What Happened to Audio High Fidelity
In the light of this modern digital age, you are wondering why I ask such a question.
Why, we live in the computer age with fast clean digital technology capable of holding millions of songs..
My response to this is not one of a flattering nature. Digital!?
Now you are wondering if I'm one of those esoteric hifi nuts with tube amps and LPs, you-know, which century am I from. Maybe I’m from the 1950s with its reels of bulky fragile magnetic tape or those easily scratched LP records with annoying pops and ticks. Maybe I'm back in the 1960s when 8-track hiss and a 10 kHz bandwidth became all the rage. Or it's the 1970s when the garbled sound of Dolby cassettes was praised. No, like you, I am here in this new century, in the continuing digital age.
Yes, I was here in the 90s when CDs finally…well, sort-of, cleaned up their act. But, now here I am in the modern era of the many compressed mega-song formats such as MP3, a definite step backwards.
My disappointment with the new digital formats began in the ladder half of the 90s when I bought one of those giga-marvels, Sony’s MZ-NH1 1GB Mini-Disc Recorder. This disenchantment continued when I purchased an Apple iPod that came with a 160GB hard drive. Then the digital age sealed its fate in my eyes when I sampled satellite radio such as Sirius, & XM and local AM & FM HD Radio. Finally, I gave up when I messed around with the PC with music and Internet radio, all, audiophile-wise, major disappointments.
Alright, before I go into my tirade, I need to explain some basic fundamentals about audio for non-audiophiles.
First of all, as we all know, audio is sound which is waves of changing air pressure. These pressure changes are generated by a rapidly moving object and are easily perceptible by the incredibly sensitive human ear. This change of pressure which occurs at a given instance of time is called a sound wavelength. This length vs. time is used in calculating frequency.
The human ear can detect sound frequencies as low as 30 changes in air pressure or cycles per second and as high as 19 to 20,000 cycles per second. This, of course, varies with gender and age.
For some reason, women can hear higher frequencies than men. Also, the older we get, the less higher the frequencies we can hear.
At age 57, I can barely hear 13,000 cycles per second. I use to be able to clearly hear a television’s high voltage fly-back transformer, (the 15,000 cycles per second whistle of a TV set.) In fact, I could almost hear 18,000 cycles per second or 18 kHz.
‘k’ signifies kilo or 1000. Also, cycles per second is signified as a Hertz, after a fellow named Heinrich Rudolf Hertz who clarified and expanded the theory of electromagnetic waves, hence frequency.
Anyway, to get a sound, something must rapidly give and then take or push and then pull on the air around us. A guitar string vibrates back and forth there by vibrating the wooden box or in the case of an electric guitar, a magnetic field.
In recording sound, the sounds are usually picked up by a microphone.
Microphones work pretty much the same way as the human ear. Sound waves vibrate a tiny membrane within the microphone and it either mechanically moves a coil through a magnetic field or in the case of the modern day microphone, a metal membrane diaphragm is vibrated inside of an electrically charged or electrostatic field. These sounds are then converted into an alternating electrical current which is amplified and can then be saved (recorded).
Most of us don’t have to worry about this process of recording sound because we leave this up to the professional recording studios to purchase and operate the exceedingly expensive and sophisticated microphones and recording equipment. What we need to concern ourselves with is the medium by which sounds are stored and transported. This media use to be analog: vinyl LP records or magnetic tape. But now it is digital: CDs (Compact Discs) Music DVDs (DVD Audio Discs), MP3s, etc, etc.
My concern with today's distributions is the methodology by which the digitized audio information is given to us.
Records…are rather passé; prerecorded reel to reel quarter inch tape…can’t get it anymore; and cassettes are just not worth considering. By the way, these miserable things have finally been put out of their misery, music wise. That leaves the digital format. In fact, the cleanest most high resolution form that we as consumers can get music on now is the ubiquitous CD. Yes, I know, there is MP3 and those other technologies. That I’ll talk about in a minute.
But I need to get back to audio dynamics. Not to get too technical, I need to explain how sound is stored, preserved and reproduced.
As I said before, sound energy has frequency. In fact, sounds have many frequencies all mixed together. Low sounds, mid sounds, and high sounds…all hopefully harmoniously combined together. This is called audio or sound in the audible range.
In the world of high fidelity or hifi, the capabilities of a sound system of saving and reproducing sound is measured by the manufacturer and is then stated in the system's specifications. Three of the most common “specs” are:
The ranges of frequencies a system is capable of reproducing is called the frequency response. A system capable of reproducing 20-20,000 Hz is said to be a very good system.
Next there is the quiet to loudness of sound. From no sound to very loud sound; this is called the dynamic range. This is measured in Decibels or dB and is the measurement of the pressure or amplitude of the sound wave. The smallest unit of measurement is called a bel. The deci is equal to 10,000 of these. Anyway, without getting too technical, a decibel measurement is a logarithmic quantity, meaning for every unit of measure, the sound goes up by an exponential amount.
