In the late nineteenth century, Guglielmo Marconi invented the radio and Thomas Edison invented the phonograph. The radio takes an acoustic event in one place and reproduced it in another place. The phonograph takes an acoustic event in one time and reproduces it at another time. The unifying aspect of these two technologies is to take a sound event and make a copy of it somewhere else and somewhen else.
It is worth going over the point of high fidelity, as I see it, so we can understand its post-modern failure. Or you can skip to the nasty part.
Over time, audio technology developed another goal, another quest, the search for audio truth. Not only do we want to be able to recognize music or speech from another time and place, we also want to get a faithful reproduction of the original event. For about a century, scientists tore apart acoustic events into their essential elements and calculated the combinations of those elements, the molecules of sound. The joy of audio science is that it makes us investigate what physical realities make music what it is.
We learn about sound. Sound is traveling waves of high and low pressure (compressions and rarefactions) in a physical medium, almost always air in our experience. We can describe an acoustic event, musical or not, by its air pressure as a mathematical function of time, a specified pressure for every instant. A sound source, a musical instrument for example, produces pressure waves that propagate through the air and bounce off of reflective surfaces to form the entire acoustic event. We can describe that air-pressure function of time as a different mathematical function of frequency, how often the compressions and rarefactions occur. Frequency is measured in cycles per second or hertz (Hz). Over a period of time, each frequency has amplitude and phase. The amplitude tells us how "strong" that frequency is and the phase tells us the time-relation of that frequency to other frequencies.
We learn about human hearing. The human ear is decidedly biased towards frequency in its interpretation of sound. We hear tiny differences in frequency as changes in pitch while it takes fairly large amplitude changes to be heard as loudness changes. There are conditions where our ears are highly sensitive to phase differences and other times when they are not. Our ears not only sense pitch and loudness in sound, they also locate sound sources in space.
We learn about space. As we listen, a musical ensemble is broken down into its individual players and each player is assigned an acoustic position in the listening soundstage. Our two ears somehow manage to create a full, three-dimensional picture of an acoustic event. This picture is called an image or soundstage in hifi jargon.
We learn about music. Music has many components including rhythm, melody, counterpoint, harmony, and tonality. The forms of these components over the course of a piece define the musical experience. Rhythm, melody, counterpoint, harmony, and tonality take us back the basics of hearing, pitch and loudness in all its forms. These take us back to the basics of sound, frequency, amplitude, and phase. And these take us back to the physics of air motion, pressure and velocity.
Artisans who craft musical instruments do not share this quest. Their science takes them down a different road, one that shares many of the same principles, perhaps, but not the same goal. The design of a violin or a trumpet requires unique and wonderful insights that are not the same as the insights required to reproduce those sounds from a piece of machinery.
As I see it, the hifi goal goes beyond the correct reproduction of a musical event. Having a single reproduction of a single event is not enough. The medium is successful when it can bring a large volume of musical events to a large audience. The volume of music should be large enough to include the standard genres of classical, jazz, folk, and popular music and should be large enough to include some other areas as well. Hifi media should allow for the production of thousands of copies.
The architecture of music reproduction has been remarkably consistent from Thomas Edison's time until the present. A musical event is recorded on some kind of master which is copied into a medium that can be mass-reproduced and distributed. Thirty years ago, a recording was made onto a master tape copied onto a record lacquer with a cutting lathe. The lacquer was used to make metal stampers that created vinyl records. Today a recording is usually made on a digital medium, stored and processed on a computer disk drive, and compact disks are mass produced.
Microphones: The science of sound reproduction starts with detecting it. A microphone system is a pickup and some kind of preamplification to bring the signal to line level so it can be recorded. The most common forms of pickups are dynamic, condenser, and electret. There is a whole science of microphone placement to record a musical event the most natural way.
Tape: The study of reel to reel tape is vast. There are various tape formulations and tape decks must be biased and equalized to get the most performance out of the tape being used. Using wider tracks and higher tape speed uses more tape, of course, but reduces the relative background noise level, the tape hiss that we hear in the background. Higher tape speed also gives more resolution of what makes music what it is.
