Lenard Audio proposes to establish a code of Audio Standards to achieve seamless integration across all music replication platforms to, potentially, change the future of the audio industry.
No two brands of studio monitors sound alike. So, it cannot be predicted how movies will sound until screened in a cinema. Likewise, competing brands of home Hi-Fi and live performance systems reproduce sound in a dissimilar manner.
The purpose of the Lenard Audio Application Standards (LAAS) is to promote consistency and transparency in just how amplified sound and music are heard, with the need for minimal adjustment to existing technology and minimal cost to the industry.
Standardization has not been achieved — until now...
— Lenard Audio
This Web site is intended to become the home for the long-awaited and needed global audio standards for music replication. Everyone is invited to participate and contribute. We will add membership and forum pages shortly.
The principle ideas put forward to achieve this objective (post the universal acceptance of the Thiele-Small parameters) emerged in Sydney, Australia in 1978, when John Burnett, Richard Priddle and Michael Dixon proposed ways to standardize sound reproduction for the audio industry.
The standards described on this Web Page are Copyright © 2015 Lenard Audio. Reference to these standards requires reference to the source. The standards will be available for commercial application shortly.
There is a bewildering variety of sound systems in today's world market, each with its own idiosyncratic sound. The problem is that "no point of reference exists" as to how amplified sound/music should be heard.
"Loudspeakers are the single most important element. They influence the art as it is being created. And in that process loudspeakers need to be neutral. If they are not neutral then they become part of the art " as stated by Dr. Floyd E. Toole (Vice President Acoustical Engineering, Harman International Industries, Inc.) Sound Reproduction: the State of our Science.
Recording engineers generally master music and film sound tracks knowing that the end result will mostly be heard on small speakers, headphones or ear-buds.
The problem begins with recording engineer/producers selecting a monitor speaker system as a compromise between, 1) what they assume is representative of the average speaker system in most homes, 2) and that which makes their favored recordings sound best to them. Similarly, a consumer will choose a sound system as a compromise, often with budget restriction, that enables their favorite music to sound acceptable to them.
Recordings that are Equalization (EQ) modified and compressed to compensate for the anomalies of one speaker system will not sound the same on a different speaker system. In each case, the outcome is often governed by the unique differences in how a given sound system modifies and colors the music.
Therefore there is very little consistency between the various speaker systems in studios, and minimal association between the sound of recordings created in studios and the listening experience of consumers.
Compensating for small and/or highly colored speaker limitations contaminates the integrity of the recording.
Many recording engineers waste endless hours caught in a "circle of confusion" by trying to evaluate the sound quality of speaker system "A" by listening to it with a colored recording that was created by compensating for the limitations of speaker system "B", as defined by Toole in his book "Sound Reproduction Loudspeakers and Rooms".
The lack of standards in how speaker systems are to be defined for recording, mastering and reproducing music has resulted in other areas of the audio industry being, effectively, "out of control".
Contentious issues surrounding the confusion between MP3 data compression and hyper-dynamic range compression DRC (two completely separate and independent types of "compression") can only be resolved after application standards are in place to achieve, as close as possible, neutral/transparent reproduction from sound systems.
MP3 data compression is often blamed as the primary cause of music degradation. MP3 may cause smearing of music to varying degrees, which is mostly inaudible on small, low fidelity speakers, but ....
Hyper-dynamic range compression (to exaggerate loudness) has a greater effect in degrading music than all other problems combined.
Dynamic range compression DRC is applied as an artistic effect on individual tracks within a recording. A whisper, a scream, loud instruments and soft instruments can be compressed or expanded to the same level. However, extra DRC is often applied to many recordings, in post-production, to further reduce the dynamic range to enable the recording to be more easily reproduced on small speakers.
When music is hyper-dynamic compressed, the average level of the recording can be increased, thereby exaggerating its loudness. Hyper-DRC is used extensively by recording companies, radio, TV shows and advertisers competing to be heard above each other.
Hyper-dynamic compressed music sounds flat, lifeless and highly distorted. Voice articulation can be so low that is becomes difficult to understand words in songs and the dialogue in films.
Hyper-DRC forces recordings to sound similar on all sound systems, this is sometimes used as an argument to negate the need for sound systems to attain being neutral/transparent.
