THE AUDIOPHILE's Principles

10 guiding principles that underpin our best practices
 


Every interest group has developed its credo to define its cause. The audiophile industry is no different. With each generation, the culture and value system is being developed, strengthened, and ingrained within the community. I am sharing what I know to be self-evident and confident that the following list reflects our shared principles.


1 - A burning desire - runtime or some may say soak-time

Burn-in (aka bedding-in) is the application of special test signals (e.g., sinewaves, pink noise) introduced into the system, intending to "charge" electronic components and eventually stabilize their electrical parameters. Some quarters believe that new audio components which comprise resistors, transistors, and capacitors, have some form of magnetism due to DC leakage that tends to orientate the electrons in some direction. Therefore, the burn-in process hopes to re-orientate them in a preferred direction so that they will ‘smoothen’ out the signal path. Others believe the burning-in process has to do with effecting changes in the internal conductance of FET or bipolar transistors or the dielectric in capacitors (including cables) positively altering their properties. The same is true for the ‘clocks’ found in the DACs which are said to require at least a week of burn-in time for the clock oscillator to stabilize. 

For loudspeakers, the electronics in crossover units as mentioned earlier may be affected in the same manner, moreover, woofer surround and spider need to be exercised to stabilize their mechanical performance. Loudspeaker cabinets like fine musical instruments may require ‘playtime’ to sound their best. For cables, they say that the manufacturing process and testing of the conductors and insulator causes mechanical and electrical stress on the cables. This is also the case some would argue for cables that were coiled when purchased or transported that must be "relaxed" with a burn-in process to stabilize them when they are laid out. The burn-in process requires that an audio signal (ideally pink noise) be fed into the system continuously (on low volume for practical purposes) between 50 to 100 hours. The actual effective time for break-in (burn-in) will vary since solid-state devices, loudspeakers, and cables are different. Many audio websites offer free customized downloadable burn-in digital files and one manufacturer (Nordost) even offers a stylish but dedicated electrical burn-in device (see illustration).

Nordost burn-in device

It is also pertinent to note that electronics when switched on need to warm up (settling time) between 1 to 2 minutes for the components to reach their optimum operating temperature for an increased current flow. This is especially true for tubes when the temperature of the cathode is increased, more electrons are emitted, hence current flow increases. Suffice it to say, for modern electronic components, a residual amount of current is left continually flowing when turned off (on standby mode) to keep the components warm which serves to justify the warm-up time required. Many manufacturers recommend burn-in time for their audio products before realizing their full potential. Whatever, the reasons are, we know that a new product or system matures over time like fine wine, and the sound changes for the better. This may also explain the perception as to why a new audio component usually sounds harsher and gradually smoothens over time. 

XLR Balanced Connectors

2 - Underrated but appropriately balanced – extraneous noise defender

True Balanced circuitry is preferred over the single-ended topology to cancel off noise-induced distortions. True-balanced means that there are 2 identical circuits at the signal input and output side thereby increasing the manufacturing cost. The component chassis is typified by the XLR balanced signal connectors. Noise picked up by the cable would be present in both positive and negative conductors. A phase comparator in the component at the receiving end subtracts both signals and accordingly gets rid of the extraneous noise that may be present. This is possible because the noise will be in phase on both conductors while the music signals are in anti-phase.

Therefore, a source needs to present the audio signal to the output in a balanced format. Some components use true differential circuitry which is ideal but expensive, but most convert the single-ended signal to a balanced form using active circuitry or transformers. The balance connectors utilize 3 conductors where one is the ground return, the second signal conductor carries the same original audio signal but inverted, or 180° out of phase from the third conductor to cancel off the noise that is caught outside the reference signal. Balanced connections are commonplace in the pro-audio industry with the use of long cables to mitigate external noise caused by Radio Frequency and Electromagnetic interferences. The hostile working environment at live events and even in recording studios is usually cluttered with lots of signal and electrical cables, especially at itinerant events. A balanced topology is therefore a tool for the battle against spurious noise interference. The balanced proposition is not meant to influence the audio signal, but to keep extraneous noise out, thereby improving sound quality, especially in areas of detailing (micro-dynamics).

