Putting the Science Back into Loudspeakers

[phase_accurate]

I have never seen the Cabar form close range so I don't know how it works in detail.

The Legends are omnidirectional laterally. In the vertical axis they are omnidirectional at the lower end only.

Regards

Charles

 

[SaSi]

Thanks for reviving that old but interesting thread.

I read the John Watkinson article, twice.

He writes about a lot of things, but the main concept he promotes is fairly simple and acceptable from an engineering standpoint:

If you want to measure a quantity then you need a measurement instrument of known accuracy and a measuring error sufficiently small for the quantity to be measured.

You can't go on measuring a DC voltage of a few mV with a 3 digit digital voltmeter with a 1V as the smallest scale AND you can't measure mV voltages with an instrument that has a 200mV error.

I will also agree that modern loudspeakers appear to be in a standstill compared to what they were 30 years ago. Minor improvements here and there, widespread use of some new material, but no real advance in design concepts or specification setting.

Continuing the "flash" analogy, I will say that loudspeakers are (for audio) what the lenses are for photography.

Cameras may have gone digital and we may argue about the pros and cons of digital sensors vs film or one sensor against another, but a good lens is certainly needed for all.

Lenses seem to have improved heavily over the last 30 years. Zoom lenses were considered "second class" back then but are showing very respectable performance now. All was possible by use of computer models and simulation software that helps the designer improve and correct a design, as well as use of previously very costly to make aspherical lenses.

Exactly the same could have happened with loudspeakers. Instead, we are seeing 5.1 systems with paper cones in plastic enclosures.

 

[Andy G]

Exactly the same could have happened with loudspeakers. Instead, we are seeing 5.1 systems with paper cones in plastic enclosures.

You can still buy disposable cameras too, y'know!

 

[graaf]

You can still buy disposable cameras too, y'know!

well
but where are "heavily improved" loudspeakers?
where are "HiFi" loudspeakers?
what is HiFi?

well, in reality anything goes and Hi-Fi is in practice a meaningless marketing term - so it is truly in a dumpster with the rest of audio industry as Dr Geddes has observed:

I can't talk brands or people, but one guy (VERY famous) said "I have no illusions about audio today, its in the dumpster."

but why? has it failed in fulfilling its promise?

for me the most interesting thing in JW article is how He insists on importance of loudspeakers' performance in time domain, something commonly neglected in the industry

on the other hand sound source recognition research shows that time domain is at least equally important for human auditory system as frequency domain: http://sound.media.mit.edu/Papers/kdm-phdthesis.pdf

so there cannot be HiFi quality - in the sense of creating realistic experience (the listener hears virtual sound source and thinks it is real) - when the loudspeakers err in time domain - the transient is partitioned and waveform deformed

why do most musicians laugh at HiFi and audiophilia? especially classically trained musicians?
as Yascha Heifetz put it many years ago - "high phooey and hystereo"

IMHO the problem is that "HiFi" loudspeaker design is fundamentally flawed, it is unscientific
it is based on false generalizations inherited from nineteenth century science that hearing only "analyzes sound waves in terms of sinusoids - a Fourier spectrum"

