A listing of sites and pointers to references which are of some particular interest to
design, to sound reproduction in small spaces and to listening enjoyment.
Bruno Putzeys on
Amplifiers and Class D
In Square CAD audio interview #15 Bruno Putzeys of Hypex Electronics
and Grimm Audio talks about his power amplifier designs and the undeserved
reputation of negative feedback. Stability of higher order and nested
feedback loops is often difficult to maintain and consequently they are
avoided by practically all amplifier designers. Easy to understand solutions
are preferred even when their performance is limited. In some ways this is
similar to loudspeaker design, where a flat on-axis frequency response and
sufficient output volume are the prime design objectives. (And room response
equalization with DSP will supposedly compensate for overlooked radiation
Boyk on Piano
A peek into the mind of a concert pianist and renowned teacher: http://jamesboyk.com/
(See also James Boyk below)
The Etymotic Audio File from May 2013 carries a well done educational video:
Watch the Ear in Action.
Publications of Linear Audio are found at www.linearaudio.net.
Volume 5 is now available. Jan Didden's personal website is www.linearaudio.nl . Of
particular interest to the LX521 DSP
Challenge is his 6-channel volume control for the Behringer DCX2496,
which could make this DSP unit a contender.
of Time-Frequency Analysis Methods and their Practical Application
On the LISTENINC website's Literature pages I found an interesting
AES Convention paper by Steve Temme et al. with the above title. It
shows the development of time-frequency-energy analysis of loudspeakers from
the Short Time Fourier Transform to Cumulative Spectral Decay analysis to
Wigner-Ville Distribution, Smoothed WVD, and Continuous Wavelet Transform.
The graphical presentations for the different analysis methods are exemplary
in showing relationship and resolution of the three variables for the same
test object, a wide bandpass filter with resonances, and for interpreting
KLIPPEL - Online Training
The KLIPPEL website
is a rich source of educational material about the measurement, analysis and
and detailed behavior of electro-acoustic transducers in the form of papers
and application notes. Now it is also possible to take online training. The
first of seven sessions has been posted.
#1 Linear Lumped Parameter Measurement
#2 Vibration and Radiation Behavior of a Loudspeaker Membrane
#3 Loudspeaker Nonlinearities
#4 Loudspeaker Distortion Measurements
#5 Predicting the Nonlinear Loudspeaker Behavior
TO YOUR EARS
The January 28, 2013 issue of THE NEW YORKER carries on pages 32 to 39 a most
interesting article by Adam Gopnik about "The quest for 3-D recording
and other mysteries of sound". It puts in perspective Edgar Choueiri's cross-talk-cancellation experiments
for 3-D listening at Princeton University with the hearing and sociological
research at McGill University by Albert
Bregman (Auditory Scene Analysis - The Perceptual Organization of
Sound), Daniel Levitin (This Is
Your Brain on Music) and Jonathan Sterne (MP3: The Meaning of a Format). The
article highlights the changes in music consumption from the audience in a
concert hall to the audiophile sitting in front of his speakers, in the
sweet spot, and to the ear-bud generation busy with other activities.
"The notion of a pure musical experience is, for Sterne and his
cohorts, the last sad effort of a nineteenth-century cult of attention that
placed the solitary alienated (and almost always male) listener in a temple
of silence, the concert hall."
MUSIC, NOISE AND HEARING: HOW TO
PLAY YOUR PART
The BBC has published "A Guide for Musicians" on how to protect their
hearing. The information has relevance to the audiophile. My technical
quibble is with the lack of definition for 'Sound Intensity' and 'SPL' and
their use. Sound Intensity is sound power per unit area. A +3 dB change
means doubling the power. The corresponding change in sound pressure level,
SPL, is also +3 dB. Doubling the SPL, which is a +6 dB change, requires +6
dB or 4-times more power. Our ears respond to and can be damaged by the
sound pressure level, SPL, at the eardrum.
Dynamic Space Stereophony
I am very much interested in the spatial rendering of stereophonic program
material, in the illusion created in the listener's mind. Peter Huebner
goes much deeper than that when he talks about dynamic space
stereophony and how music is heard in the mind of a classical
composer. It is not clear to me how that could be recreated as a listening
experience, but the thought is intriguing. Maybe specific multi-channel
program material and a combination of multiple real and phantom sound
sources with specific directional properties would approximate the
composer's experience. Anyways, the pages provide a profound view of music and
its possible effects upon a listener.
sound waves become a listening experience
An animated illustration on the Sennheiser website shows the transfer of acoustic
vibrations to electrical stimuli, which are then processed in the brain. The
phantom sources that we hear between the loudspeakers of a stereo system are
a construct of the brain. They are derived from the superimposed sound waves
of left and right loudspeakers and the room at each ear. Cross-talk is
essential for producing timing cues, which the brain uses naturally for
directional hearing below 800 Hz in the Interaural Time Difference (ITD)
range of hearing. Frequency response shaping by the outer ear and head
blockage of sound are essential for directional hearing above 2 kHz in the
Interaural Level Difference (ILD) range of hearing. Cross-talk cancellation
in this frequency range can improve the spatial detail of the perceived
acoustic scene and particularly its proximity. With HRTF derived CTC the two
loudspeakers can be heard almost as whispering into just one ear from
close-up. But maintaining a spatially uniform spread of the phantom scene in
front of the listener becomes problematic when CTC thins out the center of
the phantom scene. I have not heard CTC processing that I prefer long-term
over plain stereo.
The Science and Art of Listening by Seth S. Horowitz
The Opinion Page of the New York Times Sunday Review carried
on 11 November 2012 an article about listening vs. hearing by the author of
the book. This subject is of particular interest to me,
because in sound reproduction we ideally only pay attention to the direct
sound from the loudspeakers, while reflections, reverberation and standing
waves due the room's boundary surfaces are background sounds that can be
tuned out. I am convinced that this ideal can only be achieved by using
loudspeakers with frequency independent directivity. But this type of
loudspeaker is not common and requires special effort to design. Instead,
some people listen in acoustically dead rooms, which must be specially
designed, rather than in more normal living spaces.
