Review and Measurements of NuForce DAC80

[The Mule]

Newbie here. I am just a pig in sh*t with all the reviews and info here. (I am a big fan of the Audio Critic, so I am a very big objectivist. Amir: you just plain rock. A buddy turned me on to you, a long time ago during the initial HD DVD/Blu-ray format war. )

This is just funny, I have been looking at DACs lately, and I in fact just got a Topping D30 based on the review here (sounds great, using it with a WA7), and I had come across this NuForce DAC simply looking at what was for sale on eBay within certain price ranges. I thought it looked cool, so I looked up reviews (and measurements). ... Too bad it doesn't perform as well as it looks.

 

[amirm]

Newbie here. I am just a pig in sh*t with all the reviews and info here. (I am a big fan of the Audio Critic, so I am a very big objectivist. Amir: you just plain rock. A buddy turned me on to you, a long time ago during the initial HD DVD/Blu-ray format war. )

Wow, you and I go way back then. Welcome to the forum and thanks for the kind words and donation!

 

[Jimster480]

Newbie here. I am just a pig in sh*t with all the reviews and info here. (I am a big fan of the Audio Critic, so I am a very big objectivist. Amir: you just plain rock. A buddy turned me on to you, a long time ago during the initial HD DVD/Blu-ray format war. )

This is just funny, I have been looking at DACs lately, and I in fact just got a Topping D30 based on the review here (sounds great, using it with a WA7), and I had come across this NuForce DAC simply looking at what was for sale on eBay within certain price ranges. I thought it looked cool, so I looked up reviews (and measurements). ... Too bad it doesn't perform as well as it looks.

Glad you ended up with a D30 over a NuForce.
Welcome to the forum!

 

[El_Konrado]

Glad you ended up with a D30 over a NuForce.
Welcome to the forum!

I will be honest with you guys, I realise some of the measurements suggest that D30 is a better DAC than NuForce. Unfortunately this doesnt reflect reality. I am not 100% certain on NuForce DAC80 as the one I am using is UDH-100, its pretty much the same thing but it only has USB input and nothing else, other than that it is 99% the same thing. before this I was using Topping D30 for nearly a year, listening every single day and really being happy with the sound especially bearing in mind the price. I can tell you though that UDH-100 sound-wise simply destroys D30. The difference is massive in favour of NuForce, I would say similar step forward like going from decent integrated soundcard on your motherboard to D30 ... All my recordings sound like a new thing now, way more details, now I realise how poor the midrange of D30 sounds. These SINAD values, etc., they do not tell you everything, this is really valuable to know this figures, Amir is doing really good work, but do not think you take 10 units, compare the graph, pick one with highest SINAD value and you found the best DAC on the planet. Only comparing units side by side is the right way to do this and results of NuForce vs D30 are the best example. Anyone who listened to both these DACs already know how much better the NuForce sounds, how much more neutral and real, the difference is absolutely massive yet if you would choose only numbers D30 looks like a much better unit but it is not. Always make sure to listen to many units, compare in your system, best if you can rent from your local shops and return if you dont like them, you will be amazed how great some of them sound

 

[sbo6]

Just bought one used before reading this review and it sounds outstanding! Much fuller bodied, more detailed and more dynamic than my previous AQ Dragonfly red! Highly recommended.

 

[peanuts]

this review is a doozy. completely incompetent design
http://nwavguy.blogspot.com/2011/02/nuforce-udac-2-drama.html

to top it off, nuforce`s responses at headfi is hilarious.
https://www.head-fi.org/threads/nuforce-udac-2-drama-detailed-measurements.542281/

 

[Jimster480]

I will be honest with you guys, I realise some of the measurements suggest that D30 is a better DAC than NuForce. Unfortunately this doesnt reflect reality. I am not 100% certain on NuForce DAC80 as the one I am using is UDH-100, its pretty much the same thing but it only has USB input and nothing else, other than that it is 99% the same thing. before this I was using Topping D30 for nearly a year, listening every single day and really being happy with the sound especially bearing in mind the price. I can tell you though that UDH-100 sound-wise simply destroys D30. The difference is massive in favour of NuForce, I would say similar step forward like going from decent integrated soundcard on your motherboard to D30 ... All my recordings sound like a new thing now, way more details, now I realise how poor the midrange of D30 sounds. These SINAD values, etc., they do not tell you everything, this is really valuable to know this figures, Amir is doing really good work, but do not think you take 10 units, compare the graph, pick one with highest SINAD value and you found the best DAC on the planet. Only comparing units side by side is the right way to do this and results of NuForce vs D30 are the best example. Anyone who listened to both these DACs already know how much better the NuForce sounds, how much more neutral and real, the difference is absolutely massive yet if you would choose only numbers D30 looks like a much better unit but it is not. Always make sure to listen to many units, compare in your system, best if you can rent from your local shops and return if you dont like them, you will be amazed how great some of them sound

This is just fake.