Since Helmholtz, there has been a figurative tug-of-war between proponents of his "spectral theory" of musical sound and researchers who recognized the importance of sound’s temporal properties. Analysis-by-synthesis research, by trying to discover methods for synthesizing realistic sounds, has revealed several critical limitations of purely spectral theories. Clark demonstrated that recordings played in reverse - which have the same magnitude spectra as their normal counterparts - make sound-source identification very difficult. Synthesis based on Fourier spectra, with no account of phase, does not produce realistic sounds, in part because the onset properties of the sound are not captured (Clark et al.,1963). Although most musical instruments produce spectra that are nearly harmonic - that is, the frequencies of their components (measured in small time windows) are accurately modeled by integer multiples of a fundamental - deviations from strict harmonicity are critical to the sounds produced by some instruments.
For example, components of piano tones below middle-C (261 Hz) must be inharmonic to sound piano-like (Fletcher et al., 1962). In fact, all freely vibrating strings (e.g., plucked, struck, or released from bowing) and bells produce inharmonic spectra, and inharmonicity is important to the attack of many instrument sounds (Freedman, 1967; Grey & Moorer, 1977). Without erratic frequency behavior during a note’s attack, synthesized pianos sound as if they have hammers made of putty (Moorer & Grey, 1977).
So Helmholtz’s theory is correct as far as it goes : the relative phases of the components of a purely periodic sound matter little to perception. However, as soon as musical tone varies over time - for example, by turning on or off – temporal properties become relevant. In the real world, there are no purely periodic sounds , and an instrument’s magnitude spectrum is but one of its facets.

best regards,
graaf

 

[phase_accurate]

but why? has it failed in fulfilling its promise?

This is not a technical problem but a sociological one. I.e. the interests of the average consumer have shifted during tha last 20 years.

Regards

Charles

 

[graaf]

This is not a technical problem but a sociological one. I.e. the interests of the average consumer have shifted during tha last 20 years.

people not interested in listening to music anymore?

are You sure?

I disagree
my response is here:
http://www.diyaudio.com/forums/showthread.php?s=&postid=1509671&highlight=#post1509671

I agree with Mark Seaton:

Having sat in some interesting meetings and demonstrations with very influential people in the much larger realm of consumer electronics, it's most certainly clear there is a market for sound quality...
The hurdle is to package and present it in a way that people will buy it.
Consumers really do want it! They just don't want it to be more difficult or complicated...

...and with critofur:

That's funny, just the opposite is true, in terms of audio reproduction our capacity is still in the "infant" stage as far as I'm concerned. ALL speakers basically suck. There is just a wide range of suckyness from mediocre to terrible... Some of them may be pleasurable to listen to, but, by no means, are they capable of a realistic reproduction of all ranges of music. Some speakers are very good at some parts of music production, but they all have flaws, compromises, areas where they fall short.

yeah, they suck!
and they suck because their designers don't really know what they are doing

unscientific approach

best,
graaf

 

[graaf]

and they suck because their designers don't really know what they are doing

perhaps they know - they are making money on audophiles - perverted people who love that suckiness of sound, who are addicted to it

and at the same time best consumers - loyal shopaholics

and it is very scientific marketing approach

and as I noted in the other thread:

turn from this niche [audiophile oriented] marketing to mass marketing will not happen because it is not desirable, not good for business
higher investments and lower income (because we have to lower the prices)
nah! parasiting on obsessive-compulsive consumer is best thing for business

it is so clever business strategy but unfortunately a short-term thinking - a "short-termism" - "the most virulent form of business cancer"
parasite dies when its host dies
therefore audio is in the dumpster now because most people feel uncomfortable doing silly things like all this audiophile shopaholic and tweaking stuff

on the other hand perhaps many designers, especially in hi-end audio are honest in what they are doing
they are sincerely believing in what they are doing because they are audiophiles themselves

as JW put it, let me quote once again:

From a linguistic standpoint, the term audiophile is derived from the Greek philos, meaning love. However, in the conventional English usage, it can mean love of an unwholesome or perverted nature. I think there should be a register of audiophiles so we can keep our children away from them.

best regards,
graaf

 

[SaSi]

Well,
I would agree that in terms of consumer habbits, listening to music in the old sense is out of fashion.

The "fashionable" consumer products are projectors, large flat panels and AV processors/amplifiers.

5.1 surround allows most of the speaker shortcommings to be forgiven. Lots of drivers provide impressive SPL and we all know from the past that the untrained listener tends to prefer the louder sound as better.

People are nowadays "listening" to DVD and Blue Ray movies. Sound effects and sheer SPL appear to be the prominent discussion. Linearity and faithful reproduction of voice and natural instruments isn't really discussed in modern "HiFi" magazines.

