The next time you sit down to listen to music, not as “background” music and not on your “personal listening device,” but to really enjoy music, hopefully on a reasonably high quality audio system, I would like you to do yourself a professional favor and to critically listen to what you are hearing.

First of all if it is music it must contain; 1. Melody, 2. Harmony, and 3. Rhythm. Without all three of these, by definition it is NOT music, just sound.

I might further suggest that another component of naturally created music is its natural frequency spectrum and it’s dynamics. The difference in amplitude between loud and soft passages. This is what gives music it’s excitement...it’s natural energy. Further refining these components will require some basic physics.

The History
In the late 1930's, as a result of the oncoming war, Bell Telephone Labs was involved in some research work that set the standards and groundwork for most of our knowledge of audio perception and what has become known as electro-acoustics. (Much of the lore and wartime results of this can be found in a recently published book Secret Soldiers.) The scholarly discoveries were published in a 1934 copyrighted document by the American Institute of Electrical Engineers titled Symposium on Audio Perspective. I was introduced to this document by the late Paul W. Klipsch, founder and owner of Klipsch and Associates, makers of the famous Klipsch Horn, who became a personal friend and presented me with this document as part of gift of a series of audio papers. I personally consider this document one of the most important audio document ever written. For those interested the document it is now available as a .pdf file download from the Audio Engineering Society at:

http://www.aes.org/aeshc/docs/bell.labs/auditoryperspective.pdf

Some of the data found in this document is foundational to what I intend to present. However, the entire document is important to anyone who performs or records real music.

It is generally accepted that the dynamic range of human hearing is 120 decibels, from the threshold of perception to the threshold of pain. This is an extraordinary capability and it in itself and when fully understood leaves me no doubt of God’s “intelligent design.” (But that is another subject.) A intuitive understanding of this great dynamic range could be gained by the analogy that a 3 decibel increase in acoustic power (just perceivable) requires a doubling of amplifier power. It is important to note that the CD is the only “consumer delivery medium” capable of delivering this kind dynamic range and frequency content to the consumer. Black Vinyl cannot do it nor can replicated cassettes, even CRO2.

Human frequency perception is accepted to be from 20 to 20,000 Hertz (Hz.) or cycles per second. (Recently, some have said that this extends to around 40,000 Hertz, but this is not generally accepted nor scientifically supportable for the general populace; nor is it practical because of air attenuation at high frequencies.) The perception of these frequencies is both frequency and amplitude dependent. We perceive mid and mid-high range frequencies at higher amplitude than we do low or high frequencies. It is also sex dependent as woman perceive upper mid range energy as higher in level for the same measured amplitude than do men.

It’s Dynamically Correct!
With this foundation in place we will now look at the natural characteristics of un-amplified acoustic music, what humans have been listening to since our creation until the early 1940's, only about 70 years ago, when sound re-enforcement was introduced to allow crooners to compete with a big band.

Shown below is both the amplitude and frequency content of a totally acoustic ensemble, a symphony orchestra, in a sample of music from pianissimo to fortissimo. This is the last two minutes and four seconds of the Paul Paray 1^st Symphony, 4^th Movement - Allegro scherzando-GROTTO GP-0009. This recording was done directed to stereo digital in a single session at Assumption Grotto Catholic Church in Detroit using an Aachen Head with minor supplemental microphones. No processing of any type was employed.

The plot below shows the wide dynamic range of natural acoustic music.

Note the high peak to trough ratio exhibited in un-amplified and un-processed naturally produced audio.

The data below shows the nearly 40dB of natural dynamic range of this orchestra. It was derived from the actual audio data file, excluding the fade to infinity which would give false dynamic range information as the room noise is fading also. The data below gives more details:

• Left Right
• Min Sample Value: -30417 -30248
• Max Sample Value: 26962 28405
• Peak Amplitude: -.65 dB -.69 dB
• Possibly Clipped: 0 0
• DC Offset: -.001 -.001
• Minimum RMS Power: -46.28 dB -46.46 dB
• Maximum RMS Power: -11.52 dB -10.55 dB
• Average RMS Power: -23 dB -22.22 dB
• Total RMS Power: -21.57 dB -20.66 dB
• Actual Bit Depth: 16 Bits 16 Bits
• Using RMS Window of 50 ms

With about .7 of a dB before the clip point this natural program has a scientifically verifiable 40 decibels of dynamic range.

