Review and Measurements of Nobsound NS-DAC3 Pro DAC & Amp

[Johnny2Bad]

It didn't meet spec. That means either the manufacturer can't be trusted or it's defective.

If the manufacturer can't be trusted, there is no point in listening to their products. Let them be sold in the Dollar store and the Cellphone Unlocking strip mall retailers, where they belong with the other marginal electronic products of the world.

If the unit is defective, there is no point in auditioning it.

The net result is null plus null equals "life's too short" to invest another minute on it. It's bad enough that as amir's taken the time to setup a measurement regimen and feels obligated to report what he's done, that he has to compose a post on it.

If the manufacturer is serious, they will investigate themselves what the problem is and correct it, and contact amir and suggest he repeat his test from a random sample. Until that day, it's over.

 

[Jimster480]

I wonder how much of this just comes down to economies of scale? Once the CD format really hit its stride, CD players were sold in large numbers. I imagine that standalone DACs of the sort we're discussing here are much more boutique items. It's easier to invest heavily in engineering if you can amortize the design cost over a large number of units sold. Even with the earliest CD players being lower volume affairs, I imagine that Sony and Phillips were very committed to seeing the format succeed and would likely have been fine with taking a loss on that early hardware in exchange for building a solid revenue stream from CD sales (much like today's Sony takes a loss on Playstation hardware but makes it up on the games/licensing side).

I don't think they took a loss on it, back then these units were quite expensive.
But as time went on the performance could be had on smaller and smaller devices, some even running on AA batteries.
Recently (in the last decade) many companies have been studying how to make things worse to make more money (which is why the stock prices of companies that have figured this out have soared) because the average person either "doesn't know any better" or "its just good enough out of the box" or "they can buy a new one in 2 years".
All of this market research has gone into making universally lower quality components at a much cheaper price, while the cost of said components has been rising year after year (not the cost of the base components, but the finished product).

Audio is just one market that this has happened in, take a look at home appliances or even cars... Most are just built to last only around 5 years (and most coincidentally have 3-5 year financing) meaning that you will just get new ones as soon as your old ones are paid off.
Moreover this poor quality is subjected more to the USA than any other market, as Americans are some of the LEAST INFORMED consumers on the planet and are also the quickest to spend money.
Looking at phones today... the batteries have been mostly soldered in for the past couple years (so they know most people will replace their phone when the battery acts up in ~2 years) with the performance/specs of cheaper models actually falling (in respect to a few large brands) while these same brands are marketing "increased affordability".

 

[garbulky]

That is not what I said. I said that it is *provably* inaudible. Everything is honest about that. To be dishonest is to use a number like THD+N which is not based on psychoacoustics and saying "80 dB is good enough." Based on what? You can't prove that.

So you can hear a -80 db harmonic or similar distortion during music? That's what I'm getting at. I'm not arguing that there are better measuring devices which clearly there are.

 

[restorer-john]

I wonder how much of this just comes down to economies of scale? Once the CD format really hit its stride, CD players were sold in large numbers. I imagine that standalone DACs of the sort we're discussing here are much more boutique items.

In some ways yes. The investment was enormous, concerted and carried out with little compromises. The compromises came when target prices were to be missed and later when aggressive competition kicked in.

The retail prices in Australia of those machines in September 1983 were as follows:

I have some numbers someplace where the unit costs and key parts costs of early CD players were discussed by Sony engineers (many years later). I can't find it right now, but I recall the CDP-101's D/A converter chip itself was around $50 (due to lowish yields on the fabrication) and the decision was made after much audible testing that the interchannel delay was inaudible. So they used one D/A and time shared it. That was a very significant cost saving. Within a few years, there wasn't a CD player on the market without twin D/A converters as the costs had fallen massively and the 'benefits' had been labored by manufacturers who utilized two (or more).

These Chinese dac-in-a-box manufacturers are simply using off the shelf chips, with zero investment in R&D to design those components and surrounding them with the bare minimum of supporting circuitry, often straight out of the application notes on the datasheet. Even then, they strip out adequate rail regulators and filtering etc. Some of the internal shots Amir has posted and ones I have pulled apart are terrible. Then they make up specifications or just copy them from the D/A's datasheet.

Calling them 'boutique items' kind of intimates quality and exclusivity and they sure don't fit that criteria do they?

 

[amirm]

So you can hear a -80 db harmonic or similar distortion during music?

Who is "you?" If you mean me, how about your ears? Or someone else? And what music? There are 20 million albums out there. And what harmonic distortion? What is it made out of? Noise mostly? Harmonics? Which order? And why -80? How about -75, -85, -100, or 60???

If we are going to make a statement of inaudibility, it better be true of all people, all content and all conditions. Otherwise it is not inaudible.

We once ran controlled, independent third-party double blind tests that showed that listeners could not hear the difference between CD and just 64 kbps. How? We got lucky and the organization that ran the test picked content that was very friendly to lossy codecs. So at even 64 kbps (21:1 compression), average joe could not hear the difference. On the other hand, I can tell the difference at even 320 kbps. Where is the reality in that? Answer: lossy codecs are never transparent. It doesn't matter if you can't hear the artifacts at 256 kbps. It is not universal truth.