See also This Is Your Brain on Music
Par, founder of swissaudec,
presents at tmt27 for the first time a paper about his
ground-breaking and patented new methodology for audio signal
The paper will be delivered in English on
Saturday, 24 November 2012, Spatial Audio, 18:00 - 19:00, Room R1.
Undiscovered Treasures in Spatial Audio: Inverse Problems and Invariant
Retrieving higher order (i.e. multichannel) signals from a lower number
of input channels is a well-known problem (generally referred to as “upmixing”),
which may be satisfyingly solved with stereo or multichannel content, thus
providing a high level of accuracy regarding localization and sound
envelopment. Such methodology, however, fails when mono input signals should
be processed in real-time – prior art then will not achieve sound source
separation in real-time, particularly with sound sources showing the same
frequency! At the same time any upmixing algorithm implies comb filter
effects when the signal order is to be lowered again (“downmixing”) –
an annoying effect, for instance, with HD television broadcasts generated by
means of prior art upmixing tools, which are subsequently being played back
on a stereo device. This paper presents an entirely new approach by applying
inverse problem theory to the world of audio (to which we subsequently refer
as “inverse coding”). Inverse problems were described first by
Russian-Armenian physicist Viktor Ambartsumian in 1929, then solving, for
instance, complex linear integral equations by a simple numerical solution.
Generally speaking, lossy data of lower order may be recovered to finally
represent a data set of higher order e.g. by means of specific matrix
inversions. Additionally, algebraic invariants (a property of a class of
algebraic objects that remains unchanged when specific transformations are
being applied) may serve as “tags” to speed up signal calibration. The
result is highly efficient and comb-filter-free real-time upmixing of any
given stereo, Surround or 3D audio content for the sake of restoration,
troubleshooting, postproduction, broadcasting and content distribution. In
particular, inverse coding may help the sound engineer to avoid colouration
effects with spot microphones. In audio coding spatial bit rates may be
lowered up to 200 times with multichannel content for the sake of efficient
transmission over existing studio networks - and (the good news for 3D audio
content providers!) future low bandwidth transcoding consumer devices will
be able to cope with genuine 3D audio content of highest computational
complexity. Clemens Par (swissaudec GmbH)
See swissaudec below
Auro 3D -
Principles in Surround Recordings with Height
This 2011 White
Paper by Theile and Wittek outlines the possible directions for 3D audio
sound, where 3D means direction, distance and height. It reviews the
limitations of 2.0 stereo and 5.1 surround due to phantom source instability
and practical setup implementations. Described are different 3D rendering
systems for playback and microphone setups for recording. This is a very
interesting survey of what we have today and what might be the n-channel
future. I am still amazed of what can be extracted in natural sounding
spatial information from a properly done 2-channel recording. See Watson.
A German version of the paper is available at www.hauptmikrofon.de
and much more.
Shaping Sigma Delta Based Dacs & Digital
Jitter and Volume Controls
Mallison, Chief Technical Officer of ESS Technology, lucidly presented and
explained the technical issues at RMAF 11.
Watch it on YouTube.
I do not doubt that people can hear Sigma-Delta noise shaping differences or
digital jitter, though I have no personal experience. But I question his
explanations as to why they are audible. Data measured in the frequency
domain were not related back to the signal's nature and its amplitude in the
time domain and vice versa. In general the vertical axis of the measurements
was not discussed. In Fourier analysis the axis is calibrated in sinewave
amplitude, but it is also used to read the amplitude of impulsive signal
spectral components and of noise. Both are broadband signals that often
exceed the bandwidth of the Fourier analyzer. To understand their amplitude
presentation one must know the analyzer's impulse bandwidth and noise
bandwidth. The spectral components of an impulse may be of very low
amplitude, but in the time domain they add to a large, very short duration
pulse with a repetition rate corresponding to the spectral line spacing.
Similarly the amplitude of the same noise in the time domain is a function
of the hearing bandwidth, e.g. 20 kHz. We hear both in the time domain and
the frequency domain so we should also analyze physical processes in both
domains. The audibility on the Rebecca Pidgeon track apparently is due to
the impulsive signals from metal beads.
Jitter is different. Jitter causes non-linear distortion, i.e.
intermodulation of broadband signals. I doubt that this is what we hear
because of its low amplitude. But jitter has a spectral distribution, which
should have been shown in addition to its PDF. I believe that it is the very
low frequency spectral content of the jitter, which typically has the
highest amplitude, that accounts for audibility as phase modulation of the
audio signal in the time domain. Again proper amplitude calibration of the
analyzer is important to arrive at a meaningful interpretation of the
assessment of audio quality - The means and methods within the EBU
The European Broadcasting Union defines In section 3.2 of
this EBU-Technical Review article from 1997 a set of
subjective parameters, which cover the technical quality of “acoustic
music”. This set should be very useful for recording engineers in
assessing their work. The demonstration CD PEQS, described in EBU-TECH 3287
(October 2008), gives 66 examples of different
recording quality. The LX521 Monitors
readily show the differences.
More: The Design Ethos of Dieter Rams
I have admired the exterior design of Braun products under Dieter Rams'
leadership. The SFMOMA has currently (January 2012) an exhibition of his
work and also of his influence upon other designers (e.g. Steve Jobs &
team). Follow the link below to his 10 Design Principles as shown on a
poster in the museum bookstore and along the walls of the exhibition.
One, Two, Three
Stuart Yaniger describes the methodologies and pitfalls
of listening tests in Volume 2 of Linear Audio, p.161-174.
Myself and others, notably Don Barringer, have done a lot of listening tests
to fine-tune the ASP response of the ORION, which was derived from
free-field acoustic measurements. This is clearly a subjective process
because what I and others hear is not merely a function of the sound
pressure streams at our ears, but also of our expectations and the sonic
attributes we pay particular attention to. We may come to agreements, if we
have the same reference sounds in our minds, but this does not necessarily
guarantee that we have developed a neutral transducer. Stereo is
particularly tricky because we create a phantom sound scene that only exists
in our minds. It is a good magician's trick, if we can hide the loudspeakers
and the listening room in the process. And when have we reached YOUR
perceptual noise level, where you are no longer sure that this way or that
way is more accurate? I only know when I have reached MY perceptual noise
level, but it is not fixed in time and can change with new expectation and
attention. Do I then start over? In the end could I fool a child or dog?
