Interview with Hervé Deletraz of DarTZeel


When and how did all started for DarTZeel? 

The full answer is in the next question. It is almost the story of my all life. The name darTZeel is the anagram of my family name "Delétraz". I introduced my first commercial product in 2002 at the Stereophile show in New-York, and in 2003 I started to sell machines in USA, my first country where darTZeel products were distributed. The rest came as time went ahead!

Can you kindly tell us about your background?

Well, that's quite a long story indeed.

At 10 I received my first Philips Electronic kits (the series 1000 for those who maybe remember) and I built my first AM 2-band super heterodyne radio. At 12, I was already found of music reproduction, and I was lurking about all audio gear I couldn't afford, through specialized magazines or various audio shows.

At 15 I started my degree in electrical Engineer School of Geneva (E.I.G.), and I ended in 1984, realizing my very first audio amplifier, actually a Class-D amplifier, which sounded pretty good!

Anyway I wasn't totally satisfied, so I went back to full analog, and I started the NHB-108 project. It took me 16 years to finalize the NHB-108. I got several patents since the circuit I designed was never achieved before, and at the time it was one of the "simplest" designs regarding the signal path, which still offered very wide bandwidth without using any global negative feedback. Of course it was not 16 years of full time designing. Yes, we are quite slow in Switzerland, but here the reason I also had to work in order to earn some money… :)

Anyway taking so much time helped me keeping away from choosing designs which already existed. I had all my time, nobody was pushing me. It was really holy moments, since pressure is a thing I have to cope with in the present time. I like pressure and challenges, but sometimes I think about how sweet the beginning was!

Did you have any strong influences or people you looked up to (or still) do?

Oh, yes, for sure! Here in Switzerland were quite some nice brands like Nagra, Thorens, Swiss Physics, and especially REVOX – STUDER. I was especially found of Willi Studer, the founder of REVOX, and at the time the biggest audio show was the FERA in Zurich, organized by Willi himself. I owned almost every gear REVOX produced, save for reel-to-reel machines which were too much expensive for me. Only 20 years later I started to collect those wonderful tape machines… :)

Does DarTZeel name have a special meaning (TZ letters) or it just look nice with capital letters?

Yes! As you can see in previous answer, the name REVOX has the middle letter in big capital. I chose to use same thing in a kind of y tribute for the guy I admired in my teenager years. I also took the opportunity to dress the "Z" like a Zener diode, a basic electronic component.

In your venture you explore digital domain seriously long before anyone else, but didn't stay there. Why?

At the time I designed my first digital amplifier, I had to build all the electronics for the analog to digital domain, since such kind of converters didn't exist off the shelf, especially with the level of performances I needed. For a long time – several years - I tried to develop the "perfect" digital amp, but the biggest problem was the residual noise due to the clock jitter, which I wouldn't want to correct using a global feedback – my original 1984 digital amp was already an open loop design – and the biggest problem was the frequency response, dramatically affected by the speaker impedance, due of the use of the output filter.

Anyway I am still thinking about a digital amp, just in order to know what I would be able to do in this new, promising century…

Is digital audio coming close in these past few years?

If you are meaning "when will darTZeel release a digital source", frankly I do not know the answer myself… Very recently I heard a couple of truly fantastic digital players, yet at very high price/quality ratio. Designing and – trying to – beating these already excellent competitive products would be quite difficult, if the price enters into consideration. Furthermore, digital players can still greatly improve in the next 5 to 10 years to come. Maybe when we will be closer to the asymptotic curve of possible improvement, I will think about a digital player.

On the other end, even if I am sometimes tempted into this kind of venture, I would like to concentrate on pure analog products for a while still, since there are plenty of things I have in my silly mind…:)

If a digital product comes, it will be only – almost – by chance!

You're not the biggest fan of balanced topology. Double the electronics, double trouble?

Well, yes, and no… It mostly depends on which topology you will use for which kind of application.

I won't enter into deep, technical and physical explanations here, but in the very nature of sound propagation into the air, the waves are asymmetrical, not balanced. Balanced topology was primarily used in pro studios because they needed running very long cables without picking up noise, and they also took this opportunity for providing the known +48 volts phantom supply for microphones, a thing you can't do in single ended connections. However, if you look inside a mixing console, you will see that the signal is translated into single-ended mode, balanced mode only be used for carrying the signal outside (in and out) the console. If you look at loudspeakers, too, you will see that internal crossovers are - 99.9% of the time – single ended, not balanced.