The consumers are interested in solving other problems. How to project a 150" display with adequate luminance and no pixelation. How to share music and video in the house from a media server. How to carry our whole music library in an iPod like machine listening from mediocre earplugs *LOUD* so that traffic noise doesn't bother us.

I have read some 25 years ago the following as the "virtues" of a well engineer loudspeaker:
-It should have spherical or cylindrical shape to eliminate difraction and secondary emissions.
-If it has to be a box, the edges must be shaped in cylindrical form so that secondary emissions are reduced.
-Must use few drivers and their centers must be as close as possible in all axes. This means that the tweeter should be located backwards and not flush with the woofer's cone lips.
-Steep crossover curves generate phase distortion and must be avoided.
-Active speakers are prefferable as they eliminate passive crossover deficienscies and improve efficiency.
-Large coils with large travel are better so that they are employed in their linear mode travels only.
-Separate subwoofers allow for small satelites that can be appropriately shaped to reduce or eliminate diffraction and secondary emissions.
- Ferric magnets should not be preffered as they are not linear.
- The coils should be cooled as normal operation will generate so much heat their resistive behavior will change them altering the performance of crossover circuitry
-Vented (bass reflex) enclosures are efficient but should be avoided for decent bass reproduction.
- Cones/domes/ribbons should be as lightweight as possible to improve transient performance (How can one measure transient performance? What kind of graph shows that even in modern simulation and measurement s/w)?
-Enclosures should be made of inert material and constructed with enough braces and thickness so that they don't resonate.
-Passive radiators are preffered over vents as they provide better control over the "vent" and resonance frequency.

I maybe wrong in some of the above statements - I wrote them from memory. I would welcome any corrections.

However, my point is: Where are these (and many other) "findings" in the current designs? Were are the new findings of the last 5 years to allow further improvements?

 

[gainphile]

I'm not sure about the iPod generation don't know good sound. Headphones are inherently superior than speakers in many ways. A modest headphone can outclass expensive speakers.

I listen to my Nokia headphones and with their limitations (bass etc) the naturalness and resolution outclass my Linkwitz Pluto at times

We also need to observe the resurgent of commercial open-baffle products. I read the RMAF report here
http://www.6moons.com/industryfeatures/denver07m+h/denver.html
and the reporter noticeably surprised with so many of them around. Perhaps this is another example where people no longer longing for high SPL that vented/non vented boxes provides but rather transparency and naturalness.

 

[graaf]

I'm not sure about the iPod generation don't know good sound. Headphones are inherently superior than speakers in many ways. A modest headphone can outclass expensive speakers.

I listen to my Nokia headphones and with their limitations (bass etc) the naturalness and resolution outclass my Linkwitz Pluto at times

good point!

in fact Linkwitz Himself compares the sound of Shure E2c (an IEM of <100 USD price) to His flagship Orions
I have followed SL's recommendation and bought E2c - they are now my reference for good, HiFi sound

I would also say that good IEM is inherently (by design) superior to a conventional headphone - that is SL's opinion too

I have made numerous comparisons with hi-end conventional headphones

check "reference earphones" section at Linkwitzlab website

best,
graaf

 

[phase_accurate]

Although headphones can be quite good, there are many recordings that sound unnatural with them.
Or the other way around there are recordings for headphones and recordings for speakers.

Regards

Charles

 

[graaf]

I would agree that in terms of consumer habbits, listening to music in the old sense is out of fashion.
(…)
The consumers are interested in solving other problems. How to project a 150" display with adequate luminance and no pixelation. How to share music and video in the house from a media server. How to carry our whole music library in an iPod like machine listening from mediocre earplugs *LOUD* so that traffic noise doesn't bother us.

those fashions (and problems) are created by the industry

are You sure that those are consumer preferences and problems?