3 - Discrete is indeed concrete – customizing design

There are 2 electronic design topologies: Discrete and Integrated. Discrete circuitry places individual electronic parts onto a printed circuit board in the traditional way, whereas Integrated is a modern-day design that compresses and configures electronic components parts into a semiconductor e.g., IC chip or Op amp in pre-configured designs. The Discrete topology instead allows the designer to choose available electronic components from the marketplace with good absolute precision between channels. Integrated designs have poor absolute precision due to process variation because it's difficult to get components of absolute values into a chip. Discrete designs can use large capacitor and resistor values which are not practical in Integrated designs. Discrete designs can be made from scratch with flexibility for breadboard testing to perform (sound) the way they want whereas Integrated designs rely solely on simulation instead. Discrete designs can carry a higher voltage and are not susceptible to interferences, and control open loop gain and feedback which are important factors for quality sound.

Discreet Circuitry

Integrated Circuitry

The only advantage of Integrated designs is in their manufacturing efficiency because they have fewer parts and are normally built-in smaller casings which reduces the cost of manufacture and storage. Integrated designs are limited to available IC chips and Op-amps that are readily available in the marketplace that is used as a standard foundation to produce economical electronic products. You will find examples of Integrated designs in Class D amplifiers, DACs, and inside active loudspeakers because of their smaller form factor. Discrete is without a doubt a more concrete design because the product can be tailor-made from scratch to sound as the designer intended. 

4 - Insulate and isolate whenever possible – vibration and resonance control

The chassis of quality audiophile components are usually made of heavy brush aluminum with few parts and joints for resonance control. The thicker they are, the better the shielding from EMI and RF interferences. In Hongkong, audiophiles there are known to utilize a knock test on the chassis using their fists, if it rings or sounds hollow, they won’t purchase the item. Suitable heat sinks for power amplifiers and vents are preferred over cooling fans which induced noise and vibration. The micro-vibrations on the chassis if not controlled may be induced into the circuit boards to become microphonic thereby creating hysteresis. UV meters or fancy LED displays are best avoided if there is no ‘off’ switch as they are a source of noise that may migrate into sensitive circuitry during playback. Most serious audiophile component manufacturers have this feature to enable ‘pure audio’.

Transformers are an integral part of any electronic audio product to supply voltage. However, some part of the magnetic field escapes the transformer aka “stray flux,” which is a primary concern for audio equipment. When this AC magnetic field leaks from the power transformer and interacts with a wire or PCB trace, it will induce a small current, akin to a secondary winding on the transformer. The resultant noise voltage produced is generally very small, but in audio equipment, even a few millivolts of noise in a sensitive circuit can be audible. Therefore, out-board (separated) power supplies are preferred for noise isolation. Moreover, rectifiers switching, and transformers vibrating are also noise-generating components. The transformers should be ‘beefy’ to handle high current as and when required by the recordings and not saturate which would lead to amplifier clipping. Covered transformers are preferred for noise isolation and radiation control. Large or multiples of smaller capacitors (preferred) are used in multiple rows to smoothen out voltage spikes and store energy acting as a power bank for backup against any voltage swings.

