THE QUEST FOR AUDIO NIRVANA (1/3)
WHAT PRICE GLORY
At the urging of many friends and associates over many, many years, I have finally decided to write this series of 3 articles covering a wide range of audio subjects. I intend to bare some truths and debunk many myths that have been around for too long. For sure, these writings will undoubtedly unleash a lot of fervor by smashing some belief systems that have been awry for too long as well as hopefully shedding some light in areas that have been either obscured or just plain ignored. Ever since the beginning of the stereo era, the audio industry has been on a zig-zag course instead of heading for the truth. To be sure, there have been, and probably always will be, far too many charlatans, dilettantes, con artists, and incompetents in our industry which should be based much more on scientific truth and a whole lot less mysticism. Starting here I have prepared 3 articles of which the first is on electronics and the 2nd and 3rd are on recording and playback respectfully. So without further adieu, let the slings and arrows fly.
SEARCHING FOR THE HOLY AUDIO GRAIL
Audio is the only industry known to man that requires little or no scholarship to be a practitioner. At times it seems that any totally incompetent quack working in their garage can come up with a product and get it to market without any qualifications whatsoever. The only requirement seems to be a belief system. Unfortunately, belief systems don’t cut it because that’s all they are—just belief systems. They don’t have to be right and most of the time they aren’t. Of late, we have a very polarized academy in the world of audio. On one side are the subjectivists who tend to cast a suspicious eye on scientific doctrine. On the other side are the objectivists who tend to believe that science is the only answer and that the so called “golden eared” types are living in la-la land. Both of these groups are, to be sure, lacking in comprehension of a great deal of information regarding the bigger picture of audio. I, of course, have always taken the middle road inasmuch as I use scientific procedure to obtain results BUT am always concerned about how the final product performs in its assigned task. In other words, I know that there is correlation between the two even if we have not yet defined all of the parameters. So let’s start with some truths.
There is no such thing as “good” distortion. Period. End of argument. All distortion is bad and that’s final. Of late, some have had the misguided belief that if we go back to the horse and buggy days of the mid 1920’s with 8 watt single ended tube amps with no feedback and higher levels of low order distortion, then all of our sonic prayers will be answered. This is just plain nonsense. Are people to believe that all of the last 75 years of scientific progress is just a bunch of hogwash? I think not. I think the descriptive word “better” should be replaced with the word “different”, when discussing the sonic attributes of these products. Does anyone know a single person that was alive in 1925 that can state that back then these things were high fidelity? Can you imagine the auxiliary equipment that was used then such as electromagnetic speakers and solid iron needles, etc. I could go on, but I won’t.
Anything and everything that is wrong with an electronic or acoustic device can be classified as distortion. In no particular order these are amplitude distortions, frequency distortions, time distortions, phase distortions, power supply anomalies, flux radiation distortions, etc. One of the things that has yet never been done is to establish any kind of weighting to all of these various kinds of distortions. In addition, it is known that the range of human hearing is in excess of 130dB’s. Therefore, to state for example, that an amplifier has distortion low enough that it is below the threshold of hearing is just plain wrong. The human ear is an incredibly sensitive instrument and actually acts like a spectrum analyzer with much more dynamic range. Signals, that are buried in noise for example, can be heard by the ear although we cannot consciously identify them as such. There is one other kind of distortion that may be the most insidious of all and that is convolution distortion. This occurs when you have a string of devices, starting with the microphone, followed by whatever multitude of stages in the recording console, followed by the stylus and or laser, followed by more processors and/or an RIAA stage, followed by a line stage, followed by the power amplifier, and finally ending with the loudspeaker. Every one of these devices adds its own intermodulation products, and then in series gets convoluted all on top of each other at every step of the way down the chain. By the end what you have is an absolute mess. UGH! After all is said and done, this may be the unfortunate final frontier in that if every device in our playback system were perfect, we would still have all of the convolution products that preceded it.
