Page 1
WHITE PAPER: New PC Sound Requirements
Attacking MP3 File Compression and PC Fidelity Input/Output for HOLISTIC Listening Improvement
December 9, 2002
Traditionally, sound in personal computers has not really been
about listening quality. Beeps, chirps and other simple tones
used in personal computers were originally created to
augment various commands, not to support an alternative
listening experience. Used primarily as audio cues, personal
computer sounds were tied to games, productivity applications
or used as a single element in a brand, as in the Microsoft®
“chord” that plays upon the opening of the Windows®
operating system.
Today, however, PC sound is going through a renaissance.
Because of the proliferation and increasing sophistication of
personal computers – and the continued migration of digital
media into the home in general – consumers now want to
listen to their computers in much the same way they listen to
their stereos or TVs. Now, consumers want to improve the
quality of entertainment experiences that are uniquely and
conveniently provided by their computers.
Manufacturers are responding. New computer introductions
are emerging onto the market that position the PC as the
staging area for home entertainment. In addition to enhanced
music capability, these PCs have TV tuner cards, Personal
Video Recorders (PVR), online program guide, etc.
Philips White Paper: New PC Sound Requirements
Page 1 of 9
Page 2
Similarly, in the past the standalone sound card market was
relegated to serving the needs of vertical segments like
audiophiles and gamers (who were willing to pay quite a bit
for additional sound capability). Today’s market is driven by
mainstream consumers.
The Importance of MP3.
It should be noted that the PC entertainment trend has been
greatly accelerated by the MP3 file format. In fact, MP3 files
have revolutionized consumer music. Recent research*
indicates that one-fifth of Americans, or about 40 million
people, have downloaded digital music from file-sharing
services.
The obvious benefit of MP3 files is that they are compressed
to make Internet transmission easier. Using advanced
compression algorithms, the file shrinkage is significant, usually
at least by a factor of 10. With MP3, Internet transmission
times for a standard song track on a CD were reduced from
hours to minutes.
However, since considerable data is removed in the copying
process, the MP3 format is usually described as “near CD
quality.” Conventional wisdom says that the data loss does not
affect the listening experience of the average user (i.e., sounds
that the human ear can’t hear, louder noises are emphasized).
This, however, was a generous viewpoint established in light of
the significant offsetting benefits of reduced file size and the
accepted usage models of PCs a couple of years ago. The fact
is, the sound quality of MP3 files is substandard when
compared to CDs and home entertainment systems in
general. New standards are emerging for PCs based on the
consumer electronics usage model.
New Requirements.
The emergence of a mainstream market for CD-quality PC
sound should be good news for sound card manufacturers. But
most sound card manufacturers remain focused on optimizing
the processing capability of their cards; and, given the state of
sound card technology in 2002 (16-bit/48KHz data streams
remain the common standard), it’s clear they are struggling to
make noticeable improvements – even to their traditional
audience of audiophiles and gamers. Focusing on changing bit
depths and sampling rates has met the laws of diminishing
Philips Electronics White Paper
Page 2 of 9
Page 3
returns, and this approach no longer improves sound quality in
a meaningful way.
Most surprisingly, this traditional view of sound card
functionality has not attempted to address or compensate for the
loss of data in MP3 files. Clearly, the MP3 format (including the
new mp3Pro format) will continue to play a major (if not the
central role) in digital music for years to come. Add to this the
new standards evolving in pricing, ease-of-use and the net
effect of the listening experience in general, and you have a
market in which consumer sophistication and expectations
have outpaced product innovation.
Improving PC sound is no longer a matter of simply optimizing
the sound card itself – it’s a matter of re-defining the listening
experience associated with a personal computer. Sound card
manufacturers need to get out from under the hood of the
car, as it were, and consider where the car is going.
Consumers bring many more variables to the decision
purchase mix vs. vertical markets, each of which must be
addressed. These are:
1. Ease-of-use
2. Reproduced sound quality, especially as it relates to
MP3 files
3. Price
4. Input types
5. Output variables
6. Environment
Clearly, managing all of these variables is difficult. Philips
Electronics has devoted one of its research and development
facilities to this challenge. Located in Tempe, Arizona, the
Philips Audio Lab is a centerpiece to the development of
sound products for the personal computer and digital home
marketplace. One of the latest innovations from those
development efforts are new sound card products based on
the concept of Holistic Sound Management, which is described
below.
New Approaches.
To achieve real innovation – or more specifically, audible
sound differentiation – and put the productivity-oriented
personal computer on a par with home entertainment or
Philips Electronics White Paper
Page 3 of 9
Page 4
powerful game-focused systems, a new way of thinking about
PC sound was required.
Managing a wide range of variables is called Holistic Sound
Management. While MP3 is important, there is still a
surprisingly large number of digital file formats, each with its
own unique advantages and disadvantages. There are an
infinite number of speaker and headphone combinations. And,
whatever the speaker arrangement, the nature of the room
plays an important role.
Holistic Sound Management.