Direct to disk: A few records have been made without the intermediate stage of tape or digital recording. These direct-to-disk recordings are mostly audiophile exercises. The recording environment is very tense with the cutting lathe operating "in real time" during the performance. That tension robs the musical performance of its natural flow and beauty. There are a few exceptions and these are spectacular examples of what the vinyl medium can do.
Digital recording: Recording directly onto a bit stream eliminates the mechanical speed-control issues of wow and flutter and the introduction of medium noise such as tape hiss. Digital recording study involves intense mathematics with complex quantization and sampling issues. Quantization science affects dynamic range, how loud and soft sounds can be recorded, while sampling technology affects frequency bandwidth, resolution of high frequencies. The lowest level signals are determined by a dithering algorithm.
The recording and manufacturing part of the process is outside the realm of most hifi enthusiasts, so let us concentrate on the part of the process that converts mass-produced medium back to sound.
The architecture of hifi is also consistent. We have sources, a line amplifier, power amplification, and speakers. Sources such as tape decks, turntables, and digital disk players turn medium into electrical signal. Line amplifiers allow the listener to select a source and to make adjustments, typically volume level. Power amplifiers boost the line signal to voltages and currents for speakers to turn into sound. The proliferation of boxes in most hifi setups are mostly an increase in the number of sources. My own system can play two kinds of reel to reel, vinyl records, compact disks, and compact cassettes for a total of seven inputs and ten separate gadgets.
Sources: Each source medium has its own raison d'être with its own advantages and disadvantages. Open reel tape sounds magnificent but is expensive to reproduce as it requires a tape-to-tape copy. Vinyl records are a lower-cost compromise as they can be stamped out far more cheaply and they offer "random access" as the user can easily select any starting point. Compact disks are cheaper, lighter, and smaller than vinyl records and CDs can withstand harsher conditions. Compact cassettes are small, light, and cheap and they can be recorded easily at home, although CD burners are taking over for home-copying.
Line Amplification: The hifi listener has to be able to select a source and to adjust the playback volume. This usually involves a stage of gain so the control boxes are are line amplifiers as well.
Preamplifiers: The most popular phonograph cartridges are magnetic, moving field or moving coil, and these require a stage of preamplification. The control boxes of yesteryear came with phono preamplification circuitry as their most complex and sophisticated circuitry. We called the entire control box a preamp in those days, but it is inappropriate to call a control-unit a preamp if it has no phono preamplification stage in it.
Power Amplification: The voltage from the line amp has to be boosted to speaker levels. The speaker requires not only voltage but current as it is doing physical work converting electrical energy into sound. The power amps are often separate from the line amplifiers because they are a separate function packaged separately and because it is often better to have them near the speakers. Amplifiers come in various circuit-design flavors, Class A for the simplest and most linear design, Class B for the electric-bill and heat efficiency, Class AB to mix the blessings of the two, and other, more original concepts floating around high-end audio. The gain can come from tubes or solid-state devices (transistors).
Speakers: Speakers are the visible part of a hifi. They usually sound best when they are kept away from a wall, so it is hard to hide them in listening room decor. There are various ways to convert electrical energy into sound and several are used in speaker designs. The most common is a paper, metal, or plastic cone vibrated by a voice coil in a magnetic field. The best speakers in my listening experience have been large-area dipole designs, panels that vibrate with reasonably uniform thrust over their area. The Magnaplaner designs were my first introduction to these kinds of speakers, but most of the planar designs have been electrostatic, using changing charge in a static electric field to move a membrane.
Crossovers: Few speaker elements can reproduce the full ten octaves of human hearing from 20 Hz to 20000 Hz. So we have woofers for low frequencies, tweeters for high frequencies, and mid-range units for frequencies in between. These frequencies are separated by crossovers, electrical circuits that send different frequencies to different components. These crossovers are usually part of the speaker system, but they can be used to separate the line-level signals so separate amplifiers handle separate frequency ranges in a bi-amped or tri-amped system.
Cables and A/C power: The wires that connect one hifi component to another seem to have their own contribution to the sound as does the power coming out of the wall plugs. As trivial as these items seem, experience has shown us that these items require some design attention as well. Several meters of wire can change the shape of high frequency audio signals. The signal coming out of the wall plug has a lot of radio frequency and electromagnetic interference (RFI/EMI) coming from digital equipment such as CD players and personal computers (PCs). (Try listening to your hifi with all the digital stuff in your house turned off and then turned on. You may hear a dramatic degradation.)