MP3 and AAC explained
of Fraunhofer Institute and MP3 fame explains the perceptual coding and
decoding of sound in this 17th AES International Conference paper from 1999.
Though audiophiles often malign these compressed and lossy data formats I
have found that many of today's Internet Radio broadcast stations have very
satisfying sound quality and provide interesting classical music and culture
programming. I use a Squeezebox as signal source and an iPod to control it
What looks like a complete list of suppliers of high-resolution music files
can be found on the auraliti website.
2L - the Nordic Sound
"Norway is blessed with many churches and cathedrals and most of 2L's
recordings are made in these wonderful spaces. The music captured by 2L
features Norwegian composers and performers, and an international repertoire
reflected in the Nordic atmosphere."
Check out www.2l.no for
downloads of high-resolution sound files.
David Chesky - Composer, musician, producer,
"Founded by David and Norman
Chesky of the audiophile-record label Chesky Records, HDtracks
is a high-quality music download service offering a diverse catalog of
independent music from around the world. HDtracks does not believe in
DRM, and as a result, we offer a selection of unencrypted files that
play on any computer or portable device.
We have several formats available for fans of all
types of music. If you would like genuine CD-quality sound, HDtracks
offers both AIFF and FLAC lossless files. For those who wish to carry
their music in a portable player, we have 320kbps files that far
exceed the quality of standard MP3 files.
Finally, audiophiles take note. HDtracks offers select titles in
ultra-high resolution 96khz/24bit files. This is true DVD-audio sound
quality for music lovers that demand the very best!"
Listen - LIFE WITH CLASSICAL MUSIC
"North America’s classical music magazine covering people, places and
events; recommendations of recordings, books, and film; and all the many
ways our lives are touched by classical music. Published by ArkivMusic,
the Source for Classical Music."
Introduction to 3-D Audio
Fundamental aspects of spatial hearing in humans are studied by Edgar Choueiri in the 3-D Audio and Applied Acoustics
Laboratory at Princeton University. One of the projects deals with optimal cross-talk cancellation of loudspeaker stereo
signals at the listener's ears. The process works optimally in a room
without reflections (anechoic) and for binaurally recorded material.
Coincident and near-coincident 2-microphone recordings benefit also.
Coherent spatial information is usually lost in mixed-down multi-microphone
recording and cannot be recovered by cross-talk cancellation. I guess that playback
in reverberant spaces is optimized when using loudspeakers with frequency
independent radiation patterns such as omni, cardioid and dipole, together
with a symmetrical placement of the loudspeakers relative to large
reflecting surfaces and at least 1m away from them. This would be the same
setup as for traditional stereo playback in acoustically small spaces.
in on the listening brain
A report about research into the auditory brainstem response to complex stimuli (cABR) in
the June 2011 issue of Physics Today. Hearing involves a bi-directional
stream of information - from inner ear to brain and from brain to inner ear.
Note that the brainstem response to a 0.1 ms impulse shows activity for about
7 ms. Listen to three examples of naturally produced sounds contrasted with
their brainstem responses
at the end of the report.
While not much about audio www,khanacademy.org surely has something to teach or to
catch up upon for everyone. See for example his 10 minute talks about Specular and Diffuse Reflections or about Wealth Destruction.
The Well-Tempered Computer
Vincent Kars' introduction to computer audio provides a wealth of
information and links to suppliers for setting up a playback system where
your music collection resides in computer memory. Seeing the different
acronyms and terms explained and compared is very helpful. For many
audiophiles it is like learning a new language.
See also the dCS
Guide to Computer Audio for assembling a playback system.
vs Monopole Bass
Elias Pekonen measured and compared the response from a dipole bass
loudspeaker to that of a monopole (box) loudspeaker in an acoustically small
room using constant Q periodic wavelets. He observed that the dipole source
better preserves the modulation envelope as shown very clearly in the
frequency-time-amplitude data presentations on his website.
The work confirms my own conclusions from shaped tone-burst tests (Publications #21) and those of Holland,
Newell, Mapp in "Modulation Depth as a Measure of Loudspeaker Low Frequency
Performance", Proceedings of the Institute of Acoustics, Vol.26, Pt.8,
With the Toneburst CD you can hear
changes in bass envelope modulation for different locations in your
You can also sift through the questions, opinions, misunderstandings,
irrelevancies, insights and selfassertions of the related diyaudio.com forum discussion.
I hope that the time response in the discrete mode frequency range of a room
will eventually get as much attention as the steady-state frequency
response. Bass sound perception is derived from both.
Audio - Volume 1
The second issue of the semi-annual bookzine has just been published and is
available from www.linearaudio.net.
It continues the collection of thought provoking and practical articles
written by experts in various branches of audio. For example, I found the
article on "Low-noise microphone preamp design" and "The
F-word - or, why there is no such thing as too much feedback"
particularly interesting. Check out the content. What is published here is
too practical for the scientific paper ambitions of the JAES and far outside
the range of Stereophile and TAS. Decades ago Wireless World was the place
for this kind of audio engineering thought, but not that magazine's main
Graeme Huon from HUONLABS
in Australia presents a sound reproduction system that conveys not only
direction but also distance of acoustic sources. It is based on the premise
that we hear the curvature of the arriving wavefront. I am not convinced,
but here is a series of short articles that explain the concept. Also shown
is a loudspeaker design with omni polar response obtained from controlled
diffraction plate (CDP) structures, which are claimed to accurately combine
the outputs from multiple drivers on a spherical surface.
