The original text is written in Chinese and the English version is translated by Google.
If in doubt, please refer to the original text.

The quality of audio equipment

No matter how much budget you have prepared for the purchase of audio equipment, it is still a limited resource. Every audio enthusiast hopes to choose some high-quality equipment from limited resources to form an audio system. What kind of equipment can be called high quality? In fact, this question has an answer in everyone's mind. All answers are also closely related to everyone's expectations of these devices. Therefore, the so-called high quality means that the benevolent see benevolence, and the wise see benevolence. If you ask me, as an enthusiast and designer, what do I expect of this high quality, I think it should be able to maintain its own attributes and assume the function of playing music. If you feel the same about this, I hope the following content will help you.

High-end audio has experienced decades of vigorous development, and audio circuits have become very mature. Various circuits used in audio equipment are carefully calculated by designers. They should all be able to produce high-quality products. However, when we browse the audio market, we can still see devices with different sound quality. Moreover, even with the same circuit, the sound quality of equipment produced by different companies may vary greatly. This makes us exhausted when buying equipment.

In addition to technology, there may be many factors that affect the sound quality of audio equipment. For example, in terms of components, in order to meet the needs of some customers, there may be some unreasonable choices, such as gold-plated components. For this problem, we cannot blame the designer or product manufacturer. As a market, it always has to meet the needs of various customers. From the customer's point of view, shiny things are very beautiful and dazzling. However, from a process point of view, the bimetal structure of gold plating has nonlinear characteristics under certain working conditions. Therefore, it is necessary to use gold plating appropriately. However, we still see this kind of gold plating applied to resistor pins (resistors are typical linear components). In fact, this is nothing worthy of our criticism, but whether these resistors are used in the right place, this is a problem that needs our attention.

When a user opens an expensive device and finds that it contains some of the cheapest components, he may wonder whether the device is of high quality. It doesn't matter whether you will use this vision to check the materials of audio equipment, or whether you will use the number of components inside the equipment to measure its value. Various customer identification factors will still affect the trend of market products.

Although the products faced by consumers have excellent technical indicators, they still may not find products that meet their requirements. Many times, they feel helpless, especially audio enthusiasts. Electronic technology is an area that is difficult for people outside the industry to understand. Audio enthusiasts usually have to rely on product features to purchase equipment. Experienced consumers will use their knowledge to evaluate the applicability of these functions, rather than relying on these introductions to guide their choices. Yes, you are a savvy consumer, because you will analyze various things in a scientific way. We don't need to think deeply about the term "science", here it is just a synonym for systematic analysis of physical phenomena. No matter how advanced electronic technology is, it will not deviate from the basis of physical principles. Therefore, through some understanding of middle school physics, we can also find what we want through a large amount of collectible materials, and can reduce a lot of trouble.

Technical indicators can help us distinguish the pros and cons of commodities, but in some cases they are not feasible. The digital music format has been increased from 16-bit/44.1KHz to 24-bit/96kHz. This is a leap in technical indicators that has brought us a more perfect daydream. If you compare these numbers with 24, 96 and 16, 44.1, it is easy to understand how much improvement the new format can bring, which seems to be taken for granted. However, for the actual finished product, due to the structural quality of the data carrier, it is still not ideal in actual use (see the sound characteristics of the audio components of various encoding systems). Usually, each product introduction is introduced to people from an ideal perspective. To know how big the gap between the actual product and the ideal product is, we still need to collect data for analysis.

The audio circuit is very mature. Product manufacturers only need to use the same circuit to produce a variety of different sound quality equipment, and even they can still retain almost the same technical indicators. In today's technological progress, the main difference in the quality of audio equipment may sometimes lie in whether it has been precisely adjusted. In order to create an actual working environment for the equipment to make this adjustment, more manpower and environmental resources must be invested for this, and the price of the equipment rises accordingly. Under modern and advanced automated production conditions, the manufacturing cost of high-quality equipment will depend on manpower and technology, rather than material prices. If it is based on technical rationality, some components on the circuit also need to use the most suitable material type, regardless of the price. On the other hand, the meaning of "high quality" includes high reliability, and reliability decreases as the number of parts increases. The above situation makes it more difficult for us to see the actual working quality of the equipment from the internal and external structure.