What makes one sound system better than another? A sound system is better when it reproduces more of the musical event more correctly and resolves more musical information than another. Since the full sound system consists of so many stages, from microphone to speaker, it makes sense to consider components separately. So we ask, instead, what makes one sound system component better than another?
The obvious answer is that one component is better if it sounds better than another component. That is useful, perhaps, for audio reviews, but scientists want to know the conditions of superiority and engineers want design criteria for superior design. So we sought a system of measurements we could apply to each audio stage to establish superiority.
The measurement technology for each stage is different, but certain features remain the same. We do not measure the output of a speaker the same way we measure the output of a phonograph cartridge or a tape deck, but we want them to have the same virtues.
Frequency response: The distribution of frequency going into a piece of audio gear should be correctly represented in the output. The simplest test is to run a sequence of pure sine-wave tones into the unit and to measure the level coming out. We can plot this on a piece of graph paper and note how many decibels (dB) the output-level line wanders from its intended trajectory. It does not take a lot of imagination to visualize the importance of correct frequency response. Musical tonal balance is affected if some pitches are emphasized while others are suppressed.
Harmonic distortion: Once we determine the amount of signal as a function of frequency, we ask whether any new frequencies are added. When a pure 100 tone goes into a component, what comes out should be 100 Hz, but sometimes harmonic frequencies come out, 200 Hz, 300 Hz, 400 Hz, et cetera. We measure the total harmonic distortion (THD) as a percentage of the signal. Musical timbre is affected when harmonics are added to the signal.
Intermodulation distortion: Two frequencies can interact with each other to produce extra output at their sum, their difference or other, uglier frequencies. We can put two high-frequency tones into the equipment and see what lower frequencies come out. We measure the intermodulation (IM) component of the reproduced signal. The interaction of musical instruments is affected when extra interaction is added to the signal.
Transient accuracy: All of these tests are steady-state, continuous signals are applied and measurements are taken. Music consists of starts and stops and these transient moments often define the musical and physical shape of the music to our ears. We can apply a single pulse to kick the hifi component and see what happens. The signal can rise too slowly, it can overshoot the pulse level, or it can exhibit all kinds of strange gyrations. There is are measurements of voltage slew rate and transient intermodulation (TIM) distortion. The rhythm and space of music is affected when the initial waveform of a musical sound is altered.
Limits of measurements: The good news for those of us who believe in science and engineering is that equipment that measures well generally sounds better than equipment that measures poorly. The lab tests do an excellent job separating bad hifi from good. But once we have two components that measure well, refining the measurements does not do a good job separating good hifi from better. It is like using the SAT to detect child geniuses. An amplifier with 2.0% THD is almost certainly going to sound awful compared to one with 0.2%, but the leap from 0.2% to 0.02% is less certain to result in sonic improvement. Frequency response errors of 2 dB (60%) are almost certainly going to sound worse than a frequency response within 0.5 dB (12%), but refining the design to get within 0.1 dB (2%) is, again, less certain to result in better sound.
Back to listening: So the final discrimination among the best hifi components goes back to listening tests. One may argue that we should have been using our ears all along in the hifi design process. One may also argue that a component that contributes gross distortions to the measured signal is not going to present a musically correct reproduction and that auditioning such gear is an exercise in politically-correct open-mindedness. The proper course is probably to stick to the good stuff almost all the time and occasionally to try something that looks questionable. When an audiophile voice that has been reliable tells me that this awful-looking gadget sounds great, I sit up, pay attention, and try the gadget for myself. With occasional journeys into the realm of deliberate open-mindedness, I spend most of my time staying in the realm of science and experience.
Listening experience:
The ability to evaluate audio equipment
is not limited to golden-eared audio reviewers.
It takes time to sit down, to enjoy a piece of music,
and to let an audio component's sound sink in.
There are two aspects that make a listener a better critic.
The first is objectivity,
and I agree with Corey Greenberg
that listening to live, unamplified music is good
but that experience trying to record and reproduce music is better.
The second is practice,
the ability to listen for fifteen seconds
and to know what one will think after half a hour.
I call that being in "listening shape."
