The Four Classes of Device Banding

The information in this article applies to:

• Microsoft Windows Software Development Kit (SDK) for Windows versions 3.1 and 3.0

SUMMARY

In the Microsoft Windows environment, banding is a common technique to divide a page into a number of sections, called bands. When an application uses banding, it requires less memory to produce a page of printed output at the potential expense of some additional time. This article discusses the four types of banding devices that can be used in the Windows environment: nonbanding devices, simple banding devices, "Full Page Banding for Text" devices, and BANDINFO devices.

MORE INFORMATION

Nonbanding Devices

These devices do not perform any banding. PostScript printers and plotters are the most common nonbanding devices. These nonraster devices accept primitive output directly from Windows without any need for intermediate rasterization. An application can identify a nonbanding device by performing the following three steps:

1. Create a device context (DC) for the output device and obtain a handle to the device context (hDC).

2. Call the GetDeviceCaps function, specifying the hDC and RASTERCAPS for the iCapability parameter.

3. Check the RC_BANDING bit of the GetDeviceCaps function. If the RC_BANDING bit of the GetDeviceCaps return value is clear, the device does not perform banding.

Simple Banding Devices

These devices divide both text and graphics into a series of bands starting at the top of each page and running to the bottom. Under versions of Windows earlier than 3.1, these devices may provide a full-page band when available memory permits, but they usually do not. Under Windows 3.1, these devices generally provide a full-page band when memory permits. Dot-matrix printers are the most common simple banding devices. An application can identify a simple banding device by performing steps 1 and 2 above. The RC_BANDING bit is set for simple banding devices.

"Full Page Banding for Text" Devices

The driver for a page printer (commonly referred to as a laser printer) typically uses one band covering the full page for text followed by additional smaller bands for graphics. This technique is an optimization for a printer that allows text to be placed at any position on the page but does not have the graphics capabilities of a PostScript printer. Typically, a driver does not provide any graphics bands unless it detects a graphics command in its text band. While this technique allows the driver to optimize printing of pages containing only text, an application that prints graphics must call a graphics primitive in the text band to work correctly with this driver type. Most new page-printer drivers are developed for the fourth driver category, discussed below.

BANDINFO Devices

With a BANDINFO device driver, an application can optimize its printing and eliminate calling graphics primitives in a text band by requesting graphics bands only when required and identifying which band is the text band. This BANDINFO escape provides this capability. An application issues the BANDINFO escape to inform the driver that the application will print graphics and to determine which bands contain text and which bands contain graphics.

An application can use the QUERYESCSUPPORT escape to identify a device that supports the BANDINFO escape. If the driver supports BANDINFO, the application can determine when to print text or graphics as appropriate. If the driver does not support the BANDINFO escape, the application should print text and graphics in each band.

The Hewlett-Packard LaserJet driver is an example of a printer driver that implements the BANDINFO escape.