See also: Graeme Huon and Zeljko Velican, 'Spatially
consistent reproduction of the reverberant sound field', 116th AES Convention,
Berlin, 2004, Preprint 6109
Graeme Huon and Zeljko Velican, 'Low frequency optimization and
non-bass masking effects for sound field re-creation', 115th AES Convention,
New York, 2003, Preprint 5989
The Inaudible Loudspeaker - Unmasking
audio sound information by applying monaural stereophonics
Leo de Klerk, a Netherlands artist and recording engineer, proposes a vertical array of 4 loudspeakers, (Lbottom +
Ltop) & (Rbottom + Rtop), to produce a center phantom scene from a stereo recording that will
not shift for off-axis listening angles. The loudspeakers should be omni-directional. Bottom and top loudspeakers
receive the same input signal. I have not heard such configuration. An
excited listener to my Pluto 2.1 demo at the TMT26 was trying to get Leo de
Klerk to come and meet me, but our schedules prevented it from
Eric Weitzman started a new
website and blog describing his Orion experience and modification, his
Hubble Super Regulator, Balrog Turntable, and Curcio Phono Preamp to which
you can leave replies.
Baffle Dipoles - How they work and how to use them well
Rudolf Finke has on his website DIPOLE+ an article "Open baffle 1.pdf" that
on-axis and off-axis frequency responses of drivers on differently
shaped baffles. In particular much attention has been paid to the transition
region from 'dipole' response to 'dipolar' (beaming) behavior of practical
size drivers. It is a well done study of trends and limitations.
RCA Victor presents 'Sound and
the Story' (1956)
Here is a historic video that clearly states RCA's sonic goals for
their recordings in the days of mono and early stereo:
“A planned illusion … the best seat in the concert hall”
At 23:10: “ It’s traveled a long way – for an
illusion. The last stop home, and the best seat in the concert hall.”
No wonder the Living Stereo series got such legendary reputation. Check out
the placement of the listeners and especially of the monitor speakers in the
recording studio, well away from all walls (for example at 06:10 to 06:50).
The entire room, which features dimensions larger than the typical studio of
today, resembles acoustics more characteristic of a large living room than
an over-damped studio (but often with lots of very early reflection from the
huge mixer console in front of the typically flush mounted wall speakers).
RCA placed the main microphones above and in front of the orchestra. What
were the problems with recording directly from "the best seat"?
Could such recording be done with modern microphones like a pair of Schoeps SuperCMIT
and capture the timbre, localization and spaciousness as heard from
"the best seat"? What is the recording engineer's goal today?
- Mikrofonaufnahmetechnik und Tonstudiotechnik (Recording
& Studio Techniques)
Eberhard Sengpiel was honored in 2010 by the Verband Deutscher Tonmeister (VDT) for his contibutions
to the Tonmeister (~recording engineer) profession. His website contains a very extensive
compendium of lecture notes (example)
for recording engineers. But I found nothing about issues or solutions to stereo playback in reflective spaces, which the vast
majority of consumers have to deal with. Why are
there no lectures about the all-important "Wiedergabetechnik"
(Playback Techniques)? Why no discussion of the influence of the "Mono
Recording Angle", which captures sound outside of the stereo recording
Clemens Par creates
professional three-dimensional sound from one single audio channel. He
thereby saves 50% up to 88% of the audio signal bandwidth and, vice versa,
enhances any audio data to a professional three-dimensional sound image. I
was impressed by his demo at TMT26,
but how could he do
- BBC Research & Development Reports
Many interesting reports
about loudspeaker design and test & evaluation methods from 1957 to 1999
- Some factors
in loudspeaker quality
H. D. Harwood from the BBC Research Department talks in this landmark Wireless World article from 1976
about observations and unresolved issues related to loudspeaker design. On
page 48 he questions the optimum on-axis frequency response and mentions a
shallow dip between 1 kHz and 3 kHz. On page 54 he pleads for a non-uniform
axial frequency response. Experiences with equalizing an initially flat
on-axis response to obtain the realistic sounding ORION-3.2
have convinced me that 'flat' is not correct for a stereo system. Here we must create believable phantom sources from
two real sources at +/-30 degree angle to the listener, at the 'sweet spot',
in the room. Nor is the BBC dip the answer. Instead, the response must be
flat above 100 Hz and then gradually drop to a lower level at high
frequencies. There are general physical and psychoacoustic requirements for
such a response, which I will explain at a later
time. A specific on-axis
frequency response curve is usually the secret behind a good sounding
loudspeaker. The marketing department likes to label this curve
For Stereo, neither the two monitor loudspeakers in the
recording studio, nor the playback loudspeakers in the home, should have a
flat on-axis frequency response. A corresponding power response and
placement of the acoustically small loudspeakers in the room at least 1 m
away from large reflectors, hides both loudspeakers and room acoustically
from the listener in the 'sweet spot'.
- Acoustics Handbook
Application Note 100 from Hewlett Packard Co. marks the company's
short-lived entry into the acoustic measurement field in 1968. The flagship
product was the Zwicker Analyzer for loudness measurements. AN100 contains
fundamental and useful information in its 116 pages (9MB) such as physical
acoustics, physio-psychological acoustics, survey
of measurement methods, instrumentation, calibration of microphones, sound
level meters, filters, frequency analysis, standards.
- Application of linear-phase
digital crossover filters to pair-wise symmetric multi-way loudspeakers
Part 1: Control of off-axis frequency response, PP presentation 1
Part 2: Control of beamwidth and polar shape, PP
Ulrich Horbach and D.B. (Don) Keele, Jr., AES
32nd International Conference, September 2007
Here is a truly ground breaking, sensible and practical
application of DSP to the design of crossover filters and the polar response
of large multi-way active loudspeakers. Very exciting work! It avoids lobing
vertical polar pattern.
The Audio Toolbox by Dr. Ulrich Brüggemann (((acourate))) now
includes these crossover filters.
