Warm tube sound lurk. Posts from This Journal by “Tube Sound” Tag. What is "quality sound"

"at the request of the workers, I decided to highlight the topic of warm tube sound a little. I will not give graphs and other figures, all this will already look like a scientific work, and not a review article.

This concept originated a very long time ago, back in the days of the inception of semiconductors. Since the transistors in those days were, to put it mildly, not very high quality, and circuits on germanium devices had just begun to appear, then, as a matter of course, a subject was formed. Plus, let's add here an incomplete understanding of the operation of the transistor, the lack of circuits and the high cost of the components themselves. Radio amateurs used transistors in the same way as lamps, but as you understand nothing good came of it, either the circuit did not work or worked very disgusting. Also, do not forget about the powerful output transistors, if anyone remembers, there were such P4E rare shit. Later, P213 and P214 appeared, which slightly improved the situation. In the preliminary stages, transistors MP14 were used, and later MP40-41-42. There were also low-noise devices in this series, if I am not mistaken, there were P28 and MP39B, which were a terrible shortage, and therefore, if they managed to get it, they put it very first, which, in principle, is very correct. And do not forget that the transistors of that time had a low gain, which led to an increase in the number of stages and the complexity of the circuit.

And as one of the factors, you can add the psychological aspect of the polarity of food. As you know, the first of the transistors had P-N-P transition, which meant that the scheme was turned upside down. How so, plus on the mass ?! radio amateurs were indignant and continued to use simple and reliable circuits on radio tubes.

But progress did not stand still, transistors began to become cheaper, it became fashionable to have a small radio receiver on batteries, and the absence of warming up, and instant switching on and efficiency are also not unimportant factors.

In fact, the concept of Lapovy Sound has survived to this day. Whereas at the beginning of the history of transistors there were no trite circuits, but there were details of mediocre quality, now this is a phenomenon of redundancy, fashion and promotion of the concept of "Vintage"

Although, in fact, the sound of tube equipment differs from that on semiconductors. A simple example, if the average music lover turns on a vintage radio on vacuum equipment, he will be surprised. Yes, indeed, it does not sound like a transistor, somehow not familiar, somehow in a special way. After a while, the enthusiasm fades away and an understanding of the situation comes. It is known that transistor amplifiers have pronounced not even harmonics, while the tube is the opposite: even. Thus, tube amplifiers, as it were, mask an initially bad recording, give it a tube color, so to speak. No, after all, high-quality semiconductor equipment significantly surpasses the parameters of the lamp counterparts. So what's the deal? Let's try to understand this interesting phenomenon of warm tube sound. So:

OOS. The absence of deep, or generally as such, negative feedbacks in tube circuits. Of course, there is a rational grain in this, because lamps are characterized by higher linear characteristics than semiconductors. It is for this that OOS is introduced. But let's not bend our hearts, often linear circuits without OOS can provide much less intermodulation distortion, which we all do not like so much.

And here we have homemade people who do not know what they are doing on their own trying to make the coolest tube amplifier. There is not enough knowledge to build a more or less decent apparatus, and therefore schemes taken from radio hooligans are used, which the latter use as modulators for their AM transmitters. The circuit of such an amplifier is very simple, usually only a couple of lamps are used: 6N2P and 6P14P, not many details are needed for the lamps themselves. And now the diagram is assembled, curly mounted installation and an ugly heap of rubbish lies on the table. If the circuit started working from the first turn on (why not work there?), Then the magical selection of lamps in certain cascades begins, and you can often see lamps in the pre-amplifier that were not designed for this at all, the author personally saw how the 6P13S lamp was used in the first stage. In advanced cases, the use of finger lamps is in no way allowed, but only with an octal base, because they are older, larger, more lamp-like and warmer. Most often it is a 6N8S double triode and a beloved legend by all, the 6P3S pentode. And all that remains of the original sound must be fed by all means to speakers the size of a three-door cabinet, with one single full-range speaker. And all this abomination is fed to the speaker system through:

Transformer Output. A funny thing really. Has a pumped and very important skill: " Slice treble which arose due to self-excitation as a result of the installation described above»Has a large weight and dimensions, comparable to a power transformer. A high-quality output transformer costs approximately the same as a Russian-made car of average wear. But there is no money for such purchases, and therefore TVZ from tube TVs and radiogram is used. In the end, our hero understands that this transformer already "not enough" and needs to be replaced. But what for? Of course, on a power transformer, where its primary winding is connected to the anode of the lamp, and the filament to the speakers. Having thus received "oversaturated bass" the transformer is rewound countless times. And it doesn't matter that not all the plates are assembled back into the package, and it doesn't matter that all this disgrace begins to ring and plague the ear with disgusting rattling to the beat of the music. But, nevertheless, the transformer is very well suited for matching the high output impedance of tube stages with low impedance loads. And at the very beginning of the era of germanium semiconductors, transformers were also used in transistor circuits.

Next is a new discipline with power transformers. Initially, they are removed from obsolete equipment, and are used in their designs without any alteration. But one day, a young lover of Warm Lamp, collects the second channel and here the problems begin. There is not enough anode current and under the doubled load the voltage sags very little, which does not have the best effect on the sound quality. The filament voltage also sags, and the lamps begin to work not in mode. (By the way, from excessive and insufficient voltage, both in the anode circuit and in the heating circuit, the lamp wears out very quickly, although it continues to work.) In this case, our heroes either rewind the transformers, which does not significantly help, because the transformer is higher it will not give up the set power, or two power transformers are installed, which, in combination with two sound ones, make the unit stationary and not movable.

