The Speaker, Part 1:
How we got from LoFi to HiFi
I
May 10,2020
In a previous entry, I said live sound starts with the microphone. At the other end of the signal chain is the speaker. These two acoustic-mechanical to electronic transducers at the opposite ends of the signal chain are the weak links.
When I was young, there was no real high fidelity sound. Most of the music I heard came out of AM radios with 4” speakers. Because of the limitations on how AM radio signals are broadcast, there was no incentive to have HiFi quality speakers.
The other place I heard music was from 45 rpm records. Compared to the old 78’s, these were a big improvement in sound quality, but again, did not merit HiFi speakers. Gradually, in the 50’s and into the 60’s, two new sources of music became available: the 33 rpm record and FM radio. These sources of music did justify having HiFi stereo speakers. By today’s standards, these HiFi speakers were … well certainly not highly resolving, and, in fact, more medium fidelity than HiFi. Most of the improvement was to add a bass speaker and to pay a little attention to the midrange design. Later, tweeters were added. Still my old Webcor was a big improvement in sound quality.
In the meantime, live sound was starved for power. Today having a 1000 W amplifier is no big deal. A vacuum tube amplifier capable of a few hundred watt output was a big deal. Why? Short answer: Because they produced lots and lots of heat and required high voltages.
By the early 60’s transistor amps were on the scene, but there were issues with early transistors and high power amps.
You needed to fill the room with sound and only had limited amplifier power to do so. The solution was high efficiency speakers. In reality, even a very high efficiency speaker is really extremely low efficiency in terms of electrical power in vs acoustic power out. With low power amps, speakers had to have a premium on efficiency and not so much on HiFi. If you needed more sound power for a given amplifier power, the solution was to use horns. Horns are significantly more efficient than cone speakers, but there are two problems. First, for a horn to reproduce bass it has to be big, very big, and so the hybrid speaker was born … horn for the mid and high frequencies and a cone speaker for the bass frequencies. The other issue with a horn is they can sound tinny. The exact geometry of the throat of the horn is critical. There were computers in the 1950’s, and these computers could work out the complex equations needed to design a HiFi horn, but there were only a few electronic computers, and they were very very expensive. At the time, the term computer usually meant a person using a mechanical calculator or a slide rule to solve long complex equations. As this took a great deal of time, this form of computer was also very expensive. So horns were mostly designed by trial and error and some nifty approximations. However, once again, the results are more what we would call medium fidelity.
By the late 60’s, electronics and other factors came together, and true HiFi speakers started to come on the market for both audiofile home use and live sound.
Still there were limits. This next statement is going to get me in real trouble: 33 rpm vinyl records were the primary source of high quality sound. Tape recorders, still mostly reel to reel, but also cassette tape recorders, were available. Actually reel to reel tape recorders had been around since just after WW II, but until the 60’s were just too expensive for most users. Where I am going to get in trouble is the following statement: while we may really love the sound of vinyl, by any objective standards they are very limited. Tape recorders also had their limits. There are a number of limits such as flutter and wow, and signal to noise, but perhaps the most significant limit is dynamic range. Why make a speaker with a dynamic range far in excess of a vinyl record or a tape recorder? The folks at places like Dolby sound made circuits to try to compensate for these deficiencies.
Another issue is the crossover between the low frequency driver and the mid range driver, and again from the mid range driver to the tweeter. These crossovers should be clean … no bump in either the dynamic (volume) output or in the phase of the output. You need a crossover filter that passes the mid range to the mid range driver and the bass to the bass driver. The crossover filter circuit also needs to handle power and not create heat. This is not a simple task, and such circuits were, and often still are, a compromise.
Returning to live sound … the “source” of the signal is the microphone. If you were running a studio with a big budget, there were excellent mics to be had. But if you were running a local show, on a limited budget, you had to make a compromise, particularly if the event was to be outdoors. Mics came in two types, those that required a local preamp (ribbon and condenser mics) and those that did not (dynamic and piezo-electric mics) . There was also the issue of robustness … condenser, and even more so, ribbon mics tended to be fragile.
The other problem had to do with tube preamps. Specifically they were big. Yes, people today do make mics with a built-in tube preamp, usually based on a 12AX7, but back in 1955 it was not just the tube that was big, so were all the other electronic components. Few mics had a builtin tube preamp. There were also no electret mics … you had to apply a voltage to the diaphragm of the condenser mic. For the signal from the condenser mic to be large enough to drive a reasonable length cable to the large preamp, you need high voltage, often hundreds of volts. Now you understand my comment about outside … it can rain outside … water and high voltage don’t mix. All this changed with first bipolar transistors, followed by the FET transistors, and the invention of a practical electret mic.