The regulatory body for the global broadcast industry has implemented Loudness Standards (referred to as K-Weighting). This initiative is intended to discourage the use of hyper-DRC as a means to exaggerate loudness. This is a link to an editorial by Hugh Robjohns The End Of The Loudness War?
"Unsafe at Any Speed" published in 1965, described an automobile industry that was "out of control", with no standards for performance and safety. As a result, the global automobile industry was transformed and standards for: 1) brakes, 2) steering, 3) suspension are now mandatory. Cars today compete on styling, size, application and power, as a result of these standards being implemented.
Every industry in the world has application standards -- except the audio industry.
Now is the time for the audio industry to adopt standards (order of precedence) for sound systems to attain being neutral/transparent and compete (solely) on styling, size, application and power.
Almost every manufacturer of sound systems says "Our systems are the best", irrespective of the quality of their products. How can a purchaser ever know the true quality of any given sound system?
The only solution is to apply a rating-system based on standards.
A rating value, applied to a sound system must be determined fairly, by an aggregate of standards-approved submissions. This process is presently being developed. The aim is for the standards to be utilized commercially, across all music replication platforms, studio mastering, home systems, cinemas and concert PA systems. These can be adapted for headphones and ear buds. The Standards are similar to the energy ratings we are all familiar with, but understandably a little more complex.
Below are 5 attributes, in order of precedence, by which all sound systems can be assessed when referenced to a benchmark sound system (described in section 3). Each attribute shall be assessed as to its value, in the order listed below. The primary attribute is intelligibility.
The Lenard Rating for a given sound system is based on a performance-attribute rating system. There are 5 attributes Each of the 5 attributes is first qualified by a 5 star rating in 1/2 step increments (0.5 1 1.5 2 2.5 3 3.5 4 4.5 5). Then each attribute is multiplied by a weighting number to create a sub-total. The sub-totals are added to create an LR value, with the maximum being LR 100 as in the table below ↓ LR Lenard Rating.
|Attributes||Star rating||Weighting||Sub total|
|1. Intelligibility||x 6||30|
|2. Spectral balance||x 5||25|
|3. Harmonic detail||x 4||20|
|4. Bass discernment||x 3||15|
|5. Directivity||x 2||10|
Attempting to replicate the sound-of-nature from a man-made-sound-system is an asymptotic task (that is: we can approach but never reach) therefore LR 100 is unachievable. The reference sound system described in the next section may approach a value of LR 80 as in table below ↓
|Attributes||Star rating||Weighting||Sub total|
|1. Intelligibility||x 6||24|
|2. Spectral balance||x 5||20|
|3. Harmonic detail||x 4||16|
|4. Bass discernment||x 3||12|
|5. Directivity||x 2||7|
A typical audiophile system may achieve a value of LR 30 to LR 50. A small domestic sound system or PA sound system may achieve LR 10 to LR 30 as in table below ↓
Small home system
|Attributes||Star rating||Weighting||Sub total|
|1. Intelligibility||x 6||12|
|2. Spectral balance||x 5||7.5|
|3. Harmonic detail||x 4||8|
|4. Bass discernment||x 3||3|
|5. Directivity||x 2||1|
1. Intelligibility is the prime objective of a sound system; the greater the articulation, clarity and separation between voices and instruments, the greater the star rating. Poor intelligibility causes listening fatigue. A film may have great CGI (Computer Generated Imagery) effects, but if we have to strain to hear the dialogue, then the story's dynamic flow is severely compromised. We may hear the words of a solo voice, or phrasing of a single instrument, but when two or more voices or multiple instruments are playing, intelligibility can be reduced to a smeared mess. All sound systems, small and large, can achieve high intelligibility with minimal increase to manufacturing cost.
2. Spectral energy balance refers to the bass, mid and high frequencies maintaining consistent balance at all loudness levels - the greater the consistency, the greater the star rating. Many small sound systems sound thin and trebly at low level - the lower frequencies only become more audible when the level is increased. The opposite effect is often heard from cinema sound systems and large concert PAs. A performer talking or singing at low level can sound full and rich, but when the level is increased, the sound can be shrill and harsh.