In my own designs, yes I naturally chose single ended mode, since I wanted to have as low as possible components in the signal path. Balancing would have effectively doubled the part count.

Sometimes it's hard to explain analogy of musical instruments, but you're using the same analogy and I do often. There are always choices to make and places to explore when it comes to an instrument. Real artisans know about behavior o materials and how to implement them into specific sounding instrument. Would you agree with this? Please elaborate...

It is not always simple to make simple comparisons between music instruments and electronic gear, especially when one states that electronic components do not have moving parts. Too much often we can be misled by the fact an electric signal is two-dimensional only. This is almost true at first glance, but in reality everything in this human world – we won't talk about quanta here :) - is three-dimensional, even four if you add time. This is this fourth – time – factor which is key parameter in music reproduction.

An acoustical instrument generates its music in a very complex field, which is supposed to be omnidirectional, but actually it is even more complex than this, since the sound waves are not equal in all directions. When you try to reproduce the music by electronic means, you have to accurately reproduce the timing of the recording, without altering the relationship between two consecutive signals.

The human ear-brain system is much more sophisticated than the best available microphone. Electrically speaking, frequency is directly related to speed. In our human world, we can discriminate frequency and speed. We have different "sensing captors" for different sensations.

A very old man, for example – just like me :) – cannot hear anymore frequencies above 10 kHz, but he will easily recognize the difference between a triangle, a cymbal or a bell, while a microphone cut at 10 kHz will dramatically round the edges to the point the transients will look almost the same…

So, when it comes to audio amplification, if we can carefully handle transients and time relationships between tow very close, consecutive music "bits", the ear-brain system will "accept" it as being natural and life like, without needing any specific listening "effort".

This is what I tried to realize, and I think that it is one of the reasons why people do not hate our machines…

One of the things that I'm trying to push with my reviews is musicality. I don't care about price tag; if specific audio component cannot convey music with it's natural emotions, then it simply fail. There might be lots of approaches to this, but few succeeded. Would you describe your way...

The biggest problem, especially when you are in a "never-ending-design-process", is stopping at a point in order to actually produce the machine you created. I am sure most designers feel the very same, trying to improve further in the last minute, just before the marketing guy shouts "Stop now!"

As darTZeel is still quite a small to medium size company, I have a bit more latitude in this department, and generally I can stop "just before it is too late", yet being quite comfortable with the result I was looking for. At the end the price is inevitably a key factor, and I perfectly realize that our machines are not the most affordable ones. I can swear you anyway that I do not design machine to be expensive just for the sake of luxury, but for musical pleasure first.

When a product does not fulfill my goals, I do not leave it go into production. When the first final prototype eventually works, then I call all my favorite friends and I glue them to the listening chair for hours until they tell my if they like it or not…

Back in 1999, I was lucky enough to find a quite simple design which worked beyond all my own expectations, and most of darTZeel products are directly derived from this original design, which seems to have proved for itself. Fortunately, I do not keep on what was an already good design, I always try to go ahead, and as long as I solve a 2 + 2 operation, I will think this way… :)

There's a saying, that music can be a window of the soul. Would you agree?

Oh, yes, I do even more than just agreeing. When people ask me how I do making my components sound this way, it is always very difficult to give an answer. Simply because I actually do not know! And as I do not know, I generally say that I put a tiny bit of my own soul in every machine. This is not immodesty, but only the closest way I could describe my life time devotion into music reproduction.

So definitely yes, music is a big, full, sensitive and magical window opened to the soul.

What do you think about resonant imprint of elements, enclosure, etc? How much those affect the sound?

For sure the enclosure is a key factor in audio components. Almost all electronic parts are highly "micro phonic" and transmit vibrations into the circuit. When we perform the initial test measurement of our phono stages in the bench test, they are not already installed in the preamp housing. If you clap your hands, even at more than 5 meters, you can see the sound wave on the FFT analysis gear! So the reason we keep very quiet at this moment… External unwanted magnetic fields also are of quite big influence on the circuits, since any piece of metal is subject to react with such kind of noise disturbance.

Of course, a phono stage is more sensitive than a line stage, and the latter more sensitive than a power stage, but anyway at every step you have to be careful. On the other end, a power stage uses a bigger transformer which generates more field than a line stage power transformer, and so on.