The mics that did not need local preamps were dynamic and piezo. The piezo (acoustic) mic was inexpensive, rugged, and decidedly low fidelity. For live sound, with a limited budget, you had dynamic mics and dynamic mics of the time were typically not so good in the high frequency range.
The game changer was the CD. I first heard CD’s in the late 70’s. OK, I knew people who swore they could hear the digits in a CD, but let's not go there. The CD had a dynamic range of just over 65,000 to 1 (96 dB). Now it did make sense for speakers to be more HiFi, and to have greater dynamic range.
Starting in the early 70’s, significant developments in electronics allowed for many advancements. The first was an affordable op amp (operational amplifier). The Op Amp had been developed by George Philbrick way back in the 1950’s. At first, they were vacuum tube based, then discrete transistor based. These were good op amps but they were expensive and big. Then in 1971 Bob Widler introduced the uA709. It was not the best op amp on the market, but it was an integrated circuit … small and low cost. The uA709 would lead to the development of much better op amps. What the op amp allowed was for analog signal processing such as active filters. The other development was a bit slower to become a viable audio design tool … the digital signal processor. uProcessors (the 4004) also came on the market in 1971. However it would be years before they were fast enough, and cheap enough to be used in audio. Still the CD was based on digital signal processing (DSP). Perhaps the first use of uProcessors in audio reproduction was for digital reverb units.
Developments in DSP and uProcessors allowed designers to do the complex calculations required for speaker design (both drivers and their enclosures), and to model how a speaker would perform in a real room.
By the late 70’s all the pieces were in place for true HiFi outside of some acoustics lab, and speakers had to rise to the challenge.
Speaker development split into many branches. The two main branches are live sound and home sound.
The home sound branch further branched into three “twigs”: basic audio, artificial room audio, and highly resolving audio.
Basic audio is (or certainly can be) HiFi. Its job is to present the listener with HiFi quality sound, but not the ultimate in sound reproduction. Paul McGowan (of PS Audio) calls this background music, and perhaps it is.
Virtual room sound (aka Surround Sound) takes this a step further. The sound you hear may be emanating from a speaker, but what you hear is the product of the speakers and all the reverberations off the walls, floor, ceiling and … The room you listen to music in may be acoustically great, but I very much doubt it really is. Surround Sound uses a number of speakers, and a lot of signal processing, to create a virtual room that does have warm lush acoustics. A form of Surround Sound 4 has actually been around for a long time, since about 1970. It was called quadraphonic, and 33 rpm records recorded in four channel sound were released. The idea did not catch on, at least in the 70’s. There was also something called ambient recovery which maybe should be called Surround Sound 2 ½. Simple passive filters were used to create two rear channels from a stereo recording. Again this did not catch on, although I found it pleasant. Starting about 1982 and over the next decade, Dolby and others developed the modern Surround Sound.
And then there are the true audiophiles and their highly resolving audio . They pursue excellence in sound (and sometimes go to extremes to obtain this excellence). They work with every aspect of a sound system from the power cord, to the speakers, and the room itself. There is a sweet spot in their listing rooms in which the stereo sound is imaged. The goal is, if you are seated in the sweet spot, if you closed your eyes you could not tell if the sound was being reproduced or if there was a live performance in the room with you. In order to accomplish this goal, their speakers are often di-poles … speakers that project sound both directly to the listener and also out the backside to the rear wall of the room. As we shall see, this is not something done in live sound.
If audiophiles have a spokesman, it is Paul McGowan. He has an ability to take complex acoustic (or electronic) issues and distill them down so the listener, or musician, or whoever, can understand the principles, and yet his answers are also good (if simplified) engineering answers. I am about live portable sound, and I am not a true audiophile, yet live sound engineers can learn a great deal from audiophiles and their obsessive pursuit of excellence.
The live sound branch has at least three sub-branches: permanently installed live sound systems, portable acoustic live sound systems, and extremely enhanced live sound systems.
Live sound is different from home sound in that the space you need to fill with sound is much larger, and the number of people listening, and thus background noise, is also much larger. The audiophile has a sweet spot in his listening room that consists of a chair or two. A live sound event needs a sweet spot that has 200 chairs, or maybe 2000 chairs, and sometimes even more.
Because of the much larger space that needs to be filled with HiFi sound, live sound speakers need to project sound in a limited spatial pattern and need to be more efficient in terms of electrical power into acoustic power. Thus, for live sound, you use compression horns, and spatial arrays, and phased arrays to project the sound into the hall and not just everywhere.
A permanently installed sound system can have a huge advantage. The acoustic properties of the room can be analysed, modified, and speakers placed in near ideal locations.