Recently, Philips announced the availability of innovative PC
sound technology capable of dramatically improving the sound
quality of personal computers. This technology is now available
to personal computer OEMs as well as current PC owners,
and was developed to address new requirements of fastgrowing PC entertainment applications.
The new Philips technology is called Sound Agent 2. Sound
Agent 2 uses sophisticated algorithms to analyze sound along
two dimensions: the source of sound, which can be any source
type, and the room or headphones where the sound is going
to be played. Combining input and output with the sound card
itself results in high-quality total holistic sound optimization.
Importantly, the end-user also has a means of easily identifying
and manipulating variables that affect sound quality in context
of their particular listening experience. So, while Sound Agent
2 will automatically optimize sound quality along the two
dimensions described above, the user also has the option of
tweaking and adjusting sound to suit his or her unique needs –
via an easy to use graphical user interface.
Sound Agent 2 performs real time logic functionality that
determines where audio streams are coming from and where
they are going at any given moment. This control center
“brain” acts as a media traffic cop – analyzing, processing and
directing multiple audio streams independently and
simultaneously. The end result is fully optimized PC audio
fidelity, in any listening environment, and with any equipment.
Sound Agent 2, and the underlying Intelligent Media Processing
(IMP) technology, is available to PC users bundled with two
new Philips retail sound cards, the Dynamic Edge 4.1 with
quad surround, and the Sonic Edge 5.1 with 5.1 Channel
Philips Electronics White Paper
Page 4 of 9
Page 5
Surround. Both cards include three governing feature sets for
managing sound holistically:
1. Intelligent Media Processing (IMP):
The Philips software
foundation that constantly monitors the type of sound
input and renders it according to output.
2. Compelling 3D virtualization: Technology developed by
QSound Labs for Sound Agent 2 creates the absolute
best 3D audio experience for any headphone, stereo
speaker or multi-channel speaker system. State-of-theart algorithms eliminate the need for cross-talk
cancellation found in many solutions from other
manufacturers. The end result is an ultra-wide “sweet
spot” for strong positional perception regardless of
head movement and positioning, allowing listeners to
enjoy a true 3D surround experience with two, four,
or more speakers in any environment.
3. Comprehensive API support:
Both cards include
extensive compatibility to existing standards, such as
DS3D, EAX 2.0, and A3D 1.0.
The Sonic Edge 5.1 also includes:
1. 5.1 Channel Surround from any source material:
Stereo
sources from CDs, MP3s, and PC games are
transformed into 6 full channels of true surround
sound using QSound’s exclusive QMSS 5.1 technology.
While other solutions simply mirror the front stereo
speakers in the rear channels, QMSS analyzes each
incoming signal and automatically approximates the
location of where each individual sound should be in a
surround environment. The result is a full 360-degree
wrap-around sound experience from all sound sources.
2. QSizzle and QRumble to supercharge MP3’s
: Uniquely
provides an energy boost to MP3 files that helps
restore compressed digital music to its original fidelity,
by compensating for lost harmonic detail (see
expanded technology description “Attacking MP3
Format Limitations”). MP3 files are have a notoriously
flat sound. Truly adaptive, this technology features
active multi-band filtering that continuously monitors
input signals and makes adjustments automatically.
Philips Electronics White Paper
Page 5 of 9
Page 6
3. Virtual 5.1 Channel Surround: From 5.1 channel sources
such as multi-channel software DVD players,
QSurround virtualizes each channel for dynamic
reproduction through stereo speakers. In addition,
QSurround enhances playback through multi-channel
5.1 speakers for a seamless and more immersive
surround experience.
Attacking MP3 Format Limitations.
QSizzle and QRumble are audio frequency response modifiers.
This technology is in the same general category as tone
controls and equalizers (i.e. filters). However, whereas
conventional filters are static, QSizzle and QRumble are active.
Static filters apply a fixed change to the input signal according
only to their user settings, without regard to the nature of the
sound input signals. For example, if a bass tone control is set
by the end user to provide a given boost amount, this boost
amount, and the shape of the filter curve (boost vs. frequency)
remain constant over time, regardless of the instantaneous
characteristics of the audio input signal itself. In a sense, the
filter “ignores” the signal and blindly makes its assigned
modification.
Static filtering must be judiciously applied. With more dramatic
filtering, i.e. high boost levels, a truly artificial tonal characteristic
is imparted. This is perceived by the listener (consciously or
unconsciously, according to their level of sophistication) as
being systemic – that is, separate and independent of the
content. Second, increased boost also carries a higher risk of
overload distortion, requiring more frequent user adjustments
to deal with extremely variable content.
QSizzle and QRumble are active filters – their response varies
dynamically, in real time. Unlike the classic bass and treble tone
controls, both QSizzle and QRumble are composed of
multiple independent processing channels. Each channel
applies variable dynamic emphasis in accordance with the
audio input signal. In direct contrast to static filters, user
settings for QSizzle and QRumble provide guidelines for the
desired effect, rather than fixed rules. The instantaneous
degree and frequency weighting of the emphasis are
controlled by the input signal itself. This use of content to
control emphasis results in highly correlated, natural sounding
Philips Electronics White Paper
Page 6 of 9
Page 7
enhancement. Furthermore, it permits greater levels of boost
without objectionable side effects. Obviously, QSizzle and
QRumble substantially reduce the need for user readjustment.