At this point some of you may be wondering about basic noise floors and how is it possible that masking is not taking place. The reason is that noise, as long as it is low enough not to be consciously intruding, gets basically rejected by the ear. This is so because as stated earlier, the ear-brain link acts like a real time spectrum analyzer. Since noise is fundamentally random and has no correlation to the desired signals, it is ignored by the ear. However, all of the grunge lying below the noise can have a tremendous effect on what we perceive acoustically.
Before we go on, let’s talk about some numbers so that we have a basis to work with concerning distortion. Virtually everyone relates to distortion numbers incorrectly. For example, if an amplifier had say 1% THD as read on a distortion meter, is it really one percent? Actually, it isn’t. The distortion amounts that are read on a distortion analyzer are in Volts and the ear doesn’t hear Volts, it hears power. So, if we convert the 1% to power we get a totally different perspective. Let’s say a nominal 200 watt per channel amp is producing 1% THD. As read on the analyzer this would be –40dB’s. But –40dB’s in power is 1/10,000th or 20 milliwatts. On the other hand, since the ear’s hearing curve is logarithmic, 40dB’s represents a RATIO of only 16 to 1. In other words, for every doubling (or halving) of the loudness it takes 10 TIMES (or 1/10th) the power. 20dB’s equals 4 times (or 1/4th) and 30dB’s equals 8 times (or 1/8th) and 40dB’s equals 16 times (or 1/16th). Therefore, if you divide 100 by 16 you get, you guessed it, 6.67%. Now that’s a long ways from 1%. You can scale this up or down in either direction and the RATIO of the results will be the same. So when you apply these true numbers to the human hearing curve it is easy to see why anomalies and distortion artifacts that are buried in the noise can be detected by the ear. There is no black magic here. The so-called golden eared types actually can hear this stuff. The descriptions and verbiage associated with this situation may be a whole different can of worms. I have prepared a chart that shows the true relativity of all these factors.
|THD Reading||Db’s||DistortionPower||DistortionWatts||HearingRatio||True Distortion|
|< 10%||< -20||< 1%||< 2||< 4:1||< 25%|
|< 1%||< -40||< .01%||< 20mW||< 16:1||< 6.67%|
|< .1%||< -60||< .0001%||< 200uW||< 64:1||< 1.56%|
|< .01%||< -80||< .000001%||< 2uW||< 256:1||< .39%|
|< .001%||< -100||< .00000001%||< 20nW||< 1024:1||< .098%|
|< .0001%||< -120||< 10(-12)||< 200pW||< 4096:1||< .024%|
You will immediately notice the staggering difference between what the distortion meter is reading verses the true distortion numbers. The first line shows that the true distortion is actually 2½ times greater. The second line shows that the distortion is actually 6.67 times greater. The third line shows that the true distortion is really 15.6 times greater. The fourth line shows that the true distortion is actually 390 times greater. And the fifth line shows that the true distortion is actually almost 1000 TIMES greater, etc. It should be quite obvious to everyone now as to the fact that humans CAN hear artifacts down in the mud. Quite simply put, it is the RATIO of loudness factors that determine what we perceive acoustically and not the THD numbers off of the distortion analyzer. As an aside to this, I have said for years that phono cartridge separation numbers are a mixed blessing. In other words, 30dB’s is just not enough as this only amounts to a ratio of 8:1 loudness. That’s pretty awful when compared to today’s CD’s for example. However, it is a mixed blessing in that this lower separation actually helps to “fill in” the middle somewhat. It is however, quite amazing to listen to the difference between two otherwise identical cartridges wherein one has 30dB’s of separation and the other has 40dB’s. The difference is absolutely obvious. Unfortunately it is no mean feat to manufacture cartridges with such a consistent high level of separation but this should be the number one main concern all other factors equal.
