Audio-gd Master 7 Singularity Review (DAC)

[filo97s]

as already seen in a great number of other Audio GD products, their output stages continues saturating. Great.
I think it's difficult to design a product worse than an Audio GD, i've seen way better designs out of some university students

 

[pma]

Technically horrible, it might be interesting to test in ABX if the poor engineering is audible. RCA high H2 harmonics is most probably reduced/cancelled with the balanced XLR out.

 

[Addicted to music]

Seriously, when people say it’s well built, really! What planet are you on when screw heads are protruding above the panels? Even some cheap DIY kits use countersunk screws.

 

[tikky]

I am in the business for almost 25 years and there are very few who can handle working on a track for more than 4 hours without losing the perspective. If the statement says "the best state will be after 1000 hours burn in" it's 250 working days for average usage. If you don't work weekends which those who have family, kids, and stuff won't, you will need at least 8 to 10 months to get this DAC working properly "according to the manual". It may be true but if you are going to wait 8 months for a DAC working at its best, this would be the product everyone should avoid at any cost for professional work.

This is where you should feel how professional they are.

 

[Gorgonzola]

I am in the business for almost 25 years and there are very few who can handle working on a track for more than 4 hours without losing the perspective. If the statement says "the best state will be after 1000 hours burn in" it's 250 working days for average usage. If you don't work weekends which those who have family, kids, and stuff won't, you will need at least 8 to 10 months to get this DAC working properly "according to the manual". It may be true but if you are going to wait 8 months for a DAC working at its best, this would be the product everyone should avoid at any cost for professional work.

This is where you should feel how professional they are.

I guess it take 1000 for listeners to become inured to the distorted sound and other quirks of behavior. I believed that others have made the point that many hours of "burn-in" is really the time for audiophile to accommodated his subjective preference to the sound of the component in question.

I to think that I briefly considered a Audio-GD DAC a few years ago Fortunately I decided on a Schiit Gungnir Multibit instead. The Gungnir sounded pretty good, however I discovered after a year or two that a Topping DX7s was as good or better and 1/3 the cost.

 

[KxDx]

Dunno about you all but I find it strangely entertaining to read reviews where Amir reports abysmal measurements from an ultra-expensive solid state component. I'm started to get bored of the Topping reviews with the -150dB or whatever SINAD.

We've pretty much reached peak DAC. It's entertaining to see how the other side lives.

 

[AudioSceptic]

What a waste of unobtanium PCM1704 chips!

Can someone explain what all that other stuff is and does? It is just a 2-channel DAC, isn't it?

 

[AudioSceptic]

@amirm needs a panther that is just a pile of broken, pink ceramic shards to classify devices that are overly-engineered with such stunningly poor performance. The headless panther just doesn't cut it in a case like this.

Yes, the cost needs to be factored in. This would be bad for any price, but for $2.5k? Is this actually the worst DAC ever?

 

[sarumbear]

By the way, there is such a thing as a "Class A voltage regulator" but it is usually not named that way; the more common electrical engineering term is SHUNT REGULATOR. They offer low source impedance, good regulation and make lots of heat.

From reading your words you know what you are talking about but will you call a DC regulator Class-A, which is a specific name for an audio amplifier (hence an AC amplifier) while power supply is a DC regulator? It is the same electronics they teach in Chinese schools!

It is obvious that they want to fool the gullible audiophiles who think Class-A is a good thing. This company is either using class A drugs or are frauding consumers.

 

[Tks]

Well, at least the interior looks expensive.

It's the ultimate fail. Where you actually look like you did something, but then seemingly made it too complex and it fails utterly.

So I'll say this (mentioned it years ago).

I'll never understand the sort of buffoon (though I know one, currawong of somewhat YT fame, loves these guys for some idiotic reason) one needs to be to be drawn to audio-gd products. They remind me of totaldac, but imagine if totaldac went straight to the East and to the worst possible manufacturing and design source.

SO, as a reminder. The first and foremost: Worst thing about this particular design, is the company logo. Looks like someone straight up ripped off a default script-type font from a photoshop bundle of fonts, and just went with it, with not an ounce of custom typography work. It looks literally something I'd expect out of a middle-schooler. I wouldn't insult any other Asian audio company (I mention them due to their sometimes silly choice of font on product pages of websites) to assume they'd ever put such a garbage looking thing on their product as a logo (especially at this price).

Second issue, and this is less superficial: This company straight up lies. They've been caught before as Amir can attest to. Yet no indication this has ceased AFAIK. This doesn't demand insults like I'm bringing, this demands lawsuits.

