[SoupRKnowva]

one thing im curious, and hopeful about nwavguy. i love your very techincal measurements, but do you think you could add a Phase versus frequency chart for all your forthcoming reviews? and maybe even to the old one...pretty please. one thing not many people look at is phase distortion in sources/amps

 

[nwavguy]

Quote:

According to Stereophile, definition of clipping is '1% THD+N'.

The 1% figure isn't really the edge of actual clipping, it's just the way power amplifiers are tested to be consistent. And that number is generally used where you increase the output level until 1% distortion is reached and then you report the power at 1% THD+N. That's how I try to rate the maximum output power of headphone outputs in my reviews.
 
The actual edge of clipping is where the distortion suddenly rises a large amount with a very small increase in the level. And it's also generally where the 3rd harmonic becomes greater than the 2nd harmonic (circled and shown in some of my distortion spectrums in the review).
 
But the kind of clipping I'm concerned about with the uDAC-2 has nothing to do with the output power of the uDAC-2, or with the setting of its volume control. It's the internal circuitry before even the volume control that's clipping on the input signal. And that's just abnormal. You don't want any internal clipping in a digital device--even at 0 dBFS input. There's no good argument for having it.
 
What can be argued is if the approximately 0.7% of distortion I measured is actually audible. That's harder to answer. Some would say yes, some would say no. Lots of papers have been written on distortion and how audible it is and they don't all agree with each other.
 
It depends a lot on the kind of distortion. Odd harmonics, as from the clipping with uDAC-2, are generally considered far more audible than even harmonics. And the type of music being played makes a big difference. Already distorted heavy metal, for example, will mask lots of distortion in the playback equipment. But well recorded acoustic music, especially certain instruments, can be much less forgiving. And if you're listening to cheap $5 earbuds, you probably won't notice 0.7% distortion because the earbuds probably have 5% by themselves. But if you have a nice pair of $200 headphones, that's very different. So it's far from being black and white.
 
 

 

[nwavguy]

Quote:

This will be our only response to the uDAC-2 measurement. We received the email about your measurement on Friday afternoon. And our measurement lab is in Taiwan, close to our contract manufacturer.  That's why we have to travel to another company's lab to take measurement, in order to check out your claim. We have design staff in US and Taiwan and the engineer who is responsible for uDAC-2 works in Taiwan, which is already Saturday. 
After all the drama as you called it, we reached the following conclusion. And I am copying it directly from our email without any change.
Again, this is the way we have decided to design uDAC-2 and overwhelming majority of customers love and rave about the sound quality. If someone choose to listen with their measurement equipment, there is nothing we can do.  I will conclude with another example such as our OPPO BDP-93 Nuforce Edition. We designed it so that it sounded like solid state triode, with a FR curve that looks WORST than the original BDP-93 and some people were also complaining about it. Our BDP-93NE has other measurements that are much better and I won't go into it here. My point is that we can not simply rely on measurement to make decision for good sounding audio products.
uDAC-2 volume imbalance at low level is a big issue for a few people and they complained loudly to us. A digital volume control would fix this problem. But we insist on using this rather expensive volume pot imported from Japan because IT SOUNDED GOOD!  Believe me, we have tried to switch volume pot, but to us, better sounding product still win hands down over better measured product.

 
Many manufactures of "esoteric" equipment fall back on some version of "better sounding product still win hands down over better measured product" as NuForce is doing above and I talk about that in my review. Sometimes their equipment changes the sound in ways some people like. Some people like, for example, the sound of vinyl LP albums with all the noise, ticks, pops, etc. The HiFiMan portable players roll off the high frequencies for a warmer sound and some like that. Some like single-ended triode amplifiers with their very high levels of distortion. The science behind this has been well documented. Here's a presentation on "the sound of distortion" (check the slide show): http://www.pmillett.com/etf_sod.htm
 
But sometimes "esoteric" equipment doesn't change the sound at all, or it changes in ways most would consider to be worse than equipment that measures better. And with something like the high 6 ohm output impedance of the NuForce uDAC-2, the sound will change in ways that entirely depend on the headphones it's used with. And the uDAC-2 clipping distortion will depend on the signal it's driven with and the volume settings on the PC. So these are not predictable sorts of changes in the sound. And if they're not predictable changes, they're probably not good changes. It would be like a gun where you never know which way the bullet will go when you pull the trigger. Predictable is good while unpredictable is usually bad.
 
How about this: If NuForce wants to emphasize the sound of their products, I will attempt to make proper recordings of the sound of the NuForce uDAC-2, Behringer UCA202, and my Benchmark DAC1 Pre (as a reference) via the line outputs at a typical listening volume and I'll make excerpts available from my blog for downloading. That way anyone can listen for themselves and decide which DAC sounds best to them. I will carefully level match the signal from all 3 and record using the $1800 Benchmark ADC1 which is capable of extremely high quality transparent recordings. I'll document the exact set up so someone else can reproduce and verify what I did if they desire.
 