Science back into Loudspeakers
John Watkinson from Celtic Audio Ltd. presents his observations and theories
in a thought-provoking paper. I see many parallels to my
work, in particular his emphasis on minimal diffraction, i.e.
omni-directional radiation and the consistency of the reflected with the
direct sound spectrum. He emphasizes the necessity for linear phase response
and waveform fidelity for localization and spatial accuracy of phantom
sources. I have not found or heard convincing evidence for this, if phase/
group delay distortion is kept low. Linear phase can only matter for
instantaneously broadband transient signals. The highpass nature of
loudspeakers, though, requires 12 dB/oct low frequency response for minimum
group delay build-up, not vented enclosures. The Cabar
stereo monitor loudspeaker system bears many similarities to a pair of Pluto 2.1. I designed Pluto because I wanted
to compare a low diffraction, acoustically small source to Orion, which led
to the rear tweeter addition in Orion+.
beam-width transducer (CBT)
D. B. (Don) Keele has done extensive work on line
array loudspeakers for sound reinforcement applications, where it is
important to control the vertical radiation pattern and to avoid lobes,
while maintaining wide horizontal dispersion to cover an audience. In a
curved array with proper shading the delay of individual drivers is obtained
mechanically and the output level from groups of drivers in the array is changed
independent of frequency and only according to their placement in the array. The
joint paper with Douglas J. Button describes straight and
curved arrays and compares their vertical radiation pattern to that of a
2-way loudspeaker over a ground plane. The paper is rich in graphical
An introduction to the CBT and its application is given in Don's Boston
AES-ASA-BAS presentation from January 2010 with overheads and accompanying audio/video (Don
I have heard Monte Kay's center speaker CBT and was quite impressed with its
acoustic performance in his home theater application. It seems the CBT in
its various forms presents a business opportunity for highest quality sound
reinforcement, where especially the mid and high frequency portions of the
spectrum must be properly dealt with. It would be the end of the
horn and compression driver in quality critical applications.
- Microphones and
& Kjaer Library has a collection of their Application Notes,
Technical Reviews, Primers, Handbooks, Conference Papers and a Dictionary.
It is a treasure chest of information about acoustic and vibration
measurement methods and tools. I desired but could not justify a product
purchase from B&K, but I was an avid collector of their publications.
Here now is everything in one place and conveniently accessible.
Physics and Psycho-Acoustics of Surround Recording (2005)
places emphasis on the streaming behavior of sound as he explains perception
of direct and reflected sounds in a surround
recording context. Intelligibility,
reverberance, distance, and mud are his major concerns. Unlike for Theile
and Wittek below, the placement of phantom sources in a coherent spatial
continuum is of less concern. He states, though, that "Recordings stand
or fall on their ability to satisfy the needs of human hearing. This is
constant among people, even among recording engineers." He may not be
aware how much is revealed about the recording technique when a loudspeaker system
like the Orion is used to evaluate the plausibility of the auditory scene
that is heard.
Natural Music Recording based on Psychoacoustic Principles (2001)
Theile discusses techniques
for recording music when natural reproduction over 3 front loudspeakers
and 2 surround loudspeakers is desired. The perceptual effects that are
generated by 5-real and n-phantom sources depend upon the microphone
configurations used. Curiously, the potential influence of loudspeaker
radiation pattern and room reflections upon spatial sound reproduction is
not considered. Wittek below, makes the same
differences between wavefield synthesis and stereophony
Chapter 3 of the above PhD
Thesis from Helmut
Wittek contains a wealth of information about the properties of stereo,
about phantom sources and perception models. Chapter 2 is an introduction to
the perceptual attributes that apply. Dr. Helmut Wittek is the Managing
Director of Engineering for Schoeps
Mikrofone in Germany.
- A Case for
Blauert discusses perception, reality and virtual reality in
this paper from the 2nd International VDT
Symposium on 'sounddesign', 2007, in Germany. The paper is also
interesting in relation to my proposed tests for the accuracy
of a stereo system by concentrating on the plausibility of the reproduced Auditory Scene and
the abstraction level of the character of the reference that is implied in
Seductive (Yet Destructive) Appeal of Loud Music
Even though it is well known
that prolonged exposure to loud sounds produces hearing damage, it is very
common that people listen at dangerously high volume levels. Barry
Blesser attempts to explain the phenomenon in
- Audio Musings
by Sean Olive
An informative blog
about loudspeaker evaluation and test methods at Harman International (JBL,
Infinity, Revel). Sean is the Director of Acoustic Research.
- Enjoy the Music - Review
A timely overview of what is happening in the audiophile world. Some
interesting articles from BAS
Speaker and other magazines. It could be worthwhile to be on the mailing
- A sensibly
On the exhibit floor of the 129th AES Convention in Munich I saw a
loudspeaker that impressed me by the logic with which it was designed, the GRIMM
AUDIO LS1. I could not listen to it closely because the surrounding
noise level was far too high, but what I saw and heard from the exhibitors
convinced me that the LS1 is a remarkable loudspeaker design. The wide
baffle yields a forward pointing and smooth polar response over a very wide
frequency range with a low frequency transition region to omni-directional
behavior. The inside of the box and the panels appeared to be properly
designed. The speaker is active and uses intelligently applied DSP for
response correction. A subwoofer can be correctly integrated if needed. One
can only hope that the LS1
breaks bad habits in recording studio and home.
A very closely
spaced loudspeaker pair with head related cross-talk cancellation that
produces a virtual acoustic environment. This and other interesting work
from the Institute of Sound and Vibration Research at the University of
- Monte Kay's Home
construction project that has been in the works for a number of years. I
am already certain that when finished it will have outstanding sound
- SOUND REPRODUCTION
- Loudspeakers and Rooms
Floyd E. Toole, Focal Press, 2008
In this book Floyd Toole summarizes and explains conclusions from a lifelong
involvement with audio. I highly recommend it to anyone interested in
factual information about loudspeakers and listening rooms, about
measurements, listening observations and their practical implications. It is
lucidly written in easy to understand language, extensively illustrated and
referenced. It deals with the reproduction of sound - which existed in a
space - inside another space. My only regret is that the potential of
2-channel playback in doing so has not been fully explored and
misrepresented. This is
understandable because the conventional box loudspeaker with its frequency
dependent directivity index has been used for almost all of the observations
that are discussed. In fact, the particular interaction of a box loudspeaker
with the listening room makes it more difficult for our ear/brain perceptual
apparatus to hear the recording venue's space and acoustics, provided that
such information has been captured in the recording process. Floyd focuses
on multiple loudspeaker surround sound. He considers this approach to
spatial sound reproduction as much more rewarding and he provides extensive
practical information for that. The book is very comprehensive and in my
opinion a "must-read" for loudspeaker designers, recording and
mastering engineers, room-acoustic consultants, audio reviewers and
audiophiles. The book is about theory and praxis. It debunks much of the BS
that seems to have permeated the audio industry and many of its customers.