With a histogram of the data we discover even more:

The natural majority of the program is -20 dB below the clipping and less than 5% is at or near 0 level.

What is just as important is the natural frequency distribution in this program material, a full symphony orchestra:

The natural roll off of high frequency energy is a character of live acoustic music as is the lack of extreme high frequencies because of the attenuation in the air and is the result of humidity, absorption and other natural acoustic conditions. (The Aachen Head used in this recording has virtually flat frequency response in the frontal free field to well beyond 20, 000 Hz.) The closer any instruments are miked, the more high frequency content will be presented. This is one reason why “tight miking” orchestral musical instruments sound un-natural to us.

Now lets take a look at traditional acoustic jazz music. In this case the program material is a live recording of Clarinetist Dave Bennett’s “Remembering Benny” WJS Discs #003 Track 2 - Moonglow. Hugh Leal recorded Dave “on the road” with two SONY Mini-disc recorders, one sitting on stage with a stereo X/Y pair in front of the ensemble, the other receiving a line output from 2 line feeds from highlight mikes on stage, usually the piano and bass. I then used a common transient peak to find a lock point and “locked up” the 4 channels of audio from 2 mini-discs and mixed the program to stereo. This particular track was recorded live at the Illinois Central College of Performing Arts Center. Again no processing was employed on this project other than what is present in the mini-disc data reduction system.

The amplitude v. time plot below is from the first two minutes and 31 seconds of an over 4 minute track. This project was chosen as it is traditional jazz and was not multi-mike and represents natural dynamic range characteristics and natural spectral balance.

Note, as before, the large peak to trough ratio and the natural, un-clipped peaks. Once again the natural wide dynamic range of music is evident. This program material again exhibits nearly 15 dB of natural dynamic range even with the mini-disc data reduction dynamics processing.

Again, traditional jazz, naturally produced music has naturally wide dynamic range.

What about Multi-Track Recording?

Now lets do further investigation into the characteristics of a non final dynamically processed multi-track session. In this example electrically amplified contemporary jazz.

One was recently at hand in the form of a live recording which I was privileged to do with Dennis Coffey and friends at Baker’s Keyboard Lounge in May of 2006. This was a FUNK-JAZZ project with non-processed contemporary instrumentation. We were in the final mix-down stage of this project. The session was recorded to multi-track digital with Electric Bass, Electronic Keyboard and Guitar with Dennis’s famous electronic and wah-wah effects going direct with custom DI boxes of my design and the drums miked with 3 AKG D190 microphones and an Electro-Voice DO54 on the Kick Drum. Some dynamics processing was employed on the Kick Drum and Bass for effect. No other dynamics processing was employed.

The results of a two minute and 40 second sample of a near 11 minute performance of “Little Sunflower” are presented below:

Again, even with some minor effects processing the program exhibits wide dynamic range.

A full review of the data shows:

• Left Right
• Min Sample Value: -32397 -31348
• Max Sample Value: 28321 27771
• Peak Amplitude: -.1 dB -.38 dB
• Possibly Clipped: 0 0
• DC Offset: -.017 -.017
• Minimum RMS Power:-40.92 dB -41.5 dB
• Maximum RMS Power:-9.32 dB -9.45 dB
• Average RMS Power:-20.52 dB -20.4 dB
• Total RMS Power: -19.37 dB -19.25 dB
• Actual Bit Depth: 16 Bits 16 Bits
• Using RMS Window of 50 ms

A histogram of the data shows:

Again observe that naturally some 90% of the program material is -19 dB below full 16 bits and that the natural program material has over 20 dB of dynamic range. Certainly not that of classical music but naturally wide in dynamic range.

Further spectral examination show a natural fall off of high frequency energy above about 1800 Hz as may also be observed in the classical and traditional jazz productions:

Conclusion
The conclusion of this rather verbose and scientific examination points us to the undeniable common components in natural reproduction of music:

1. A wide dynamic range naturally from 20 to 40 dB in program material without solo instruments and even wider when solo instruments are presented in their natural dynamic perspective
2. A gradual fall off of high frequency energy
3. The majority of program material -15 to -20 dB below the peak program level
4. A wide range frequency range

This is what the ear desires to be presented with because for centuries this has been the “diet” of human perception of music. Conversely, when these components are not present music will not sound “natural” and listener fatigue is natural result.