But it also happens, the author read an article about how a person assembled a tube amplifier according to a very good circuit, but it did not work with the power supply. He did not have the funds to buy a two kilowatt transformer, and the size and weight characteristics crossed all reasonable boundaries. And then it dawned on the person: "Pulse PSU" Despite all the prejudices and forum protests, the power supply was built. And naturally he gave excellent results, no voltage drop under load and, despite the shit that we call electricity in sockets. But, in the end, having discovered high-quality semiconductor ULF with lamps, it was done.

Connoisseurs of Warm Lamp Sound are recognized only as wall-mounted installation. More than once, the author noticed statements that fiberglass spoils the sound, I don't know about you, but I can't even imagine this. Printed montage is evil, it does not breathe and does not have a soul, and it is also desirable to solder copper as it was implemented in battery tube radios. Although some logic can be traced here, soldering with ordinary solder has a large transition resistance, which is tens, and sometimes hundreds, times higher than the resistance of the printed conductor. So, in fact, brazing with copper as such is not, it is rather welding, alloying metals.

So, what can we take away from our conversation. Lamps are certainly good, they glow beautifully in the dark, they will warm you with real, physical, warmth and, in the end, they are fashionable, cool and, in our time, unusual. In no case will I dissuade you from building a lamp apparatus, on the contrary, it is very interesting and informative. Just remember, there is HIGH VOLTAGE on the anodes of the lamps! It happens that it is much higher than in the mains, do not forget to discharge the capacitors in the anode voltage circuit. Also, do not forget about the temperature of the vacuum devices, it is high enough to get burned. But from a practical point of view, for daily home listening, I don't think it is advisable.

About "warm" tube sound June 27th, 2017

What is tube sound? There are many myths about him, and fierce disputes, and honest attempts to figure it out. I will try to tell you as simply as possible, so that non-engineers understand what is at stake. And if quite figuratively, then the sound of the tube is like a film photograph. On the one hand, it is just a certain stage in the development of technology, where each subsequent stage, as a rule, is more perfect than the previous one. For example, it is difficult for a digital photographer to imagine the problem of calculating the amount of film needed to shoot. One cassette contained films of only 36 frames. Ten cassettes is already a bag, but only 360 shots, and until the moment of development, you do not know what you have done. And the printing process itself was a non-trivial problem. Digital, on the other hand, radically simplified everything and technologically gave the photographer opportunities that even professionals could only dream of in the film era. But on the other hand, for some reason, “filters” are very popular for giving a “digital” picture a “film” look. What's the matter here? Why and why do people spoil technically “better” shots?

The point here is that a person (for now) is an analog system, full of distortions and conventions, however, like the rest of the world around us. If we feel something "discrete", "symmetrical" and "refined", then we subconsciously do not "believe" such. For us, it becomes "imitation" or "inanimate." And it doesn't matter if we are talking about glossy "club mucks", digital photos or transistor sound. We can hardly express the emerging feeling, but we well feel the "wrong" of the right one. And therefore, for example, with age, you begin to appreciate the female beauty a la Playboy of the 60s much more than similar options of the 2000s (if only because you already know exactly how it really is). The same thing happens with sound, with color, with taste. Everywhere, noisy and incorrect, most subconsciously like more than refined. We are made like that.

But back to the "tube" sound. Amplifiers built on lamps, during their operation, objectively introduce into the original signal, significantly "more" measurable distortions, consume more electricity, heat up more, are less powerful, more difficult to operate and require regular replacement (adjustment) of lamps. But at the same time, compared to "transistors", the tube sound is perceived better. Why?
The answer is simple "tube" sound: even distorted, it is more reminiscent of natural natural, recognizable by "our" our senses, and the rest can be easily corrected by our adaptive perception. At the same time, how much, from the “rational” point of view, the tube sound is “worse”, and its “warmth” is fictional, you can see here:

The author explains everything theoretically very well and correctly. Competently and convincingly. But this has about the same relation to real life as mathematics. On the one hand, she is the queen of sciences, and on the other, Gödel's theorem on the incompleteness and impossibility of description with the help of mathematics of sensory perception.

So how does a tube amplifier work? Why do they continue to "rush" with him, although, from a technological point of view, he certainly loses to the "transistor" in almost everything?

Firstly, the "lamp" is able to squeeze the entire dynamic range of the signal into certain frames, without "clipping". Probably everyone heard how strange the "cymbals" sound through a transistor amplifier? Why is that? Whatever the range of the transistor amplifier, the original signal will still be wider. Therefore, "transistors", everything that does not fit into the range of the amplifier, is cut off and then they work with a "castrated" signal, which is why the unnatural sound of "cymbals" or plucked ones in modern technology arises. In such a situation, a "lamp" behaves fundamentally differently, and although it usually has a very narrow range, it is capable of "squeezing" (carrying out a kind of analog compression of an audio signal) the entire range into the existing framework. The result is a "dense" "juicy" sound, which by its nature resembles a "real" one, although, in a mathematical sense, it is strongly distorted.

Second, the "lamp" does not disassemble the whole into its components, so that, after amplification, it can be reassembled "approximately" as it was. Instead, she works with the signal "in general"... Yes, objectively, the “lamp” distorts the signal more strongly, but at the same time preserving its fundamental nature, whereas with transistor amplification at the output, the signal has a different (in terms of “harmonics”) “nature”. Therefore, although in a mathematical sense, the "transistor" signal is closer to the original, but our senses will feel its greater "distortion"

Third, we live in a digital, discrete world, but our senses are still "analog" and operate with "continuous" signals. A discrete signal from a file, after being converted to

Their signal contains a small amount of harmonics (dominated by the second, third and fourth), due to which there is a softer sound, or as it is often called - "warm", "tube".

A number of authors believe that the reason for the "transistor" sound is not the transistor itself, but negative feedback, which is characteristic of the circuitry of transistor amplifiers. This argument is very controversial, since a significant part of tube amplifiers (and almost all industrial amplifiers) also have OOS.