- The first open
Chester W. Rice and Edward W. Kellogg: Notes on the Development of a
New Type of Hornless Loudspeaker,
Transactions of the American Institute of Electrical Engineers, issue 44,
1925, pages 461-475.
From page.464: "Provision was made for boxing in the
instrument, and an interesting experience in this connection was that
of placing the box over the back, which had the same general effect on sound
quality as applying a short horn
to the front of the diaphragm. Both helped to bring out the low tones and
gave rise to some resonant effects.
Bringing out the low tones was due principally to preventing circulation of
air between the front and back of the diaphragm.
The resonance was in the horn in one case, and in the box in the other.
A peculiarity of devices employing very flexibly supported diaphragms is
that resonant air chambers behind the diaphragm
do about as much harm as resonant cavities in front of the diaphragm, the
diaphragm usually taking part in the resonance.
Attempts to damp the interior of the box with felt were not entirely
A happy solution of the problem of preventing circulation was obtained by
employing a flat baffle-board,
at the suggestion of Rice, who was the first of the group to recognize the
importance of the circulation factor
in preventing the radiation of low tones. With the flat baffle, no air
resonance occurs and both sides of the diaphragm
give useful radiation, the total power radiated for a given diaphragm
amplitude being nearly four times as great
as that radiated when the back of the diaphragm is enclosed."
This is probably the first description of an open baffle loudspeaker.
The laboratory model is shown on page 474 and is described as:
".... cabinet set containing rectifier, amplifier and
loud speaker. The front of the cabinet acts as a baffle.
To prevent air resonance in the box, the sides and back are vented by
inserting panels of perforated brass."
The first active loudspeaker as well?
(Source: Ralf Gerhard Ehlert, Musikwissenschaftliches Institut der
Universitaet Koeln, Medienstimmen)
- H-frame loudspeaker
dicke mit Mineralwolle isolierte Doppelwände.
Die Leitflächen vor und hinter der Membran vermeiden akustischen
und ergeben eine günstigere Schallabstrahlung" Abb. 61 aus [Günther
PA loudspeakers from Telefunken at the 1936 Olympics.
auf der Berliner Funkausstellung 1939
"Die Entwicklung des
dynamischen Lautsprechers scheint schon seit mehreren Jahren so
ziemlich zum Abschluß gekommen zu sein. Man sieht deshalb wenig Neues.
Nach den Angaben der Aussteller bemüht man sich zur Zeit vor allem
darum, die Feldstärke im Luftspalt zu erhöhen, um die
Empfindlichkeit zu steigern und um, was erheblich wichtiger erscheint,
die Haupteigenschwingungen der Membranen zu dämpfen. ...."
From a German website that has collected
historical information on loudspeakers, amplifiers and
- This Is Your
Brain on Music - The Science of a Human Obsession
Dan Levitin was the Keynote Speaker at the 123rd AES Convention in New York,
2007. He gave a fascinating talk on why we can get goose bumps when we
listen to music. What is the true impact of powerful music on the human
brain? As a former rock musician, producer and now professor at McGill
University Laboratory for Musical Perception, Cognition and Expertise in
Canada, he combines art and science in this most
stimulating and easy to read book. (Dutton, 2006)
- Spaces Speak, Are You
Listening? - Experiencing aural architecture
A book about the science, art and culture of acoustic
spaces written by Barry
Blesser and Linda-Ruth Salter. In light of my observations with the ORION+,
for attempting to reproduce one acoustic space inside another one, chapters
such as "2 - Auditory Spatial Awareness", "6 - Scientific
Perspectives on Spatial Acoustics" and "8 - Auditory Spatial
Awareness as Evolutionary Artifact" are of particular interest. (MIT
A paper about audio
measurements by Henning Moller of test instrument maker Bruel &
Kjaer. The information is technically sound, but the measurement
applications are dated. The paper is very educational, particularly for
those in the DIY community who lack measurement capability, especially for
free-field acoustic measurements, and instead rely on computer simulation
and their ears. That is a very limiting method for designing loudspeakers
and guarantees sub-optimal results if the goal is accurate sound
Placement for Optimized Phantom Source Reproduction
|A joint paper by J.
Gerhard, B. Theiss, M. O. Hawksford from the 1996 AES Convention in
Copenhagen (Preprint 4246).
Their loudspeaker layout is quite different from
the ORION++, but similarly based on
psychoacoustic observations. I suspect that their listening experience
could have been even more convincing, if omni-directional loudspeakers
had been used. Those would have increased the reflected energy from
the wall in front of the listener over the energy coming from the side
walls. That would have reduced the need for micro placement of
speakers and listener to 0.5" (12 mm) accuracy without affecting
A setup of loudspeakers parallel to the long
wall of the room and close to the listener would seem to be well
suited for small rooms and the PLUTO.
It could also easily fulfill the minimum distance requirement of 3' (6
ms) from any wall. Close listening distance to the speakers also
reduces the magnitude of floor and ceiling reflections if that should
For general background see: Jens Blauert,
Spatial Hearing - The Psychophysics of Human Sound Localization, MIT
- Bob Pease is a legend in
electronic design. In the Pease Porridge column of Electronic
Design magazine he often talks about audio related electronics. For
all this best stuff, anyhow?" shows an easy way to measure
operational amplifier distortion and noise by analyzing the amplified error
voltage. In "What's
all this Sallen-Key stuff, anyhow?" he points to the influence of
the frequency dependent output impedance on filter roll-off performance. Bob
highlights many design issues and deepens understanding of circuit and
- The Acoustics and Psychoacoustics of Loudspeakers
in Small Rooms
Floyd Toole presented a tutorial on this subject at the New York AES
Convention in 2005. Here is a summary as taken from the March/April issue of
Multimedia Manufacturer Magazine. It appears that his latest findings match
more closely my own observations and experiences which had convinced me that
uniformity of directivity over a very broad frequency band, such as from
omni-directional and dipole radiators, is optimal for accurate sound
reproduction in typical living rooms. See also related earlier publications below
and my Letter to the
Editor of TAC.