Compared with other electronic products, audio equipment has its special side. If we consider the attributes of the equipment itself, high-quality equipment cannot be found with the naked eye. Current technical indicators may not fully and effectively reflect the sound quality of the device. In order to find suitable products in a highly commercialized market, we still need to use some knowledge to evaluate the negative effects of device characteristics and analyze the applicability of product characteristics. This is the only way we can reduce our frustration when choosing equipment.

Below, I will give an example of evaluating the impact of equipment functions. This is to analyze the various effects of a certain function to evaluate the positive or negative changes in the overall quality of the audio system after the function is introduced.

There is a functional product on the market called "Up-Sampling". This function can change the sound quality. After adding this feature, whether the change in sound quality is positive or negative depends on the overall quality of the audio system.

When playing a CD, this function can double the original sampling frequency of the CD from 44.1kHz to a higher 88.2kHz, or 176.4kHz, etc. This is equivalent to increasing the number of supersampling of the digital filter. In this way, the cut-off frequency of the analog filter can be raised to a higher position to allow the analog signal to obtain better frequency response and phase characteristics (we are difficult to determine whether this is the intention of the producer, but apart from this, we have not seen other significance). Under ideal circumstances, this feature can bring the above benefits. However, in actual operation, it may not be able to achieve this goal.

To analyze the benefits of this function, we must start with the role of digital filters. The CD players we currently use all have a built-in digital filter. The original intention of adding this digital filter is to push the frequency band of D/A conversion noise to a position far away from the audio band. The analog signal after D/A conversion only needs to use a simple filter to fully filter out the residual digital components and obtain the same frequency response and phase characteristics as a straight line. If the digital filter chip has no operating speed limit, in theory, we can achieve this goal by continuously increasing the number of oversampling of the digital filter. In fact, the 8x digital filter commonly used in CD players has achieved this effect, and the higher number of supersampling has almost no greater meaning. Therefore, the maximum operating speed of digital filters and D/A chips commonly used at present stays in the region corresponding to 8 times oversampling.

If nothing special, the advanced CD players produced in post-production are all equipped with an 8x digital filter and a D/A chip that can cooperate with 8x oversampling. These chips configured in CD players usually use 8 times oversampling to obtain the maximum functional benefit, that is, the working speed has reached the limit allowed by the chip. In this case, if the CD player uses the Up-sampling function, the number of over-sampling of the digital filter must be reduced accordingly to adapt to the maximum speed limit of the D/A chip. In this way, the D/A conversion noise still stays in the frequency band position before the Up-sampling. That is, the cutoff frequency point of the analog filter can only be maintained at the original position. (We only need to check whether the clock frequency of the D/A chip is increased by multiples accordingly, and then we can know whether the conversion noise has been pushed up to a higher frequency band.)

Up-sampling cannot improve the accuracy of the analog signal after D/A conversion. The CD format standard is 16-bit quantization and 44.1 KHz sampling. No matter what method we use or use more digits and higher sampling frequency to process the data before CD production, we still need to use a standard format to create the CD in the end, otherwise it will not be able to be played on a CD player. The number of bits and sampling frequency of A/D conversion determine the accuracy of the quantized analog signal. When the CD is produced, its format has been determined, and the highest accuracy it can achieve is also determined. The previously improved accuracy is now in vain. Similarly, after the CD is made, any form of data conversion cannot improve the accuracy of the analog signal, and the accuracy may be reduced due to errors in the data conversion process.

Using multi-sampling instead of digital filters, it may be that some devices have the same function but have different names from other devices to show their differences. Up-sampling or digital filter oversampling, these two processing methods have the same result, that is, 8 times sampling is equivalent to 8x digital filter. Therefore, when other aspects remain the same, they are just different names. In an ideal state, the advantage of this type of CD player with different names is that you can freely choose the equivalent of 2x, 4x or 8x digital filters to meet everyone's preferences. However, for split CD players, this feature will encounter another problem.