Strictly speaking, the adherents of the "tube sound" adhere to different points of view on this topic: scientific and esoteric. Proponents of the scientific point of view argue their arguments by the physical features of signal amplification by vacuum and semiconductor devices. The proponents of the esoteric point of view, as a rule, ignore the physical features of amplifying devices, and argue the advantages of "tube sound" by appealing to the auditory experience, musical preferences.

Having experienced an unprecedented rise in popularity in the 90s - 2000s, "tube sound" today is going through hard times and its future is very vague.

Scientific rationale and criticism

Recognizing the fact of obsolescence of electronic tubes as amplifying devices, large weight and size characteristics and low energy efficiency of tube devices, supporters of "tube sound" usually make the following arguments in favor of the superiority of amplifiers for electronic tubes:

Vacuum tubes, especially triodes, have a very wide linear section of the I - V characteristic, which makes it possible to abandon negative AC feedback or reduce its depth. Transistors, especially bipolar ones, have greater nonlinearity, which is why they are most often used in audio equipment with negative feedback(DFB) or with local DFB, covering one stage, but, as a rule, with a general DFB, covering the entire amplifier.
The CVC of electronic tubes is practically independent of temperature the environment(since the temperature of the heated cathode is significantly higher), therefore, they do not need deep OOS for direct current to stabilize the cascade mode.
The presence of OOS in the amplifier leads to distortion of the dynamic characteristics of the signals, which is especially noticeable when playing percussion and string instruments. In this regard, tube amplifiers, which are usually built without feedback, have advantages.
Vacuum tubes, especially pentodes (beam tetrodes), are characterized by very high gains, which makes it possible to build amplifiers with a small number of stages (2 - 3), which reduces the overall level of distortion.
In tube amplifiers, an output transformer is almost always used, the use of which makes it possible to optimally match the final stage with the load and thereby reduce the level of distortion introduced by the final stage. The exception is relatively high impedance tube headphone amplifiers that do not require an output transformer.
Less level of intermodulation distortion. From the perspective of tube sound advocates, intermodulation distortion is a key disadvantage of transistor amplifiers.

Opponents of tube sound provide counterarguments to each argument:

Transistors do not have such a long linear section of the I - V characteristic, but they can operate at lower voltage amplitudes than triodes, which eliminates the indicated disadvantage of transistors.
The temperature regime can also be stabilized for the transistor stage using a cooling system.
There is also no fundamental need for OOS in transistor amplifiers. It's just that the circuitry of tube stages was developed back in the 20s - 30s, when the OOS theory was still insufficiently developed. Transistor circuitry arose later and all knowledge of the OOS theory was already applied in it. However, transistor (especially on field-effect transistors) cascades without OOS are quite efficient.
Pentodes and beam tetrodes are characterized by high gain, but their linearity is much worse than that of transistors. Therefore, fans of "tube sound" rarely use multigrid lamps in their designs or use them in triode switching. And triodes have significantly lower gains than transistors.
There are no fundamental restrictions on the use of the output transformer in transistor amplifiers. Moreover, transistor amplifiers with output transformers are made by amateurs and are mass-produced.
The theory of intermodulation distortion appeared after the end of the era of tube sound technology and is currently being actively developed specifically for transistor amplifiers. For tube amplifiers, this issue has hardly been studied. Therefore, it is almost impossible to compare tube and transistor amplifiers by this criterion.

Additionally, the following disadvantages of vacuum tube amplifiers are indicated:

Main streams and branches

At the end of the first decade of the 21st century, "tube sound" can be regarded as a well-known phenomenon. In the world, tube amplifiers are produced using both classical and new circuitry, new literature on tube circuitry is being published, there are Internet resources dedicated to this topic. Nevertheless, the environment of fans of “tube sound” is not homogeneous, just as the types of tube sound technology are not homogeneous. Therefore, a number of main ideological currents and branches from them should be singled out here.

Commercial realizations

In the 90s of the 20th century, in various countries, primarily in Japan, the USA, Germany and Russia, and later in Taiwan and China, a number of companies were created specializing in the production of tube sound-technical apparatus and acoustic systems for it. These products are produced in various editions and have a wide range of prices, from low-cost solutions from Chinese manufacturers (in particular, under the Music Angel trademark) to piece products costing hundreds of thousands of dollars, for example, Ongaku from AudioNote (Japan). To complete such products with electronic lamps, the production facilities of a number of factories, including the Svetlana Production Association, were re-launched. The development of new types of vacuum devices was started, for example the SV572 lamp. The global economic crisis of 2008 greatly reduced the demand for such super-expensive products. In addition, a number of manufacturers of elite semiconductor equipment have launched on the market fundamentally new products that are significantly superior in sound quality to tube sound, and the consumers themselves, having faced in fact with “tube sound” and realizing that, by and large, nothing outstanding in it began to lose to interest to him. As a result, many manufacturers of new tube equipment went bankrupt or repurposed. The production of electronic tubes also fell into final decline. Their new types were never mass-produced. The future of this industry is very uncertain. It is quite possible that after a certain period of time another surge of interest in the "tube sound" will come, but it is most likely that this will not happen, since the interest of the 90s - 2000s was, for the most part, fueled by a generation of people who still caught the "lamp era". The scarcity of vintage components also calls into question the possibility of commercial success for tube projects in the future.

Hi-End

Representatives of this direction consider tube amplifying systems as a means of achieving best quality sound reproduction. However, this current is also heterogeneous, and a number of branches can be distinguished in it, differing mainly in the criteria for the quality of sound reproduction. Here it should be understood not numerical values quality indicators, and the very set of these indicators. In particular, a number of designers of sound engineering (for example, Yu.A. Makarov) prioritize such a factor as the reduced rate of rise of the output signal voltage and the value of its lowest cutoff frequency, as well as the output impedance (the so-called dumping factor). Other authors (for example, Japanese: H. Kondo, S. Sakuma) pay more attention to the harmonic composition of the output signal. At the same time, almost all followers of the Hi-End direction agree that the power of the output signal is not a determining factor.