3. Harmonic detail refers to the frequency spectrum above the voice - the greater the detail discerned within the harmonic region, the higher the star rating. In early sound system history, adding a tweeter defined a sound system as being Hi-fidelity. In an ideal sound system, a tweeter would be applied above the voice i.e. above 5kHz. But most low-frequency cone-speakers tend to behave chaotically from approx. 2kHz upward. Therefore most two-way speaker systems are crossed-over to the tweeter between 2kHz to 4kHz. This splits the upper region of the voice between two different speaker components (where our hearing is most sensitive to capturing detail) which can cause the voice to sound distractedly unnatural. Also tweeters crossed over at frequencies lower than they are designed for, may result in the harmonics sounding shrill and smeared.
Note: It's not our ears that hear, it's our sense of hearing.
4. Bass discernment refers to the ability of a sound system to evenly reproduce bass notes in the lowest 1.5 octaves - the more even and distinct the bass notes are, the higher the star rating. Many small sound systems, speakers in laptops and small home cinema sound systems (bar speakers), may have little to zero bass. Some larger sound systems, particularly with separate sub-bass cabinets, might appear to have an impressive bass response. But close listening attention may identify a pronounced resonance, sounding as a single note regardless of what notes are being played in the bass scale. This problem is most noticeable with car sub-bass systems and concert PAs, which can sound particularly annoying if the resonance peak is in the upper register of the bass scale.
5. Directivity control refers to the physical area in which 'the stereo sweet spot' can be experienced - the greater the area, the greater the star rating. Lower frequencies tend to be omni-directional whereas higher frequencies can be so narrow (beaming) that the stereo sweet spot can only be experienced from a single point. A small change in listening position, left or right, can result in the closest speaker dominating. Directivity index is a technical term referring to energy dispersion.
All good sound systems should already have a flat response, but may not perform well in some of the 5 attributes - why? A frequency response graph shows Sound Pressure Level SPL across the spectrum, but does not show of what that energy consists.
Inter-modulation IMD is the major distortion produced by speaker systems and is the primary reason differing sound systems sound different. IMD is not revealed in a frequency response graph. One of the questionable trends in the audio industry is to regard the room and speaker system as a singular entity. Then with the aid of computers (DSP Digital Signal Processing) to use room reverberation or distortion as a filler to achieve a flat response or a specified response. This practice further increases IMD, reducing intelligibility. Many professionals within the audio industry are embroiled in arguments about this approach, which is another key reason why the establishment of audio standards is essential.
Application standards enable everyone to better appreciate sound systems and how they can be qualified. For this approach to have efficacy, however, requires attendance at seminars that are run as interactive workshops, where everyone can be methodically taken through the necessary steps. Certification is based on peer achievement in attaining agreement about standards ratings as a self-correcting process. When audio reviewers become certified using this approach then their critique will be similar to the following examples.
Example 1. A renowned brand that specializes in small sound systems released a new (expensive) model marketed as having outstanding performance. However when qualified by the 5 attributes, applied to a weighted rating system, the real performance is better understood. The impressive bass energy was astounding for its size but highly resonant, sounding as a single note regardless of what notes were being played. Intelligibility of voice articulation was lacking, making it difficult to understand words in songs which became inter-modulated when two or more voices were singing. Higher harmonic frequencies were harsh and cymbals did not sound natural. The spectral energy balance was very different between low and high power. A stereo image was difficult to hear, as it always appeared as if the music were coming from the nearest speaker.
Example 2. A less renowned brand, similar model to Example 1, but a lower price. Bass energy was less by comparison, however the bass notes were even and not resonant. Intelligibility of voice articulation was good — it was easier to hear individual words in songs. Harmonic frequencies were minimal, not harsh, and cymbals sounded more natural. The system was less powerful by comparison, however the spectral balance remained consistent over the limited power it had. The stereo image was also good, the music appeared to be evenly spread between the two boxes over a reasonable listening area.
Many small sound systems try to be everything for everyone, but often perform poorly in most attributes. An example is a video bar-speaker that has deliberately-introduced distortion to make it appear louder in comparison to other brands. The reviews describe the bar-speaker as being delightfully bright. The response graph shows a rise between 1kHz - 6kHz, but does not show what that energy consists of.
In this example, the introduced distortion is in the middle of the intelligibility band, which creates a psycho-acoustic effect, similar to the sound of fingernails on a school chalk-blackboard. For someone watching sports and motor racing the added distortion might contribute to the excitement. Whereas for someone watching a TV drama series, trying to understand the dialog can be a tiresome task.