That's what I'm getting at. I'm not arguing that there are better measuring devices which clearly there are.

That has nothing to do with the argument I made. I said that at some level of distortion, we can guarantee inaudibility. At higher levels of distortion and noise, such a proof simply does not exist. It may very well be that audiophiles would flunk even a 50 dB test (Ethan has a test that shows that at even higher levels of distortion). But I can't prove that to be universal.

The only proof point is one that analyzes how good an audio "channel" needs to be to be transparent. That calls for a distortion and noise free dynamic range of 20 bits or 120 dB. You could derate this based on listening to lower levels, etc. But the more you do, the more you are on your own proving this.

Any other approach requires large scale studies that are not going to come about, especially since correlation between THD+N and audibility is very weak.

The only alternative left is getting devices as close to that ideal as possible. This, we can control.

 

[Headphonaholic]

I personally support Amir's decision to stop testing if the gear performs as poorly as this one does. Do you really think there are going to be riveting revelations in the other tests that redeem the device? At the end of the day there are better engineered products you can get for roughly the same cost of this device... so why bother with it? If you can't hear any issue with it and you like it, then by all means be happy with it. But as always, the point of what Amir is doing is to cut the crap and weed out poorly designed products. With many of the well engineered and measured products the differences are purely academic. If something fails to perform well out of the gate, we can imagine how the rest of the tests will go. Lets not waste Amir's time.

 

[Krunok]

That is not what I said. I said that it is *provably* inaudible. Everything is honest about that. To be dishonest is to use a number like THD+N which is not based on psychoacoustics and saying "80 dB is good enough." Based on what? You can't prove that.

Well, you can try to listen to it. My tube amp has THD <0.2% and SNR 70dB but I don't hear any difference in noise or distortion when I switch between my tube amp and my SS amp.

 

[Blumlein 88]

About the early CD players, yes they were over a $1000 (when $1k was equivalent to $2350 now). I remember as a poor working schmuck, saying, "when one hits $200 I'll buy it." Well I think it was a 2nd gen or early 3rd gen Philips unit. Roughly 18 months later. There was a Labor day sale of the Magnavox brand for $199. This had the 14 bit DAC chips (one per channel TDA1540 I think) with 4x oversampling. Opening that up you see a tremendous number of components and multi-layer boards. So these two behemoths had invested heavily to make digital happen.

 

[Krunok]

If we are going to make a statement of inaudibility, it better be true of all people, all content and all conditions. Otherwise it is not inaudible.

Don't get me wrong, but it's a good thing you don't work in pharmaceutical industry because you would be developing a single product for a cure for quite some time if you would want it to work without harm on "all people, all content and all conditions".

Answer: lossy codecs are never transparent. It doesn't matter if you can't hear the artifacts at 256 kbps. It is not universal truth.

You can say the same for RBCD format, or for any other digital format. When you convert original analog signal to any digital format that is not "universal truth", but a compromise. As was this device you were testing. The only thing that matters is if the compromise is acceptable or not, and to whom. If you have perfect hearing and speakers that cost $25000 that play in acoustically treated room than this device is not a good compromise for you. But if you are a student using this device on your table where you learn for exams paired with $200 desktop speakers, it might be.

 

[Blumlein 88]

Well Dolby labs might tell us we need a spurious free dynamic range of 129 db. I personally would be happy to declare a spurious free dynamic range of 120 db inaudible under all conditions.

More practically I would be happy to say spurious free dynamic range of 100 db is so close to always inaudible we need not worry. Any difference is so small as to be swamped by any other factor. Plenty of gear even inexpensive gear gets all over this. So I agree with Amir something with 80 db may not be a fatally flawed product, but we don't have to settle for it and I wouldn't put any more effort toward it. Heck phones a couple generations old can manage that.

I personally would think anything that can do a full level sweep, and a full level twin tone sweep with nothing above -100 db is going to sound fine for any use. The only other thing worth checking is flat frequency response. Anyone know of some condition that might pass this test and show other problems audibly?

 

[garbulky]

Who is "you?" If you mean me, how about your ears? Or someone else? And what music? There are 20 million albums out there. And what harmonic distortion? What is it made out of? Noise mostly? Harmonics? Which order? And why -80? How about -75, -85, -100, or 60???

If we are going to make a statement of inaudibility, it better be true of all people, all content and all conditions. Otherwise it is not inaudible.

We once ran controlled, independent third-party double blind tests that showed that listeners could not hear the difference between CD and just 64 kbps. How? We got lucky and the organization that ran the test picked content that was very friendly to lossy codecs. So at even 64 kbps (21:1 compression), average joe could not hear the difference. On the other hand, I can tell the difference at even 320 kbps. Where is the reality in that? Answer: lossy codecs are never transparent. It doesn't matter if you can't hear the artifacts at 256 kbps. It is not universal truth.


