home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Shareware Breakthrough: Utilities & Productivity
/
SharewearBreakthroughUtil_ProdCol.cdr
/
util
/
bench
/
readme.win
< prev
next >
Wrap
Text File
|
1990-05-31
|
22KB
|
705 lines
PC Magazine Labs
Laboratory Benchmark Series
Software and documentation (c) 1989,90 PC Magazine
==============================
GRAPHICS PERFORMANCE TESTS
Microsoft Windows version
Release 1.1 -- June 1990
==============================
Installation notes:
The PC Labs Graphics Performance Tests for Microsoft Windows
(Release 1.1) is distributed on one low density floppy (5.25"
360K diskette version also includes the Graphics Performance
Tests for OS/2 Presentation Manager; 3.5" 720K diskette version
incorporates the Hardware Performance Tests Release 5.5) or as
an archived file on PC MagNet, our on-line service associated
with CompuServe. The following three files are located on this
diskette in a directory named "WINBENCH":
README.WIN - This file: a text file containing documentation.
WINBENCH.EXE - The Graphics Performance Tests benchmark program
REGISTER.WIN - Form to send in to register or to make comments
In order to install the test simply copy the WINBENCH.EXE file
to the directory of your choice. The program can be executed
from within any current version of Windows. We also recommend that
you keep a copy of this file along with the test program.
The Microsoft Windows and OS/2 Presentation Manager versions of the
Graphics Performance Tests (which are both included on this diskette)
are quite similar in design and implementation in order to provide
comparative tests for the two environments. However, there are some
tests that are implemented slightly differently under each environment;
also, the Presentation Manager version contains some tests that are not
included in the Windows version. The differences exist because there are
graphics functions in Presentation Manager that are not available in
Windows, and therefor cannot be tested. For the specific details read
this README.WIN file and the README.PM file in the PMBENCH directory on
this diskette; these text files contain documentation about each of the
features of the two versions.
- Page 1 -
TABLE OF CONTENTS
Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
BitBlt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Alignment . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Overlap . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
ROP's . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Screen/Memory . . . . . . . . . . . . . . . . . . . . . . . . 5
All . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Partial Arcs. . . . . . . . . . . . . . . . . . . . . . . . . 6
Full Arcs . . . . . . . . . . . . . . . . . . . . . . . . . . 6
All . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Single Lines. . . . . . . . . . . . . . . . . . . . . . . . . 7
Polylines . . . . . . . . . . . . . . . . . . . . . . . . . . 7
All Directions. . . . . . . . . . . . . . . . . . . . . . . . 7
Wide Lines. . . . . . . . . . . . . . . . . . . . . . . . . . 7
All . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Polygons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Alternate Fill. . . . . . . . . . . . . . . . . . . . . . . . 8
Winding Fill. . . . . . . . . . . . . . . . . . . . . . . . . 8
All . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Rectangles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Boxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Random. . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Patterns. . . . . . . . . . . . . . . . . . . . . . . . . . . 9
All . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
- Page 2 -
Text . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Image Fonts . . . . . . . . . . . . . . . . . . . . . . . . . 10
Background Mix. . . . . . . . . . . . . . . . . . . . . . . . 10
Greyed Text . . . . . . . . . . . . . . . . . . . . . . . . . 10
Mixed Fonts . . . . . . . . . . . . . . . . . . . . . . . . . 11
All . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Misc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Set/Query Current Position. . . . . . . . . . . . . . . . . . 12
Stretch . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Image Data. . . . . . . . . . . . . . . . . . . . . . . . . . 12
All . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Dialog Boxes. . . . . . . . . . . . . . . . . . . . . . . . . 13
Rubber Banding. . . . . . . . . . . . . . . . . . . . . . . . 13
Scroll Window . . . . . . . . . . . . . . . . . . . . . . . . 13
Clear Window. . . . . . . . . . . . . . . . . . . . . . . . . 13
All . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Clipping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Region Clipping . . . . . . . . . . . . . . . . . . . . . . . 14
All . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
System Information. . . . . . . . . . . . . . . . . . . . . . 15
Modify Options. . . . . . . . . . . . . . . . . . . . . . . . 15
Display results . . . . . . . . . . . . . . . . . 15
Spreadsheet Output. . . . . . . . . . . . . . . . 15
Iteration Multiplier. . . . . . . . . . . . . . . 15
Run All Tests . . . . . . . . . . . . . . . . . . . . . . . . 16
Demo Mode . . . . . . . . . . . . . . . . . . . . . . . . . . 16
- Page 3 -
Introduction
This document describes the PC Magazine Labs Graphics Performance tests for
Microsoft Windows (Release 1.1 -- June 1990). The Benchmark program
provides a set of benchmark tests that determine overall performance
information for Windows display drivers and graphics display boards.