A comprehensive review of this subject has now been published in the June 2006
issue of the Journal of the Audio Engineering Society by Floyd Toole, "Loudspeakers
and Rooms for Sound Reproduction - A Scientific Review". Highly
recommended study material for what is known and what needs further
- Auditory Scene
A relatively new field of study which should lead to further
understanding of the listening room's role in audio playback. I highly recommend to take a
close look at the work of Albert S. Bregman, "Auditory Scene Analysis -
The Perceptual Organization of Sound", MIT Press 1990. In a different
application this material has already led to greater audio data compression
rates but with
fewer artifacts. There is an Audio
CD available with 41 demonstrations of Auditory Scene Analysis. The
included booklet gives a very clear introduction to the subject and detailed
descriptions of the demonstrations.
in the Age of "Good Enough"
Blair Jackson, MIX
Magazine's senior editor writes about the recording business in the May 2005
and listener positions for optimal low-frequency spatial reproduction in
of Lexicon argues and demonstrates that realistic bass reproduction requires
more than a flat steady-state frequency response in the room. He points out
the importance of the randomly varying pressure gradient between the ears,
the velocity field vector, for spatial envelopment. The now widespread
listening experience with dipole woofers, which set up strong velocity
fields, confirms his observations. See also my Publications
(21), (1) and Issues in speaker design - C.
You Hear What I Hear?
Learning to listen in a mediated world. An interesting article in the June
2005 issue of MIX
magazine which talks about our auditory references.
Tank (2-D Waves) Applet
Paul Falstad has written and collected many interesting Applets. His Ripple
Tank Applet shows numerous acoustic and electric wave propagation phenomena
in an infinite plane and in a bounded plane. Of particular interest are
visualizations of the wave fields for point and dipole sources when walls
introduce reflections. Note that for acoustic waves the Fixed Edges check
box must stay unchecked at all times. Select a very low source frequency so
that the wavelength is in the order of the wall length to see what happens
in a room at low frequencies. Place the source in different locations. Vary
the distance between the (+) and (-) dipole sources. Focus on a fixed point
in the plane - the listening position - and observe the sound pressure
variation with time relative to that of the source. Check out the 3-D View
and rotate the plane. All this should give you an idea about the complexity
of sound propagation in acoustically small spaces where room
modes dominate. Then increase the Source Frequency and Resolution to
their maximum to see the essentially random reverberant field behavior above
the Schroeder frequency.
the Science Back into Loudspeakers - John Watkinson
A thought provoking article about the information capacity of loudspeakers
and how it might be measured. The mp3 encoded tracks on my Demo
CD give you an idea about your own sound system's information capacity
and/or your hearing ability.
The books by John
Watkinson are highly recommended. 'The Art of Sound Reproduction' should
be required reading for every audio equipment reviewer.
and the Brain - Norman M. Weinberger
The November 2004 issue of the Scientific
American carries an article about what happens in the brains of
listeners and musicians.
we Localize Sound - W. M. Hartmann
A very readable and thorough review from Physics
Today , November 1999, of how we use our ears and mind to localize sounds.
Audio - The art and the science - Bob Katz
If you are interested in knowing what can be done to the recorded sound
before the bits are committed to a CD, then you should read a few chapters
in this book. It was written by Bob Katz, a well known and respected
mastering engineer. Critical technical content has been checked by Jim
Johnston, formerly of Bell Labs/AT&T. We tend to judge speakers by our
favorite recordings without fully realizing how much the signals may already
have been processed to suit commercial interests. Highly recommended reading
and very educational.
a midrange driver from the cabinet - Andrew Jones of TAD/Pioneer
Different techniques have been used to reduce the transfer of vibration
energy from a loudspeaker driver to its cabinet. I like to mount the driver
by its magnet and to let the driver basket rim touch the baffle only lightly
so that little force is transmitted. It is effective and easier to implement
reliably than compliant mounting from the rim. The PHOENIX
and ORION Revision 0.1 driver attachments are examples of this.
In a box design I had rested the magnet in a cradle and clamped it with a
metal strap, which could be tightened from the outside using long
Loudspeakers at Revel - Interview with Kevin Voecks
Note the description of the listening tests. Trained listeners are used, but
what is their reference? Neutral sounding pink noise? Live acoustic sounds?
Speakers are compared against each other in double blind preference tests.
How accurate is the preferred speaker? Note the listening room with
diffusers. Certainly not what most people will use. Kevin Voecks clearly
states the importance of timbre due to direct frequency response, 1st
reflection and power response, the importance of polar response to 60
degrees off-axis horizontally. He points to the listening room as the
biggest problem. The need for multiple woofers to combat modes. These are
areas where an open baffle speaker like the ORION
has clear advantages due to the figure-of-eight polar response down to the
lowest frequency. Thermal compression in drivers and the associated response
changes due interaction with the passive crossover elements are non-issues
in an active speaker. No mention is made of other nonlinear distortion
tests. The importance of reducing resonances (stored energy) is highlighted,
but no mention of how Revel tames the cabinet vibrations or the sound inside
the cabinet. Revel certainly has all the resources one could ever wish for
and their top speakers probably define the performance plateau that can be
reached within the box speaker paradigm.
See also http://www.reed-electronics.com/tmworld/article/CA475937.html
for related speaker design targets of the Harman team.
Music: The Impossible Task - James Boyk talks from a performing musician's view
point about the difficulties
of recording music and reproducing it accurately. (See also James Boyk below)
psychoacoustics of multichannel audio - Robert
Stuart of Meridian Audio Ltd gives a tutorial about the different
psychoacoustic parameters that influence sound perception in single channel
and multichannel sound reproduction. Little of the material is
immediately useful other than to point out the complexity of the processing
that happens between our two ears. It can be expected that multichannel
sound has less masking than mono or stereo, though this depends heavily on
how the different channels were recorded and processed. Given the great
variability of stereo recording, which should be a stable technology by now,
it needs to be seen how well multichannel will do when there are no
standards for recording technique nor playback equipment performance. For
example, neither the SACD or the DVD-A consortium, nor any multichannel
receiver manufacturer, specifies the low frequency roll-off characteristic
of the multiple loudspeakers to be used, which is absolutely necessary to
know for proper bass management. Following merely the installed Home
Theater practices will hardly reveal the full potential of the new sound
recording media. See also my comments
and Publication 19.