At this juncture I would like to digress for a moment into a little musical escapade. There are those out there that believe that harmonic distortion is meaningless. These persons are, of course, idiots. The following shall prove my point. Firstly, some musical tidbits are in order. In nature’s natural world, the most complex musical structure, no matter how dissonant or polytonal, can only exist within a span of 2 octaves that is, say from C to the second octave C. In other words, 15 natural notes. It doesn’t matter how complicated the harmonic structure is. It still can only occur within 2 octaves. Any notes that are outside, either above or below these two octaves, are merely repeats. For example, the third D above our starting C is merely the II note, etc. Any notes below the C are called suspensions. Any notes above are called octaves. Secondly, human hearing is leading tone sensitive. This is an extremely important human recognition ability. It allows us to always hear the melody on top no matter how complex the structure below the melody is. You may ask, why is this so important. It is important because in nature, the musical scale is tempered. I’m sure you have heard the expression “the tempered scale” regarding the tuning of instruments. What’s happening here is that the frequencies of the notes are not mathematically perfect hence “tempered”. As the musical notes go up in frequency, the pitch is ever so minutely bending sharp. In musical language we call this “cents” bending. Let me give you a couple of examples. In the early days of manufacturing electronic organs, they had 12 master oscillators which were divided down by flip-flops to produce the succeeding lower octaves. Unfortunately, these divisions were mathematically perfect and not tempered. The effect of this is that you could not have one of these electronic organs work with other musical instruments that were tuned according to the tempered scale. It was an absolute definite clash. In my early career, I was working in a club once when a female vocalist came in and wanted to sing with my duo (organ and drums). She had never sang with an organ before and for a whole set she sang everything SHARP. I heard this and she heard this and the audience heard this. Not pleasant. Later in my musical career when I switched from organ to piano I had to re-acclimate my voice because I was singing everything FLAT. It was an experience to say the least. It took me a couple of months to get my ear-brain-voice linkage working properly. Today of course, virtually all electronic organs are made with a DSP processor which produces EVERY note exactly according to the tempered scale so that problem doesn’t exist anymore. What am I getting at here? That third C in our previous example is the XV note or the fourth harmonic. It is the highest harmonic that if it were distortion would still be tolerated by the ear because it is enharmonic within the two octave natural order. The third harmonic is the 2ndoctave G which is the XII note or the perfect 5th raised an octave. Now lets talk about the next octave, the third. This is what I call the killer octave which shall become painfully obvious as we continue. The fourth C is the 8thharmonic. Now then, the third octave contains all the harmonics between the 4th and the 8th!! That is the 4th, 5th, 6th, 7th, and 8th. All clustered together within one octave. WOW. How many of you readers have ever thought about that or realized that. I’ll bet that very few engineers ever thought about this situation. Distortion products created by say our amplifier, are NOT tempered. Therefore, all of those harmonic distortion products, that is the 5th through the 8th, will definitely clash with the natural harmonics created by the real music. And all of this clashing happens within ONE octave namely, the third. There is absolutely no question that the ear can hear this discumbobulation. I believe that this is one of the reasons for the so-called “transistor” sound. I’m sure that the editor doesn’t want an engineering course to be presented here. It should therefore, be obvious to everyone that this is a major problem that needs to be addressed.
I have always stated throughout my career that I am more concerned with the character of the distortion performance rather than just the numbers. It is incredibly difficult if not virtually impossible to make the distortion products behave in a declining manner as the frequency goes up. This is obvious because we are running out of bandwidth and consequently losing feedback at the same time. Coupled with the fact that under most circumstances of solid state engineering over the last 40 years, we have not, for the most part, yet learned how to treat semiconductor devices in the natural manner of which they are comfortable – operating wise. Another equally difficult task is to try to design circuitry that operates in the current mode as opposed to the voltage mode. Initially, this can be done rudimentarily if one also accepts throwing bandwidth and noise right out the window. Fundamentally, tubes do not exhibit the same kind of mechanisms that solid state devices have and therefore do not have the same kinds of problems.
In the last few years, some extremely powerful simulation programs have become available that can now allow us to delve deep into the problems of circuit design to the degree that was simply just not possible with normal test equipment. I’m talking about being able to go down into the mud so to speak nominally, to –180dBs of range. The very best analytical test equipment on earth can’t, at present, go much below about –120dBs or so. There are some special painstaking methods that can be used but they take forever to get results wherein the same results, and more accurate results, can be achieved with simulation in a matter of minutes. One of the most formidable tools is from “Electronics Workbench” and is called “Multisim”. Using these tools has allowed us to devise completely new circuit philosophies that have distortion products that are 1000 times lower than what we have previously achieved as well as having bandwidth much greater than we have had in the past. Of course it will be a few years before most of these concepts become a reality and show up in products, but the writing is on the wall and we can definitely see the future.