To run the benchmarks you may select each test separately, or you may run
a set of tests by selecting the "All" menu item from each of the function
menus on the menu bar. You can run all the tests consecutively by selecting
"Run All Tests" from the "Options" menu or by pressing the "Ctrl+R" key
combination on the keyboard. A Demo mode is also provided. To start the
demo mode select the "Demo Mode" menu item under "Options" or press the
"Ctrl+D" keys. To exit demo mode you must strike an ESCAPE, SPACEBAR
or ENTER key or terminate the timing dialog box using the keyboard or
mouse before it is destroyed. If a key is struck to exit demo mode, the
current test will be finished before the benchmark actually terminates the
demo mode, however the timing dialog box for that test will not be displayed.
In demo mode, the timing dialog boxes will be displayed for about 5 seconds
and then will automatically be destroyed before proceeding to the next test.
Dialog boxes, which display timing information after each test, must be
cleared before executing subsequent tests. Alternatively, you may turn off
these dialog boxes by selecting "Modify Options" under the "Options" menu.
Set the "No" radio button on "Display Results". "Modify Options" may also be
reached by pressing the "Ctrl-M" key combination on the keyboard.
Timing results may be saved in a file. To save the timings bring up the
"Modify Options" dialog box and set the "Yes" radio button for "File Output".
This will enable the "File Options" box and allow Text, Excel, or Lotus
compatible file output to be specified. If a file with the specified filename
already exists in the current directory (e.g. from a previous run of the
tests) the existing file will remain unchanged AND THE NEW FILE WILL NOT BE
SAVED. Note that a file name along with a 255 character file header can be
entered if the default name and header are not acceptable. Be sure to give
your files new names each time (or move the files to another directory) to
insure that the filename does not conflict with an existing file.
- Page 4 -
BitBlt
Description:
These selections test the performance of BitBlt using various alignments, 15
of the most common ROP operations, and random bitblts from memory to screen,
screen to memory and memory to memory.
Menu Items:
Alignment
This selection bitblts 32x32, 64x64, 128x128, and 256x256 rectangles
from destination to source using the following alignments: Source Aligned
Destination Aligned, Source Aligned Destination Not Aligned, Source Not
Aligned Destination Aligned, and Source Not Aligned Destination Not Aligned.
Overlap
Overlap tests the speed of blting overlapping rectangular bitmaps.
The test overlaps bitmaps by eight pixels in both horizontal and vertical
directions.
ROP's
ROP's tests 15 of the most common Raster Operations (ROP's) whose
source and destination are Screen to Screen, Memory to Screen, and
Memory to Memory. The ROP's tested are SRCCOPY, SRCPAINT, SRCAND,
SRCINVERT, SRCERASE, NOTSRCOPY, NOTSRCERASE, MERGECOPY,
MERGEPAINT, PATCOPY, PATPAINT, PATINVERT, DSTINVERT, ZERO,
and ONE.
Screen/Memory
This menu item bitblts randomly placed 64x64 monochrome bitmaps from
screen to screen and from memory to screen.
All
Selecting the "All" option will run each of the "Bitblt" menu items
consecutively.
- Page 5 -
Curves
Description:
This menu selection tests output of partial and full arcs.
Menu Items:
Partial Arcs
This benchmark test draws two sets of arcs. Each set comprises 2 arcs
defining an ellipse, in either the x major axis, or in the y major axis.
Full Arcs
Full arcs draws filled and outline ellipses using the same base
parameters as Partial Arcs. The routine draws 2 ellipses with different
major axes.
All
Selecting the "All" option will run each of the "Curve" menu items
consecutively.