Musical instrument amplifiers, particularly guitar amps, are intended to be highly colored, enabling characteristic types of distortion (overdrive) to be inclusive of the art. However when the guitar amp is recorded, it is hoped that the studio-monitor and home sound-systems will faithfully reproduce the recording, which includes the artistic sound character of the guitar amp.
Some companies, a minority, make exceptionally high-quality sound systems. But a large percentage of poorly performing sound systems are promoted by marketing hype and unethical reviews that describe them as being otherwise. This makes it difficult for customers (without technical expertise) to discern what is real and what is not.
Application standards will enable customers to identify the legitimate performing sound systems and better understand what they are purchasing. Also, Application standards will help those who design sound systems to adopt an agreed discipline to achieve an ordered focus in design quality.
It is often argued that people have various preferences in how different sound systems "sound".
When making this argument we should consider that no one would prefer to read a book with smeared-unintelligible text, or choose to watch a movie that is out of focus. The same principles apply to sound. No one would prefer listening to a sound system where the intelligibility is so low that it is difficult to understand the words in songs or the dialogue in a movie.
However there are positive arguments for personal preferences in sound system performance.
When listening to a live band playing acoustic instruments, we might prefer sitting closer to the piano, bass player or percussionist. These choices are associated with the artistic experience of the music and not to the quality of reproduction. For example, a personal preference to have the bass frequencies at higher or lower energy than the recording was intended to be heard at, is a legitimate choice, providing the attribute of the bass notes is evenly reproduced and not sounding as a single resonant note.
To achieve standardization, a sound system is required that stands as a point of reference. The word "reference" in this text does not mean absolute. “Reference”, in this context, refers to a starting point from which measurements and comparative assessments can be made, similar to how we use a ruler.
The reference sound system must be:
A neutral sound system is a paradox because it can only be defined by its ability to reproduce that which is not neutral. For example, a large full-fidelity sound system can reproduce the sound of a smaller lower-fidelity sound system, but not the inverse.
Therefore a reference sound system that enables us to qualify the attributes that attain to neutral/transparent reproduction is defined by its structure.
How can this be understood?
Many people, including some designers of sound systems, have not heard (up close) a 15-inch speaker in an 8 cubic-foot enclosure. Nor played a Fender Stratocaster or a bass-guitar into a Marshall quad-stack, nor directly heard (up close) a large compression-driver horn used in cinemas.
Common sense enables us to understand that a small 2 to 6-inch speaker (regardless of specifications) could not reproduce the equivalent dynamic energy of a real double bass, violin or saxophone. Also a bass guitarist in a rock band would not use a small speaker box with a 4-inch speaker on stage, they would use one or two cabinets with 12 or 15-inch speakers.
Commercial cinemas use large compression-driver/horns for mid range that are capable of reproducing the same dynamic energy as real instruments. Therefore, such 15-inch speakers and compression-driver/horns are used as a reference for defining sound-system attributes for reproducing bass, voice and voice articulation.
Understandably, it is essential to use appropriate, high-quality driver components.
A reference sound system must have the capacity to replicate the equivalent fidelity and dynamic power of each instrument within a real orchestra (that is: it be larger than any single orchestral instrument). It also must be capable of replicating the performance of any sound system to which it is compared, in reference to its application, as listed below. It is only from this reference that recording, mastering, creative effects, analysis and other standards can thus be established.
The Reference System must be:
Understandably, certification will require sound engineers, audio consultants, system designers, and reviewers to attend seminars and have access to a system of this structure, so as to obtain a perspective for qualifying the 5 attributes.
Four-way active sound systems of this stature are not readily available in the marketplace so require independent assembly. Anyone conversant with the points below can achieve this task.
This reference system was originally based on JBL professional drivers. Unfortunately only two models of these drivers are still available — 2446 Compression driver and 2226 15-inch. The JBL 2405 slot radiator (tweeter) is no longer available. Other driver brands, however, are available, of which some, it can be argued, are either equivalent or superior. The parts list below is for a stereo pair.
The reference system will give an experience of:
The driver components vary greatly in efficiency dB/mW. These efficiency differences need to be measured, calibrated and compensated for. There are various approaches as to how this can be managed. All that matters is the result. The points below are a guide, beginning with the lower-voice 15-inch speaker. Using two people, working together to achieve this, is a recommended approach.