Finally, as far as correlation is concerned, we don’t arrive at answers until we get there. Therefore, all these new revelations that we uncover must be carefully evaluated lest we end up chasing the proverbial ghosts. Next are a few more audio tidbits for you to ponder.
So you think that your 200 watt amplifier is really 200 watts, eh! Guess what folks: it isn’t. Actually, it’s far from it. Let me explain. It would only be 200 watts when producing a SINGLE NOTE. Now if you add a second note, it isn’t 200 watts anymore. Some examples are in order. Let’s say a note is being played that is 2 octaves above concert A or about 1760 Hz. Now, 200 watts is represented by 40 Volts RMS across 8 ohms. Let’s assume that the voltage for that note is required to be HALF of the available output or 20 Volts. This 20 Volts is 50 watts. Now let’s have another note at say low A at 55 Hz. also at 20 Volts or 50 watts. Guess what! The voltages are directly additive and in fact, the 1760 Hz is in effect modulated by the 55 Hz. Now the average sum of these two signals equals 40 Volts which should be 200 watts but, it isn’t. Each frequency is only amounting to 50 watts for a total of 100 watts. Already, our 200 watt amplifier has been reduced to half its size with only TWO NOTES. Now lets add in a 3rd note sufficiently away from the other two notes. We must now divide the available output voltage by a factor of 3 which gives us 13.33 Volts per signal. This results in approximately 22 watts per note. That sure is a long way from 200 watts. Now can you imagine what happens with very complex musical structures? The true available UNDISTORTED output power is but a mere fraction of what the amplifier is really rated at. If you don’t believe me just hook up an oscilliscope to the output of your amp and watch the clipping. Those of you with amps that have true peak indicators can surely know what I’m talking about.
Lastly, let’s examine a situation that will become increasingly important as we develop new circuit concepts and ideas. For most of the last many decades in audio, the standard topology for power amplifiers has been the half bridge. In other words, we have a hot output and a ground. As we delve into the depths of trying to improve performance, this topology has a built in automatic brick wall that is virtually impossible to get through. I’m talking about the ground. In this type of circuit all of the signal power on alternate half cycles must pass through ground and up through the transformer windings and through the rectifiers and finally completing the loop through the filter capacitors. It is absolutely impossible to make and guarantee the integrity of the center tap on the transformer to better than –120dB’s. What this means is that in this concept the bottom line as far as power supply anomaly distortion will hit this brick wall. The ONLY way that we can get around this problem is by using a full balanced bridge design. The reason is quite simple. In a balanced bridge circuit all of the signals are circulating and NO POWER goes through ground. Ground is merely and only a reference. All of the anomalies previously mentioned actually cancel out provided that the design is done properly. In other words, in our search to improve performance beyond the –120dB range, this is the first major hurdle that must be overcome. There have not been a lot of balanced bridge amps manufactured but it is the only way to go in order to push the envelope forward.
One other note of caution is in order. I am ONLY referring to power amps here because we are dealing with just that –power. I am NOT advocating balanced circuits for any other audio components and as a matter of fact, trying to utilize balanced lines in home audio is a stupid idea to say the least. There is absolutely NO ADVANTAGE to using balanced lines and as a matter of fact, there are some very serious drawbacks. I will explain this at a later time however, Walt Jung wrote a very good and definitive couple of articles a few years ago in one of the major electronic trade magazines of which the exact one escapes me at the moment. He really nailed the rationale.
In closing, I probably have opened some Pandora’s boxes however, the main goal here is to try to get the public and the engineering community on a course of more understanding so that in the end result we can have a better and more believable acoustic situation. If any of you want to rail at me or even discuss some of these issues, you can contact me through my e-mail at firstname.lastname@example.org