Internal layout is also very important, for the very same reasons described above.

What are benefits of open-loop input and output stages?

The big advantage of an open-loop circuit is that its frequency response does not depend of any "retro-active" compensation. It does mean that the phase and the internal gain of the circuit will virtually not vary versus frequency. In a closed-loop circuit, of course you can compensate a lot of things like distortion, input/output impedances, but all these corrections are not applied the same way versus the frequency to be reproduced. A bit like a Father with his Son. If the son always asks his father what to do, it is a good thing for most of the time, but the son will not feel free to do his own way. If the father does not tell his son all the time what doing, the son will better exhibit his own character and personality.

An open loop stage will not try to "correct" an error, but will just amplify the signal it receives. Of course if the circuit is poorly designed the result will be poor. But is the circuit is quite okay, the signal will be just amplified without being "corrected" all the time. Furthermore, a closed-loop (negative feedback), will never act in real time, but always with a time delay, no matter if this delay is very short, supposedly beyond human perception. I do personally not believe that anything in the music reproduction is "beyond of perception". The biggest difficulty is designing an open-loop circuit with wide bandwidth, since these are two contradictory things, but it is possible anyway… :)

Lately everyone is talking about power supplies. How important is a power supply in your view? What differ in your designs from other audio manufacturers?

I do not think I am very different than others in designing my power supplies. I wouldn't simply say that "a power supply is a power supply", but here again, I think that when you know just enough about Ohm's law, you have an idea which kind of power supply to use for which kind of circuit. Just let us say that in power supplies I also favor simpler design, with open-loop regulation when possible, and as less as possible components in every power supply stage – which doesn't mean I do not use multiple regulation steps when required.

Oldest debate. Tubes vs solid state. You probably went through the both schools and obviously concluded something. What and why?

I wouldn't be so categorical about which is the best. Tubes are very fast devices, and are still used in broadcast emitters for the carrier frequency. They also are capable of very big power – still in broadcast application – and they can behave as well as transistors do. I personally believe that I could design a tube amplifier with a sound very close to my machines, or vice-versa. It is true that when I went to school, we concentrated on the transistor, and I am more comfortable with them. Transistors have the advantage of being smaller, more reliable, less expensive, and much cooler. But tubes are not dead to my point of view, especially in music reproduction. I actually think that solid-state amplifiers generally sound much less good because it is "so simple" to use a lot of them and applying all various artifacts in order to compensate their supposed flaws, rather trying to think about what we can do with a handful of transistor part count, exactly what our grand fathers tried when they only had a couple of tubes to play with…

There are lots of speculation with NFB (negative feedback). No global negative feedback with DarTZeel?

Negative feedback is a closed loop. Degenerative feedback is a kind of compensation made by a resistor directly connected to the end of the transistor, and it is not like a real negative feedback (the "loop" is only created around the resistor in this case).

There is no any design which can work without any kind of negative feedback.

In the darTZeel circuits, we only apply a very small degenerative feedback at the input stage. The second stage set the voltage gain, and this stage uses a small negative feedback. Actually we use two half-negative feedbacks; one for each polarity leg, in order not the positive part will directly influence the negative part, and vice-versa. The third, output stage is fully open-loop, and do use any kind of negative feedback. And of course we do not use any global feedback. A global negative feedback consists in taking a part of the output signal and feeding the first input stage in reverse phase in order to compensate any flaws; this is the worst type – but the easiest one – of feedback one could use in audio reproduction.

Life span of products comes with knowledge and applied technology. Your philosophy address this profoundly...

No big mystery here, just some experience over time… When you know that the lifespan of a component is halved each time the temperature rises by 5° Celsius, you naturally try to take care about internal thermal dissipation… :)

Specific warm yellow/gold aluminum front plate finish with red enclosure radiate with unique, distinguished, dandy kind of timeless appearance. What drove you to this harmony?

Thank you for calling this "harmony". It was just about what I thought, but not everybody liked these colors right at the beginning. I actually wanted to make my machines differently not only in their sounding, but also in their looking. I chose these colors because I liked them, and also because I found that they would nicely match old wooden furniture. Funny enough, this is only several years after that I realized that red and gold (or yellow) is the official colors of the Geneva flag… Maybe I was influenced unconsciously…

Now, after quite some years, our colors are now a trademark, and almost anybody can recognize our brand just looking them at a glance. Actually all our "old" customers like them and say they love listening to music in red and gold… :)

There are many ways to incorporated volume. You choose opto-modul for gain control and I'm guessing starlights wasn't the only attribute for that. Kindly elaborate?