Its called expectation bias.
I would say that Linearity is what makes the biggest difference in his measurements.
Before I found this site I had compared many DAC's by my own listening and "ranked" them.
Later on as Amir tested them it basically confirmed what I had originally ranked.

Linearity is the ability to render samples to a specific bit depth (with accuracy) and those DAC's that scored better on the linearity test also were the ones that sounded more detailed to my ears.
This DAC80 fails the linearity test harder than the D30.
So either you like distortion or you are living in a world of expectation bias.
The D30 is worse in SPDIF but the DAC80 is even worse than the D30 when it comes to SPDIF so...

 

[sbo6]

This is just fake.

Its called expectation bias.
I would say that Linearity is what makes the biggest difference in his measurements.
Before I found this site I had compared many DAC's by my own listening and "ranked" them.
Later on as Amir tested them it basically confirmed what I had originally ranked.

Linearity is the ability to render samples to a specific bit depth (with accuracy) and those DAC's that scored better on the linearity test also were the ones that sounded more detailed to my ears.
This DAC80 fails the linearity test harder than the D30.
So either you like distortion or you are living in a world of expectation bias.
The D30 is worse in SPDIF but the DAC80 is even worse than the D30 when it comes to SPDIF so...

I guess anytime anything sounds better that isn't justified with numbers = expectations bias.... Net - don't trust what you hear trust what someone else tells you is "better' based on some measurements...

 

[Jimster480]

I guess anytime anything sounds better that isn't justified with numbers = expectations bias.... Net - don't trust what you hear trust what someone else tells you is "better' based on some measurements...

Yes, you cannot trust what you hear because your mind can deceive you with expectation bias.
Research Psychoacoustics and you will find this to be true.

 

[sbo6]

Yes, you cannot trust what you hear because your mind can deceive you with expectation bias.
Research Psychoacoustics and you will find this to be true.

I'm fully aware and understand expectation bias, but at what point can you trust your ears? Do the majority in this forum avoid and omit what they hear (even if better) and only go by numbers? At what point do you listen and enjoy what you hear as better? Or do you continue to only go by numbers? IME both matter.

 

[amirm]

I'm fully aware and understand expectation bias, but at what point can you trust your ears?

You can trust them all the time. Just do the testing blind and with levels matched so that your brain doesn't override what is coming into your ears.

 

[Jimster480]

I'm fully aware and understand expectation bias, but at what point can you trust your ears? Do the majority in this forum avoid and omit what they hear (even if better) and only go by numbers? At what point do you listen and enjoy what you hear as better? Or do you continue to only go by numbers? IME both matter.

I mean my ears told me that better linearity = more clarity.
Some people "go by their ears" in that more bass = better.
So they like muddy music with lots of bass and therefore claim that worse measuring products are better because they produce more muddy bass that they like to listen to.

But I can tell you that with all the DAC's I have listened to; the differences have always been small. You have to do controlled listening tests with good headphones back to back to discern any difference.
Also you need to listen to music with high / good dynamic ranges and high levels of detail.

If you want to really compare two DAC's you need to level match them and get an input switcher or use an amp with an input switcher.

 

[audimus]

Expectation bias works the other way too. There are people who hear only “what is measured”. It measured very clean and so it sounds good. Just like it cost me $3000 so... etc. Human auditory system isn’t a machine and it varies from person to person. Even if you eliminate psychological aspects somehow.

This debate will go as long as we assume we can hear only what we measure or that we can measure everything that “we hear”. That is an axiom which is difficult to prove. I suspect there is even a left brain vs right brain thinking differences between people who go by what they hear and those who go by what is measured. People who have been musicians or attend live concerts “hear” differently than people who have never experienced either.

Eventually, we have to go by what we hear because we have to live with it. If something does not sound good to me, it does not matter what it measured. If something measured bad (or unusual) within limits we cannot necessarily eliminate it from consideration. The measurements are just numbers. They don’t tell you whether the difference between two numbers will necessarily make a difference to you in particular.

These measurement tests are good and necessary as ONE of the inputs to the decision process. They might also explain something one may have experienced in listening tests but aren’t the final word on what one can expect to hear subjectively to use it as the selection criterion.