Representatives of this direction mainly develop the circuitry of single-ended output stages, but there are also adherents of push-pull. Nevertheless, in practice, representatives of this direction adhere to the presumption of objective characteristics over subjective assessments. This, in particular, determines the choice of lamps and other components not according to the sound signature, but according to the data of instrumental studies.

Quite often, developments made by representatives of this direction are sold (including from auctions) or are given to order. But, more often than not, these are constructions that their authors implement for themselves and do not plan their commercial success. In the overwhelming majority of cases, after commissioning, the devices are continuously modernized by their authors.

"Warm Sound"

Representatives of this trend a priori do not reject the high fidelity of sound reproduction, but at the same time they believe that the main task of the equipment is to involve in music. This determines the main approach to the construction of equipment by representatives of this direction - the components are selected not only according to technical characteristics, but by "sound". At the same time, the authors often use components, for example, radio tubes, in modes other than those recommended, often exceeding the maximum permissible parameters.

This direction also has a number of branches. Often, representatives of this direction, misunderstanding the physical and psychoacoustic features of the "tube sound", begin to use their lamps in those nodes of audio equipment where the use of lamps either does not affect the passage of audio frequency signals at all (for example, in power stabilizers of filament circuits of other amplifier lamps ), or where the use of lamps is impractical due to the high level of the microphone effect, and their linearity does not play any role (for example, in the input stages of microsignal circuits: RIAA correctors, amplifiers for playing tape recorders). There are also completely absurd solutions, such as the use of tube signal generators to clock digital devices, such as CD players. As a rule, such solutions are offered by technically incompetent authors.

There are also radical trends, whose representatives completely ignore the circuitry aspects of using lamps and other components, putting in the first place the subjective sensations of listening. These persons operate with such pseudo-scientific understandings as "the direction of the guide." Among the representatives of the radical direction, vintage electronic components released in the 1920s and 1930s by such firms as Western Electric, Klangfilm, Telefunken, etc. ". These authors and their fans try to mask their technical incompetence with their “fine ear for music”, “dedication” and other subjective arguments.

Amateur radio designs

Amateurs' tube sound-technical devices are usually created for the purpose of experiment - "touching history" or getting a tube sound - "for a reasonable price." This direction is popular all over the world. It is also important that the amateur construction of a tube amplifier is much simpler from the standpoint of circuitry compared to semiconductor devices, which require a much larger number of elements and accurate calculation of all circuits, which is often a determining factor for a radio amateur. Often, the characteristics of homemade devices are very modest in comparison not only with factory Hi-End tube amplifiers, but also with similar homemade semiconductor amplifiers. Often radio amateurs set themselves the task of creating an original circuit design, without much regard for the sound quality: for example, with the control of the pentode not according to the first, but according to the second grid, or, for example, a circclotron or the use of an electronic light indicator ("magic eye") as an amplifying lamps.

Since the mid-1990s, the Russian radio amateur A.I. Manakov (known in the amateur radio Internet community as Gegen) has published a description of a number of amateur amplifiers based on vacuum tubes, built according to circuits that are very different from the classical ones and have rather high characteristics. Outside the Internet community, these designs were popularized in the book by MV Toropkin "DIY Hi-Fi Amplifier".

In 2005, interest in simple amateur tube designs was fueled by the publication in the Radio magazine of a series of articles by S. N. Komarov devoted to the circuitry of push-pull amplifiers. After this cycle articles, publications in the magazine "Radio" dedicated to tube sound technology have become regular.

It should be noted that interest in homemade tube devices in the mid-2000s sparked a surge in prices for vacuum tubes, transformers, winding wire, vintage speakers and other related products. As a result of this, and also due to the fact that all these products have not been mass-produced for a long time and have become scarce, in the early 2010s, interest in lamp circuitry among radio amateurs fell again. The decline in interest was also facilitated by the fact that it is already practically impossible to create municipal new circuit solutions on electronic tubes. Therefore, many technical forums on "tube sound", popular in the mid-2000s, have either already been abandoned by users, or repurposed to aesthetic-esoteric and commercial-consumer topics, or turned into message boards.

Vintage audio equipment

A number of tube sound lovers prefer only serial vintage equipment, released during the heyday of tube circuitry. Usually this category includes lovers of musical works of past years (30s - 60s of the twentieth century). Their main reasoning is, in general terms, "60s music should be listened to on 60s equipment." Representatives of this direction, as a rule, do not modernize the equipment and limit themselves only to its repair.

Links

The audio portal is one of the largest multidisciplinary resources of this topic
Favorite lamps - the home site of the radio amateur Sergei Komarov, entirely dedicated to lamp sound and radio technology, both historical and modern. Basically - amateur radio designs.
Hi End - in Russian! - Sergey Sergeev's home site.
DIY vintage electronics - independent creation of hi-end tube amplifiers by radio amateurs.
Our audio portal is a Ukrainian resource for tube sound
DIY HiFi & HiEnd - Mikhail Toropkin's home site is one of the oldest in the Russian Internet on this subject
Welcome to ALTOR - home site of Alexander Torres - tube and semiconductor sound engineering.
ClassicAudio - mainly discusses vintage hardware and hobbyist designs on vintage components
Through the asphalt - the site of Anatoly Markovich Likhnitsky. Author's articles in the field of sound engineering. The early ones are scientific and technical, the later ones are aesthetic and esoteric.
AML Forum - Forum of A. M. Likhnitsky. Mostly esoteric in nature.