5 - One monkey at a time – only one variable in the audio chain

Whenever we evaluate any product, we should ensure that all other things remain constant to make a value judgment on the review item against a known reference. We need to ensure that only one variable is in play even the software and volume level should remain constant for an accurate comparison. The A/B/A evaluation method should be adopted to enable retention and recency effects. This is where you listen to product A, then product B, and back again to product A. Ideally, blind testing utilizing the A/B/A method is considered if you have a friend to facilitate the change to avoid observational bias. A/B/X testing is not recommended because you are introducing a switching system in the process that negates the quality of the sound that we are trying to evaluate. Double-blind is worse because the person making the switch cannot know what is being tested, so how do we know which one is better? We need to toggle between products (A & B) during evaluation to affirm a finding. Nevertheless, it is said that findings that are made when making any comparison in a blind methodology need to be made at least 16 times to achieve an appropriate sampling size for the test results to be reliable. Before any evaluation for new equipment, a 7-day burn-in time should be observed, and you would need to be familiar with the review item played in your system for at least a month to identify its peculiarities thereby making any informed judgment later during the assessment. On-the-fly snap evaluations are not recommended in high-resolution audio. Remember, only one monkey at a time.

6 - Power to the people – the engine that drives performance

The power supply is an integral part of an electronic component, and the power transformer is instrumental in its design. The goal for a transformer is to perform close to 100% efficiency i.e., all the energy in the primary winding is coupled to the secondary and is not wasted by heating the core or windings or leaking magnetic flux outside the core. This kind of transformer is costly to build. In a power transformer, the “primary” winding is driven with AC line voltage that comes out of the wall. The voltages required for the rest of the electronics are generated in the “secondary” windings. All windings are placed on a “core” made of an iron alloy. This is done because the permeability, or magnetic conductance, of iron, is much higher than that of air, which allows a transformer to work more efficiently.

EI-Core

Toroidal

R-Core

The EI-core transformer design is the most common type and the least expensive. However, there are air gaps between the laminations that provide an opportunity for stray flux to leave the core. Its mounting location inside the chassis and its core orientation if not carefully identified can also produce unnecessary noise. A toroidal transformer (doughnut) on the other hand will provide superior performance and is generally preferred because of its greater efficiency and lower noise floor over the conventional EI-core designs. Generally, toroidal transformers are half the size, run quieter, and are said to be 15% more efficient than comparable EI-core transformers. However, it is more expensive because they are difficult to manufacture. The R-core type transformer has the potential to become the predominant choice for use in affordable high-end audio. They provide all the benefits of toroidal transformers but at a lower cost.

 Upgrading component power supplies

Audio component power supplies like power conditioners, protect against power surges, filter out noise, and additionally, the transformer isolates AC noise when it is converted to DC. The front-end audio components such as the Phono stage, DACs, streamers, CD players, Turntables, etc., benefit most from a better-designed power supply instead of the stock switch mode types that are provided as standard with modern-day electronics.

Switching mode power supply

A ‘switching’ type power supply is universal in application i.e., they can be connected directly to any 220 or 110-voltage ac mains without the need for voltage adjustments. The power supply though compact in size is power efficient. Switch-mode power supplies regulate output voltage through a process known as pulse width modulation, the switching process however generates high-frequency noise. Therefore, these power supplies are a source of noise that would find its way into the signal path and may be modulated with the audio signal.

An economical upgrade or a step up would be to get a better switch mode power supply that offers a higher current rating (amperage) than that of your stock power supply for the same output voltage. This beefier power supply will have more headroom with higher-rated electronic components that would not be pushed to their limits hence emitting less heat and operating more quietly. Companies like ‘iFi’ offer a switching power supply that is available in different output voltages and boast an active noise cancellation technology like those used in modern noise-canceling headphones. Otherwise, use the laptop power supply bricks that may be sourced online with higher amperage and are more robust in their build quality. For music streaming and DAC products that are used in conjunction with your laptop or computer, the USB power supply from your laptop or computer is the worst-case scenario regarding noise. Using a mobile phone charger instead is better in this respect, and even better still, use the Apple iPad charger which is better built with a lower noise floor for the Streamer or DAC component.

