|
The MediaStones(TM) benchmark was released at WinHec 95
by Providenza & Boekelheide, Inc. This
benchmarking program was the first to use full motion
video to explain what is happening while fully testing the
subsystems utilized by multimedia applications.
MediaStones was designed for Windows 3.x computers running
Microsoft Video for Windows 1.1. While MediaStones is a bit out-of-date, it does run well under Windows 95
and is a good example of some of the types of products and
services P&B is capable of delivering.
The original white paper is available below.
MediaStones(TM) --
Benchmarking in the Motion Video World
By Tim Roberts,
March 1995 (TimR(at)ProBo.Com)
When a group of computer industry professionals
get together, either professionally or socially, the
discussion invariably turns to benchmarking. Few topics
offer the opportunity for such a wide variety of opinions
and interpretations as do the current (or not so current)
suite of popular benchmarks. One of the problems
with benchmarks is that, because computer workloads vary
so widely, no existing benchmark will ever be able to
exactly predict the performance characteristics
experienced by an actual computer user. Even so,
benchmarks can be a useful tool in the comparison of
computers, computer systems and computer peripherals, as
long as the limitations and target areas of the benchmark
under consideration are well understood.
One area where benchmarks have the potential to be
particularly useful is in evaluating motion video quality
and performance. A number of advances have been made in
the recent past in the development of motion video
hardware and software. The release and rapid adoption
Microsoft's Display Control Interface (DCI) as part of
Video for Windows 1.1 has produced a crop of video and
audio subsystems with sophisticated capabilities for
displaying digitized movies. It has also led to the
development of related software that can take advantage of
these sophisticated capabilities. Unfortunately the speed
with which these changes have taken place did not allow
time for current benchmarks to adequately address the
special considerations and performance profiles of these
multimedia systems.
Making things even more difficult, the performance of
personal computer systems has increased to the point where
raw performance is not necessarily the most important
factor in choosing a multimedia subsystem. Despite claims
made by video chip vendors, the difference between an
instantaneous screen repaint and one which is 50 percent
faster is simply not relevant. Both systems are "fast
enough." It is more important to measure and quantify
the quality of a multimedia system, not just the
performance.
We at Providenza & Boekelheide, Inc. (P&B)
recognized the advances being made in multimedia systems
and the lack of adequate tools to measure and compare
these systems. We have constructed a new benchmark tool
that measures not only the performance but the quality of
multimedia systems aimed at the Video for Windows user.
MediaStones(TM) is the result of this effort.
MediaStones(TM) provides quantitative measurements in
an educational and entertaining way. It features a
full-motion video "talking head" that explains
what is happening and gives the viewer hints of what
performance issues to observe and evaluate. MediaStones(TM)
addresses the shortcomings of even the most commonly
referenced existing benchmarks, including the widely used
Winmark benchmark.
If you would like to have a complimentary copy of the
MediaStones CD mailed to you, please send a
self-addressed, postage paid (we will add about 3oz of
material to the mailer) CD-ROM mailer to:
MediaStones Request
7460 SW Hunziker Road
Suite A
Tigard, OR 97223
|
