Review and Measurements of Totaldac d1-six DAC

[Joachim Herbert]

.... just that his measurement basically discord with Vincent's

I was under the impression that Vincent is using non standard methodology, to put it politely. His -120 dB chart is a joke.

 

[MC_RME]

Amir, I wrote about why we see an incorrect rise at HF for both curves while they show the (more or less) correct relationship to each other (the ZOH drop). It's the same type of measurement error that we sometimes see from measuring D/S-DACs with their strong rise of modulation noise above 20kHz which gives incorrect values for S/N and THD+N measurements wrt to what's actually relevant for perception. Or if you measure PWM amps and don't mask the influence of the carrier residual on the noise part.
Your new measurement looks better but I can tell you the only way to measure amplitude and phase response with great robustness is the Farina log sweep. I don't know how steep the bandlimit filter is in your AP model but I know it's not steep enough in my own personal 2322 and the 2722 which I have years of experience with (besides other analyzers, like DScopeII, R&S UPL/UPD, etc).

I'm not blaming you for making an error (we're all human and things happen), rather that you just took everything the AP spits out for real without questioning what you see when it seems to be too odd, before publishing. And yes, the THD+N vs frequency graph is completely off as well for the same reason

Trying to explain what KSTR tried to say: The AP has different methods to measure a frequency response. The old school version slowly steps the sine from lowest to highest frequency, while the analyzer measures the full range result. NOS DACs with their multitude of images therefore often show a rising fr at the higher end. It needs a very steep tracking filter on the analyzer to only get the right frequency measured with that method.

AP mentions this exact problem/use case as reason for adding the 'chirp' based measurement of frequency response:

https://www.ap.com/technical-librar...d-frequency-sweep-for-converter-measurements/

And here is the reference to Mr. Farina:
-------
The logarithmically-swept sine—also called chirp or continuous sweep—test signal has been an integral part of APx for over a decade and is based on the pioneering work of Professor Angelo Farina, a leading Italian theoretical and applied acoustician, of the University of Parma. The log-swept sine chirp is a brief, broadband stimulus signal that moves continuously across a specified range of frequencies (e.g., 20 Hz – 20 kHz) and is extremely advantageous in its ability to provide multiple high-resolution measurements—frequency response, harmonic distortion, rub and buzz—in a very short period.
-------

 

[FredYves]

Off topic

About Mahler, maybe you are interested: Gustav Mahler - The Nine Symphonies - Georg Solti (1975), Vinyl x15, Decca, Germany

https://www.discogs.com/Gustav-Mahler-Georg-Solti-Die-Neun-Sinfonien/release/13703087

https://www.diyaudio.com/forums/mus...39254.html?s=1559756620cac9dff6ae5a12c2e1a826

Thank you Maty. I like very much the ones from Kubelik and Bayerischen Rundfunks, or Jansons with Concertgebouw. The last recordings of Abbado with the Lucerne Festival Orchestra are exceptional in my view. Now, for a very good technical recording and good play as well, you have Inbal with RSO Frankfurt.

 

[Hephaestus]

I wonder if you are not making with one case a generality. You would argue that it is not one case but a bit radical no ?

My main message was to get more well engineered and sensibly priced products in the market. Totaldac doesn´t belong in either category.
Once manufacturers are aware of these measurements here and their possible effect on sales numbers.... maybe they will design better products?

P.s. I have bad sense of humour anyway...

 

[Frank Dernie]

I think no one is comparing LP and waves amplifier in that thread. You speak about Chord, do you have one of their products? Did you hear it and compared it to others? How? I have or had DAC or streamers or CD/Bluray with DAC from Naim, Linn, W4S, Topping, Pioneer, Oppo, Chord, RME, TotalDac. As I bought them, I could listen to them on quite long period of time. I am curious of the products which provides the same at a fraction of the price of Chord? RME ADI for example is excellent but doesn’t cost of fraction of Hugo2. Benchmark DAC3 is for me similar and in Europe is costing a bit more. If you consider Topping, that’s a fraction of the price indeed, but I have had issues with all the one I had in my hands. I am considering Okto but not so easy to listen - just have the excellent measures and again is not a fraction. Very much interested to benefit from the list you recommend.

