Review and Measurements of Totaldac d1-six DAC

[gvl]

Trying to defend a NOS DAC on a measurement-centric forum is an uphill battle. To Vincent's credit he is at least not invoking the typical audiophile arguments.

 

[maty]

Before, the old R-2R TDA1541A, NOS, 16/44

https://www.audiosciencereview.com/...ments-of-totaldac-d1-six-dac.8192/post-204630

Now, a modern DIY R-2R, Soekris dam1021, up to 24/384

https://www.diyaudio.com/forums/ven...iscrete-2r-sign-magnitude-24-bit-384-khz.html

It's a DAC module based on a discrete R-2R sign magnitude DAC design, with FPGA based reclocking and custom digital filters, < 1 ps jitter clock generator, with 28 bit resolution so there is headroom, oversampling up to maybe 3.072 Mhz. Up to 24 bit / 384 Khz input from SPDIF, I2S and USB (via USB to I2S interface board), with isolation on the I2S interface. The board will be very flexible, with digital volume control and filter parameters that can be downloaded, a possibility is also to add digital crossover filters, but that will probably not be ready for first production lots, although the board is firmware upgradable over a simple serial connection...

FFT at -1dBFS

FFT at -60dBFS

 

[maty]

Totaldac d1-six, at 0 dBrA

 

[FredYves]

As many others among of us, I went thru your process many times.
Placebo is a HUGE factor in audio, you know that, right?

I have just learned something : -) Thank you! In fact, I do care of measures but I make my final judgement on what I hear and I am conscious of biais. I even knew that to compare two DAC, I have to secure the same environment (I have appreciated the questions which suspected I could have listened to one DAC with loudspeakers, and the other with headphones and from that come to conclusions...). Knowing that I am even careful to only compare audio products when they are in the same room because the room maybe the #1 acoustic element...
As said before, I perfectly admit that I could/can have a bias. I have listened to music during 30 years with products from Naim - I like(d) them so much. They may have influenced what I expect with some specific color or attributes. As an example, I do like the Chord H2 with Focal Utopia headphones better than the ADI2, but I like ADI2 with Focal Stellia as much as Chord H2 with Utopia.

 

[Joachim Herbert]

http://www.totaldac.com/USB_gigafilter-eng.htm

If one needed an illustration for "snakeoil"...

This just keeps getting better and better. Highend audio is the most entertaining business you never thought might exist.

 

[gvl]

Maty, you're a very prolific poster of marginally relevant content. My apologies, but I have to put you on my ignore list as your posts induce unnecessary scrolling.

 

[maty]

@FredYves

It is not worth discussing about it in these forums.

I do believe that you appreciate differences, as happens to me but that they do not want to believe me either (they do not even bother to download the files to do their own tests).

 

[Haskil]

If you would answer our questions with something else than other questions or one-sentence replies, and take a bit of time to explain, that would be nice.
If the signal transmitted by your DACs is as good as your communication skills, no wonder it doesn't take the transparency test.

Vincent Brient never takes the time to explain things when asked about what his products do. On a French forum, I tried to know what was doing his "bass boost", he never answered clearly, directly, sincere. It was necessary to shoot him the worms ... as they say in French. And the worms did not come! And all this long before Amirm measures the Totaldac. And it continues visibly.

 

[maty]

@gvl

If comparing the measurements of other R-2Rs, old or modern, is getting out of the way ... I simply add references to compare with Totaldac amirm measurements.

 

[THW]

This just keeps getting better and better. Highend audio is the most entertaining business you never thought might exist.

I’m curious, how is this sort of shit even allowed?

Like, those probably don’t even seem to do anything regarding actual audio quality.

 

[amirm]

No you don't get ML. He despises the whole approach of sites like ASR and thinks only his subjective listening results matter.

What he hears? Who knows what he hears? You have been in his head to know that what he writes is only influenced by the sound that went into his ear and no other factors like looks, discount or loan status of gear, pre-bias by the marketing list of products, improper comparison without level matching, etc., etc.?

If people truly reported what they "hear" I would pack up and close this site!

Fact is that he doesn't report on what sound waves hit his ears. He provides a subjective impression that is completely unreliable and untrustworthy. It is not like he has taken at least one audibility test where we know the answer for sure to evaluate his. Without any type of evaluation, you are telling us he reports on what he hears? I think not.

To his credit, he doesn't claim that anyone else will hear what he hears.

If that is true, then he should stop writing and doing videos. Because as long as he does that, he absolutely wants to convince others that what he "hears" is what others would "hear."