When I designed the NHB-108 model one power amplifier, I decided not only to use a very "simple" audio circuit, but my goal was also to preserve the signal path the best I could. So the reason the NHB-108 has no any contact or relay or switch in the signal path; when you plug the AC cord off, the machine still plays for about 20-30 seconds, just draining out the capacitor bank.

The big challenge was to design a preamplifier with the very same design philosophy. My big problem was how could I select a source without physically switching it, and even more important, how could I vary the loudness level without involving a potentiometer wipe or relay, or even analog switched arrays (which involves MosFets).

After thinking quite a lot about this, a… light went on in my head. Why not using the light for activating light dependent resistors? Well, the idea was not completely new, since I know other designers already thought about the idea. However, this technology has a big drawback: these components are not linear at all, and they do not act the same from sample to sample… Then the God of software entered. We were able to calibrate and control these special analog optocouplers the way we wanted.

The result is the NHB-18NS, the first and only preamplifier in the world which is a truly "zero contact" signal path design from input to output. Avoiding any contact greatly improves transparency, and as far as I know, customers like it very much…

Your obvious choice are bipolar transistors and you're not even generous with used amount, but output power could make many amplifiers envious. Why?

Here again, I chose bipolar transistors because they are easier to drive with very low part count. I also love their sound which is – to me – tighter and more full bodied than MOSFETS. The latter are very good for class-D amps, however, but this is another story….

What was with the "Noose" happy face design of NHB-108?

Well, as you certainly know, I do not take myself too much serious. Music must be fun, after all. As the 2 big LEDs reminded me Eyes – I called them this – I naturally called the central switch "Power Nose". The same way I called the loudness "Pleasure Control" in the preamp.

Only two bipoloar transistor per channel?

Yes. I wanted to use very few components. If not made extremely carefully, multiple parallel pairs could lead to unfocused sound, because of different propagation delay time. With only one pair, this problem simply does not exist! (The just released NHB-458 use 4 pairs instead of one, but the circuit topology is different here, benefiting from nine more years of research…)

Usually you cannot obtain more than 50wpc with a single pair. Thanks to my specially designed Monitor Control circuit, I was able to monitor all parameters of the transistors, and could make them work very close to their maximum ratings, still being always in the safe operating area. So we could obtain a high as 230 watts per channel into 4 ohms (claimed 160wpc), as measured in Stereophile magazine. People who do not own an NHB-108 might think it is a low power amplifier, but it is not at all. It sounds quite powerful, more than its power rating might suggest.

DarTZeel CTH-8550 made big impact in media and on high-end audio shows. Would you say that is a DarTZeel signature sound or a lack of?

The CTH-8550 marked a truly new corner in darTZeel products. Not exactly as good as the separates, but very, very close, at a much lower cost. CTH stands for Close To Heaven, and "8550" was for 85% of the quality sound of the separate, at 50% of the price. The sonic signature is the true darTZeel one. You have to listen to it to believe your ears. It is only when you directly compare it to separates that you will notice the difference.

Is NHB 458 just a pure statement or true cost no object design, cut no corners, speak no word approach?

None of the above. Actually there were my distributor who asked to offer a big, big amplifier. I replied "Yes". But only if I could make it sound even better than the NHB-108, which is the case at the end. I would never design it if the sound wouldn't be at the level of performance I wanted.

It is very close to a real no-cost-object machine. Right now, it is my no-cost object machine, actually… :)

In recent years there is a lot talk about price barrier of 10.000$. CTH-8550 is round double that price. DarTZeel components are object of choice for those who recognize and respect fine craftsmanship. But, where is the line or price border that you think is necessary?

It is true that all darTZeel components are quite expensive. But it is doubly true by now.

The past 3-year of economical crisis pushed the Swiss franc to very high value, and we truly suffer from this. Only eight years ago, 1 US dollar was 1.7 Swiss francs. Today, the dollar is almost on par with the Swiss franc. Almost all European currencies are quite weak too, and I can only deplore this fact.

Anyway, when you open one of our machines and look inside, you understand quite quickly why the value is high. Our quality of building is often compared to Swiss watches, and I do not talk about Swatch here… :)

Where does hi-fi stops and high-end comes in for you?