Third issue, what is this finishing work? Is this a higher end product or not?

Fourth, no volume control.. I know why, but.. why?

Fifth, everything looks two decades old, but with a hint of DIY of someone with enough money and charging just enough to opt for some milled aluminum.

Sixth, can't hold steady clock... They expect someone to buy a master clock or something lol? This is a disaster more than any THD, and more than any lack of filtering.

Seventh, say goodnight to your tweeters. If this doesn't do it, idk what will..

Eighth, sorry to the person who sent this. I mean you no disrespect, but... what the heck were you thinking?! Like if you got this used, perhaps in a time of ignorance before anyone tested one of these. I can somewhat stomach it if you got it 75% off or something (just based on looks, why would anyone get this over a Matrix Audio X-Sabre Pro for instance?). Otherwise, thank you for your loss. I don't say this in gest, it really was a treat to see another one of this company's garbage put in it's place definitively and not assume the prior was a fluke.

 

[AudioSceptic]

I have owned a few audio-Gd products, in fact I sent Amirm a DAC and a DAC/amp to test, and the results were miserable. Their build quality is kind of over-the-top which I thought would result in good sound. It does not.

By the way, there is such a thing as a "Class A voltage regulator" but it is usually not named that way; the more common electrical engineering term is SHUNT REGULATOR. They offer low source impedance, good regulation and make lots of heat.

Audio-Gd pays a lot of attention to power supplies. Their power supplies are quite good. Unfortunately, the rest of the circuitry and firmware leaves much to be desired as we can see from Amirm's analysis.

As I have said before, I would like to see a blind listening test to see how readily we can hear the difference between something like one of these "broken" designs and a decent design like the various nice Topping DACs. I think we would be surprised at the results of such a study. While it is my belief that some people could hear the difference, I don't think many of us audio hobbyists are nearly as good at hearing as we like to think.

This DAC is typical of Audio-Gd products, it seems to me. People buy their gear and report being happy with it. Surely if typical users were able to readily hear the flaws in these products, they wouldn't sell very many. But Audio-Gd enjoys fairly robust sales- and this makes me think that many people can't hear these flaws.

There is a logical problem here. If these poorly measuring devices sound as good to the average listener as well measuring ones, then the benchmark can be set really low, and spending $1000s on a DAC is even more of a waste and any old junk will do. So what are you paying for? A nice looking case and impressive but malfunctioning internals?

 

[Deleted member 36131]

It's not ears but the brains of these people that come to this consclusion.

Yes, even still....the purpose of measurements can be said to establish a correlation between sound and what is accepted as good measurements.

Here we have a device that measures horribly -yet- does not sound horrible. Even professional 'listening' tests conclude at minimal 'good' sound.

This is a contradiction of all that we expect.

Whats next? An audio designer using a cheap $400 amp and tweaking it to make it sound like a $20,000 Audiophile grade amp?

 

[raif71]

The google algorithm might work differently for them. Your google results are the produce of your footprint online and his translation by their algorithm.

Can't seem to find reviews of this DAC...I found 1 non-english video. Maybe not many people bought it.

 

[doug2761]

There is a logical problem here. If these poorly measuring devices sound as good to the average listener as well measuring ones, then the benchmark can be set really low, and spending $1000s on a DAC is even more of a waste and any old junk will do. So what are you paying for? A nice looking case and impressive but malfunctioning internals?

The good news is that the benchmark today, from a technical perspective, is really high AND the costs are really low. $100 will get you there. For enthusiasts it might be wise to try something like a D10. If you decide you really want a multi $k device then that $100 won't be consequential.

 

[douce]

Years ago I bought one of their headphone/dac/pre combo's... the left channel failed while using it as a dac/pre and almost destroyed my speaker after about 2 weeks of use. Tried to return it for a warranty replacement but the process was so difficult that I had to just take a loss.

needless to say the absolutely glowing reviews at that time were misleading, particularly now that I can see how they actually perform.

 

[KxDx]

I like products whose "burn-in" time exceeds the return window.

 

[blse59]

Vestige of the dark days of audio when everything was inside a black box and you had no idea what was inside. You only had to go with what the manufacturer said or what your imagination thought. This company has been hiding in the shadows for far too long.

 

[garbulky]

Can someone explain what all that other stuff is and does? It is just a 2-channel DAC, isn't it?

There is a transformer for every channel. I believe it also uses Krell's CAST tech. There is a third transformer that is a regenerative power supply that feeds the other two. Not sure what the middle circuit board is but left and right circuit boards are the left right channels separated out. Basically each channel is sort of dual mono.