Usually, the CD transport of a split player uses 44.1k sampled data to encode and output digital signals. Now, the Up-sampled CD transport needs to output data with a higher sampling code than 44.1k. That is to double the density of the data. As the data density increases, the period of each digit will be shortened accordingly. As a result, the data stream jitter caused by the data transmission process will also increase relative to the digit period (although the absolute jitter value in the unit of time has not increased, but the relative jitter compared to the digit period has been increased. It is increased by multiples). This will make the data more prone to errors (in the digital signal transmission, the jitter of the data stream leads to the data error is also caused by the time domain deviation of the bit clock and the digital. For its mechanism, please refer to the relevant content of CD disk data jitter). Therefore, a split CD player that uses the Up-sampling function needs to use a higher-quality connecting cable to transmit digital signals to ensure that the relative jitter will not increase due to the shortening of the digital cycle.

For a split CD player, if the data stream jitter value during the data transmission process cannot be reduced accordingly, then the Up-sampling function will be equivalent to a low-pass filter to a certain extent. As the data stream jitter increases (data errors will increase as a result *), this filter will gradually narrow the music signal band from the high end of the audio segment (some players have multiple filter bands for users to manually select). In this way, If the audio system can handle the bandwidth in real time is narrower, the Up-sampling function can bring about the effect of reducing the music background noise ( For the real-time processing of the music signal, and reducing the music background noise, please refer to the particularity of the audio system).

* Based on the data transmission mechanism, the jitter of the main clock required for D/A decoding will be very closely related to the jitter of the data stream. When the time domain factor is combined, the jitter of the D/A decoding master clock will be equivalent to an error in some audio data.

From the above analysis, we can know that only the Up-sampling function, when it reaches the ideal level (no data error occurs), the sound quality of music playback should not change. If there is a change, it is of course that some of the data has errors. Can the analog signal changes caused by these errors compensate for the defects of your audio system? Or it can be said, do you need to narrow the bandwidth of the music signal to make it easier for your audio system to handle it? When choosing this feature, we need to add this consideration.

As for reducing music background noise, it only refers to equipment other than CD players. The background noise of the player itself has increased. Which one is more important between the two? It may vary from system to system, and it is difficult to predict.

Above, we only use Up-sampling to analyze the functional benefits of the equipment, not to evaluate the quality of the equipment. In fact, there are still some equipment we are still using that require other equipment to do some mutual compensation to have a good sound effect. This is another reason why we can see that there are still various quality equipment in the audio market. The emergence of various quality equipment, it was necessary for their existence in the historical period at that time. Similarly, some old equipment can still be circulated in the second-hand market, which of course is that they still have the value of being able to complement other equipment.

Let me use audio CD to give another example. Audio enthusiasts may own some audio CDs produced using various technologies. Have you noticed how the sound effects of these CDs are related to the music details (for music details, please refer to the system settings for sound field restoration). Have you found that compared with standard audio CDs, some CDs with less distortion of the sound state (the emotional state of the sound) have a narrower frequency range. Or have you noticed that CDs produced with a certain technology rarely have classical music performed by large orchestras and have a wide frequency range. This is because a signal with a narrower frequency domain will be easier to obtain accurate processing on the audio system, and its sound state will be less distorted.

The structural quality of the disc is the root cause that affects the sound quality of audio CDs. XRCD disc data has less jitter and can have better sound quality. This is closely related to the fact that its production directly uses the master disc to press CD discs ( Usually, the production of CD discs is to first make multiple sub-discs with the master disc, and then use the sub-discs to make the finished CD discs ). Although, its relatively narrow frequency domain (which is not related to the compaction process of CD discs) also makes it lacking and leaves some regrets. However, there are still many audio equipment that cannot effectively process broadband music signals. It is not necessarily a bad thing for these equipment to use a sound source with a narrow frequency domain. Therefore, XRCD combined with various improvement measures can still bring different degrees of sound quality improvement when playing on a large part of the audio system.

HDCD is very different with XRCD.