- Interesting material on acoustics and psychoacoustics that is well
presented by Ralph Glasgal. This type of reproduction system requires a
dedicated room and is capable of a convincing sonic experience. Yet, it does
not fit my life style and there is something about it that I find tiring
upon extended listening.
- Reference earphones - My
motto is "True to the original",
which means true to what has been placed on the storage medium be it CD, DAT
or whatever. Every electro-acoustic transducer introduces distortion, with
loudspeakers being the worst offenders. Etymotic
Research ER-4S earphones come the closest to reproducing what was
recorded and they can serve the very useful purpose of identifying the areas
where a given loudspeaker is failing. The accuracy of these earphones can be
significantly improved by inserting a simple passive equalization
network into the electrical signal path. It is very
illustrative, but possibly disconcerting to hear, how a loudspeaker is distorting in frequency, time or
dynamics. For example, take track 2 of Deedee Bridgewater's Live
at Yoshi's CD and listen to the range of her voice, the drum set, the
background sounds, etc. If you are not using dipole bass speakers you will
be surprised to notice the sound that your boxes and room contribute to what
you hear. The earphones can be a reliable reference for comparing frequency
response, transient response and dynamics. I have some trouble with unnatural in-head localization, the lack of any tactile
sensation of sound, and pushing a plug into my ear canal, but at 10%
of the cost of the ORION they are a
most worthwhile investment in accurate sound reproduction.
- MP3 interview
with John Sunier for Audiophile
Hear me talk about some of the loudspeaker features that are most important
for realistic sound reproduction.
Artistry - The company builds a line of open baffle, dipole loudspeakers that I
The Beethoven-Elite and Beethoven-Grand are 4-way active systems that
incorporate my latest thinking in speaker design. The two models evolved
from the bi-amplified Beethoven. A review of the Beethoven is available in
The beginnings of the Dvorak series are described in my 1992 AES paper about a
compact dipole design.
The Vivaldi is a 3-way system and demonstrates the high level of performance
that can be obtained even in a fully passive dipole design. The frequency
dependent cancellation of front and rear radiation is a good challenge when
designing for a flat on-axis and off-axis response down to the bass region.
Audio Artistry Team
Siegfried Linkwitz, R&D
Marshall Kay, AA President,
Operations, Marketing, Finance, Purchasing, Personnel
Tom Hoffman, Distribution, Sales
Kurt Pasquale, Manufacturing
Vivaldi Loudspeaker introduction, 1996
- Shannon Dickson
||Shannon Dickson interviewed me for Stereophile
Magazine, (Vol.19, No. 4, 1996). I give a little history of how I got involved in
audio and what I am trying to accomplish. His article is reprinted
on the Audio Artistry web site.
Here he is in his home in Hawaii with his Audio
- Here are loudspeakers developed by Jorma Salmi. His observations and thinking
about the room-speaker interaction are very parallel to mine.
- Measurement and modeling of large signal parameters of drivers and
simulation of their effect upon distortion. Highly educational material that
applies especially to drivers for use in open baffle woofer designs, which
require large volume displacements. See, for example, Wolfgang Klippel's
109th AES Convention paper , "Diagnosis and Remedy of
Nonlinearities in Electrodynamical Transducers".
- "Rules of the
Game" by James Boyk - Lecturer
in Music in Electrical Engineering, California Institute of Technology.
Is Digital Audio perfect? This article is an excellent look at human perception
when trying to evaluate new versus old technology.
"....Then we must judge the New, not against the Old, but against the ultimate purpose of both. In audio, this means either a bypass test with a high-quality direct feed identified to the listener, or, if possible, a comparison with the live
of Sound" is about differences in listening and a reminder that a
loudspeaker must be able to reproduce the most subtle sonic details in order
to convey music's emotional message.
Griesinger's work on sound envelopment is required
reading for anyone interested in a music surround system. The popular home
theater setups for movie viewing provide mostly pan-potted five channel mono
sound, which works effectively because of the dominant cues from the picture
screen. Music surround requires different recording techniques and playback
processing to create a convincing illusion of spatial relationships. For an
introduction to the issues involved study the preprint
of Griesinger's talk at the Nov. 2000 Surround Conference in Paris.
Also, study the insightful MS PowerPoint slide presentation,
"Perception of mid frequency and high frequency intermodulation in
loudspeakers, and its relationship to high-definition audio - A
physicist meets the twilight zone". In addition to the ultrasound
findings, I was particularly interested
in his discussion of distortion in loudspeaker reproduction of massed chorus
and orchestra, beginning with slide 39. These are the two types of signals
that I find most critical for assessing speaker performance. They involve
naturally occurring distortion in the ear, which is also present at a live
You On The Road To ... Audio Hell? The problem of reviewing and
comparing audio components in the absence of a reliable reference is
discussed in this article from Enjoy
I maintain that an accurate system will give you sonic surprises. Too many
systems, though, impart a sameness to everything played through them and the
sound experience becomes predictable.
Animations - Here are nice illustrations of the sound fields radiated by
monopole and dipole sources.
baffle, full-range circular array source - A box-less speaker construction for
those with extra room for the rear radiation. I have no doubt that this
approach has the potential for very accurate sound reproduction. Look
at the many pictures if you do not understand Swedish.
Optimum number and location - Based on the assumption that a smooth
steady-state frequency response in the room is most desirable this
investigation by Todd Welti from Harman
International develops practical monopole subwoofer layouts. In my
experience the steady-state frequency response does not explain the qualitative
difference in bass reproduction between monopole and dipole woofers. The
low frequency transient behavior due to the active number of modal
resonances in the room response must not be ignored.
Additionally, the white paper "Getting
the bass right" by Floyd E. Toole will give insight into practical
ways for dealing with the steady-state response of the room.