OK, enough math. A…good hifi system has a dynamic range of 100dB or greater. That’s from full quieting to very loud. This measurement is taken over the whole audio range. A flat response from 20 to 20,000 KHz is what is sought after…
…but is nearly impossible to achieve without significant distortion or cost.
The next broad issue is distortion or the misshaping of the sound. A good system has less than 1% distortion over the whole frequency range.
Now, by system I mean, it is the resulting sound that ultimately emanates from the speakers.
Oh yes, let’s not forget the issues of:
OK, OK. Anyway, audio researchers have been searching for the ultimate technology of storing and reproducing sound accurately since the cylinder player. Even though they have come far, they still have a way to go. These audio aficionados have a tough job because they are constantly fighting to adapt audio high fidelity or hifi to the ever changing trendy market place.
For example, the following is a list from an historical prospective of the most common end user consumer technologies:
Did I miss anything? Well, there were other mediums that were a flash-in-the-pan such as RCA’s huge quarter inch cassette like tape cartridge system, the Elcaset (1976) similar to the RCA Sound Tape Cartridge (1958 – 1964), miniature 8-track looking 2-track PlayTape cartridges (1967 - 1969) sold by Sears, and the DCC or Digital Compact Cassette (late 1992). There are others, but they are a rarity and hardly worth mentioning.
Today…over 30 years after the innovation of the CD, there is the anything-digital-goes computer technology, such as MP3, Sony Atrac, and a dozen other formats.
Yes, this is what I really need to talk about.
It’s the digital age!
For some reason, the word DIGITAL gets everyone all happy inside.
But I'm not so impressed with this format in its present form. Let me explain.
Digital Audio: is the sampling, at regular intervals, the amplitude of a sound wave and storing it as data. This sampling for a CD is usually 44,100 times per second for each channel, each sample being a 16 bit number representing a range from -32,768 to +32,767 units of amplitude measurement. This is called the polyphony or amplititude resolution.
This has been the Red Book standard for CDs since 1983.
The Red Book standardization specifies:
The bit rate is 1411.2 kbit/s: 44,100 samples per second × 16 bits per sample × 2 channels = 1,411,200 bit/s = 1,411.2 kbit/s. or 176,400 bytes per second of music or 600 mb for 70 minutes of music..
As each sample is a signed 16-bit two's complement integer, sample values range from -32768 to +32767.
On the disc, the data is stored in sectors of 2352 bytes each, read at 75 sectors per second. Onto this the overhead of EFM, CIRC, L2 ECC, and so on, is added, but these are not typically exposed to the application reading the disc. By comparison, the bit rate of a "1x" data CD is defined as 2048 bytes per sector × 75 sectors per second = 150 KiB/s, or approximately 9.2 million bytes per minute.
It must be noted that when the CD was first introduced, most people could hear a certain harshness caused by a strange harmonic. Later on, the audio CD player manufacturers came up with an internal computer algorithm to take the 44,100 samplings per channel and through some real time calculations internal to the CD player, convert each sample into 8 times as many samples. The CD player’s tiny embedded computer processor would look ahead to the next several samples and calculate (guess at) 7 additional points on the wave form or sign wave giving it a smoother more realistic shape.
That is all well and good. In fact, CD audio is rather good for 1980s technology. Today’s CD players are fairly good at reconstructing the waveform and is hardly indistinguishable from the original source, for most people. But many audiophiles…like me, can hear the difference.
The problem is simple. The maximum frequency that can be accurately represented in a sampled waveform is termed its Nyquist frequency, and is said to be equal to one half the sampling rate. Thus, for example, a waveform sampled at 44,100 times per second can represent all frequencies up to its Nyquist frequency of 22,050 cycles per second. But a problem called aliasing occurs when a signal to be sampled contains frequencies near, at, or above the sampling Nyquist frequency. The next figure illustrates how aliasing will occur when the sampling rate is much too low for the frequency of an input signal.
To solve these problems, a sampling rate of four times is required. To get LP sound, eight times is required. LP - 33 1/3 RPM long playing vinyl record.
Setting aside various issues with the pops and ticks associated with misused records, I really did like these things.
Yes, you heard me correctly. In 1979, I had one of the best audio system ever constructed by human kind. It had a homemade preamp and two Mac mono-block class A tube amplifiers pushing two Quad ESs, all driven by a direct drive Denon turntable with precision micrometer like tone arm and a Grace F9-E cartridge. It was perfection.
So, what happened to my high end stuff? It got old and was sold. The Quads finally quit working and the amps were sold for $1,000 each. I still have the turntable. "With today's technology, I thought I could easily replicate this system" But, I was so very wrong.