For portable sound, you usually place the speakers by seat of the pants heuristics, and the acoustic adjustments you can make to the room is limited to adding draping on the stage and walls of the hall. Your goal is to create as big an acoustic sweet spot as you can. In any hall there are areas where the sound is less than HiFi quality. These are normally near the corners of the room, especially the back corners.
There is also a severe limit imposed on the live sound engineer … if you have a speaker and a microphone you also have the potential for feedback. This is another reason to focus the sound … have directional mics and directional speakers such that the mic cannot hear the speaker.
Another severe limitation is that the performers want monitors, and this requires the sound engineer to create two sound stages, one for the audience and one for the performers. When you consider feedback, the performer's soundstage is the most dangerous.
If you want to project light across the room you can use a simple lens. If you look at the edges of the light beam, there is a rainbow. The lens suffers from chromatic aberration, it bends different wavelengths of light slightly differently. The same is true of directional microphones and speakers … when you get near the edge of their sonic beam you “see” acoustic aberration … different frequencies of sound are “focused” differently. A really good directional speaker or mic should have as little acoustic aberration as possible. Directional speakers are a compromise, however, both in terms of acoustic power required, and to avoid feedback, the live sound engineer needs to use directional speakers.
For portable acoustic sound there is an additional requirement: being heard at the rear of the hall. Sound power falls off as the square of the distance. To a very large extent your ear compensates and perceives sound volume to fall off more or less linearly with distance. For a small to medium room the solution is simple. Place speakers flanking the stage. However if the room is large, this approach has a problem … in order for the sound to be loud enough in the rear of the room it needs to be too loud in the front of the room. There is a solution … use distributed speakers where some of the speakers are in the room itself. The issues are setup time and complexity: more speakers, and if the sound is to be HiFi, each speaker needs to be balanced in terms of both volume and time delays.
There is another branch in live sound I call extreme enhanced sound. This branch discards the concept of HiFI. While certainly not my “cup of tea”, this approach is very popular and has been for the last 40 years. I would say my dislike of the approach is personal preference except for one factor … it is causing real and irreversible damage to our hearing! The approach is simple: full frontal attack. Speakers flank the stage. The average volume in mid house is often well over 95 dB, sometimes well over 100 dB. There are other characteristics … extreme bass enhancement, extreme amounts of synthetic reverb, and very often a high degree of compression.
The popularity of this approach to live sound has caused the development of new types of speakers … speakers that can output volumes well over 130 dB and subwoofers that can output extreme amounts of bass.
Since the invention of electronic speakers there have been twin goals, higher fidelity and higher volume. Starting with the development of the 33 rpm record, much effort has gone into the development of HiFi speakers. With the advent of transistor amplifiers, and later very high efficiency class D amplifiers, even on a moderate budget, we had the power to make speakers louder. But this too requires a change in speaker design. A speaker is say 5% efficient at converting electrical energy into sonic power. Where does the other 95% go? It goes into heat. If the average power consumption of a speaker is 100W (which is extremely high for a home system, but not so high for a live sound system) then the coil at the cone of the speaker has to dissipate 95 W of heat.
In the last 40 years, both for live sound and also for in-home (or more likely in-car) speaker systems, HiFi has taken a back seat to extreme volume and extreme bass. We have two divergent paths: highly resolving speakers that are so highly resolving I very much doubt my old ears (or even my ears of 20 years ago) can detect the subtleties. At the other extreme we have speakers that don’t attempt to be HiFi; they attempt to be acoustic cannons. I do need to apologize to Bose, as they once sold a very good product called an acoustic cannon, that provided subsonic sound to movie theaters.
Along the way, we developed some speakers (mostly for electric guitars) that were technically not very HiFi, but we liked the sound of their distortion and so made clones of these speakers; they have become a part of the sound of the instrument. At the other extreme, for home use, we (again this charge being led by Bose) have semi-HiFi speakers that are very small. I have a pair of miniature speakers. They are so small you can put your hand around them. While Paul McGowan would certainly not consider these speakers highly resolving, nevertheless they are much better than the original HiFi (12” 3 way) speakers I bought when I was in college.
Paul’s Highly resolving speakers, designed for use in a listening room, would not make for good acoustic HiFi speakers at a live sound event. Live sound directional speakers would not make for good speakers in Paul’s listening room. And neither of our setups are suitable for surround sound or for that matter, a high energy rock band.
My son has a speaker; it is a tube about 6” in diameter and 16” long and has sound ports at both ends. It is waterproof, battery operated, and Bluetooth. Like my miniature speakers, the sound quality is remarkably good. In the last 20 years we have developed many types of specialty speakers, some of which will literally blow your eardrums out, some of which are extremely small, some are waterproof, some are battery operated, some are … But all these speakers are unimaginably better than the speaker in the AM radio of my youth, and will put to shame the original speaker system in my old Webcor HiFi .