- Page 6 -
Lines
Description:
Lines tests the drawing of single lines, polylines, single long lines drawn
in all directions, and wide lines.
Menu Items:
Single Lines
Single Lines draws 800 horizontal, vertical and diagonal solid lines.
All 800 lines of each type (horizontal, vertical, or diagonal) use the same
beginning and end points. Therefore, only a total of three lines will be
visible in the window.
Polylines
The polyline test look exactly like the single line test. PolyLines,
however, use the PolyLine call to output 5000 horizontal, vertical or
diagonal solid lines for each polyline call. Again you will only see
three lines on the display.
All Directions
This test draws lines in all directions. A large box, centered in the
middle of the display is filled by drawing radial lines around the perimeter
of the box from its center. For each box, the line colors are changed on
each pass around the perimeter of the box.
Wide Lines
For each of line widths 3, 5, 10, 20 the wide line test draws 100 random
lines.
All
Selecting the "All" option will run each of the "Line" menu items
consecutively.
- Page 7 -
Polygons
Description:
This set of tests test the speed at which alternate and winding area fills
are done.
Menu Items:
Alternate Fill
This test draws a hour glass shaped polygon and then fills it using
Alternate fill attribute, repeating for 75 iterations.
Winding Fill
Winding fills draws two overlapping triangle polygons and then fills it
using Winding fill attribute. Winding Fill also repeats 75 iterations.
All
Selecting the "All" option will run each of the "Polygon" menu items
consecutively.
- Page 8 -
Rectangles
Description:
This menu selection benchmarks square and rounded cornered filled rectangles,
random sized rectangles and patterned filled rectangles.
Menu Items:
Boxes
This item first draws 100 square rectangles, which are filled and
outlined, from smaller to larger sizes. Then repeats itself by drawing
the rectangles from large to small. The test then draws 100 rounded corner
rectangles in a manner similar to the square cornered rectangles.
Random
This item draws randomly sized rectangles on the display in random
colors.
Patterns
Patterns tests 2 types of brushes in the windows environment. It first
draws full screen rectangles using grey-scale pattern brushes. It then
fills the screen using hatch brushes.
All
Selecting the "All" option will run each of the "Rectangle" menu items
consecutively.
- Page 9 -
Text
Description:
The Text menu tests image fonts, background mix modes, greyed text and mixed
fonts output.
Menu Items:
Image Fonts
Image fonts draws the string: "The quick brown fox jumped over the lazy
sleeping dog's back then sat on a tack." The text is first drawn in the
system font and then is drawn using Times Roman in point sizes 21 and 27.
It next draws the string in Helvetica in point sizes 20 and 24.
Background Mix
Background mix draws the above text string in the system font using the
background attributes of TRANSPARENT and then OPAQUE. The first test
fills the entire client window with a pale gray color. Then it sets the
color back to default and mode to TRANSPARENT. 200 lines of text are
output to the display. The second test again fills the client window with
pale gray but this time it changes the color to yellow instead of default
and background mode to OPAQUE. Two hundred lines of text are output.
The background of the text should be yellow.
Greyed Text
The string "Grey text" is randomly placed on the display using a cyan
background, a black foreground and the halftone attribute.
- Page 10 -
Mixed Fonts
Mixed fonts writes the following four strings to the display:
The quick brown fox jumped over the lazy sleeping dog
The quick brown fox jumped over the lazy sleeping dog
The quick brown fox jumped over the lazy sleeping dog
The quick brown fox jumped over the lazy sleeping dog
The first line is written in the default font. The second line is
written with the default font with some of the words underlined. The
third line has some words bolded and the fourth line has some italicized
print.
All
Selecting the "All" option will run each of the "Text" menu items
consecutively.
- Page 11 -
Misc
Description:
Misc tests miscellaneous functions: set and query current position,
stretch/compress bitblt, and image data.
Menu Items:
Set/Query Current Position
This test shows nothing on the window. The test calls the
GetCurrentPosition and MoveTo 5000 times each.
Stretch
Stretch test the speed of doing Bitblt compression and expansion. A
64x64 monochrome bitmap is loaded and then compress or stretch to the
following sizes: 32x32, 48x48, 80x80, 96x96, 192x192, 32x48, 48x96, 48x32,
108x96, 21x21, 105x105.
