Optimizing Printer Throughput

U.S. patent application Ser. No. 918,414, entitled IMPROVING PRINTER THROUGHPUT, filed Oct. 14, 1986, and having J. W. Malcolm as inventor.

1. Technical Field

This invention relates generally to printing from a print buffer, and more specifically to printing based on a match of the characteristics of a data stream in the print buffer and available print velocity options, with maximum print velocity being a goal.

2. Background Art

In the prior art are daisywheel printers where variables such as character selection time, required impact force for a selected character, print hammer firing timing, and carrier velocity are taken into account to obtain a desired print quality and to maximize throughput. Impact force, firing timing, and carrier velocity are continually varied during the printing of a line. However, these printers are generally expensive and are primarily tailored for printing high quality text.

With the generally low cost matrix printers contemplated for use with the instant invention, print velocity is selectable by an application programmer (user) from a finite number of velocities capable of being handled by the printers. The user builds a selected velocity into a program for a particular print job. For another job in the program, another velocity will be selected. Illustrative of wire matrix printers contemplated for use with the instant invention are the IBM Proprinter and the IBM Personal Computer Graphics Printer.

The primary problem with the above application driven printers, from a throughput standpoint, is that no variance is permitted while running the application program. Generally, a worst case velocity selection is made on a job basis by the programmer during the writing of the application. With a worst case being selected, there will be many instances, due to the characteristics of the data stream, where the printer is being inefficiently utilized.

With the above in mind, the advance presented by the instant invention is improved printer utilization. This advance is provided by storing a plurality of head movement velocities which are selectable by the application program, and having the application program select a velocity based on the characteristics of a line of graphics data to be printed. The advantage of this advance is that throughput can be increased without materially affecting costs. The same low cost printer can be utilized.

A unique method of optimizing printer throughput is provided by incorporating available head movement velocities into an application program and causing the program to select a velocity based on data stream characteristics for a line of graphics data to be printed. As each line to be printed is buffered, the line is scanned or analyzed against the available printer head movement velocities. Based on the analysis, a maximum workable head movement velocity is selected, and the line of graphics data is printed according to the selected velocity.

FIG. 1 is a block diagram illustration of the major elements of an application program that causes data to be printed in a graphics format.

FIG. 2 is a pictorial representation of a portion of a line which can be printed at a fast head movement velocity.

FIG. 3 is a pictorial representation of a portion of a line which must be printed at a slow print head movement velocity.

FIG. 4 is a pictorial representation of a portion of a line having no two consecutive dots in a row and which can be printed at a fast head movement velocity.

FIG. 5 is a pictorial representation of a portion of a line having no two consecutive dots in a row and requiring a slow head movement velocity.

FIG. 6 is a flow chart illustrating a determination of head velocity based on criteria which includes the characteristics of the buffered line of graphics data.

Matrix type printers generally include a printhead attached to a carrier. Illustrative of low cost matrix type printers are the IBM Proprinter and the IBM Personal Computer Graphics Printer. Included in the printhead is a vertical column of a plurality print wires. Selected ones of the print wires are activated or fired as the printhead traverses a print line on a page to be printed. The firing of each print wire will result in a dot on the paper being printed. Selective firing of the wires will ultimately result in the construction or printing of a character and/or a graphics object. For example, the character "A" may be made up of various dots in a matrix 5 columns wide and 8 rows high.

With the above noted IBM printers, there are four graphics printing formats available to an application programmer for three separate densities of printed dots and two separate head movement velocities. These formats will be described in detail later herein. The criteria for selection of these formats relates to the density of the dots in the graphics data. The density of the dots in the graphics data directly affects the frequency of firing the print wires. If consecutive dots have sufficient separation, then the carrier may move at a faster rate. Also, with less density, a reduced amount of data may be sent to the printer using certain formats.

For any particular job, one of the available velocities is selected by the user and embedded in the program. As such, outside of modifying the program, no alternative velocities are available to an operator using the application program on a stand alone, or host connected, device controlled display computer workstation or system. In many instances, the system will be a keyboard/display computer system, and as such, the device is a keyboard. Included in the system will be a printer such as a matrix printer described above.

In FIG. 1 the major elements of application program that causes printing of graphics data is shown. The application program is made up of a minimum of two processes. A process is defined as a section of code that performs a particular function. Data creation process 1 determines what data is to be printed, and places that data in a buffer 2. When the buffer 2 contains all the data deemed necessary by data creation process 1, printing process 3 takes control and sends the data in buffer 2 to the printer.

Refer next to FIG. 2. Graphics data containing dots no closer than 1/60 inch can be printed in one format requiring 60 data bytes for every inch of printed output. In this instance, a fast head movement velocity is to be used by the printer. For the above noted printers, the identifier that is sent to the printer to specify this format is an Escape character (Esc) followed by the letter K.

Refer next to FIG. 3. Graphics data that contains dots no closer than 1/120 inch can be printed in a second format requiring 120 data bytes for every inch of printed output. In this instance, a slow head movement velocity is to be used by the printer. For the above noted printers, the identifier that is sent to the printer to specify this format is an Escape character followed by the letter L.