This article is, as it were, an informal continuation of the previous one: "Analog vs Digital: a fight that never happened." Since the above article received a wide response (and still has it;), I decided to develop this topic further. I will not give any graphs, as in the previous article - as it turned out, for the general reader it is somewhere even harmful. I’ll talk about digital audio in a more casual manner, trying to cover topics that were not well covered in the previous essay.

In what follows, I will use the word "audiophile" in large quantities. Note that this word is used mainly as a diagnosis - this article will be no exception. A person who appreciates high quality sound and understands it, it is usually called a music lover. But audiophilia is an addiction to supposedly "quality" sound, based on myths, legends and, as a rule, the absence personal experience and knowledge.

What is quality sound?

The funny thing in the entire history of disputes over various sound reproduction technologies is that there is simply no exact definition of "quality sound".

Let's start with the fact that one and the same sound can be of high quality for one individual and completely poor quality for another. For example, someone loves bass more and suffers from a lack of it. And someone, on the contrary, likes "strong" tall ones - and if they are "soft", then there is discomfort when listening. Even more interesting, these preferences for certain ranges can change over time, even for the same person. Everything happens because the human ear is a rather subjective instrument for perceiving sound. The ear can "adjust" to the sound, thereby deceiving its owner rather robustly (here one immediately remembers the cables of pure gold made according to the latest nanotechnology).

The "hearing tests" that audiophiles rave about are, in fact, subject to wild errors and generally cannot be seriously considered as reliable evidence of the "bad" or "goodness" of sound. It is impossible to enter the same water twice - it is equally impossible to hear the same sound, even from the same speaker.

Further, any sound reproducing system will distort the original sound a priori. The sound was distorted during the recording, then during processing, and then in the amplification paths and the acoustic system. It cannot in any way be 100% consistent with the original for the simple reason that the ideal recording / reproduction technology does not exist (and it is unlikely that it will ever appear). Moreover: after recording, the sound is distorted deliberately, in order to obtain one or another effect. The number of treatments that sound goes through in modern recording studios is in the tens. As a result, everything turns out beautifully - just like in the picture of a Hollywood movie, which is 99% far from reality. But nevertheless, everything sounds very good (unless, of course, the sound engineer was a layman). Therefore, you should hack into your nose: the sound in the final track is refined, refined. And refined, not with the aim of worsening it, but vice versa.

As a rule, the required sound reproducing system is selected very simply: by sound. You turn on the system and hear the sound that you either like or not. Seeking out "transparency", "warmth", "bulk" is pure audiophilia, which does not lead to anything good in this case. Either you like the sound of the system - it's simple. And interestingly, as the cost of the system increases, the sound usually improves. Is it strange or not? It seems to me - not very much.

Of course, people with increased sound requirements choose the system in more detail. For example, for this case, I have several tracks with me - a couple of auditions - and everything becomes clear. No amplifier has an ideal frequency response, which means that you need to choose the one that sounds the most pleasant (ultimately, it all comes down to how acoustic system reproduces well certain frequencies necessary for an individual for comfortable listening). Moreover, an amplifier with an ideal frequency response in a subjective test is likely to lose to an amplifier that reproduces certain frequencies with a higher gain (or, conversely, suppresses them) - as they say, to whom what.

The audio world today is dominated by digital technology. This cannot surprise a specialist in this field: digital is an excellent way to preserve and reproduce sound. The method is much more perfect than the methods that existed before him. However, as is the case with all relatively new technologies (although the figure is no longer "new"), digital technologies still receive less than deserved criticism. "Critics", basically, are divided into two camps: people who are grounded in theory - and, accordingly, those who are not grounded and have no experience at all. The former (apparently due to pathological conservatism and personal predilections) invent myths that can influence the latter. The latter are happy to spread these myths and argue to the point in conferences, not understanding the essence of the subject as such. With all this, without changing the fact that everything around is digitized and will not be translated back into analog.

In general, I am not an ardent defender digital technologies sound recording / playback (at the moment, this is not required). I have heard both analog and digital. Naturally, they sound differently. But who said that analog sounds better? This is completely unprovable. The main advantage of the number is its replicability and eternity, huge post-processing possibilities. And the figure sounds, forgive me audiophiles, is no worse than an analogue. More precisely, it sounds better.

In the previous article, I did not cover some of the "cherished" topics with which typical audiophiles try to "crush" the number. Topics, in general, hackneyed and sucked from the finger. I will try to disassemble them here in more detail and hope that I will be able to arrange the most understandable educational program.

Dynamic range

"Dynamic range !!!" - the first cry with which an audiophile rushes into the loophole of disputes. Absolutely all audiophiles with whom I happened to talk about sound, they called these two words. And absolutely all of them did not really know the true meaning of these words and the real picture of the case.

Roughly speaking, dynamic range is the difference between the quietest and loudest sound. In general, the larger it is, the better: it means that the system can record equally high quality and very loud sound, and extremely quiet. The dynamic range calculated for CD "by mathematics" is about 96 dB. The dynamic range of the best analog media (without noise reduction) is 50-60 dB. In total, it seems like it turns out a 30-40 dB gain from the figure (which is extremely large), but everything is not so simple. The fact is that below the range of 50-55 dB at CD, the nonlinear distortion coefficient increases. That is, the analog's dynamic range is limited by noise in which sound is lost. And the figure (in its CD-shnom version) - permissible distortions. It turns out that the dynamic range in both cases is approximately the same (and the figure does not lose, even at this stage of reasoning). However, there are several nuances.

The first nuance. What is better: when the sound is completely hidden in the noises, or when it is still heard through the noises? However, it is better to have a sound than not.