That has nothing to do with the argument I made. I said that at some level of distortion, we can guarantee inaudibility. At higher levels of distortion and noise, such a proof simply does not exist. It may very well be that audiophiles would flunk even a 50 dB test (Ethan has a test that shows that at even higher levels of distortion). But I can't prove that to be universal.

The only proof point is one that analyzes how good an audio "channel" needs to be to be transparent. That calls for a distortion and noise free dynamic range of 20 bits or 120 dB. You could derate this based on listening to lower levels, etc. But the more you do, the more you are on your own proving this.

Any other approach requires large scale studies that are not going to come about, especially since correlation between THD+N and audibility is very weak.

The only alternative left is getting devices as close to that ideal as possible. This, we can control.

Yes I get your point.
But...I think you are ignoring the issue by saying who. You aren't going to be hearing the coloration of what you consider to be poor performance of this unit and neither will I. Those measurements with those harmonics are just too low for audibility during music listening.
Though you worded it correctly - provably inaudible, you can see all the dissing this unit gets due to its distortion when it's fine. Just not as well performing as another one. But during actual use, what difference does it make if that 2nd or 3rd harmonic is -80 db or -150 db to your listening? Unless you are actually hearing a difference - sure then let's talk about it. Otherwise it's a data point. ( Makes a listening test useful imo. )

Also there's not a ton of tube gear that do a whole lot better. I've seen a tube amp with a THD of 3% ! So it's in this case worth giving consideration as to what kind of amp it is.

 

[Krunok]

. So I agree with Amir something with 80 db may not be a fatally flawed product, but we don't have to settle for it and I wouldn't put any more effort toward it.

What you have to settle with depends also on your budget, not only on your crtieria what is better.

To continue car analogy tradition of this forum: I'm fully aware that Aston Martin DB9 cabriolet beats my Volvo C70 in every aspect and I would gladly drive it, but budget doesn't allow it.

Btw, IMHO amps frequency response is pretty much always good enough even with poorly engineered products as it is anyhow masked by speakers nonlinearities. The same goes for THD/IMD.

 

[restorer-john]

Those measurements with those harmonics are just too low for audibility during music listening.

Next amplifier test I do, I'll upload a .wav of the THD+N residual at 0.01% (-80dB), notched out fundamental and normalised to 0dB for you to listen to. Residuals are some of the nastiest sounding grunge you will ever hear. Buried or not, I don't want it.

 

[Blumlein 88]

What you have to settle with depends also on your budget, not only on your crtieria what is better.

To continue car analogy tradition of this forum: I'm fully aware that Aston Martin DB9 cabriolet beats my Volvo C70 in every aspect and I would gladly drive it, but budget doesn't allow it.

Btw, IMHO amps frequency response is pretty much always good enough even with poorly engineered products as it is anyhow masked by speakers nonlinearities. The same goes for THD/IMD.

On the speaker masking thing........not so fast partner!

A speaker could have 1% distortion. You under the right conditions could detect a difference between 1% and 1.1%. So you need the amp to get to .1% or less even with speakers that aren't that good.

 

[Blumlein 88]

Pharmaceutical comparison is crap. Analogies are usually only for the un-educated and usually wrong.

Stick to the relevant(audio} science.

Want to reconsider your position on analgous waves and digital?

BTW, one of my good friends always warns me about analogies.

 

[garbulky]

Next amplifier test I do, I'll upload a .wav of the THD+N residual at 0.01% (-80dB), notched out fundamental and normalised to 0dB for you to listen to. Residuals are some of the nastiest sounding grunge you will ever hear. Buried or not, I don't want it.

Are you saying the -80 db is amplified to 0 db? I imagine that would sound terrible!

 

[Blumlein 88]

Next amplifier test I do, I'll upload a .wav of the THD+N residual at 0.01% (-80dB), notched out fundamental and normalised to 0dB for you to listen to. Residuals are some of the nastiest sounding grunge you will ever hear. Buried or not, I don't want it.

I started to give you a like on this. But I reconsidered.

For a total non-sequitor, and mixing car analogies how about this:

I've recorded the low frequency resonances of my Corvette. Related to the torsional stiffness of the entire car. I think it was 21 hz. Anyway, having measured a few different cars this way I can speed them up and amp them so you can listen to chassis stiffness along with lots of grunge. What does it mean? Ask @Frank Dernie as I don't know.

 

[Blumlein 88]

Are you saying the -80 db is amplified to 0 db? I imagine that would sound terrible!

That is exactly what he means.

BTW, whats with the generic Avatar pic. Come on man, get with the program. Pick a gooffy audio format logo, or a pretty girl. My preference is the pretty girl.

 

[Krunok]

On the speaker masking thing........not so fast partner!

A speaker could have 1% distortion. You under the right conditions could detect a difference between 1% and 1.1%. So you need the amp to get to .1% or less even with speakers that aren't that good.

I'll admit this is true although I have deliberately chosen so high figure to make a point, as with this amp we are talking about THD of 0,0085%.

 

[Krunok]

Want to reconsider your position on analgous waves and digital?

BTW, one of my good friends always warns me about analogies.

Was this aimed at me?