So no measurements, just 3 different DACs playing the same music, connected to the same transparent ADC, and recorded the same way. If NuForce is confident the $129 uDAC-2 really has been tuned for the best sound, it should at least have an advantage over the $29 Behringer, right?
 
I haven't tried the above with these particular devices, but I have in the past in other situations and I can say the results can be rather interesting. Sometimes differences show up where you're not expecting them, and sometimes the opposite happens. I'm willing to do it here because the Behringer measures so much better than the uDAC-2 I'm confident it at least won't sound any worse. I'd also suggest the actual test, and the results, are more appropriate for Sound Science forum as that opens up more options in the ways people can compare the sound files.
 
It would be nice to also run the same test using the headphone outputs driving real headphones, but I can tell you right now the results you'll get will depend mostly on which headphones I choose. The output impedance of the Benchmark is essentially 0 ohms, the uDAC-2 is 6 ohms and the Behringer a high 50 ohms. So the only one of the 3 that won't be affected by my choice in headphones is the Benchmark DAC1. If I succeed in my planned $5 upgrade of the Behringer's headphone amp to lower the impedance, that might make for a more interesting challenge to the uDAC-2 using headphones.
 
So does this seem like an interesting idea? Are there Head-Fiers here who would like to download the files and listen for themselves?

 

[129207]

Ofcourse I'd like to hear it! Problem is, I'll have to run your downloads through my own DAC which will then colour the sound. Differences should be noticeable though. Look forward to it! 

 

[Bacci]

Great stuff! How about testing some DIY DACs (GrubDAC, AMB Gamma 2, ...)? Contrary to commercial companies, the DIY community would be very grateful for suggestions made by someone with your skill and knowledge!

 

[nwavguy]

Quote:

Great stuff! How about testing some DIY DACs (GrubDAC, AMB Gamma 2, ...)? Contrary to commercial companies, the DIY community would be very grateful for suggestions made by someone with your skill and knowledge!

Thanks for the encouragement! I've done a bit of DIY testing for others, and in the end, it felt more like I was bursting a kid's favorite balloon with a big sharp pin. I hate to say it, but DIY gear often measures rather poorly--in part because the person who built it had no way of properly measuring it and often designed it "by ear" much like NuForce argues they have done with the uDAC-2. It's sort of like someone trying to paint the inside of their bedroom mostly in the dark. I don't like being the guy who turns the bright lights on for the first time.
 
When you design by ear it typically doesn't work out well for some very well documented reasons. Those interested can read about some of them here: http://seanolive.blogspot.com/2009/04/dishonesty-of-sighted-audio-product.html So the end result often is rather disappointing when it's subjected to the "bright light" of a proper instrumented test--especially for the first time.
 
While I don't feel too bad offending a company making a profit selling stuff that doesn't measure well, it's a whole 'nother thing to shine that bright light on one person's not-for-profit personal creation and labor of love. So I would only want to measure a DIY project if nobody had a big emotional attachment to it.
 
But I agree it would be a good thing to do if several others were considering replicating the same project. It would make sense to refine the design as much as possible before finalizing the PCB layout, doing a group buy of components, etc.
 

 

[jasonl]

Your statement that we designed by ear is incorrect. We take both approach. What I said was that if we have to choose between an approach that measured good versus another approach that sounded good, when both approaches are feasible, we choose the one that sounded good. In general, when something measured very good, it does sound good.  So our approach to design is that we trust the instrument first, then we do the listening test and tweaking later.
 
Take cryogenic treatment for example, I was never a believer. There is no detectable change in measurement, whether on the DAC or high-end power amp. But my ear and many other people who have tried it tells me otherwise. We compared the cryo version of uDAC-2 with standard version, and cryo version sounded much better.  Perhaps there are other effects that can't be easily measured. Well, may be you can figure it out.
 
Quote:

Thanks for the encouragement! I've done a bit of DIY testing for others, and in the end, it felt more like I was bursting a kid's favorite balloon with a big sharp pin. I hate to say it, but DIY gear often measures rather poorly--in part because the person who built it had no way of properly measuring it and often designed it "by ear" much like NuForce argues they have done with the uDAC-2. It's sort of like someone trying to paint the inside of their bedroom mostly in the dark. I don't like being the guy who turns the bright lights on for the first time.
 

 

[Wildstar]

From what I understand, the signal is boosted by 1dB and then fed to the DAC? This may improve dynamic range slightly, but at the expense of guaranteed distortion for any signal higher than that. The point of a DAC is to convert the full range signal, not just a part of it. If the device does not have sufficient dynamic range, "extending" it like this is nothing short of deceiving the consumer (not acceptable for this kind of market), and as an engineer, I would refuse to make such a design decision, if forced by my superiors I would heavily consider resigning.
 
I also do not understand how a 24bit DAC could have "too low" output level with "mp3s" (which are rendered to 16 bit) unless the scaling logic is grossly incorrect.
Finally, I can attest that many, if not most, modern pop music is mastered to 0 dBFS, in fact often more, having outright and quite noticeable clipping.
 
 

So does this seem like an interesting idea? Are there Head-Fiers here who would like to download the files and listen for themselves?