Note the current excitement in these forums on acquiring the Okto DAC unseen and unheard based on measurements. No guarantees that it will sound great to everybody but there will also be people who will also be convinced it sounds great because it measured so well. Expectation bias ... a dual-edged sword.

 

[amirm]

Note the current excitement in these forums on acquiring the Okto DAC unseen and unheard based on measurements. No guarantees that it will sound great to everybody but there will also be people who will also be convinced it sounds great because it measured so well. Expectation bias ... a dual-edged sword.

Oh, there is a guarantee. Based on psychoacoustics, i.e. years of research, we *know* such a DAC has reached transparency for all content, and all people.

That is the power of measurements showing excellence in engineering. It erases the gray area of whether something is audibly good or not. We can prove it is.

Now, you could buy it, and using all of your emotions, preconceptions, and incorrect testing decide "it doesn't sound good." That simply indicates improper subjective listening test. It has no bearing on factual performance of the device.

Eventually, we have to go by what we hear because we have to live with it.

For that, you can look to the features of the device. What we do here, takes care of audibility issues when you select from top-tier products. Again, if you want to convince yourself you are sick, even if you are not, you can. But don't do that.

 

[audimus]

Oh, there is a guarantee. Based on psychoacoustics, i.e. years of research, we *know* such a DAC has reached transparency for all content, and all people.

That is the power of measurements showing excellence in engineering. It erases the gray area of whether something is audibly good or not. We can prove it is.

Now, you could buy it, and using all of your emotions, preconceptions, and incorrect testing decide "it doesn't sound good." That simply indicates improper subjective listening test. It has no bearing on factual performance of the device.


For that, you can look to the features of the device. What we do here, takes care of audibility issues when you select from top-tier products. Again, if you want to convince yourself you are sick, even if you are not, you can. But don't do that.

Sounds like a circular definition of “audibily good” (as an abstract quality of sound) and the measure that defines that quality and conflating that with audibly “good” as perceived.

Just to see if I understand you correctly

Let us say we run a perfect double blind listening test between the Okto and the Essence here (just using as an example, have no relation with any manufacturer or vendor).

Are you saying
1. Most from a random sample of people will prefer the Okto OR
2. The people, (even if a majority) who don’t prefer the Okto, have some deficiency to positively perceive what is “audibly good” OR
3. Both?

Let us run another thought experiment. Say we gather all the people that purchase the Okto based on the measurements here and start to rave about it. We will give them a well-designed double blind test between the Okto and the Essence.

Are you saying
1. Most or all of them will pick out the Okto as the superior one of the two for their listening? OR
2. The people who see no difference or pick the Essence were victims of their expectation bias based on measurements for their earlier exuberance? OR
3. The latter in 2 are somehow incapable of perceiving “audibly good” as perceivably good?

Your expectations of the outcome of such an experiment...

 

[amirm]

Let us say we run a perfect double blind listening test between the Okto and the Essence here (just using as an example, have no relation with any manufacturer or vendor).

Are you saying
1. Most from a random sample of people will prefer the Okto OR
2. The people, (even if a majority) who don’t prefer the Okto, have some deficiency to positively perceive what is “audibly good” OR
3. Both?

Vast majority of people would flunk this test and would not be able to tell these apart.

I cannot guarantee that in a transparency test against a source, the Essence would get there for all people and all content. Psychoacoustics says that we can find cases where transparency is not achieved.

I can guarantee that in a transparency test against a source, the Okto would, for all people and all content.

Think of it this way: put a bottle of bleach on the counter and test it a month later for any bacteria. You will find none. Now do the same for milk. Naturally it will spoil and will be full of bacteria. There is a spectrum between these two. To the extent you are at one end of the other, the answer is simple. That is what Okto is. It is the bleach. I can't tell you if a glass of water a month later has bacteria in it. That is Essence.

 

[amirm]

Sounds like a circular definition of “audibily good” (as an abstract quality of sound) and the measure that defines that quality and conflating that with audibly “good” as perceived.

I said nothing about "audibly good." Almost everything I test is "audibly good." I am talking about what I can prove based on measurements alone. If the measurements reach a certain level, then we can be confident of transparency. That is a far higher bar than "audibly good." MP3 is audibly good. Even FM radio is audibly good.

What is transparent audibly has a definition and it calls for noise and distortion being below threshold of hearing:

At mid-frequencies, the curve dips below 0 dB SPL. For simplicity, lets assume it is 0 dB. Research into how loud live music can sound shows that it can be as high as 120 dB. Using that, you need to make sure your distortion products are -120 dB. Now, turns out your hearing has a limit below that of about 116 dB. So anything that has distortions below that based on psychoustic research can be shown to be distortion-free.