This is the essential genesis of the term High Fidelity and why since the birth of the recorded reproduction of music it has been the goal of engineers to preserve it.

Digital recording and the Compact Disc is the first delivery media which has allowed this dynamic range and full frequency response to be delivered to the consumer. Analog media cannot accurately do this.

As before your comments are solicited and can be noted at the “Contact Us” link on our website at : http://www.audiographicservices.com

In our next session we will review these components in early recordings through the recordings of today and look at how “HIGH FIDELITY” has previously pointed the way to sales growth and try to discover the distorted mentality that has led us to the LOW FIDELITY of the audio production of today.

-----------------------------

When Clay first approached me to write this column, I was both surprised and honored. Having worked with Clay thirty some years ago our approach to recording was similar. I told him that I would be outspoken and at times controversial. He said that he understood that as “OLD SCHOOLED” our perspective was from the “do it right, no matter what it takes” rule.

As with all the arts and sciences God’s laws of physics do not change and what we are doing in audio recording is simply implementing those laws for the sake of art. It’s really simple if you understand the laws of Physics and now with digital audio, the laws of Mathematics. We use these laws as our tools. The problem is, from my observation, that the so-called engineers of today neither know the laws of Physics nor Math, or it at least appears that they are Hell bent on ignoring them.

Claude Elwood Shannon a Gaylord, Michigan native and Bell Telephone Labs researcher gave us the concept of ?INFORMATION DIGITIZATION.? At age 32 his ground breaking 1948 paper ?A Mathematical Theory of Communication? demonstrated the unity of all information media be it graphics, text, telephone signals, video images or audio information. He showed that all of this ?INFORMATION? could be coded into the universal form of binary digits then transmitted or stored without error. This was a breathtaking concept that has led to the digital media age of today. This was an amazing conceptual leap. This paper is basic reading for all involved in the music business as it establishes the basis for the media of today.

In 1985 the introduction of the Philips/Sony compact disc system turned this concept into a commercial reality for the music industry. For the first time audio could be directly delivered to the music consumer with fidelity equal to the original master. The capability of a bit for bit copy of the original master was here.

But more than that, the CD medium itself offered technical specification better than the analog medium that then most masters were recorded with. Frequency response DC to 20KhZ, WOW and FLUTTER-none, signal to noise ratio better than 96 dB and Dynamic Range of 96 dB as well.

No more clicks and pops of a vinyl disc; no more the variability of a laquer master and the electroforming process; no more Pressing variables; no more pickup variables; no more speed variables; no more disc warpage; no more cassette tape hiss; no more bias variability; no more tape saturation; no more cassette shell damage; No more consumer product variability at all to plague us. No more problems!!!

On first listening the system seemed perfect, but it quickly became apparent that this was not the case. The early analog to digital converters, in many cases, could not approach the 16 bit 96 dB capability of the PCM system. Many of these early converters introduced audio artifacts and distortion in conversion and the digital to analog converters in some of the poorer quality consumer CD players left much to be desired as well thanks to manufacturing “cost control.” Early digital audio got a pretty bad rap and in many cases it was deserved. But technology improved and by the early 1990's over sampling conversion technology had solved almost all of the associated problems and digital audio could be considered nearly “pristine.” Many people argue otherwise but the audio measurements speak for themselves.

It further must be admitted that analog recording techniques, what we have been doing for nearly 60 years, are not suitable for digital recording. Because of the transparent nature of the digital media, any frequency response or phase discrepancy, becomes sonically apparent. This has become a rude awaking for most recordists. Simply put, you cannot get away with what you did before. Multiple microphones and the comb filtering exhibited by unsophisticated use of this technique becomes a serious liability. Sloppy analog recording techniques, when used digitally, become apparent. But what happened in most cases is that digital has become the whipping boy.

Analog transfers to the digital domain sound “smoother” and “warmer” because the unpleasant sonic characteristics are masked in the analog original recording. Multi-track analog mixes to digital are described as “harsh” because of the overuse of equalization and processing, so common in analog recording, and in many cases were needed as compensation for the medium itself became apparent in digital mixes. This problem continues to this day. Digital is simply unforgiving. Garbage in. . . . Garbage out!

But with all of the technology improvement one might ask why has the quality of today’s musical products decreased rather than increased? What has happened to the expectation of the music buyer?

One word describes the problem. IGNORANCE!