I have been on this site a long time and gone into this before.
I have the following DACs now (some of them have other functions as well but it is the DACs I am concerned with)
Goldmund Mimesis 20, dCS P8i, Resolution Audio Cantata, RME ADI-2 Pro, Chord Hugo, Astell and Kern 240, Metric Halo LIO8/4, Metric Halo ULN2, Meridian 568, SMSL M9, Oppo 105 and maybe some others I have forgotten.
About 10 years ago I decided to try something other than the Goldmund to decode "high res" files. As a good customer of local dealers over the years I borrowed some well regarded units to see if how much better the expensive ones were. The items I tried were Linn Klimax DS, Resolution Audio Cantata, Weiss DAC2 and the Metric Halo. I carefully matched the volume levels since I was well aware even slightly louder sounds better. I was truly astonished by the fact that they all sounded the same. I could just convince myself that maybe the Cantata revealed the mechanical noise of the pianist pedalling on a piano recording, but it wasn't repeatable. One one comparison the Weiss sounded a bit harsh, but It didn't repeat on another compare. I am completely sure I would be unable to pick out which was playing in a level matched listening test.
The rest of the system, fwiw, was a Goldmund Mimesis 36+ transport (which I still use as my main source) I forget which streaming source, Goldmund Mimesis 22 preamp, Mimesis 29.4 monoblocs and Epilog 1&2 speakers.
I was certainly not expecting this result.
I have 4 record players and they all sound different and are easily distinguishable from each other on the same system btw.
I now use a Devialet Original d'Atelier in my main system as DAC/pre/power, the other stuff has been accumulated over the years for other things or on impulse.

 

[maty]

Thank you Maty. I like very much the ones from Kubelik and Bayerischen Rundfunks, or Jansons with Concertgebouw. The last recordings of Abbado with the Lucerne Festival Orchestra are exceptional in my view. Now, for a very good technical recording and good play as well, you have Inbal with RSO Frankfurt.

I have a lot of Mahler, usually old vinyl recordings. One of them: Kubelik, SBR - Mahler - Symphony No 5 (1981), Vinyl x2, Audite 1999, Germany

https://www.discogs.com/es/Gustav-M...schen-Rundfunks-Symphony-No-5/release/4095477

https://www.amazon.co.uk/Mahler-Symphony-No-5-VINYL/dp/B0000942OR

Let us leave Mahler alone in this thread.

 

[THW]

im still wondering what’s so “special” and “mysterious” about R2R DACs that require completely different test conditions in order to even be considered a fair comparison.

the other bit that still eludes my understanding is how they seem to preach that measurements are basically meaningless and that you should select equipment by ear, yet make such a big fuss when their gear measures badly and how “unfair” the tests are.

 

[Frank Dernie]

You may be right that some / many products are adding things which at the end were not « in the record » itself. I have always liked Naim products which are controversial in some forums. But it is what it is, I perceive more dynamic and I have more fun with them.
As I have always used solid state amps, I have no opinion or experience vs tubes.
When it comes to RME ADI2 which I like, it happens that I have spent a full afternoon listening to music with Focal Utopia headphones (using Roon and a list of music I know very well) while reading a book. The experience was nice, music was good, very precise. Then I moved to my Chord Hugo2 and the experience was different, still very precise but such a fun! - I was just enjoying the music better. Same source, same music, same headphones, same day. It was not a blind test, but I didn’t expect it neither and frankly the respective price didn’t come into play contrarily to what some people repeatedly say in that forum.
Then the same music on the TotalDac + Vivid Giya 3, again something else. The same music for sure, but I could almost « see / feel » the musicians because rythm and 3D location are so good. I am sharing it again because it is based on an experience of listening. Is there a bias?maybe. But I am in mathematics by education, and science by profession. Why in my listening be the biais for price be bigger than the one for measurements now that I know the results thanks to Amir and Audioscience? I respect the measures and for sure the outcomes also guide me in acquiring some products (eg RME, Massdrop THX782, JDS Labs O2), but I know as well that I am not listening via an oscilloscope. So measures are not the only element which are important - your personal experience listening to the music and especially records you know is even more important.