15-inch JBL 2226 with 2446 compression driver and horn or their equivalents. The 15-inch 2226 or equivalent can be selected as the dB reference for efficiency matching the other drivers. The 2446 compression driver/horn or equivalent is approx 12dB higher efficiency than 15-inch 2226.
Ensure correct phasing and double check. Compression drivers must have protection capacitors in series (approx 200µF). These capacitors block pops, bangs, turn on/off pulses from the amp getting to the Compression driver.
Tweeters must be crossed over above the voice range approx 5kHz to 8kHz. Slot/Ring radiators, compression tweeters are suited for far field, cinema or PA, and are approx 3dB to 6dB less efficient than the 2446 compression driver/horn or equivalent.
For near to mid-field listening, ribbon tweeters might be better suited. These can be approx 6dB - 12dB less efficient than compression driver/horn. Ensure correct phasing and double check. Tweeters must have protection capacitors in series (approx 50µF). These capacitors block pops, bangs, turn on/off pulses from the amp reaching to the tweeter.
Thiele/Small parameters are an essential guide for bass cabinet design. Sub-bass must be crossed over below voice, approx 100Hz or below. 2 x 15-inch sub-bass speakers can be installed either together in one large cabinet or separately, in two smaller cabinets.
Sub-bass speakers may be 12dB less efficient than 15-inch 2226 or equivalent. Sealed and ported cabinets have advantages and disadvantages, sealed cabinets can introduce less complications. Ensure correct phasing and double check.
Sub-bass refers to bass speakers crossed over below the voice-range. This ensures zero inter-modulation of the voice. Sub-bass speakers have strong heavy cones with extra long voice coils. Most sub-bass speakers for cars are unsuitable for this purpose because they can be excessively inefficient and have high Fs making them unable to reproduce deep bass. Select sub-bass speakers with low Fs (approx 20Hz) designed for large speaker cabinets. Experiment as needed with the crossover points used.
Special note: The recommendation is for the reference system to use the highest quality driver components available. However, the difference in performance of a system consisting of the highest quality drivers and another of lesser quality drivers could, possibly, be within 10%. This is due to size and structure of the system as well as being four-way active. Whereas, when comparing different models of small, similar-size domestic two-way passive speaker systems, that use drivers of variant quality, the auditory differences can be significantly large, despite their specifications being similar.
The reference system has 2 x 15-inch sub-bass speakers, 1 x 15-inch lower voice speaker, 1 x 2446 compression driver/horn, 1 x tweeter, per side. Any larger or smaller will result in reduced performance fidelity. This system can be trollied through an average door, easily moved in a van, large station wagon or pick-up truck and fit within most homes, yet perform perfectly as a jazz club PA. Most of all, being fully active, it would work as an excellent studio mastering system, once critically aligned.
An orchestra played through a sound system cannot be directly compared to an actual orchestra - why? A sound system may give either a poor or enhanced representation of an orchestra, but it only provides a simulation or superficial likeness, known as a simulacrum. Sound from a real orchestra is a fractal complex multi-point and multi-directional radiated set of acoustic wavelengths; and the perceived sound heard by an audience will be different from seat to seat. A stereo sound system, by comparison, simply radiates sound forward from two distinct locations.
Also, a sound system can only replicate a representation of an orchestra from the perspective of a microphone (from a position that is quite likely unknown) including additional recording EQ effects, also indeterminate. Therefore the performance of a sound system cannot be judged by direct AB comparison to an actual orchestral instrument. The difficulty in understanding this logic has, for many, been the basis of confusion.
It is possible to design a sound system that can replicate music so realistically that we can virtually feel the musicians being present. This illusion of auditory reality is magnitudes beyond what can be achieved in the visual realm. But no matter how close we approach the grail of neutral/transparency it will always be an elusive quest – remaining as a simulacrum.
A truly 100% transparent sound system could not actually be, because only the music would exist. Those of us who are captured by this quest are deeply humbled. Attempting to create something for it not to exist, is a magnificent paradox.
Paradoxes are philosophical arguments that involve apparent contradictions which are valuable for promoting critical thinking.
It is often believed that a perfect sound system can be defined by "analysis". But neutrality cannot be defined by analysis and is, therefore, a paradox in itself.