The above is a classical proof of inaudibility used in audio research in countless papers. So it is not some made up notion that can be dismissed out of hand.

 

[audimus]

Vast majority of people would flunk this test and would not be able to tell these apart.

You are then reinforcing one of my earlier points that the expectation bias can be based on reading test results. I am sure there will be plenty of people waxing poetically about how good the sound is after using the Okto purchasing it based on the tests here. And yet, if most of them are unable to tell it apart from the Essence (or whatever they replaced) in a blind test, then their perception is indeed colored by reading the measurements itself. The same way people were accused of liking a particular unit even if it did not test well. Just doing measurements has not solved the bias problem, it has introduced a new expectation bias.

I cannot guarantee that in a transparency test against a source, the Essence would get there for all people and all content. Psychoacoustics says that we can find cases where transparency is not achieved.

I can guarantee that in a transparency test against a source, the Okto would, for all people and all content.

Think of it this way: put a bottle of bleach on the counter and test it a month later for any bacteria. You will find none. Now do the same for milk. Naturally it will spoil and will be full of bacteria. There is a spectrum between these two. To the extent you are at one end of the other, the answer is simple. That is what Okto is. It is the bleach. I can't tell you if a glass of water a month later has bacteria in it. That is Essence.

Actually, I would like bacteria in my yogurt than bleach.

But seriously, what practical use is such guarantees on transparency? We already know (from devices in the chain that are far less linear) that what people like to hear and enjoy (subjectively) is not necessarily correlated with transparency. One DAC could sound warmer than others and be preferred as enjoyable music by some and allowing them to listen without fatigue for hours. Another DAC could sound bright and be preferred by people who like it for what it does for their home theater. Sure, you could possibly measure both and point out how less “transparent” both are in some metric but that is just of academic interest.

In the method of science, measurement is just means to an end not the end. Good theories build an explanation for observed phenomena and are considered useful and strong if they have predictive abilities.

It would be far more useful to build a measurement based theory on how the behavior of a device in how it is perceived is correlated and/or predicted by the metrics. Then one could say if you prefer warmer harmonics or greater detail or bright and clear conversations are whatever, X is a better choice than Y. Even more useful is predicting the nature of sound that emanates from a chain of components based on the measurements of each and have it be borne out by listening tests.

If the above can be reliably done by measurements, then one wouldn’t need to rely on the “Golden Ears” in magazines possibly with all their biases giving such pronouncements.

All these measurements to establish some perfect transparency isn’t really helping in practice other than to create another form of expectation bias (I like it - it tested great) or resulting in unnecessary overspending. I paid $1k for this because it tested great and it sounds great (but not likely to be able to tell the difference with a $40 DAC).

In practice, what these tests are saying is that a device A is introducing more “colouration” than a Device B, but you may or may not be able to tell the difference and furthermore, you may even like listening to device A more than device B for linger without fatigue. No guarantees one way or the other. Such measurements should NOT be relied upon to make equipment selection.

This is not a knock against the work being done here. It is a comment on certain interpretations made based on the measurements especially to influence purchase. That is flawed.

 

[audimus]

I said nothing about "audibly good." Almost everything I test is "audibly good." I am talking about what I can prove based on measurements alone.

I was referring to this earlier statement

It erases the gray area of whether something is audibly good or not. We can prove it is.

It was conflating “measurably good” with “audibly good” leading to a circular definition.

 

[sbo6]

But seriously, what practical use is such guarantees on transparency? We already know (from devices in the chain that are far less linear) that what people like to hear and enjoy (subjectively) is not necessarily correlated with transparency. One DAC could sound warmer than others and be preferred as enjoyable music by some and allowing them to listen without fatigue for hours. Another DAC could sound bright and be preferred by people who like it for what it does for their home theater. Sure, you could possibly measure both and point out how less “transparent” both are in some metric but that is just of academic interest.

In the method of science, measurement is just means to an end not the end. Good theories build an explanation for observed phenomena and are considered useful and strong if they have predictive abilities.

Took the words out of my head, exactly. If the goal is to measure to ensure transparency, congrats you then know what is (or isn't) transparent. How about every other sonic parameters that we consider when purchasing audio products like dynamics, warmth, bass output, spacial cues, sound stage width / depth, etc. Show me how you measure and capture these attributes in a chart or graph? Answer: you can't, measurements are only a piece of the puzzle.