IMHO it is rather that the industry tells consumer what they need and what are their problems
this is pure marketing

we all know from the past that the untrained listener tends to prefer the louder sound as better.

mass market consumer prefers essentially what he is told to prefer
in audio niche market an "untrained listener" is not a problem
it is rather a "biased" audiophile listener
as Linkwitz put it:

People who only listen to loudspeakers and thus always compare loudspeakers are poor judges of accuracy.
Very few sales people of "high end audio" ever listen to unamplified life sounds. They are highly susceptible to marketing department suggestions.
Unbiased listeners have no difficulty recognizing accurate sound reproduction, even with hearing damage or with hearing aids.
Unfortunately, marketing departments and dealers think that bass and high frequencies need to be emphasized for products to sell.
Some listeners prefer euphonic loudspeakers. Accurate, and thus neutral, loudspeakers are not that exciting unless the source material is.

audiophiles want SOUND and they are given it by the industry
but music lovers want MUSIC and... they are ignored as not too good consumers

I have read some 25 years ago the following as the "virtues" of a well engineer loudspeaker:
(...)
However, my point is: Where are these (and many other) "findings" in the current designs?

well, perhaps check Mr Watkinson’s desings, for example

BTW there are IMHO more fundamental requirements:
HiFi loudspeaker should be time-coherent (=no transient split/deformation - one single clear step on step response measurement) - not only on-axis but omnidirectional, at least in the lateral plane to make inevitable lateral reflections coherent with the direct sound
in other words it should be as close as possible to a single point source
check Mr Watkinson’s observation WRT passive crossovers (quoted above in the thread)
HiFi loudspeaker should be silent – I mean it should generate as little parasitic noise as possible - this parasitic noise caused by energy storage can be seen most easily on CSD plots – it means in case of dynamic drivers using the smallest suitable and the same in case of any enclosures
and there is yet more important requirement - optimal loudspeaker-room interface - topic discussed in another thread I started

Mr Watkinson's designs meet most of requirements from Your list and also my above formulated requirements
so do designs of Mr Yoshii for Time Domain and Fujitsu Ten

but they are limited in their absolute SPL capabilities so there is still work do do

best regards,
graaf

 

[graaf]

Although headphones can be quite good, there are many recordings that sound unnatural with them.
Or the other way around there are recordings for headphones and recordings for speakers.

well, most speakers sound unnatural on most recordings

and real recordings for stereo speakers - I mean recordings respecting stereo principles formulated by Blumlein (stereo inventor) - are in minority

and strangely enough many speakers with many recordings sound better when listen to them as if to giant headphones - with crosstalk cancelled

it is so called transaural stereo or ambiopole in ambiophonics (www.ambiophonics.org)

best,
graaf

 

[soongsc]

Mostly development of loudspeakers consists of lots of trade-offs. Ideally the reproduction soure should be one. But if we only use one driver per channel, we are limited by dynamics, and bandwidth among the other design limitations. Sometimes when a driver is improved in CSD to a certain point, other problems stand out making the product unacceptable without other difficiencies masking it. Thus we are faced with the financial issue of "how much we should spend on development and when will we be able to break even selling such product considering consumers like change?" These are very tough decisions.

Ambiophonics is one set of compromises among lots of others.

 

[soongsc]

Many techniques involve using multiple speakers and multiple mic. The sad fact is, the more speakers/mics used, the more complicated soundfield thus the less possibility of finding a solution. Both our ears hear the sound from all the speakers and the room. If one asks a person the developed a complicated model to just identify one single condition to see if the criticle aspects are covered, the response you will get is an answer trying to make things not understandable.

A very similar situation as motion base used to give people motion feeling for simulations like flying a plane. Lots of people get lost in math and forget about reality. I once asked aircraft people to do analysis on aircraft rates and accelerations during a specific maneuver, then I compared it with a motion base which I knew the motion ques were very wrong (acceleration in the wrong direction). Sure enough, the comparison revealed where the wrongs occurred, but the motion que software experts had no clue what to do.

