Along with the increased use of high resolution products, USB-DACs and DAC built-in amplifiers have become popular. Among audiophiles, a built-in DAC is one of the key factors in selecting equipment. Asahi Kasei Microdevices (AKM) is a leader among the various DAC brands. AKM has become one of the standard brands used in high-end audio equipment. AKM released “VELVET SOUND” in 2014, and since then has consecutively released new products that embody the concept of “Real Live Sound,” which goes to show that they attach great importance to the original sound. The number of products utilizing the AKM DAC is increasing.
This time we visited a major audio manufacturer, TEAC, which is utilizing the AKM DAC. The first lot of UD-503 sold out with pre-orders. We organized a round-table discussion with TEAC’s developers and AKM’s engineers to discuss the UD-503.
TEAC’s existing model, the “UD-501”, was well received in the market, but with the following model, the UD-503, AKM’s “AK4490” has been adopted as the DAC. Why did TEAC decide to switch to this DAC? We heard an interesting story about how critical a role the DAC plays in creating the final sound, and about DAC selection criteria.
We invited Mr. Kato and Mr. Watanabe from TEAC and Mr. Sato and Mr. Aniya from AKM to this round-table discussion. First of all, would you explain to the readers about what kind of work you are engaged in?
Mr. Tetsuya Kato (Hereinafter all indicated by surnames only): I am in charge of product planning at TEAC. I conduct system designs to consider, from a technical point of view, what sort of product we should make, and also supervise the quality of the sound.
Mr. Kazuo Watanabe: I am in charge of product development. I develop all electric circuitry including digital circuits, analog circuits, and power supply modules.
Mr. Mitsuru Aniya: I am in charge of the development of premium audio devices.
Mr. Tomonori Sato: I develop high-end ADC and DACs and also decide the sound for each product at AKM.
Mr. Sato has been called the “Audio Master” of AKM and has appeared in various press events. He explained to us the details of what goes into each DAC.
The theme of this discussion is the USB-DAC/headphone amplifier, theUD-503, which is a continual best seller, and the network player/USB-DACNT-503, which will be released in late October. Both products were made in accordance with the “dual monaural” philosophy, which is one of TEAC’s specialties, and utilize 2 AKM “AK-4490” DACs.
Kato: Yes. As you kindly mentioned, both the UD-503 and NT-503 have been designed based on TEAC’s traditional “dual monaural philosophy”. Of course, the DAC design is fully capable of clearly separating L and R channels, but our standards do not tolerate the slightest interference, so dual AK4490 are employed for L and R channels, two in total. Our dual monaural philosophy extends not only to the DAC but to everything. Dual power sources are provided separately for left and right channels as well. Even the analog circuits are separated. Getting the monaural components into the small casing was no simple matter, but after having experienced the high quality and true-to-original sound quality, it was well worth it.
Watanabe: Even if a chip’s internal circuitry has L and R channels but only one power source, it may lead to interference. Therefore, we divided the DAC into left and right to improve sound quality. As a set manufacturer, this point is something that cannot be compromised.
• What was the most difficult part of designing the UD-503? (Did you have any difficulties designing the UD-503?)
Watanabe: We put a great deal of effort into developing the analog circuitry including the DAC for the UD-503. The casing is small, but due to the added circuitry a considerable amount of space was needed. But even with this issue, I still felt that I was able to develop the product without needing to compromise. When sound quality is the goal, the trend has been to make everything larger, but as the size was fixed at the product concept stage, some creative and ingenious designs were required. As the DAC sound quality is greatly affected by its periphery, condenser placement was crucial.
Interviewer: I would like to ask about the specifications for the periphery of the DAC. The UD-503/NT-503 are compatible up to 11.2MHz of DSD and up to 384kHz/32-bit of PCM, so it has the highest specifications at the present. Did you, TEAC, from the onset go out to develop a product with the best specifications?
Kato: Yes. As DSD sound sources are becoming accessible, it was our intention to create a product compatible with the highest level, 11.2MHz. Our intention was to create a product capable of playing DSD, should such a sound source exist. Then the AK4490 was created to meet that goal.
Interviewer: I would like to ask AKM about the development of the DAC. I assume that DAC development requires several years per project, but 2.8MHz DSDs were not common a few years ago, right?
Sato: That’s right. When we started developing the AK4490, there definitely were not many products on the market capable of that high spec performance level. Based on results of studies conducted at universities, we were able to estimate product development down the road and came to the decision to develop a product compatible with 11.2MHz sources.
Interviewer: A filter function is provided with the AK4490. Can be used for the UD-503 and NT-503?
Kato: Yes, we included the filter option which users can turn on or off. It makes quite a difference so we hope people will enjoy listening to the difference between the sound with and without the filter.
Sato: If there is common impedance within a circuit creating mutual interference, performance falters by disturbing the sound gradation or changing the sound color. That is why we provided a separate power source to avoid such issues. We provided a dedicated power source for the clock. There are 6 power source pins. In addition, there is a power source for return. I seriously doubt that there are many examples of a DAC like this having so many separate power sources.
Interviewer: Having many power source pins means that development of the final product becomes harder.
Watanabe: Indeed. We are having a hard time maximizing DAC performance. We thought that parts placement around the DAC on the board would not be a major concern, but there were some time-consuming issues to deal with. We had to put all the ingenuity we possess into figuring out how to manage the internal arrangement and what to do to minimizing their influence on sound quality.