The paradox of neutrality requires us to see that a neutral sound system can only be defined by its ability to reproduce that which is not neutral. For example - A large high-fidelity system can replicate the performance of small low-fidelity system, but a small low-fidelity system cannot replicate the performance of a large high-fidelity system.
This approach for defining application standards, by measurement related to structure (and not by measurement alone) comes under an umbrella of a discipline described as "Bayesian analysis", further explained at the end of this chapter.
Audio frequency spectrum is approx 10 octaves (three decades) 20Hz to 20kHz. Our hearing is least sensitive at the ends of this spectrum and most acute in the decade from 500Hz to 5kHz, the range critical for speech intelligibility. This is where sound systems need to be most accurate and where technical shortcomings are most apparent.
For music replication, a basic sound system must be capable of reproducing a two decade range from 80Hz to 8kHz as a minimum requirement, and an extra octave either side is required for a sound system to be classified as hi-fidelity. Because the speed of sound is approx 340 metres / second, the bottom E on a double bass is a 14 metre wavelength and efficient reproduction requires a large radiating surface.
When listening to a live band acoustically (not through a PA system) we hear the instruments separately as coming from their various locations (horizontally). If all the instruments in the band were heard as if coming from a single point it would sound implausible much like a typical speaker system (woofer and tweeter).
A fully active sound system divides the frequency spectrum into 3 or 4 sectors with each sector amplified separately (vertically). When done correctly the resultant effect is similar to the live band in that it sounds realistic. The irony of this is that the ear equates this illusion to the separation of sectors, regardless of whether said sectors are separate instruments in the horizontal plane or separate frequency bands in the vertical plane.
With two fully active speaker systems acting as a stereo pair, the added horizontal effect produces an equivalence of auditory enjoyment synonymous with reality, creating an illusion that the musicians are actually present.
The weighted rating system described in section 2, defines an order of precedence to the attributes of a sound system. Intelligibility (voice articulation) is the most important attribute and therefore has the highest weighting. The principle being that if only one attribute can be achieved from a sound system then intelligibility is selected first. Intelligibility is the least expensive attribute to implement, often at no increased cost. The other attributes follow suit in order of precedence, similar to ace, king, queen.
The word "reference" in this text does not mean absolute. “Reference”, in this context, refers to a starting point from which measurements can be made, similar to how we use a ruler.
The principles of science state that wherever possible, measurements are best taken from a position that is external from what is being measured. Measuring from an external position enables us to have a better perspective of what it is that is being measured. For example: on a pitch black night (without an external reference) it may be difficult to know whether a spot of light is a firefly in a tree or a galaxy a million light years from earth.
The traditional method for measuring sound systems is basically self-referential. Therefore without seeing the sound system being measured, it would be difficult to know, from the frequency response and distortion figurers alone, if the system being measured is a tiny speaker in the rear of a video screen or a large cinema sound system. Understandably the two speaker systems would sound completely different, regardless of the similarity of specification.
Going to a high-end audiophile shop and listening to various expensive speaker systems (with identically flat responses) starkly reveals this problem.
Measurement based on self-referential analysis is very different to measurement based on external comparisons between physical structures. The reference sound system is a large-scale four-way active structure, which enables the five attributes to be defined and easily understood. This provides a starting point from which other sound systems can be compared.
A reference system must be based on first order principles in physics, at the boundary of what is technologically achievable, utilizing traditional technology with speaker drivers that are universally available. A reference must also be that which can be repeated and verified as well as something everyone can agree as to its behavior and performance.
The reference also represents the zenith on a bell curve where there is trade-off between the size of a system and its fidelity, diminishing either side by being smaller or larger.
For most applications, sound systems do not need to be large-scale four-way active. The attributes of the reference system are only used to qualify the performance of other sound systems regardless of the technology that is applied. In the future, a new technology might be discovered that will be superior, requiring a new reference to be established.
A single given cone speaker can only accurately reproduce a maximum bandwidth of 3 octaves (1 Decade) with minimal distortion. This being a function of cone diameter and wavelength, and therefore the reason the reference system is of a four-way active type. Compression-driver horns follow a similar rule of 3 octaves (1 Decade). This being a function of horn-length, flare-rate, mouth-diameter and wavelength.