Second nuance. The sound at the level of -50 dB is almost inaudible. Disbelievers can try to normalize anyone sound file up to -50 dB in some editor and listen (of course, you do not need to turn the volume to maximum at the same time - let it remain at the usual level). That is, somewhere out there, beyond the range of -50 dB, distortion occurs in the CD. Only it is not possible to hear them on the way - this is the catch, no one records music at this level - in this volume range you can only hear the after-sound at the end of the track. Well, the analog media has just noise there, that's all.

The third nuance. Audio science has long known about nonlinear distortion at low signal levels in CD (quantization noise). And there has long been a technology that allows these distortions to be dither. This technology is used in the process of creating an AudioCD. In fact, dither is invisible (due to the fact that it affects small levels that are already inaudible). But a fun experience can be made: dither on an 8-bit file! In this case, distortions will practically disappear (albeit due to an increase in noise), despite the low bit resolution. Thus, distortion in the virtually inaudible range of levels can also be effectively masked!

And the last, fourth nuance: all these "terrible limitations" of the dynamic range are applicable only to CDs. For a long time, studios have been recording and processing with a bit resolution of at least 18 bits (more often - 24 bits). 24-bit offers a dynamic range of more than 140 dB, leaving all analog technology far behind. Now it is difficult to say which format will steadily replace AudioCD, but we can say for sure - it will not be 16-bit resolution. However, while the majority are satisfied even with AudioCD - based on the above, I don't see anything strange in this.

Thus, tales about the limited dynamic range of a digital are nothing more than fairy tales. Which, firstly, are tied to a specific AudioCD format, and secondly, even with an AudioCD, everything is in order with the range.

In the comments, a controversy has erupted over CD's DD, so I will give additional explanations here. The fact is that digital technologies are so perfect that the practical (attainable) dynamic range of a CD (16 bit) is about 120 dB! By applying dither and noise shaping, having the original 24-bit audio file at your disposal, you can make a 16-bit file where signal levels of -100 dB and below will be listened to. The price to pay for this will be noise, which will make the recording at such a level not only substandard, but simply unsuitable for listening. But the fact is the fact: the dynamic range of a CD using tweaks is simply enormous. Another thing is that no one really needs him like that. Firstly, sounds at levels less than -50 dB are practically not found in the recordings (except perhaps the "attenuation" of the tracks or relatively rare classical pieces), because this is a very "quiet" zone. And secondly, the noise from dither-shaping is also not a gift. All the inexperienced reader needs to know: CD's dynamic range surpasses any analogue audio medium released in the pre-digital era.

Jitter

Jitter - the instability of the sampling rate. It can occur both during recording and during playback. Audiophiles have long been accustomed to frightening others with the terrible word "jitter". In fact, everything is simple. Jitter occurs in low-quality ADCs / DACs - that is, in cheap, consumer and non-professional ones. And in expensive - professional and high-quality - there is no jitter. That, in fact, is all.

Jitter is most commonly found in cheap sound cards for computers. The sound card must reproduce sound at completely different sampling rates (typically 8 to 48 kHz). Naturally, no one will insert a dozen stable oscillators for different frequencies into it. They will make one oscillator, and all the necessary frequencies will be obtained using a frequency synthesizer, which will pass part of the pulses and thus generate an unstable sampling rate (producing jitter).

Calling jitter "one of the problems" of digital audio is like calling an MK-60 cassette an analog problem. If you know what I mean. ;)

Recording level

We often hear that due to "problems with digital overloads" sound engineers "underestimate the recording level" to a margin of 12-16 dB. This naturally leads to an increase in quantization errors, corresponding to signal distortion, as well as a decrease in the dynamic range. Just a couple of nuances are enough to defeat this myth.

Firstly, nowadays no one writes in 16 bits (namely, for such a bit resolution, it will be a problem to underestimate the recording level). That is, the problem may have existed in the 90s for people trying to record something on sound card class SB16.

Secondly, even when I recorded in 16 bits, I never left such a huge margin and did not underestimate the level to such a minuscule amount. Simply because there is no need to do it: you need to adjust the recording level to -3 -4 dB and record for your own pleasure. In addition, when I worked with 16 beats, I was doing track recording: each instrument has its own track (this is the usual scheme). With such a scheme, even at 16 bits everything turns out deliciously: each instrument is recorded with a large dynamic range (since "no one interferes" with the instrument). In the final mix, the instruments were mixed and the actual dynamic range was greater than could be achieved by recording "everything in a heap".

This Kotelnikov of yours is just a theory

You often have to read arguments that practical use Kotelnikov's theorem, on the basis of which digital sound is recorded and reproduced, faces obvious problems - which supposedly does digital sound"which does not stand up to any criticism." Problems do exist: that the record, that the reproduction of the figures are faced with pitfalls. The only question is that these pebbles are the size of a speck of dust, if we take into account the final resolution of the human ear, which is simply not able to calculate these pebbles. And the description of the "problems", as always, is based on the bare format of AudioCD - as if there were no others. To complicate matters, as a rule, audiophiles present their accusatory "auditory tests" from the "JingHuang" mp3 player to Genius speakers.

When recording audio, the problem of limiting the input spectrum of the signal arises mainly. If this is not done, then the frequencies above the cutoff (22.05 kHz for AudioCD) will "creep" down when digitizing, creating low-frequency distortion (aliasing). Signal filtering is a non-trivial process and, in general, completely filter out the entire RF spectrum above the desired frequency without significant distortion useful frequencies it still won't work. However, the problem is easily solved by using higher sampling rates (oversampling) - both in recording and in processing. For example, 88.2 kHz instead of the traditional 44.1 kHz (in studios, hardly anyone in their right mind still writes in 44.1). At a sampling rate of 88.2 kHz cutoff frequency input signal - 44.1 kHz, which allows you to design filters low frequencies more "relaxed", given that the desired range ultimately is the frequency range up to ~ 20 kHz.