QSizzle and QRumble are particularly effective in the
treatment of digital audio content that has been encoded using
data compression algorithms such as MP3. Since compression
is often applied to a degree that results in audible degradation,
content can sound flat and lifeless due to the loss of frequency
content and dynamic range.
Static filters cannot help – and can even worsen – the dynamic
range problem, and tend to sound artificial (i.e. “boomy” or
“shrill”) at higher settings. In contrast, the active dynamic
spectral emphasis provided by QSizzle and QRumble
effectively compensates for the negative side effects of digital
compression in a natural, content-dependent manner. QSizzle
brings liveliness and sparkle to midrange and upper
frequencies, while QRumble restores warmth and punch to
the low end of the spectrum.
Usage Example.
To illustrate the Holistic Sound Management capability of Philips
Sound Agent 2 and its sound cards, consider this complex (but
not unrealistic) example:
Joe is playing Half Life, a game that supports 3D
positional audio via Microsoft's DirectSound3D
API. Game sounds are being accurately placed
around him using DS3D and EAX support built
into Sound Agent 2, and are being rendered
(sound is being created) through Joe’s 5.1 speaker
system with his Sonic Edge 5.1 sound card.
At the same time, Joe is playing his favorite MP3
play list, with Media Player® in the background.
Joe likes to listen to his own MP3 collection while
playing games.
The MP3 audio stream is automatically being
processed from 2-channel stereo into 5.1-channel
surround simultaneously -- using Sound Agent 2's
QMSS 5.1 feature -- giving Joe an optimized
combination of game audio mixed with music for
his own “custom” experience.
Philips Electronics White Paper
Page 7 of 9
Page 8
Joe also likes to interact with other people online.
While the music is playing, he uses his PC
microphone to talk with his online gaming friends
via software that provides IP based Internet chat.
His friends create mono voice streams that are
processed by Sound Agent's QMSS 5.1 technology
that is mixed with the other multi-channel
streams and” localized”, so Joe feels like the
voices of his friends are immersed in the actual
soundtrack of his online adventures.
In the midst of all this, Joe's friend Mike steps into
his room and asks if Joe wants to check out his
new Spiderman® DVD. Joe, not wanting to exit
his online game and playlist -- but who wants to
see the DVD -- reduces the music volume, then
launches a multi-channel software DVD player
and plays the first chapter of Spiderman. Again,
rendered through Sound Agent 2 in enhanced 5.1channel QSurround, Joe adds a reverb Theater
preset via QSound’s Environmental Modeling
(QEM) technology to the DVD playback. The
result is a custom, in-the-theater like experience.
As Joe and Mike watch Spiderman, Joe is still
monitoring his game in the background to make
sure he is “in the action”...
No other sound card manufacturer today can address
such input and output complexity.
Platform strategy extends to other products.
IMP is a software architecture that can be utilized in a
wide range of consumer devices. With powerful,
affordable products emerging for home use including
servers (PC and otherwise), wireless media, new TV
sets, web tablets or “smart displays”, the home
environment will become increasingly complex and
diverse. Sound Agent 2, and the underlying IMP
technology, represents an important and highly
extendable user interface platform for this increasing
broad range of home entertainment products.
Philips Electronics White Paper
Page 8 of 9
Page 9
Affordable pricing.
Unlike vertical markets, mainstream consumer demand
requires that the sound cards be less that $50 (less than 10%
of the average PC price). As a result, the suggested retail price
of the Sonic Edge 5.1 is well under $50 (the Dynamic Edge 4.1
is half that).
Philips.
Royal Philips Electronics has been marketing technology into
the home for over 100 years. Its core strengths in driving
standards, innovation, consumer expertise, audio and video
intellectual property and extensive display technology allows
Philips to uniquely bridge the gap between the personal
computer and digital homes and vehicles.
The Company is one of the world’s biggest electronics
companies and Europe’s largest, with sales of $28.8 billion
(EUR 32.3 billion) in 2001. It is a global leader in color
television sets, lighting, electric shavers, medical diagnostic
imaging and patient monitoring, and one-chip TV products. Its
184,000 employees in more than 60 countries are active in the
areas of lighting, consumer electronics, domestic appliances,
components, semiconductors, and medical systems. Philips is
quoted on the NYSE (symbol: PHG), London, Frankfurt,
Amsterdam and other stock exchanges.
For additional information on any of the technologies in this
article, please contact Dennis Johnson at +1-480-752-6096 or
dennis.w.johnson@philips.com.
*Ipsos-Reid, June 2002.
All trademarks are the property of their respective owners.
Sonic Edge 5.1, Dynamic Edge 4.1, Sound Agent 2 and IMP are
trademarks of Royal Philips Electronics. QSizzle, QRumble,
QSurround and QMSS5.1 are trademarks or registered
trademarkets of QSound Labs.
© Koninklijke Philips Electronics N.V. 2002
###
Philips Electronics White Paper
Page 9 of 9
Loading...