Are you saying you can confidently pronounce that the totaldac sounds better than the RME when your comparison was not only not level matched and direct, but one was with headphones and the other in a room with speakers?

 

[BillG]

what’s so “special” and “mysterious” about R2R DAC

According to Wikipedia, not much, and I certainly don't see the need to sell one at $13K US:

"An R–2R ladder is a simple and inexpensive way to perform digital-to-analog conversion, using repetitive arrangements of precise resistor networks in a ladder-like configuration. A string resistor ladder implements the non-repetitive reference network."

https://en.wikipedia.org/wiki/Resistor_ladder

 

[JohnYang1997]

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.

The measurement protocol is the same regardless of measuring any topology. Accurate waveform reproduction is accurate reproduction. There is no difference between different dacs. We can only omit something, but never introduce artifacts.
You can come up with proper set of your measurements showing no main power supply noise and acceptable imd curve. Otherwise there's nothing wrong with this set of measurements. And amir didn't say anything about audibility. The only standard here is the price should reflect to objective measurements. This is the standard here. Nothing more nothing less.

 

[soekris]

There is nothing special about R-2R DACs that require any special treatment, in this case here is more the NOS type with its aliasing products that create problems. As a manufacturer of discrete R-2R DACs, I still want my DACs to measure ok, otherwise there is something wrong....
You just need to make sure when you do measurement, that you measurements really measure what you want.... Like, when measuring frequency response and Noise, you need to bandwitdh limit it. Or when measuring THD, make sure you do it at the correct levels, preferable at multiple levels. To show aliasing products, you do a higher frequency FFT.... And those FFTs showing noise at -160 dB and then claiming 160 dB S/N ratio are bogus, noise always have to be qualified with bandwidth, typically 20 Khz.

 

[daftcombo]

I am happy that after this thread nobody will prevent their children from going on holiday, using their money to pay 13.000$ for the sound quality that they could get in a 9$ dongle.

On the other hand, it is worrying that chinese employees are paid, let's say 13.000/9 less to produce the same kind of device.

 

[Patatorz]

My main message was to get more well engineered and sensibly priced products in the market. Totaldac doesn´t belong in either category.
Once manufacturers are aware of these measurements here and their possible effect on sales numbers.... maybe they will design better products?

P.s. I have bad sense of humour anyway...

 

[JohnYang1997]

Good measured r2r dac
https://www.stereophile.com/content...ned-edition-level-3-da-processor-measurements
It's not impossible to implement good measured r2r dac. But i still don't see the reason for it.

 

[daftcombo]

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.

What do you offer?

 

[Haskil]

Bonjour, Sorry for my english. I am French. I am a great admirer of your job on Audioscience for the transparency of real audio performances of Dac and others.

I do not say in the ear of my digital file: "attention, my boy, I send you in a DAC NOS, cover up." ...

To put an end to these "stories" :

- that Amirm redo some measures of signal/bruit that put off the laboratory feeding of incontestable way.

- Let Vincent Brient do the same at home for the results and publish them here. For comparaison of the two.

- That Vincent Brient publish here the measures which according to him are those which determine, them, the sound quality of a DAC what would not do or only partially those used by Audioscience, according to recognized standards.

 

[graz_lag]

IMHO, everything is explained here-below.

Definition of High Fidelity
Source: Wikipedia
Ideally, high-fidelity equipment has inaudible noise and distortion, and a flat (neutral, uncolored) frequency response within the human hearing range.
Source: Merriam-Webster
The reproduction of an effect (such as sound or an image) that is very faithful to the original.