When playing digital sound, interpolation problems arise: it is necessary to restore the original signal as accurately as possible. Again, this is often done by software upsampling. Here audiophiles will gleefully cry out that software interpolation takes billions of operations, and that no computer can do that. Ideally - yes, but in reality - you can apply very simplified formulas, sufficient to restore a signal with such a quality that analog media never dreamed of. An example with graphs for this case is given in the previous article, which shows how accurately the signal is restored even for the AudioCD format (which is traditionally used to kick with audiophile feet). I will also clarify that I did not pull those graphs from somewhere from the Internet, but built them myself - using my own program for simulating digital signal processing systems sDCAD. A supercomputer with billions of operations, fortunately, was not required.

Flat sound

Audiophiles often hear the term "flat sound" when referring to digital. The term can vary: "plastic", "artificial", "lifeless" and the like. What exactly is the difference between analog audio and digital audio?

Firstly, the analogue is characterized by softness (blockage) of reproduction of high frequencies. Softness arises from the banal shortcomings of analog technology. In the case of vinyl, this is the inertia of the needle. In the case of magnetic tapes, gradual demagnetization (which occurs immediately after recording). In short - the analog sounds soft and delicate (nevertheless, the softness is due to "chewing" the high frequency).

The figure is a different matter: what you wrote down is what you got. If the sound reproduction path is of high quality - we hear what was recorded - and nothing is lost. Some digital tracks sound very harsh because they were recorded like that - nothing surprising here, not every sound engineer likes softness. Especially considering the trends in modern music, where it is customary to distort everything that is possible, including the vocalist's voice. But the fact is that digital is also capable of reproducing soft sound - you just need to record it accordingly.

Listeners of the "old school" are accustomed to hearing from vinyl or tape a juicy "hooting" bass, which appears due to the natural blockage of the treble and the accompanying selection of bass against this background. With the advent of digital technologies, sound engineers were able to operate with high quality throughout the spectrum, as a result of which the recordings became more saturated in high-frequency terms. And they really do sound more attractive than the old ones - if we put aside prejudices. However, to get a powerful "whooping" bass, it is enough to do a simple operation: add lows. Unless, of course, your music center is equipped with an equalizer at all ...

In general, the advent of digital sound recording technologies has changed the very sound that we hear from tracks. Is there anything amazing about this? I do not think. Is digital sound bad? No, digital sound is good. When used correctly, as with everything else.

It would also be correct to mention the fact in the end of all rantings: in disputes about sound technologies, it is customary to forget about the music itself. We still listen to, for example, early Beatles recordings and rejoice. Despite the fact that these recordings were made on washboards, it is hardly possible for an untrained person to fully imagine the progress in the field of sound technology that has occurred since those times. Every musician has a different perspective on conveying ideas, and take my word for it, the last thing we think about is warm tube sound and spherical vinyl in a vacuum. Least of all, a musician thinks that someone will listen to his recording with gold wires and speakers, near which the shaman danced with a tambourine, before sharpening the stylus of the pickup in the pyramid. The musician thinks about how to convey his message to the listener. Understanding perfectly well that in 90% of cases his music will be listened to on very budget equipment, which often does not stand up to criticism.

And then, for thirty years now, the world has been under the rule of synthesized sound. Sound that does not appear from living instruments, but from a variety of electronic devices. And the concept of "flat sound" relative to an electronic instrument cannot exist at all. Who said that the sound of the synthesizer that we hear on the record should sound somehow differently?

I seem to have covered all the topics not covered in the previous article. Have questions? Welcome to comments.

Musical addition to the article: "Vinyl" (not recommended for people without a sense of humor and audiophiles)

Somehow I accidentally noticed that 90% of articles on Habré with the tag "warm tube" talk about anything, but not about tube technology. At the same time, few publications about tube devices receive many likes of admiring comments.

I no longer remember how and when this strange idea settled in my head - to assemble a tube amplifier. Why, too, is not entirely clear - I am not a music lover, I have been ill with home theaters for a long time and quickly, the floor-standing speakers Wharfedale Diamond 8.4 remained in memory of this time, in recent years they were used exclusively as a decorative stand for flowers. Be that as it may, the thought so deeply settled in my head that a leisurely study of profile resources began, reading forums, searching for circuits of tube amplifiers "for dummies", etc. etc. The lack of any experience of communicating with lamp technology (the most modern gadget that I remember is a b / w TV in a student hostel in the early 90s of the last century) scared and attracted at the same time.

A sluggish search could go on indefinitely, if one day a wonderful resource was not discovered - http://tubelab.com/. I stopped my choice on a Tube Lab Simple Single End (SSE) single-ended amplifier, ideally suited to my interests, namely: a simple amplifier for beginners with a minimum of components, no adjustments, at the same time, it is quite universal and, judging by the reviews, has proven itself very well. The order of the board was made on the site (sent anywhere except Russia and Italy), payment via Paypal, short correspondence with the developer, fairly fast delivery of two boards (In addition to SSE, a board was also ordered for the advanced version of Tublab SE - so to speak "for growth") ... It was decided to order the components via e-bay, not quickly, but reliably and inexpensively - the delivery time was compensated by convenience (receipt by mail, leisurely search while sitting at the computer). The process took enough long time, but I was not in a hurry (it took almost 2 years from the moment of ordering the boards to the moment of successful inclusion).

The first received components

It makes no sense to describe the process of assembling the amplifier board, detailed instructions with pictures is on the project website. I was especially pleased with the disclaimer disclaimer

We are not responsible for injury, accidents, acts of random stupidity, burning your house down, exploding parts, and other undesired actions (all of which are possible) resulting from the use of ANY information contained herein.