All one needs to do is talk to the record executives of today at any of the so-called major labels. For me it was a startling experience, an Epiphany. The majority of the so-called producers had some knowledge of what constitutes a quality audio product, but the executives were totally void of knowledge. They had no experience other than the bottom line. Thirty years ago this would have been grounds for dismissal. Today, it’s considered an asset. The May 2005 issue of MIX MAGAZINE highlighted many of these problems. Even more enlightening is to compare the fidelity of audio product thirty or so yeas ago to that of today. This is exactly what happened with Clay and me a few weeks ago.

By a series of circumstances we had the opportunity to audition a 15 IPS Dolby A master against the same product in four consumer release formats, each of progressive age:

1. Black Vinyl (original release)
2. A 1987 Re-release CD of the same product.
3. A 2001 Audiophile release of the product in ?Super Audio CD?
4. A current overseas CD release of the product. (UK Release)
5. A current domestic CD release of the product.

Clay (who was also a former quality control director at Motown) and I both concur with the following subjective review when comparing these products on a rather elaborate wideband monitoring system.

The original master was of excellent quality and was done on 3M 202 1/4 inch stock so it did not suffer from ?stiction? problems. It had excellent low end and smooth top end. The apparent dynamic range was excellent. A quick review in Spectra-Plus showed nearly 40 dB of program dynamic range.

1. The black vinyl sounded amazingly good. It was a very fine pressing mastered by Bob Dennis and had nearly 35 dB of dynamic range. It was well balanced and sounded much like the master other than some extreme low end bumps (probably head bumps inherently in the analog tape process) and some high frequency loss. Some compression could be noticed when compared to the master. It had a few minor ticks and pops. All in all other than the ticks and pops it’s listenability some 30 years later was a surprise to us both.
   
2. The 1990 re-release CD was almost indistinguishable from the master. This was also a surprise to both of us. The transfer obviously was done with high quality A-D conversion as it exhibited no top end harshness. Subjectively it was a wonderful musical experience with wide dynamic range and balance identical to the master.
   
3. The Audiophile release was similar to the master but was equalized to be somewhat brighter. Dynamics were excellent and the balance was good. Dynamic range sounded identical to the original. It was a bit “edgier” at the top than the original.
   
4. The UK CD release was somewhat compressed and processed. The top end was very harsh and the bottom end was restricted. It sound as if some high pass filtering was implemented in the transfer. There was less than 10 dB of dynamic range in program material. It was very difficult to listen to and made us question how accurate the monitors were when mastering this product. But it was LOUD. Not Good, but LOUD!
   
5. The recent US release of this product has less than 6 dB of dynamic range. It was over equalized and exhibited extremely harsh top end and a low end bump which made it even more unlistenable on a high quality system. But it was LOUD. Not high quality; not at all a representation of the master; not musical. IT WAS LOUD!
   

Which brings us to the question, what has happened to quality? What has happened to fidelity? Some might say that since the multi-track product has no basis in reality there is no standard for ?fidelity.? I would counter that the final product must reflect the integrity of the original master tape since this is what the producer and engineer had as the goal and the capability of presenting this to the consumer is now here. Further, the music consumer has the right to expect this for the hard earned money he is paying. It is a matter of ethics!

The problem of ignorance of audio quality cannot be denied. It is reflected in all we hear. But what has brought us to this dilemma?

Howie Klein, former President of Warner/Sire and Reprise Records and now teaching occasionally at McGill University, in a recent audio magazine article sums it up in one word, corporatization. Howie says “At the very minimum, all the power to make decisions has gotten into the hands of people who are just corporate hacks.” Both Clay, Howie and I concur that the corporate business model cannot work with art. This has proven true time and time again. The people running both the audio and music industry today are simply “Corporate Hacks” and understand nothing of the art or science of the industry other than the bottom line. The failure of the industry pragmatically bears witness to this fact.

This is the problem and is reflected in the musical and technical products produced. The level game and concept that “LOUDER IS BETTER” is just one ill fated technical aberration of the total problem that affects what we are forced to sonically endure.

In the next column I will document the artifacts of this problem and present thought and suggestions from myself and other seasoned audio professionals regarding potential solutions to this serious problem which many audio professionals believe may have serious effects on listenability and hence sales of audio products in there present formats.

(Comments and “hate mail” may be addressed to the contact address at my web site www.audiographicservices.com.)