Image Data
A bitmap is created in memory. The image is then displayed on the screen
in all combinations of foreground and background colors in a 16x16 matrix.
All
Selecting the "All" option will run each of the "Misc" menu items
consecutively.
- Page 12 -
User Interface
Description:
Special tests for creating and deleting dialog boxes, drawing rubberband
lines, scrolling windows, and clearing windows.
Menu Items:
Dialog Boxes
Dialog boxes creates and destroys a dialog box 20 times.
Rubber Banding
Rubber banding draws a rectangular box and then rubberbands it across
the window 100 times and then back again.
Scroll Window
This test scrolls byte aligned and non-byte aligned text vertically
through a rectangular region in the window.
Clear Window
Clear window clears the display 50 times to a gray background color
using Rectangle.
All
Selecting the "All" option will run each of the "User Interface" menu
items consecutively.
- Page 13 -
Clipping
Description:
Clipping tests bitblt, line and text clipping using a disjoint rectangular
clipping path.
Menu Items:
Region Clipping
Region Clipping tests Bitblt, Line and Text clipping in a clipping region
comprised of 4 disjoint rectangles.
All
Selecting the "All" option will run each of the "Clipping" menu items
consecutively.
- Page 14 -
Options
Description:
The options menu item allows you to configure the program to your particular
needs. The default options are Display Results, No File Output, and Iteration
Multiplier of 5.
Menu Items:
System Information
Choosing this item will display pertinent information about the display;
including display size, window size, color information, and number of
system objects (fonts, brushes and pens).
Modify Options
Choosing this item will allow you to reconfigure the Benchmark program.
You may reach this option by pressing the "Ctrl-M" key combination on the
keyboard.
Display results
Enables the result dialog boxes to appear or not appear in the
window. The default is Display Results enabled.
Spreadsheet Output
Enables timing results to be sent to a text file. Text output is
formatted to be ascii text, Microsoft EXCEL compatible or LOTUS 123
compatible depending on which radio button is selected. The default
output format is ascii text.
A file name and 255 character header may also be specified. The
default file names are pmbench.txt for ascii text, pmbench.xls for
EXCEL and pmbench.wks for LOTUS 123. The default file header is
"Benchmark Results".
NOTE: If a file with the specified filename already exists in the current
directory (e.g. from a previous run of the tests) the existing file will
remain unchanged AND THE NEW FILE WILL NOT BE SAVED. Be sure to give your
files new names each time (or move the files to another directory) to insure
that the filename does not conflict with an existing file.
Iteration Multiplier
Allows you to increase the number of times Benchmark will execute
each of the tests. Choices are 1, 2, 3, 5, and 10. If the results
vary more than 1 ms. between runs increase this number. Use 1 or 2 for
an EGA or VGA and 5 or above for cards with graphics coprocessors.
- Page 15 -
Run All Tests
Picking this option will execute all of the available test in Benchmark
using the choices supplied in the Option Menu or the defaults if Modify
Options have not been changed. You can choose this option by pressing
the "Alt-R" key combination on the keyboard.
Demo Mode
Selecting Demo Mode places BENCH into a demonstration mode in which the
full test suite is continuously executed. This mode can be disabled by
pressing the space bar, return, or ESC keys, or by clicking the mouse on
the "OK" button in the dialogs displayed after individual tests.
-----------------------------------------------------------
Any comments about the tests should be submitted on the accompanying
registration form to the following address:
PC Magazine Labs
ATTN: Windows Benchmark Registration
One Park Avenue
New York, New York 10016
PC Magazine Labs retains all rights to these tests but we do grant you the
specific right to make an unlimited number of UNALTERED copies of
the tests provided you do not charge anything for them; these tests are
distributed as a public service of PC Magazine and are meant to be
widely distributed at no charge to the user. The PC Magazine Labs
Graphics Performance tests for Microsoft Windows (Release 1.1, June 1990)
were programmed according to PC Magazine Labs specifications by Graphic
Software Systems (GSS).
Thank you for your interest in the PC Magazine Labs Graphics
Performance tests for Microsoft Windows.
Sincerely,
Frederic E. Davis
Director, PC Magazine Labs