Switching Power

Linear Power

Linear power supply

However, for the serious audiophile, you should without hesitation change the stock switching power supply to a traditional linear type of power supply. Linear power supplies utilize a larger transformer to draw voltage from the AC mains and step down to a lower DC voltage through a series of rectifiers and filtering processes to produce a clean DC voltage. In addition, to their increased reliability, they are less vulnerable to noise interference making them perfect for sensitive duties like medical equipment and low-noise audio components. They have a faster transient response i.e., less time is required for the output voltage to recover from changes in the load because of the beefier transformer used. In short, linear power supplies have the lowest noise floor with better headroom if compared to switching-mode power supplies. It will however have a larger chassis, it's heavier and more expensive. Many specialty manufacturers are offering linear power supplies with various output voltages and a variety of interchangeable barrel plugs to fit your component’s receptacle. A good designed linear power supply will maximize the companion component’s performance to the next level, like an upgrade at a nominal cost that affords more power to the people.

7 - Hearing is believing - the proof is in the pudding

Audiophile system components are all about the sound and not about the price, brand, or specifications. We, therefore, make judgments based on how the product sounds instead of how it measures or how much it costs. Putting together a complete system and fashioning it to sound ‘good’ is a complex, highly subjective, and personal matter. It is complex because there are many variables such as our physical hearing ability, room size, and configuration, loudspeaker design types and placement, sitting positions, room acoustics, and integration between components. It is subjective because of the listener’s musical preferences, tone, and loudness levels. It is a personal choice because some people may prefer dynamics, laidback presentation, pace and rhythm, full range in the audible spectrum, strong midrange, sparkling highs, deep bass, or some combination thereof.

Therefore, a product that measures well at some place or with a good specification sheet may not perform to your expectations in your listening space or may not integrate seamlessly with the existing components in your system. Moreover, the calibrated measuring microphones used for recording don’t hear the same way our ear does notwithstanding that, the sound produced is relative to the listener. We should therefore make choices as to how they sound to us and in our room and with our system, and not how they measure or what someone else has recommended. You will have to eat your pudding first to know how it tastes, and I may not like how it tastes even if you do, because as they say, one man’s meat is another man’s poison.

Measurements are for the designers to figure out what works and would sell at a given price point. Therefore, all audio products would have a particular ‘voice’ as designed, and this voice whether good, better, or best is a subjective judgment call made by the designer. However, this ‘voice’ will change, even dramatically depending on the loudspeaker placement, position of the listening seat, listening environment, and its companion components. Therefore, I believe the gap narrows significantly where a modest system that is well set up will sound better than a more expensive system that does not follow the 8 rules for proper setup to mitigate the variables mentioned. If all things are equal, a more expensive component in the same reference system may sound ‘better’ notwithstanding the prevailing law of diminishing marginal returns. Therefore, we will have to use our ears to figure out the cost-benefit and make judgments based on the product's perceived value proposition in our listening environment. Cosmetics and features may not necessarily be high on the list but how it sounds is usually the priority for the discerning audiophile. Hence, they say, the proof is always in the pudding.

8 - In precision we trust – electronic components should be mirror imaged and matched in a stereo system

The discrete circuitry topology discussed previously should further include physically separating both left and right channels to reduce ‘crosstalk’ that would otherwise cause intermodulation distortion. This is best achieved by using 2 separate chassis (mono-blocks) or 2 circuit boards inside one chassis as the second-choice alternative, or one circuit board with the layout mirror-image for the left and right channels inside a single chassis as the third-choice alternative. The components used for the circuitry are measured for identical values so that the output voltage is precisely matched.

Mirror image Layout

Therefore, electronic parts are selected based on their sonic qualities (to maintain the integrity of an audio signal) and their precise electrical values of less than 1% tolerance between each stereo channel. Polypropylene signal capacitors, wire wound resistors, and air core inductors are usually chosen for their qualities to reduce ‘hysteresis’. The parts are selected in matched pairs in absolute values that are mirror-image in a stereo configuration. Higher voltage values are also favored as overkill because they believe the respective components will not be operating close to their limits (saturate) and will therefore not overheat (causing distortions) or suffer from a lower life expectancy. Teflon circuit boards with double-sided heavy gold-plated solder tracks are desired for signal flow integrity. Hardwiring (point-to-point soldered) to switches or control knobs, and tube pins directly instead of running them through the tracks on the printed circuit board is favored which improves signal integrity. The inputs and outputs connectors should be of high quality with heavy gold plating for the line-level signal or pure heavy copper for the loudspeaker binding posts in the case of amplifiers and loudspeakers. In summary, attention to detail as discussed is an essential factor of a good high-end audio design which is why we trust precision.