The most realistic sound reproduction I have experienced was doing a recording with mics pluged in my ears and playing it back through the earphones. Besides the noise of the mic and some mic rubbing, if someone spoke behind the listener in the recording, the listener actually tended to turn his/her head to confirm whether they we being spoke to or not.

 

[phase_accurate]

Yes the real thing for headphones are dummy head recordings (sorry, but they are called that way- though the expression might be out of place in the case of your recording).

Intensity stero recordings tend to give you "in-head localisation" when listened to on headphones. Though there would be possibilities to get araound the problem by processing the signal accordingly.

Regards

Charles

 

[graaf]

we are limited by dynamics, and bandwidth among the other design limitations

let's discuss it
what is problematic and what can be done and how

without other difficiencies masking it

what do You mean?

Mostly development of loudspeakers consists of lots of trade-offs. Ideally the reproduction soure should be one. But if we only use one driver per channel, we are limited by dynamics, and bandwidth among the other design limitations. Sometimes when a driver is improved in CSD to a certain point, other problems stand out making the product unacceptable without other difficiencies masking it. Thus we are faced with the financial issue of "how much we should spend on development and when will we be able to break even selling such product considering consumers like change?" These are very tough decisions.

yeah, "financial issues"... "development of loudspeakers"... "making the product"... "tough decisions"... for an "industry guy"
yeah, "consumers"

BUT are we here "industry guys"? or "consumers"?

NO - we are DIY hobbyist!

trade-offs? compromises? I say - **** compromises!

Ambiophonics is one set of compromises among lots of others.

I am not a promoter of ambiophonics, I mentioned it purely incidentally

best regards!
graaf

 

[graaf]

Intensity stero recordings tend to give you "in-head localisation" when listened to on headphones.

true stereo with coincident microphones is intensity based
no problem if there are distance cues (reverberation!) in the recording

the problem You describe is with closely miked and pan potted recordings
perhaps "pseudostereo dual mono" would be appropriate name
most of available recordings are some sort of "dual mono" "somewhat" producing "some sort" of "spatial impressions"

best,
graaf

 

[forr]

I was familiar with many of John Watkinson's ideas since his articles published in Electronics World. Just like dipoles or Linkwitz's Pluto, his recommandations of point source loudspeakers necessitate to install them away from any wall, more than three and a half feet. Not many people can afford to live in large rooms with speakers having an emission not submitted to reflections of a short neighbourhood.

I recently try baffles 2ft wide with rounded corners of 1/2' radius around the 25 mm dome tweeter (high slope crossover at 1.75 kHz). Albeit the realisation is very crude, I have been happy with the results in my relatively little room, the front baffle being at 2 ft at least from the rear wall and not parallel to it. The side walls are at 5 ft.

Compared with thin baffles (1/2 ft wide or less) with the same 25 mm dome tweeter, I have a feeling of less intermodulation, more clarity and localisation and less listening fatigue (an important criterion for J. W.).
I also have to say that I have prefered the symmetry of the baffle around the speakers.

Looking at Olson's diffraction figures, I think a less regular shape than a sphere or rounded corners can be beneficial for the linearity of the frequency response above 1500 Hz. It will be my next exploration in the diffraction territory.

 

[tschanrm]

Yes the real thing for headphones are dummy head recordings (sorry, but they are called that way- though the expression might be out of place in the case of your recording).

Intensity stero recordings tend to give you "in-head localisation" when listened to on headphones. Though there would be possibilities to get araound the problem by processing the signal accordingly.

Regards

Charles

Yes, good music reproduction starts at the source - the recording. Binaural recordings have been niche, and will continue to be, unless technologies like Optimal Source Distribution System (http://www.isvr.soton.ac.uk/fdag/vap/html/osd.html) are pushed more mainstream. Marantz has started to push this technology in recent years with the ES-150 and ES7001, but they are not priced for mainstream acceptance.