But he loves the TD sound and that is all the "proof" he needs.

Again, nothing of the sort is on the record with regard to "sound." You are throwing every bit of research in audio to make that claim.

Have you not heard the expression that we "eat with our eyes?" Why do you think high-end chefs spend time to make your food plate pretty? It tastes the same if it is ugly or pretty, right? They are pre-conditioning you to like the taste of food, even though taste has nothing to do with looks.

So no, please don't defend him. I have countless high-end audiophile friends and acquaintances. I know what makes them tick and just explaining that false notion about audio doesn't get us anywhere. Your posts make it look like none of us know who these people are, and how they go about their audio evaluations.

 

[graz_lag]

I have just learned something : -) Thank you! In fact, I do care of measures but I make my final judgement on what I hear and I am conscious of biais. I even knew that to compare two DAC, I have to secure the same environment (I have appreciated the questions which suspected I could have listened to one DAC with loudspeakers, and the other with headphones and from that come to conclusions...). Knowing that I am even careful to only compare audio products when they are in the same room because the room maybe the #1 acoustic element...
As said before, I perfectly admit that I could/can have a bias. I have listened to music during 30 years with products from Naim - I like(d) them so much. They may have influenced what I expect with some specific color or attributes. As an example, I do like the Chord H2 with Focal Utopia headphones better than the ADI2, but I like ADI2 with Focal Stellia as much as Chord H2 with Utopia.

Of course it was clear to me you would have learned something new, that was the reason for which I mentioned the placebo factor, Ha Ha Ha

Pls. pls. scroll back on page #1 and go thru Amir' conclusions for both Listening Tests & Final Conclusions.
Of course you did it already, but pls. do it one more time.
Everything is very well explained there.

 

[JJB70]

I might start a new thread - I bought an Apple dongle, got better sound and did this with the 13,000 euros I saved:

Then we can all insert 13,000 euros of stuff we'd rather buy.

I will open by saying that I would buy some Dutch and Dutch speakers and pay the deposit on a Tesla Model 3.

 

[THW]

I might start a new thread - I bought an Apple dongle, got better sound and did this with the 13,000 euros I saved:

Then we can all insert 13,000 euros of stuff we'd rather buy.

I will open by saying that I would buy some Dutch and Dutch speakers and pay the deposit on a Tesla Model 3.

I would honestly invest that 13,000 euros into some fairly safe bonds and other investment products, I can’t think of another way to spend that 13,000.

 

[JohnPM]

NOS DACs have an output spectrum that contains images at high levels, since there is no filter to attenuate them. Proponents would argue that those images are outside the bandwidth of our hearing so they are of no audible consequence (setting aside their potential to generate intermodulation later in the chain with artefacts back in the audible range). Analysers can not easily exclude those image components from their measurement results, so the figures the analyser produces may not be representative of the parts of the signal that are in the audible range. The frequency response measurement originally posted was an example of that, as are the THD measurements.

Having taken the decision to produce a NOS design there are many other technical challenges, linearity being one. It has already been noted that the design delivers linearity performance that is usually beyond the reach of this type of DAC. There have been many posts assigning some fraudulent or deceptive intent to the designer, which seems unfair. There looks to be technical merit in the design, within the realms of what is possible with a NOS architecture. I wouldn't buy one, but that isn't to say that the designer hasn't made every effort to the do the best job that can be done with that approach.

 

[amirm]

You are a little too certain of yourself and your method. People don't all "hear" the same thing. Hearing is not just a physical response, it is also how your brain responds and interprets sound. You are convinced you are the only person who can hear things? I think not.

Some of those audiophiles can hear those distortions just like you - and they like what they hear. They think that's how things are "supposed" to sound or that those sounds are more "natural" It's as simple as that.

There is no "Amir method." It is either what is accepted by formal audio research and engineering or voodoo audio mythology created by audiophiles with zero verification of validity of what they do.

We, the scientific and engineering group in the industry and researched have arrived at above conclusion through hard work and verification. It is not some random approach you can dismiss. Let me give you an example that is relevant here.

When I was at Microsoft, we were developing our next generation *lossy* audio codec (WMA Pro). We had a mode in the encoder where you could set the quality as the priority and the codec would then choose the (variable) bit rate to achieve that. At the highest level, the quality could be set to max which by definition had to be transparent. To make sure we achieved this, we decided to leverage the large group of audiophiles at Microsoft in a blind test.