HDCD uses more digits for A/D quantization, and the resulting more data is also placed in the position allowed by the CD format standard. Therefore, after decoding, HDCD can have more data expressing music information, and CD playback should have better sound quality. However, HDCD can produce these characteristics of high quality, because the quality of its physical structure has not been improved accordingly, so this more data will cause more background noise when the data is wrong, causing its music playback to produce more background noise (this Related to the functional quality of the audio system. High-accuracy audio systems can play HDCD with better sound quality). Audio enthusiasts have already experienced this.

The above audio CD examples can illustrate that there is no absolute measure of whether the equipment is good or bad. In some cases, in order to adapt to the specific environment, we need to make some coordinated concessions accordingly. Similarly, there are similar situations in hardware. So, what kind of audio system is of high quality? In a broad sense, it is a high-quality audio system with excellent linearity after mutual compensation. As far as a single piece of equipment is concerned, if it has high accuracy and does not need to rely on other equipment for compensating coordination, it is a high-quality equipment. Using this type of equipment to form a audio system will be able to accurately play various types of music.

There are differences between the above two types of high quality. The mutual compensation method needs to sacrifice the original music details ( Reduced the density of micro dynamics ) as a price in exchange for the improvement of linearity. Subjectively, this may make us have some regrets. However, we still need to look at the actual product from an objective angle that is not absolutely accurate. In fact, the audio system still needs some coordination to achieve the best linearity. Typical examples of this are speakers and speaker cables.

In general, we can comprehend the quality of the audio system in this way. As long as it can accurately play the type of music you want, it is a high-quality audio system for you. From the above software and hardware examples, we can also see such a problem. The high quality of the audio system still requires the cooperation of both signal source and signal processing. In some cases, unilateral high quality may also have an adverse effect on the entire system. We should not ignore this when choosing equipment.

Specifically, for people outside the electronic technology industry, what means do we have to evaluate the accuracy of equipment? Here I will introduce to you a simple and effective method to judge the accuracy of audio equipment. This method does not need to pay attention to the technical indicators of the equipment, which is very practical for audiophiles. I use the background noise of the music as an indicator when I'm evaluating a piece of equipment or cable for accuracy and whether it will benefit my system long-term.

Excellent equipment, it should have less distortion. Any distortion of the equipment will alter the electrical signal of the music. That is to say, when the electrical signal passes through a equipment that contains distortion, the signal will have energy that should not appear at some point in time. According to the fact that the sound that should not appear is noise, and the energy that should not appear at a certain point in time is the energy that should make noise. We can judge the accuracy of the device from the strength or weakness of the background noise of the music, so as to confirm whether it can bring long-term benefits to your system.

The so-called long-term benefits mean that when you want to replace other equipment in the system, the current benefits will not disappear. With high-quality equipment, many times you can get the sound effect you want through the system settings without relying on the mutual compensation between the equipment. Because the sound effect obtained from mutual compensation will make you need to replace a certain device in the system in the future, you also need to replace another equipment in order to maintain your current satisfactory sound effect.

I have an experience, which can also be used as a reference for you who are purchasing equipment. Regarding audio, the high or low quality of equipment will be reflected in how much detail the sound can express, and will be displayed in the micro-dynamic density of the sound. In addition to details, the improvement of sound effects can often be obtained through system settings.

We use audio equipment to play music and always hope that it can have good sound quality. Even, we hope it has the sound effect of the famous music hall. It is not too much to have such a request. Some of the music we are playing is recorded in these concert halls. To achieve this desire, we need to understand all aspects related to audio. Yes, we are traveling through a market that is open to all customers. Of course, there will be a variety of products that meet the needs of everyone. Knowing more about physics, even some mechanics, even mechanical principles, etc., can help you find suitable audio equipment in this market. Compared with other products, choosing an audio system that you are satisfied with may take more thought. Be able to let music accompany you, and in the end you will think it is worth it. Here, I wish you can choose satisfactory audio equipment, and eager to share your experience.

Created by Chen
Last revision date: Sep-2023

All rights reserved by Chen Audio Laboratory, Inc.
If any or all of its contents are reproduced, this webpage must be cited as its source.