9 - Less is more – the shorter the signal path the better the performance

This principle presupposes that every component contributes to distortion and that more components added to the signal path or audio chain will add to the distortion level. That means each member component in your system has the potential to alter the sound and or raise the noise floor. The more components you add, the worse the contribution because the signal-to-noise ratio correspondingly increases. This is because any conduit in the signal path influences the signal notwithstanding the gain structure and equalization of the circuitry which is why we say every audio component has its own ‘noise’ that adds to the signal flowing. This means lesser components in the audio chain and fewer parts (simplest) in the signal path for each component. For example, point-to-point wiring is better for the audio signal than going through a circuit board. An integrated amplifier may perhaps be better than a separate pre and power amplifier (I am not saying an integrated amplifier is better sounding than separates, but only if they are of the same quality and power).

An active loudspeaker not requiring a separate power amplifier may be preferred, however, the vibration from the loudspeaker affects the electronics, and this may be an exception to the rule where the cure may sometimes be worse than the disease. No cable or less cable is better or shorter cables are better than longer ones. A loudspeaker cable terminated with pins, plugs, or spades is worse than the bare wire connected straight to the terminals, a minimum number of joints, dry joint/crimped instead of soldered is preferred. An equalizer, DSP room correction, tone controls, and loudness switch introduced to influence the audio signal will invariably add to the noise floor and alter the default tonal balance. I must reiterate in this regard if all things remain equal (ceteris paribus) the principle of less is more is taken into consideration.


10 - Every nuance matters – in high-resolution audio, any difference counts

 The law of diminishing marginal returns for utility implies that we must pay much more just to get a little benefit. This is because audiophile products are designed and made with high-resolution sound reproduction in mind, it is, therefore, conceivable that a narrow sound performance gap exists between similar product categories. Manufacturers may also choose to use state-of-the-art components in their circuitry and high-quality or exotic materials for their designs which knocks the price way up with no further sound benefit. Moreover, most of their designs are assembled by hand, making them labor-intensive and time-consuming, or with designer styling inflating the price for the audiophile. Because of the relatively low consumer demand for audiophile products in this niche market, the manufacturing cost is higher due to the smaller production size in this regard.

 It is also probable at the other extreme, that a smart design, with functional materials, improved technology, and ingenuity, can keep the cost of production low. Moreover, selling direct or through established online merchants can keep the retail price even lower re-directing the savings to the audiophile. Today, there are more affordable audio products readily available for this privileged generation. Nevertheless, with all this in mind, we should expect to pay at least 3 times as much (new price) in my reckoning from the previous spending on the same product type and group if we want to hear a tangible difference in sound quality. This I believe is the only way to circumvent the law of diminishing returns, otherwise don’t bother. So, when upgrading, consider this final principle that may apply across the board to all product categories. In high-resolution audio, every nuance matters, contributing to the overall listening experience. However, to realize this, your system must be finely tuned, and you should be schooled in the art of critical listening (Read my article Scaffold/Rule 1 - Listening and hearing) to hear these little nuances that can mean much in a finely tuned audio system.

What an Audiophile Purist would also say:

Besides those guiding principles, the audiophile purist among us may also consider the following:


Conclusion

 These Audiophile Principles should be considered when setting up an audio system and to influence each purchasing decision. Please note that these principles should be read in the context of music reproduction to include all audio components with other ancillary equipment. Notwithstanding this, the law of diminishing marginal returns will often rear its ugly head. 

Questions?

Email

dennis@hearasia.com

 to request more information or to provide feedback