The tweeter is crossed-over between 5kHz to 8kHz to ensure zero contamination of the voice. There are many excellent compression tweeters available. Excellent ribbon tweeters are also readily available. This choice is discretionary.
The mid-range spectrum is approx 100Hz to 6kHz, approx 6 octaves and is divided into two approx 1 decade bands (lower voice 100Hz - 800Hz) and (upper voice 800Hz - 8Khz). 1 decade = 3 octaves (approx). The exact choice of crossover frequency will vary depending on the selected drivers.
A compression driver/horn reproduces the upper voice, 3 octaves, 1 decade (800Hz - 8Khz). Compression driver/horns achieve possibly the best accuracy and highest efficiency compared to every other method known. This is where our hearing is most acute for discerning speech articulation and music intelligibility. It is the reason compression driver/horns are applied in cinemas throughout the world. education.lenardaudio.com/en/07_horns.html
Comment: Regardless of what opinions some audiophiles may have about disliking horns, I have found (without exception) that the sound system they heard had a poor quality horn/compression driver that indeed did sound dreadful. Moreover, rarely did the audiophiles I spoke with mention that they were aware they were hearing a higher quality horn/compression driver when they went to a cinema.
The lower voice, 3 octaves (100Hz to 800Hz) is the body that gives voice and instruments their character. In this lower voice band our hearing is less acute for discerning articulation and intelligibility. However our hearing is attentive to the averaging effect this frequency band has in acting as a carrier for the voice and instruments.
A large speaker/horn could be used but it would be physically difficult to move and fit within most rooms. This lower voice band, for the reference system, is best reproduced with a front loaded 12 or 15-inch speaker designed to be maximally efficient and linear in this region.
The bass register is from open E 41Hz to 165Hz. As a general rule, the cone excursion of bass speakers increases 4 times the distance for each octave decrease to maintain equal energy. The bass speakers are crossed over below the voice range approx 100Hz, to ensure the lower bass frequencies do not inter-modulate the voice. 2 x 15-inch bass speakers for each speaker stack = 4 x 15-inch for a stereo pair, which is the minimum radiating area to achieve linear dynamic power of the bass frequencies.
1 x 18-inch speaker could be used instead of 2 x 15-inch speakers. However, 2 x 18-inch speakers per side is ideal for extra large areas. The technical reason is: The diameter of the combined bass radiating area must be greater than 1/10 of the wave-length of the lowest frequency, so as to maintain equal spectral energy across all bass notes, when applied to a free field environment. The reference system is free-field-referenced, not room-referenced.
Note: The above statement about bass radiating area does not apply to resonant bass systems. Examples include ports within cabinets and organ pipes.
Comment: Many bass players today have small bass amps that have 10-inch speakers. At concerts, these small bass amps often require being amplified by the PA system to enable the bass player to be heard. In the earlier era 1970s rock-blues, musicians often played directly to an audience and not through a PA system, a good bass player had a large speaker box 2 x 15-inch that produced real bass and often used two boxes = 4 x 15-inch speakers for a larger venue.
The major distortion noticeable in sound systems is IMD, which emanates from the speakers, not from electronics. Fractal cone breakup, non linearity, nodes and cross-over anomalies all accumulating as inter-modulation within the speaker system. Inter-modulation distortion (IMD) is caused by amplitude modulation of two or more frequencies in a system with non-linearities. The interaction of these modulating frequencies creates additional frequencies as integer multiples of each other, which include sum and difference frequencies plus multiples of those sum and difference frequencies.
Inter-modulation creates spurious harmonic side-bands that spread out across the music spectrum. It is these side-bands that cause the distinct coloration we hear in speaker systems, particularly passive speaker systems. For a sound system to achieve minimal inter-modulation distortion the frequency spectrum must be divided into separate sectors and each sector amplified independently.
Digital technology has transformed the audio industry. Analogue Class AB amplifiers and analogue active crossovers are being replaced by Class D amplifiers and DSP Digital Signal Processing. Some Class D amplifiers perform equally as well as Class AB amplifiers, but many do not. Many audio DSP management systems function well but also some DSP systems use proprietary or secretive software that may have complications with unresolved bugs.
When assembling the reference sound system it is advisable to begin by using traditional Class AB amplifiers and proven analogue four-way active crossovers. Then step by step, convert to Class D amplifiers and DSP management.