The files were encoded, achieving roughly 2:1 compression ratio. In other words, lots and lots of data was thrown away (in a psychoacoustically aware method of course). To my disappointment, the audiophile group severely underperformed our codec test team. They simply could not hear any artifacts.

I was busy so I had not take the test myself. My codec team manager came to my office and ask me if I could run the test. I said I did not have time. Seeing how badly he needed the results, I told him to wait and I pulled up a couple of samples. In an instant, not only could I tell the difference, but identified exact moment in the track where the difference existed. The team was very happy as they went and found the problem that caused the fidelity loss in those segments.

The conclusion is certain: audiophiles like you state love to think they have great ears. But when it comes to non-linear artifacts (as opposed to simple frequency response changes and such), they have no better ability than the general public. Which is to say, their ability is quite poor compared to trained listeners.

Fact is that people buy flawed audio products because they are totally deaf to their artifacts. This proves their inability to be critical listeners.

Faced with above, the convenient counter is: "oh, I prefer the distortions." Who says? If you can't hear the distortions, you certainly can't prefer them. What's more this theory has never been proven to be true using controlled tests. I have tested countless flawed products and as I observe in this review, the defects are hard to hear. But if you do hear them, they are unpleasant.

So please, don't try to fight facts and science with just word arguments. I have the data and science on my side plus decades of testing hypothesis like yours. There is no there there. Audiophiles need to wake up and get out of their illusions of audio. Reality is a cold, hard place for them to be sure. But it is also liberating when you can prove something, than claim it!

 

[daftcombo]

English is not everyone's first language, but I agree that many of the TotalDac posts appear to be attempts at obfuscation.

Let me tell you that Vincent's replies in his mothertongue on the french forum are no longer, no more explicative and not better written.

Last message he wrote to answer somebody was "Are you my boss"?
https://www.homecinema-fr.com/forum...elite-et-dac/dac-totaldac-t29963819-5490.html

 

[JJB70]

I would honestly invest that 13,000 euros into some fairly safe bonds and other investment products, I can’t think of another way to spend that 13,000.

That's the sensible answer, but in the context of spending 13,000 euros on a DAC, and one which isn't very good at that, I think we can assume that the train of common sense left the station a long time ago So have a bit of fun, it does put things into perspective when you think about just what you could buy if you got a perfectly good low cost DAC and had a load of money left over.

 

[Dialectic]

NOS DACs have an output spectrum that contains images at high levels, since there is no filter to attenuate them. Proponents would argue that those images are outside the bandwidth of our hearing so they are of no audible consequence (setting aside their potential to generate intermodulation later in the chain with artefacts back in the audible range). Analysers can not easily exclude those image components from their measurement results, so the figures the analyser produces may not be representative of the parts of the signal that are in the audible range. The frequency response measurement originally posted was an example of that, as are the THD measurements.

Having taken the decision to produce a NOS design there are many other technical challenges, linearity being one. It has already been noted that the design delivers linearity performance that is usually beyond the reach of this type of DAC. There have been many posts assigning some fraudulent or deceptive intent to the designer, which seems unfair. There looks to be technical merit in the design, within the realms of what is possible with a NOS architecture. I wouldn't buy one, but that isn't to say that the designer hasn't made every effort to the do the best job that can be done with that approach.

If only TotalDac could have put it so clearly.

 

[amirm]

Amir your protocole doesn't work for this NOS R2R DACs, They are too many artifacts of different kinds. You should remove this whole article, unless the aim was only to attack Totaldac.

Let's summarize my article then.

1. Device is very cheaply made with stamped sheet metal and such. No way this is justified in a $13,000 device where audiophiles look to CNC machined thick metal and beautiful industrial design. So on that front, the device is not a high-end product even though its price claims to be.

2. You have not explained, nor shown what is wrong with simple measurements like THD+N. On that front, the analyzer has deployed a low pass filter that should have been in your device, but isn't. Out of 144 DACs tested by us todate, your product finished at rank 121. Your performance in this regard is the same as the Soekris dac1421 which costs only $1,000 or about 8% of yours.

3. You have been encouraged multiple times to present your own contrary measurements. You have provided none.

4. Your DAC violates the most basic principle of sampling theory by not including a reconstruction filter. This is a fact that you agree to. Those of us in audio science and proper engineering consider such a product broken. Period.

5. I provided praise for the performance of the product where it was merited such as linearity test.

So no, the review stands. You have made no attempt whatsoever to counter the criticism in there. We go by data here, not words.