Refer next to FIG. 4. Graphics data that contains columns of dots no closer than 1/120 inch and has no two consecutive dots in a row closer than 1/60 inch can be printed using a third format requiring 120 data bytes for every inch of printed output. However, in this case, since no consecutive dots are closer than 1/60 inch, a fast head movement velocity can be used by the printer. For the above noted printers, the identifier that is sent to the printer to specify this format is an Escape character followed by the letter Y.

Refer next to FIG. 5. Graphics data that contains columns of dots spaced at 1/240 inch must use a fourth format requiring 240 data bytes for every inch of printed output. In this format, a requirement exists that no two consecutive dots in a row be closer than 1/120 inch. The printer will use a slow head movement velocity. If consecutive dots are closer than 1/120 inch, then some dots will not be printed. For the above noted printers, the identifier that is sent to the printer to specify this format is an Escape character followed by the letter Z.

With the instant invention, a program having the elements illustrated in FIG. 1. is provided. Added to the printer process 3 is code for analyzing the characteristics of the data stream for the line of graphics data to be printed. Included in the code are line buffering capabilities, and all available formats for printing graphics data. Following a determination of the characteristics of the data in buffer 2, the printer process 3 selects the optimum format from the available formats for printing graphics data, and then the line is issued to the printer with the selected format for printing at the selected velocity.

All graphics data will be placed in buffer 2 by the data creation process 1 in the format for printing the line illustrated in FIG. 5. This is a default format which is equivalent to a worst case selection by a user. All other formats are derivatives of this format. Upon analyzing the data, it may well be that a different format can be selected for printing the data at a higher velocity. The printer process 3 analyzes the data in the buffer 2 and selects an optimum format for printing the graphics data using the criteria set out in the following print mode determination table.

Set out below is an illustration of a routine to be included in printer process 3. This routine is in programming design language from which both source code and machine code are derivable.

Refer next to FIG. 6 in conjunction with the above programming design language. The programming design language defines the operations performed by the printer process 3 in determining an optimum printing format and causing subsequent printing to occur according to the format.

All data in buffer 2 must be scanned before the results can be analyzed. This is illustrated in block 11.

The format that causes printing at the highest velocity and has the fewest number of bytes to send to the printer is Esc K (FIG. 2). If buffer 2 contains one byte every 1/240 inch, since the Esc K format requires that data density be not less than 1/60 inch, all the data bytes with the exception of every fourth byte must be zero or contain no graphics data dots. This is illustrated in block 12. If bytes that are not multiples of four are all zero, then the graphics data count is divided by 4 as shown in block 13. This is since only every fourth byte will be sent. Next, the 4-byte graphics Escape K sequence header is sent to the printer as indicated by block 15. The header is followed by every fourth byte from the graphics buffer beginning at byte 0 until the count is exhausted as shown in block 17.

If the data in the buffer 2 does not qualify for the Esc K format, then the next best format is the Esc Y format (FIG. 3). The printhead will move at the same velocity as with the Esc K format. However, twice as many data bytes must be sent. In order to qualify for Esc Y, all the odd-numbered bytes must be zero as determined by block 14. If zero, the columns of data will have a minimum separation of 1/120 inch. Once this criteria has been met, the graphics data count is divided by 2 as indicated by block 16, and a determination is made as to whether a second criteria is met. Moving at the fast print speed, the printer is only able to print every other dot position. Consecutive dots must have a minimum separation of 1/60 inch when the fast printhead velocity is to be used. Therefore, if the print buffer contains no dots in the same position in consecutive even bytes as determined by block 18, then the Esc Y format will be selected. The 4-byte graphics Escape Y sequence header is sent to the printer as illustrated in block 21. The header is followed by all the even-numbered bytes from the graphics buffer 2 beginning at byte 0 until the count is exhausted as indicated by block 23.

If all the odd-numbered bytes in the buffer are zero, and there are dots in the same position in consecutive even bytes as indicated by block 18, the the third format (Esc L) may be used (FIG. 4). The 4-byte graphics Escape L sequence header is sent to the printer as indicated by block 20, followed by all the even-numbered bytes from the graphics buffer beginning at byte 0 until the count is exhausted as in block 23.

Finally, if none of above criteria are met, the buffer must be sent using the Esc Z format (FIG. 5). Here, the graphics data count as was passed to the print process 2 by the data creation process 1 is used. The 4-byte graphics Escape Z sequence header is sent to the printer as indicated by block 19. The header is followed by all the bytes in the buffer as indicated by block 22.

In summary, a unique method of optimizing printer throughput is provided by incorporating available printhead movement velocities into an application program and causing the program to select a velocity based on data stream characteristics for a line of data to be printed. As each line is buffered, the line is analyzed against the available printhead movement velocities. Based on the analysis, a maximum workable head movement velocity is selected, and the line of data is printed according to the selected velocity.

While the invention has been particularly shown and described with reference to a preferred embodiment, it will be understood by those skilled in the art that various changes in form and detail may be made without departing from the spirit and scope of the invention.