Well, electrostatic speakers are still being produced today by two or three manufacturers. And relative to what they cost in 1979, they are a bargain. As for a Class A amp, there are dozens of different tube amps being made. And, clean preamps? They're around. But, LPs are or were the best sound available to the common consumer. CDs are clean, but lack some definition and depth.
The industry did promise music DVDs (2 channel audio with 192,000 samplings per second at a 24 bit sampling resolution.) BUT, and I say but. It has been 13 years now and hope has faded. Music producers have been hesitant to adopt this format. Well, there has been some stuff, but it's not mainstream and like direct-to-disc LPs of the 80s, not very much is being produced on Music DVDs. Plus they are very expensive and may be short lived due to a lack of demand.
What I along with other audiophiles...and most violinists miss is, the sound of the bow being drawn across the strings. In a live performance, there is a certain sound that is there. But on a CD being played through the most sophisticated digital to analog converter, this sound is not there. The truth being: it is known that people who are used to hearing symphony music via a CD player, become disturbed when hearing an orchestra in a live performance. It is the strings and the trumpets. "There is this certain fuzzy noise."
Oh well, the studio must filter that out, right?
Nada…This is the limitation of the current digital sampling rate. It can be heard through analog magnetic tape, LPs and DVD Audio with a sampling rate of at least 96 kHz per channel or higher, but not CDs at 44.1 kHz per channel sampling rate and most certainly not MP3s.
Lossy vs Lossless Compression
Lossless and Lossy compression are terms that describe whether or not all the original data can be recovered from a compressed audio file when this file is uncompressed. With lossless compression, every single bit of data that was originally in the file before it was compressed remains after the file is uncompressed. All of the information is completely restored. As for lossy...
Now For My Rant
Today in this modern age, there is this new technology called Digital Audio Compression, a methodology that allows for the significant shrinking of digitized audio. Many say they cannot distinguish between CD audio PCM (lossless audio), and the compression methodologies such as MP3. Well, I am here to tell you it is definitely noticeable…even by the untrained ear.
For me, it is disturbing to my 57 year old ears. I’d probably be really sick like I was with Dolby cassettes in the 80s if I still had the hearing I had when I was 25. In fact, when I was 35, I heard a CD for the first time and said, “This thing will never catch on.”
To me, a cymbal recorded on an MP3 is a 10 kHz whistle. I also notice that the ambiance is missing. Then I hooked up my PC (MP3 compression) to an oscilloscope and looked on the screen at the waveform. And low and behold, a Zildjain Cymbal is a whistle, a high pitched jagged-edged repeating sign wave. That’s when I figured out what they were doing. Not only are they not capturing low level background sounds, they are also repeating exact waveforms that are 95 or even 90% the same. That’s how they can compress a 5 minute song into 30 seconds of storage. This of course leads to a loss of 5 to 10 percent. The human ear can easily detect this.
Question: would you buy a stereo system that had as one of its specs a 10% distortion factor? Well, that's what you are doing when you buy into MP3, WMA, AVI or ATRAC.
Anyone who appreciates real audio knows that every moment of true lossless audio is so minutely different that it cannot be compressed. It also cannot be digitized at a slower sampling rate than at lease 4 to 5 times the upper limit of human hearing. And believe it or not, humans can distinguish subtle changes in sound pressure level greater than 1/32,000ths of the 100 decibel human dynamic audio range.
Oh well, for the average banger hard rocker, this current compressed mechanical sounding digital format is fine...I guess. But for me...the CD is barely tolerable.
My Simple Audio System
Along with the CD player, I still have my LPs and reel to reel tapes and I listen to them through the old Denon direct drive heavily weighted turntable with a Grace cartridge and a Teac heavy metal reel to reel deck respectively. The signals are sent through a studio mixer to a pair of Mackie HR824s. The total cost of the mixer-preamp and speakers is comparable to most mid-priced systems. But unlike many high priced systems, this system's audio is studio quality.
The result is... Honestly!
This is a copy of the graph printed on the calibration certification that came with each Mackie studio monitor speaker.
So even though the Mackies are not electrostatics, I’m OK...for now.
I do keep looking on the web wishing for recordings recorded at 96 KHz 24 Bit linear lossless 2 channel PCM DVD-Audio and I do find some. But like prerecorded reel to reel tapes in the 60s, music DVDs of this type are very expensive and are limited. NOTE: This is not what is known as High Definition 192 kHz surround sound 5.1 or 7.1 audio. It is discrete 96 kHz per channel sound. True HD is 192 KHz 24 bit 2 channel PCM Audio.
I hope by this article you have become more informed. I also hope you continue looking or listening to the current digital technology with more sensitive ears. And I hope more people will start demanding higher resolution digitized music and hopefully the industry will start producing audio in this format. Then we will truly be in hifi heaven.
Steve II 2007