Translation

We are not responsible for injuries, incidents, acts of accidental insanity, burnt houses, exploded components and other undesirable consequences (all of which are possible) as a result of using the information contained on the site

Some recommendations received in the process of studying the materials.

Never install electrolytes "all the way", there should be a small gap between them and the board. The fact is that when soldering, the leg heats up and lengthens, and when it cools, it shortens, and, with a tight fit, it may simply fall off from the lining. Considering that in a tube amplifier the heating-cooling process occurs regularly, it is worth paying attention to this point.
Place the chassis of the output and power transformers perpendicularly to reduce mutual influence.
Isolate the audio input connectors from the chassis to avoid ground loops on the signal lines. If the wire is shielded, then the shield should be grounded only on one side.
Order spare parts with a stock in order to avoid delays in logistics and save on delivery.
And most importantly, be careful when shopping for components on ebay (more on that later).

One of the problems that I had to face was the choice of transformers (power and output) - it is quite difficult to buy a transformer with the required voltages, if the 110-volt version is usually available from American retailers, then a 220V transformer must be ordered from the manufacturer and wait 45-60 days. In addition, they are quite heavy and the cost of shipping from the USA almost doubles the cost of an order. Fortunately, a suitable version (Hammond 374BX) was found in Germany, which made it possible to significantly save on shipping and, along the way, order a choke (inductance) for use in the output filter of the power supply. The first mistake - when ordering an inductance, I selected the resistance, completely forgetting about the current, as a result, I got a coil with a current limitation of 100ma instead of the minimum required 170ma, I had to go back to a simpler and lower quality version with an RC filter and buy an appropriate wirewound resistor, change the resistor on the reel, if the desire arises, you can at any time. It was easier with output transformers, only Transcendar proved to have adequate delivery times, the TT-119 model came up in all respects.

Finally, the moment came when all the components were received, there was free time and nothing prevented me from seeing how all this would work. In violation of all safety regulations, all connections were made directly on the table in front of the monitor.

An old LG-P500 was invited to act as a signal source, speakers from music center, it took some red tape and a little courage. Taadaaaam - switching on took place, nothing exploded, the lamps lit up with a beautiful orange light ... and silence, more precisely, if you bring your ear to the speaker, you could even hear music against the background of the noise, but it was not at all the "warm tube" sound that I had hoped for hear.

The first thing I decided to check was the voltage at the output of the rectifier, and was immediately unpleasantly surprised, instead of the expected 375V x √2-27V = 503.33V (voltage on the secondary winding multiplied by the root of 2 minus the drop on the lamp), I saw almost 550V at the output of the rectifier and, accordingly, 525V B + (anode voltage). There was no desire to test electrolytes for endurance (they are designed for 500V), so I had to turn off the power. After checking the mains voltage, I was once again surprised - it turned out to be more than 240V (a further survey of neighbors confirmed that this is the case for everyone). Fortunately, the transformer can be re-wired for this voltage too. On the second switch-on, the voltages returned to normal, but the speakers were still silent, further testing revealed the absence of anode voltage at the input triode, which, in my opinion, indicated a malfunction of the only semiconductor device - the IXIS10M45 regulated current source.

Having decided that the problem was due to overvoltage and / or a Chinese ebay seller, I ordered a new pair of IXIS10M45 from England, it seemed more reliable and faster. I must say that the next inclusion was completed absolutely similar to the first and second, although the new parts looked completely different, they refused to work in the same way. Here I already started to worry, since both channels behaved completely identical, and the voltage on the 12AT7 anodes was completely absent. Since in this circuit, except for the lamp itself, the current regulator and a priori working small things, there was nothing else, suspicion fell on the lamp. The auction on ebay made it possible to buy ECC81 (the European analogue of the American 12AT7) at a very low cost, and at the same time the next batch of IXYS 10M45 (again, the Chinese seller took it with a margin just in case). The third batch of 10M45 looked (and rang) exactly the same as the second, for the purity of the experiment, immediately replaced the lamp and IXYS, disconnected all unnecessary (second stage) and for the fourth time did not find anything on the anode of the first triode.

A complete failure, the mind refused to understand how this could be. On a breadboard, I assembled a simple circuit with an LED and regulated source current (used untouched from the third batch), powered from the laptop power supply - and SHE DIDN'T WORK !!!

At that moment, the thought of a universal conspiracy began to haunt me, even what was supposed to work did not work ... and again I decided to order problematic microcircuits, only through a trusted seller (Digikey). And once again, difficulties arose even where they should not have been. The first problem that arose (in Digikey, the minimum shipping cost to my region was $ 75, even for a $ 5 order). This problem was solved with the help of an American intermediary, but the second one came to light after placing the order - a letter came to my email with a request to confirm that I was not a terrorist to fill out the BIS711 form (who is interested in goo.gl/VAkDYB). I ordered ordinary radio parts to an American address, why do I need to fill in this form When buying conventional radio components, I still do not understand. By indicating my name, my name and home address in all fields, namely: I am the end user, I am the official representative of the end user, I am the buyer, I am the exporter and indicated that I am an individual, sent the completed form to Digikey, and the very next day I received an order confirmation and tracking for the parcel.

Another batch appearance differed from all previous ones, which inspired some optimism (picture below)

The test on the breadboard made me happy, the LED happily changed its brightness depending on the resistance of the control resistor. Five minutes to replace a part on the board ...

… Another switch on and MUSIC began to sound from the speakers.

As it turned out in the process of communication on specialized forums, counterfeit radio components on ebay are becoming a big problem. Here's what the Diyaudio moderators write

- Fake parts are a real plague by now. No small chance we all get a share of those when fishing for a quick small purchase.
- I never buy semi-conductors or electrolytic capacitors on eBay for this reason.