 

[AudioSceptic]

This is a review and detailed measurements of the Audio-gd Master 7 Singularity DAC. It is on kind loan from a member. Can't quite find the cost for it but seems to have sold for $2,500 or so when it came out in 2018 (?).

I must say, this is one of the heavies DACs I have measured:

View attachment 152201

If you are paying for weight, you are certainly getting your money's worth! I was surprised that there is no volume control on the unit. The only control is to select one of 6 different inputs:

View attachment 152202

I get a kick out of the tape (upper left) that is supposed to tell them if you opened the unit because to change its settings, you have to change jumpers inside!!! Speaking of inside, here is a shot of it:

View attachment 152203

There no jumpers set in there but I can't figure out from their "manual" what that means. The English is quite poor. Owner had not modified it and I tested it as is. Company says after "May 15" settings are one way versus before. Problem is, it doesn't say which year!

Speaking of the manual, it starts this way:
View attachment 152204

Unbelievable.

Audio-gd Master 7 Singularity
Man, this DAC was an adventure to measure. I started with USB input whose jack is kind of crooked in the back. The clock in the dashboard was jumping back and forth every few seconds so I plotted the frequency by itself and got this:

View attachment 152205

So every two seconds or so, it jumps precisely from 1 kHz to 0.9993 kHz! Clearly there is a bug in the clock locking algorithm here. Fortunately it doesn't happen with the other two inputs (Toslink and Coax) which I tested. Because of the disturbance, I could not filter the FFT for USB so it looks more fuzzy than it would be otherwise:

View attachment 152206

Nominal output is around 5.3 volts or so. To make the comparison more fair I brought that back down to level you see with -2 dB attenuation on the input (it doesn't change the results hardly at all). We see that there are some 8 dB difference in distortion between one and the other channel. Averaging the two SINADs we get 85 dB which puts the Master 7 squarely in the poor category of all DACs tested:
View attachment 152210

Zooming in you can see its competitors better:

View attachment 152211

Two notches higher is the Google phone dongle!

Back to our dashboard, this is what you get when you switch to Coax and Toslink:

View attachment 152212

View attachment 152214

So much garbage and variability.

Moving on to dynamic range we get another poor result although not as bad as distortion:
View attachment 152215


I went testing IMD and for kicks, I also measured RCA out in addition to balanced which was used above:

View attachment 152216

The balanced output is not much better than a phone dongle but what on earth is going on with RCA out in green? So I went back and ran the dashboard with RCA Out:
View attachment 152217

What on earth are you doing here Audio-gd? Why produce so much more distortion with RCA as opposed to balanced?

Jitter test shows distinct difference between USB and coax+toslink:

View attachment 152218

So much for fancy PLL implemented in FPGA and such.

Considering what we have seen so far, Multitone results were much more decent:
View attachment 152219

Since the output is higher than 4 volts, let's sweep it to find the sweet spot:

View attachment 152220

Oh come on now, what is with that ditch at 1.5 volt? This curve should smoothly climb up and then maybe drop a little. It should never have this kind of nonlinearity.

Speaking of non-linearity, let's run our linearity test:

View attachment 152221

Notice how the level shifts as you get to some amplitude with a constant error. There is some kind of design problem here.

Filter in the output stage is quite slow (and hence wrong):
View attachment 152222

This is partially responsible for poor response in our THD+N versus frequency which uses wideband measurements:

View attachment 152223

The total noise and distortion literally jumps out of the chart above 500 Hz! Switching to 192 kHz sampling eliminated the effect of the slow filter (green and brown). Sadly performance is still quite poor and gets worse with frequency above 1 kHz. In that regard, our dashboard which also runs at 1 kHz is showing the best case response, not the worst.

Conclusions
Some of the earliest DACs I tested were from Audio-gd. Results were not only poor but different than what company advertised. This DAC was kind sent to me to see if their newer products do better. Sadly they do not. The Master 7 Singularity has a suite of problems in many areas. Clearly it has not benefited from proper design verification and optimization. Tons of parts are thrown at the problem which look good on paper and website but perform very poorly. Performance is simply unacceptable at any price let alone what they charge.

Needless to say, I cannot recommend the Audio-gd Master 7 Singularity.

------------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

It seems that Audio gd can produce well measuring gear. <https://hometheaterhifi.com/reviews/dac/audio-gd-as-1-dac-review/> Analysis by David Rich, no less.

 

[sarumbear]

I believe it also uses Krell's CAST tech.

Which funnily enough is not supply voltage sensitive. In other words you don't need excellent DC source if your amplifiers run in current mode.