20 years ago, we only spoke about Hi-Fi. The term high end came later – at least here in Europe – and was just another way for qualifying better audio systems, compared to those "compact systems" you could buy in the 70's and 80's. To me, there are only audio systems which make real music, and others. This is not a matter of price. While our products are among the very up of the hill, for sure you can find already very satisfying, musical systems, for much less. It is just like a car. If you want the fastest one, you will have to pay the price. If you just want to go from point A to point B, any other good car will do the trip…

What would you say is the goal of DarTZeel audio?

The primary goal is sharing a true passion for music reproduction. darTZeel is still a young company at the moment of writing, but for sure it will grow along the coming years. We will try to expand our product line step by step. One of my dreams would be able to offer outstanding quality at more reasonable prices. But the road is not so easy, and it will take some time still. Be sure that I work on this very hard.

There is vivid difference between DarTZeel and competition. What is that exactly?

I do not really know if there is a "vivid difference" between us and the competition. I think what is different is the luck being independent and not having a dozen of engineers giving a dozen of different thinking points of views. For sure I do not design anymore all products 100% myself. I still continue – and I will continue as long as I will be able to – to imagine products I dream of, then I also work with very sensitive partners who truly understand what I need and what I want. This kind of synergy allows me to keep the drive wheel the way I want, and at the same time I can listen to good advices I'm told…

What audio/music references do you use when designing the prototypes?

Very good question! Well, a lot of old records (Jazz/Pop/Rock), most of the time not well recorded. What I am looking for is the emotional feeling which is present in the music, rather than the quality of the recording itself. A lot of designers simply forgot this kind of approach, and I think it is too bad…

Is there a way to offer more affordable DarTZeel integrated amplifier, or do you plan this in future?

Yes. Please refer to my other answer above (goal of darTZeel Audio).

Many liked your phono stage in CTH-8550. Can you tell us more about it?

As I wrote above about tube versus transistors, I believe you can also design a very good phono preamp even if you do not use discrete transistors only… :)

The phono stage in the CTH-8550 is using dedicated integrated circuits, but not dedicated for phono stages… We took the microphone amplifier stage approach, and it paid off, apparently…

What are your views on whole analog/vinyl revival or is it at all?

Which revival? Vinyl didn't die, and it won't ever die. So many LP's are available in the world. Even if scratched, this media is the only one I know which can survive more than 60 years without losing a single bit of quality. Not sure it is the same about hard disks…

Should we fire up analog vs digital?

Of course not. Both will cohabit together. Each of them has its advantage and its drawbacks. I think in the future they could eventually merge (how about analog music on hard disks?).

What is your connection with Playback design? They're offering direct connection to your amplifier.

Andreas Koch, the Playback Design designer, is friend of mine. He his Swiss, too, but he lives in California. When he designed is MPS-5, he kindly asked me if he could implement a dart 50-ohm output.

I simply answered: "Yes, with great pleasure!"

How do see current state of high-end audio?

We are in some trouble years. Digital was the best ever to come, then MP3 arrived. This added a lot of confusion, misleading "digital" and "quality of sound". This our task – we, old audio guys – teaching coming generation about how great music can be when properly listened to it. I am quite confident with the future, and I am pretty sure that my grand' sons will buy vinyl's!

What holds future for DarTZeel?

Depending on how is the "future" for you? I am 48 years old in next October, and I am in quite good shape. I plan to stay in the business for these next 50 years… :)

So you will certainly see plenty of other products in the future. I mean new products. We do not like to revamp existing products with a new housing and call them "next generation". Our products are made to last "forever". They never come obsolete if and when a new "brother" arrives…

Any last thoughts for our readers?

First of all, thank you for reading everything until the end! It was a very long time I didn't almost tell my all life in an interview… :)

I would like to tell all your readers that music is more than just audio components. Music is a universal language, which conveys all possible emotion palettes. If and when you will build your audio system, please do not forget one thing: if you can feel some of the emotion through your system, just keep it as long as you will like it, no matter it is vintage or very new.

Instead of spending only in components, buy software: CDs, SACDs, Vinyl. The more you will have music, the more you will be able to know if your system needs an upgrade. Audiophile demo disks will not always tell you the truth…

With my very, very best regards - Hervé

Matej Isak
Mono & Stereo
matejisak (at)