A bunch of us well conditioned listeners
were evaluating turntable mats,
gizmos that go on a turntable under a vinyl record,
and we did not need more than ten or fifteen seconds
for each one.
When we went back later
and did more extensive listening
to our favorites,
we found our ten-second conclusions
persisted over longer auditions.
Tubes: The quest for better-measuring gear circa 1970 took hifi from tubes to transistors. They clearly measured better and amplifiers took a leap from 0.1% THD to 0.001% THD. But a few luddite listeners called attention to the more musical sound of older, vacuum tube amplifiers like the Dynakit Stereo 70 that so many of us grew up with in our living rooms. A few serious audio designers went back and applied their best hifi science to making good tube amplifiers and Bill Johnson produced the Audio Research D-75 and D-150 models. These were spectacular amplifiers that stood on top of the audio mountain for several years. I remember the D-150 presenting the entire audio range from the deepest bass to the highest frequencies with a freedom from coloration, an articulation of musical detail, and a stunning three-dimensional image. Their SP-3 tube preamplifier was similarly superior to contemporary solid state preamps. Here was equipment that measured worse and sounded better than contemporary solid-state gear, but the best tube and transistor equipment measured well.
Transistors: Around 1980, two extraordinary solid-state amplifiers appeared in the high-end audio market. John Iverson offered the Electro Research A-75 and Andy Rappaport offered the Rappaport AMP-1. Both of these were 40-volt (75 watt per channel) amplifiers with temperature tradeoffs in their design. The A-75 had a noisy fan and the AMP-1 ran egg-frying hot. And these amplifiers changed the picture of hifi by presenting a musical signal with more range, more detail, more texture, more space, and more music than the previous generation of top tube amplifiers. These two amplifiers were an interesting contrast in design and measurement. The A-75 was a high-feedback design with something like 0.0005% THD while the AMP-1 was a no-feedback design with 0.5% THD at half-power in the high frequency range. Both amps, however, were articulate in their reproduction of music. Mark Levinson made some superlative solid-state amplifiers, but not in this league. And Jim Bongiorno offered the Sumo amplifiers which were. Russ Sherwood has continued improving John Iverson's designs to make the Eagle 11 amplifiers in my living room.
There is a community in the twenty-first century that seems obsessed with the Good Old Days. They drive new VW beetles and Chrysler PT cruisers and watch television shows with "Star Trek" in the name. These are not luddites, they like new technology, but they want it to have an old-style theme. These are the people who would buy Art Deco cell phones.
I share their sentiment, but not their choices. If I could get a brand-new 1967 VW beetle, then I would drive it heartily. The beetle was a fun car because it was fun to drive and its appearance was a consequence of its design. The new beetle is a theme car, a perfectly adequate post-modern VW Jetta on a frame shaped like the old bug. I fly a 1967 airplane not because it looks cool but because it is fun to fly. I still enjoy watching the original "Star Trek" from time to time, more than the three spin-offs that put "Star Trek" in their names to bolster ratings. The retro movement has chosen the appearance of the past over its content.
We have seen the same retro movement in hifi. There is a new generation of equipment that looks like the old, excellent gear of yesteryear. Turntables and tubes proliferate.
This is mixed blessing.
I have
a terrific Camelot CD playback system
and a nice Sony CD recorder.
The sound is fine,
but not in a league with my
modified
Linn/LOCI/EK-1 phono system.
(It has to be better if it's
modified.)
The retro movement
maintains an interest in vinyl
so there is a network of record stores out there
and, of course, there is
eBay.
A convenient spin-off of the back-to-vinyl movement
is a passing interest in reel to reel tapes,
more expensive and often better sounding than vinyl records.
These reels have almost no presence in stores
and the only real reel market is
eBay.
Why do I lament post-modern hifi as a failure?
Because the combination of digital audio and tubes
has obscured the meaning of musical accuracy in audio.
The high frequency component of most digital audio systems
is absolutely awful.
While some really good engineers
are making the digital high frequencies better,
much of the high-end audio community simply designs
electronics that depress the high frequencies
and substitute their own false color.
Everything sounds musical on these systems.
The good news is guitars, pianos, and violins sound musical.
The bad news is jack hammers and steam shovels also sound musical.
These machines are themselves musical instruments
rather than devices of musical reproduction.
