases to improve the appearance of some kinds of data. Image data originally for an IBM 5152 Graphics Printer can be printed on another printer with a conversion of the 1/72 vertical data mapped to a pattern of 1/240 inch pels. The mapping pattern is a repeating 3/240, 4/240, and 3/240 for each of the graphic printer pels. This conversion process shows how to prevent any accumulated error when mapping real input data to the paper.

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 52% of the total text.

Page 1 of 3

Conversion Process for IBM Graphics Printer Bit Image Data

A conversion process is described which permits bit image data originally created for the IBM 5152 Graphics Printer to be accurately printed on other printers with (1/240 inch) vertical print resolution. There are two parts to the conversion process. First is the treatment of the general conversion algorithm. Finally is the treatment of special cases to improve the appearance of some kinds of data. Image data originally for an IBM 5152 Graphics Printer can be printed on another printer with a conversion of the 1/72 vertical data mapped to a pattern of 1/240 inch pels. The mapping pattern is a repeating 3/240, 4/240, and 3/240 for each of the graphic printer pels. This conversion process shows how to prevent any accumulated error when mapping real input data to the paper. With this general-case algorithm, it is clear that a line that is only a single pel high in the graphic printer's data stream could be printed as either 3/240ths of an inch high or as 4/240ths of an inch high. This difference may be objectionable in some applications. The problem can be solved by making single pel high lines either all 3/240ths high or all 4/240ths high. The input image data bytes represent 8 bits of vertical data. To illustrate, the bit locations of four lines of horizontal data are drawn in Fig. 1. The space between the lines is added for readability. Each of the input bits is 1/72 inch high. The input bits are numbered from 0 to 7 to illustrate that the least significant bit is the bottom bit on the page. The column of 3's and 4's is the number of 1/240ths of an inch that the input bit is mapped to. Fig. 1 shows that in order to have no accumulated error from one line to the next, each successive line needs to use a pattern of 3's and 4's that picks up where the previous line left off. This results in three different patterns of 3's and 4's. Though it is clear what sequence the patterns come in, it is easiest to implement by picking the pattern based on the amount of error between 1/72 and 1/240 so that no memory of the previous line is required. This amount of error is shown in Fig. 2. In order to decide if an input data pel requires special treatment, we must determine how the boundary conditions between 8-bit-high stripes of vertical image data are to be handled. The original statement of the problem dealt with a single input pel high line. In terms of input data that means a 1 bit with a 0 bit on both sides of it. However, in order to implement a solution using a reasonable amount of memory and processing, we need to assume that we don't know what the data bit is, either...