DSP enables us to obtain detailed control of sound systems that could not be achieved with previous analogue technology. But there has been an increase in false and misleading statements in the marketing of sound systems using DSP technology. DSP cannot fix a leaky submarine or enable a plane to fly when the wings have fallen off, nor can DSP enable a small low fidelity sound system to perform equally as well as a large hi-fidelity sound system. Most of all, in reference to marketing claims that DSP can resolve room reverberation problems ---
"Thou cannot equalize time with amplitude"
DSP cannot enable the performance of a sound system to be improved in an excessively reverberant environment. The only way to solve problematic room reverberation is to fix the room.
One of the many arguments in the audio industry is - "What is the reference by which a sound system is designed and calibrated?" Some argue that the room comes first, and a sound system should be designed and calibrated to comply with room measurement and acoustics.
The alternate argument is that room acoustics on its own, without being externally referenced to free field, is non-defined. Therefore a sound system should be free field referenced first. Then when applied to a room, using DSP technology, definitive measurements of difference can be made, enabling calibration of the sound system to the room to achieve greater accuracy.
Cinemas do not show films with lights on, nor do cinemas put mirrors on walls. Cinemas are visually neutralized (dark). Only the screen is seen, we are engaged in the story. The recorded sound for a cinematic film includes the reverberation of the environment that the story is set in, which is inclusive of the art. Therefore cinemas (already visually neutral) should be acoustically neutral as well, and not have its own reverberation that is competing with the art, or contaminating the story.
The same argument applies when listening to all recorded music. Recorded orchestral music often includes the reverberation of the auditorium, which is inclusive of the art. Therefore the room we are listening to the recorded music, should also be as acoustical neutral as possible so as not have its own reverberation that contaminates the art.
A reference for understanding sound systems in rooms is "Sound Reproduction Loudspeakers and Rooms" by Dr. Floyd E. Toole.
Bayesian analysis is an external process, as a philosophy, that when applied to the disciplines of mathematics and science, ensures theoretical concepts are grounded in reality. Bayesian analysis is included within almost every area of science, engineering, economics and actuarial studies, but has rarely been applied to audio.
Example: An audiophile speaker cable costing $1,000s is marketed as having magical qualities that transforms the listening experience of a sound system. The questionable claims are supported by unethical audiophile reviews. Any skeptic can smell snake oil and scientific analysis tells a different story.
The resistivity of copper is 1.68 x10-8 Ωm. A quality 3m speaker cable may have a total resistance of 0.1Ω and the larger audiophile cable (for the same distance) may have a lower resistance, 0.05Ω. This 0.05Ω difference in resistance is irrelevant, and the differences in the reactive components (capacitive and inductive) of the cable are equally irrelevant at audio frequencies.
No method of verification exists to prove that any human could detect a difference in the behaviour of a sound system (by changing speaker cables) at a distance of 6ft or 2m from a sound system. And the same argument applies to many other questionable claims quoted in the marketing of audiophile products.
Common sense enables us to understand the difference between peeing in a tea cup compared to peeing in the ocean. As an imagined theoretical exercise we can calculate how much the ocean will rise, which may be less than the distance equal to the diameter of 1 atom. Obviously we see the absurdity of believing a rise in ocean level equal to the distance of an atom could be detected by human perception.
Bayesian analysis ensures theoretical concepts are verified by being detectable and measurable. And in reference to audio, that verification is human perception, which ultimately is the only place that "sound" exists, being the product of the brain's processing of auditory signals. This analysis is a much needed forward step towards reining in an industry that in many ways has been "out of control". wikipedia.org/Bayesian_inference
"For a successful technology, reality must take precedence over public relations, for Nature cannot be fooled." Richard Feynman
Audio Application Standards bring technological, economic and social benefits to the world, and help to harmonize the specifications of loudspeakers and sound systems for music replication.
As part of our Action Plan for "saving the world from bad sound" Lenard Audio is committed to developing programs to increase and strengthen our support to musicians and producers throughout the world.
In addition, the Lenard Audio Application Standards (LAAS) can improve global markets, and help define the characteristics of music production and reproduction with the intention of achieving a closer connection between the musician/producer and us, the audience as consumers.
We welcome everyone to play an active role in the (LAAS) community, promoting the Application Standards and taking part in its development.
John Lenard Burnett