Kato: In addition, we had AKM engineers come to this listening room many times, and we listened to the sound of the product under development. We then asked them to make changes to meet our needs, and AKM took our requests into consideration in the making of the final product. This is what we do day to day in our technology exchange.
Interviewer: I see. When you ask them to make changes, do you ask for quite specific things?
Kato: Yes, we do. We tell them, ‘we want THIS type of sound’, and we also discuss many other things as well.
Watanabe: When I was involved in the development of TASCAM products several years ago, I told them that we wanted a short delay digital filter developed. When a delay is long, it takes considerable time for the sound played by the player to enter the equipment and return to the listener, which adversely affects player performance. Our impression is that they made their product very well by incorporating the functions we requested. I believe that such functions can be utilized for various products.
**Aniya**: I would like to add to that delay story. The reason why the delay was long back then was because chip could be made smaller. However, we were told that usability was poor, so we shortened the delay by making a system with redundancy. By incorporating the technology into ADC and DAC, we were able to realize the shortest delay in the industry.
• DAC should be as “colorless and clear” as possible.
Interviewer: You incorporate casual conversation into product development. Do you have specifics in mind when creating DAC sound?
Sato: It should be as “colorless and clear” as possible. The sound which expresses everything truly without any deviation is what we hope to create. That is “Real Live Sound”, and that concept, our philosophy, is embodied in VELVET SOUND.
Another point during new product development to consider is low-pitched sound reproducibility. We put all our ingenuity and circuitry know-how into the realization of low-pitch sound reproduction. The amount of momentary electricity in the circuit has a huge impact on low-pitched sound, so we are now looking into how much this part can be upgraded. We would like to continue further development in this direction.
• How is a DAC developed?
Interviewer: Mr. Sato of AKM listens to the opinions of various manufacturers, including TEAC. How do you reflect those opinions in the development of your DACs?
Sato: I bring a test sample to the client and ask for their opinions while listening to the test sample together. Then I convey their feedback to the engineering section or bring an engineer with me to see the client.. We talk about not only the sound but also technical matters.
Mr. Sato of AKM explains that he combines requests from companies including TEAC to decide DAC sound.
Interviewer: I am very interested in what kind of discussions you have when you do so. Specifically, what kind of sound do you listen to? Test signals?
Kato: No, I do not listen to test signals but music. While listening to music, I ask to make this part sound this way. However, what we have at that time is a test sample. Though we use a reference amplifier for sound, we are still able to imagine the final sound to some extent even under such circumstances. Therefore, we sometimes bluntly request improvement, and we sometimes make severe comments.
Sato: Exactly. (laughs) We bring the actual high-end and top products here, so we receive many requests as a matter of course, and the standards sought by the clients are naturally high. Of course, we receive many requests not only from TEAC but from other companies as well. We improve our products based upon opinions received from companies. Know-how accumulated by developing high-end DACs can be reflected in lower models, thereby raising the quality of the entire series.
Interviewer: This is a simple question. How long does it take to develop DAC?
Sato: We tried many things to reduce noise in the AK4490. We worked on the AK4495S before the AK4490 to reduce distortion.
Aniya: Therefore, it may have taken 3 to 4 years.
Sato: It was very difficult to exceed the -100dB barrier distortion characteristics when developing that DAC. It took a considerably time to overcome that barrier. As a result, we were able to reduce down to the -110dB maximum.
Interviewer: The value, -110dB, is a great figure. In the previous interview, you mentioned that sound would remarkably change when the external band noise is lowered to such a low point that going any further would be moot.
Sato: Yes. Of course we consider values, but in the end, the sound is everything to set manufacturers.
Kato: Exactly. We’ve never simply asked you to improve the numbers.
Kato: Yes, we never have. What we talk about is how the power source operates or how the LSI pattern is. Sometimes we talk about something else but not about sound. We sometimes say things like, “An electron will be difficult to run if the path for the electric current bends at a right angle. Why don’t we make the corner round?” Occasionally we’ll have that kind of conversation.
“Let’s think about it from the point of view of electricity!”
Kato: Taking the UD-503 as an example, Watanabe draws the actual circuit, but we would say things to him like, “How can we make electricity run straight? Let’s think from the point of view of electricity!” (laughs)
Interviewer: From point of view of electricity. (laughs)
Kato: We talk about things like that all the time. Therefore, we would like the inside of the DAC to be made based on that concept. I am sure that the engineers of AKM did not know what we were talking about at the beginning. As we have worked together for years, I think that they have come to understand our way of thinking, so that now we are able to make something that is mutually satisfying.
Mr. Kato says that he often tells his co-workers to think from the point of view of electricity.
Interviewer: Does sound change based on the placement of the DAC on the board?
Kato: Yes. For example, this AK4490 has about 40 legs. Therefore, many things have to be connected. The connecting method may vary depending on the maker. When making a leg closer to something, other legs become farther apart. Each engineer or company has their own philosophy as to what to prioritize under such restrictions. That philosophy will greatly affect the sound.
Sato: It’s important to achieve a certain unique flavor to the sound, of course, but we must also consider “sound quality” when developing the inside of the LSI. I believe that TEAC also share this goal. We continue to develop products by focusing on sound quality.
Interviewer: This discussion was very meaningful in understanding the behind-the-scenes process of DAC development which is something that is rarely heard. I look forward to AKM’s future developments. Thank you very much for a wonderful discussion today.
Original Article Arranged and Edited by Yusuke Kazama