As the amplifiers got more laid-back in their treble frequencies,
speakers became brighter to compensate for their deficiencies.
We now have bright speakers
compensating for dull electronics
compensating for harsh source material.
The equipment measures as bad as it sounds
with wayward frequency response
and high distortion levels.
It took me until 1984 to get the
hifi
I always wanted,
Linn/LOCI/EK-1 for vinyl,
ReVox A-77 for half-track tape,
Aiwa M700 for cassette,
Eagle 7A power amplification
for Quad Electrostatic Loudspeakers (ESLs) with Deccas on top,
Advent 300 power amplification
for a Fried "T" on the bottom.
That's pretty much what I have today.
(With the introduction of the ESL-63 in 1981,
the older Quad ESL model has been called
the "ESL-57" or, my preference,
the "old Quads.")
I added a Camelot Arthur compact disk playback system,
a Sony TC-650 for quarter-track tape,
a Sony CDR-W33 compact disk recording system,
and an Eagle 2000 control box and line amplifier.
I replaced my damaged Eagle 7A with a pair of Eagle 11 amplifiers
and my worn out Aiwa M700 with an Aiwa AD-F810.
The new Eagles do sound better than the old gear
from the same designers,
so I cannot claim to have exactly the sound I had then,
but it's pretty much a 1984 hifi rig.
I remember the excitement of hearing gear
that was absolutely and clearly better
than what I had in my living room.
From 1980 to 1984 I was "blown away" by
the Mark Levinson HQD (Hartley-Decca-Quad) system,
the Electro Research A-75 amplifier,
the EK-1 phono system (no preamplifier),
the Sound Lab R-1 electrostatic panels,
the Eagle 7A,
and, of course, my dear old Quads.
Each of these was an enormous stride closer
to realizing the emotional wonder of music
from a piece of machinery.
Since 1990 I have only had one audio mind-blowing experience,
the Weathers phono system from the 1950s,
a design that disappeared from the market
when it could not play vinyl records in stereo.
Nothing made since 1985 has given me a hifi thrill.
For almost two decades,
all these experts designing all these products
with all this new technology
have not been able, even once,
to generate the excitement of a genuine audio breakthrough.
I suspect that
the effort to make digital audio as good as analogue
has absorbed
all the engineering talent that
used to make analogue better and better.
Ed McMahon shows up at my door with a cheque,
what audio would I buy?
That was a game we audio-weenies would play
twenty years ago.
What would you buy with (U.S.) $1000,
$10,000, or $100,000?
Even though I was on a college-student budget at the time,
I had definite opinions of what I would buy with $20,000
and what I would change going to $30,000.
The world was full of new, fun, exciting stuff.
Linn Sondek turntables,
Audio Research preamplifiers and amplifiers,
and big panel speakers headed the lists,
but there were tape decks and super-deep-bass subwoofers
to be added with incremental funds.
Jump to today,
what new hifi gear would I spend Ed McMahon's money on?
I would buy a pair of good microphones
for a few thousand dollars (each).
My Nakamichi CM-701 cardioids make very nice recordings,
but the improvement going from them to a pair of
Neumanns, Schoeps, or B&Ks would probably be significant.
I might lean on Russ Sherwood
at EKSC Eagle
to make another Eagle 5 prototype for my Quads and Deccas,
an amp that sounds even better than my Eagle 11 pair.
So far as old gear goes,
I might go on a quest to find a Weathers phonograph system
and I might find the space for
a Studer, Scully, or Stellovox tape deck
to play the live "master" tapes
recorded on my portable ReVox decks.
But that's it,
the hifi thrill is gone today.
Have you noticed that the hifi has lost
place of pride in post-modern living rooms?
In high-tech, modern times the hifi was glorified,
proudly displayed in the living room.
Many young people have no more hifi
than the cute little speakers that came
with their personal computers (PCs)
or they have a "home theater" system
built around big-screen video.
Whatever it was that excited me about hifi
is not exciting the new generation
or maybe it just isn't there anymore.
It isn't me.
I haven't lost the urge.
Rather,
the hifi industry has lost the vision.
The message has been lost in the post-modern haze
leaving me with a malaise about the whole business.
Hifi is just no fun anymore.
3:31:14 Mountain Standard Time
(MST).
188 visits to this web page.