Definition of Sonic Transparency
Source: Benchmark Media Systems
... that the product has been designed to be sonically neutral and transparent. We lead the industry with our wide-bandwidth, linear-phase, low-noise, low-distortion designs. Transparency can be measured and defined by these parameters ...
Is Transparency Right For You?
If you are looking for audio products that will change or "enhance" the sound of your music, you have come to the wrong place. If you are looking to add the warm sound of even-harmonic distortion, you have also come to the wrong pace. On the other hand, if you are looking for accurate, clean, and transparent audio equipment, you will enjoy Benchmark products.
Listening vs. Measuring
If a product is transparent, this will show in the measurements. A product that measures poorly is not transparent. But, a product that measures well is not necessarily transparent.
It is very important to verify transparency by listening. Many times, we have detected problems in listening tests that didn't show up in basic measurements. Usually this was an indication that the measurements were incomplete.
We have learned that if we hear something that didn't show up in the measurements, then we haven't done the right measurements. Over the years, our listening tests have helped us to refine our measurement techniques. We have learned how to detect and quantify defects that were initially revealed by our ears.
When measurement techniques are adequately refined, defects can be quantified and design changes can be evaluated objectively. Measurements then allow us to refine the product until the defect is reduced to inaudible levels.
Transparency in the Studio, Transparency in the Home
Transparency is essential in a studio monitoring chain but it is also important in a home listening environment. Transparent home systems can recreate the studio experience. If you want to hear your music the way the artists heard it in the studio, you will need a transparent playback system.
Many audio products are designed to euphonically enhance the sound of a recording. These products produce effects that change the sound in an attempt to improve the sound of a recording at playback. With these effects, the home playback may differ substantially from the studio experience. For this reason, Benchmark products do not add euphonic effects. Our products deliver an accurate rendering of the input signal.
Transparency Connects the Listening Space to the Performance Space
If the entire recording and playback chain is transparent, the listener can be transported to the performance space. This is the magic of music. It is enabled by the magic of a transparent audio chain - from the performance space to the listening space.
Transparency Captures the Magic of Music

 

[FredYves]

Are you saying you can confidently pronounce that the totaldac sounds better than the RME when your comparison was not only not level matched and direct, but one was with headphones and the other in a room with speakers?

I will repeat again.

My main system is made of AVM SA8.2 and Vivid Giya G3. TotalDac is connected to it.
I have a second system made of Hypex NCore and JM Reynaud Abscisse.
I also listening to music with headphones, Focal Utopia. If I need an headphone amp, I am just using Rupert Neve RNHP.

I connected first Chord Hugo2 and RME ADI2 to the main system. I already shared my personal conclusion, and I am forcing no one to have the same.

Then now, permanently, the Chord Hugo2 and the ADI2 are connected to my second system. I am also benefitting from them when I listen with headphones.

I have the 3 of them, I kept them. I have my preference, I have no issue to recognize it is mine based on what I like and certainly what I have listened to during years. And another thing which can be a matter of taste and not measurements: I do prefer the ADI2 with the Focal Stellia when I prefer the Chord Hugo2 with Focal Utopia. Then ADI2 is really fantastic to me, while I know what is the difference in « color » between Utopia ans Stellia

This is an experience at home with long hours of listening. I would never compare the experience in one room vs another one, and even less a DAC listened on one system vs a fully different system. And frankly I spend time listening to the music and not to the equipment.

And I hope it is not devilish to add that. Something else I am enthusiastic about is Astell&Kern SP1000 with Massdrop x Nobel Kaiser 10. Absolutely fantastic.

 

[totaldac]

People here are mixing up R2R and NOS.

 

[BDWoody]

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.

Wow...
Just...Wow...

Good thing you stopped by to get everyone sorted out.

Hope that 'custom analyzer' keeps working out for you.