I suggest naming them with number or letters in order to keep it blind, at least for a few days. This would avoid psychological bias.
 
 

 

[attenuated 3db]

This will be our only response to the uDAC-2 measurement. We received the email about your measurement on Friday afternoon. And our measurement lab is in Taiwan, close to our contract manufacturer.  That's why we have to travel to another company's lab to take measurement, in order to check out your claim. We have design staff in US and Taiwan and the engineer who is responsible for uDAC-2 works in Taiwan, which is already Saturday. 

Thanks for the chronological clarification of the events related to this thread as they transpired from your perspective, Jason. In this day of instantaneous communications via e-mail and sometimes equally instantaneous dissementaion of those exchanges with a worldwide audience of owners and potential owners of your products on the web, it can put small businessmen in a bind. Despite how fast communciations may be, a lot of the world still runs at less than the speed of light (I know I do, especially early in the morning), and with date and time zone differences, it gets even more complex.

The whole thing reminded me of a dust-up over an Audio-gd product, where Kingwa, the effective CEO of Audio-gd (he needs a COO to share the workload with him, IMHO) posted this:

http://www.head-fi.org/forum/thread/524263/audio-gd-nfb-12/585#post_7260384

I love Kingwa (and by extension) Audio-gd, but your Taiwanese engineer has better English, as does Ryan Ping of Maverick Audio, whom I likewise "love." You were very prompt in replying to my PM inquiry about how I might go about swapping my uDAC-2 HP for the $30-more-expensive uDAC-2. I very much enjoy my unit for both its compact, rugged build quality and perceived "musicality" to my aging ears.

Keep up your good work. I'd like to listen to some of your higher-end products someday (nice range of equipment in many different categories), but don't live near any potential "meets," so have to rely on on-line research. Don't regret my uDAC-2 HP purchase at all, except for wishing I'd spent the additional $30 at the outset to get the full uDAC-2.

 

[Vitor Machado]

 
Quote:

At the end of the day, customer buys the uDAC2 to listen with their ear, not their scope.

 
People listen with their ears, their wallets, the others opinions, the brand names, the product look-and-feel, etc...
 
Honestly, I'd rather trust the scope. But that's just me...

 

[nick_charles]

Quote:

 
 
How about this: If NuForce wants to emphasize the sound of their products, I will attempt to make proper recordings of the sound of the NuForce uDAC-2, Behringer UCA202, and my Benchmark DAC1 Pre (as a reference) via the line outputs at a typical listening volume and I'll make excerpts available from my blog for downloading. That way anyone can listen for themselves and decide which DAC sounds best to them. I will carefully level match the signal from all 3 and record using the $1800 Benchmark ADC1 which is capable of extremely high quality transparent recordings. I'll document the exact set up so someone else can reproduce and verify what I did if they desire.
 
So no measurements, just 3 different DACs playing the same music, connected to the same transparent ADC, and recorded the same way. If NuForce is confident the $129 uDAC-2 really has been tuned for the best sound, it should at least have an advantage over the $29 Behringer, right?
 
 
So does this seem like an interesting idea? Are there Head-Fiers here who would like to download the files and listen for themselves?

Definitely, If you include the reference file i.e the file used for playback before any DA/AD stages - then  it would be possible to do mathematical comparisons of the different levels of deviation created by the different DACs, for those so inclined. You might also ask members to PM you their preferences initially, the problem with members reporting their preferences openly is that you get into the whole power of suggestion thing , this potentially biases results, often badly. Once two or three have expressed the same preference that is something hard to consciously ignore.

 

[nwavguy]

@WIldstar thanks for joining in here as a fellow engineer. I agree about pop music hitting 0 dBFS and agree the DAC should handle the full range of the digital input signal without clipping. It's my position that any company that "trust the instrument first", as Jason from NuForce puts it, would discover this error and correct it. Here are my expectations about NuForce and their testing:
 

• If either the CEO or VP of R&D at NuForce (the two people I exchanged emails with) already knew they made this intentional design trade off before I ever contacted them. I would have expected them to  receive my email asking about the clipping and reply with something like: "we know about that and it was an intentional design choice". But that's not what happened. First there were other questions, then they thought I had a bad uDAC-2, then only after they made their own measurements, did they come back with (paraphrasing) "we did that on purpose for more dynamic range". They only have a few products and the uDAC-2 is relatively new. So how would the VP of R&D not know about major design trade offs they made if they were intentional?
• It's important to understand 0 dBFS is the reference level most all other tests of digital devices are referenced to. If you want to measure the dynamic range, for example, you have to first know what 0 dB is. This is done by feeding a 0 dBFS digital signal to the device and calibrating the audio analyzer to that signal. This is such basic and standard practice it's part of the automated testing scripts provided by Audio Control and Prism Sound. If there was excessive distortion of the 0 dBFS signal it should be very obvious when performing the basic instrument set up and measurements. And, in any case, it would certainly become obvious in testing for distortion and dynamic range--also very basic tests. So it's hard to understand how this could be missed if NuForce really did "trust the instrument first". That's why I wondered if they might have "designed by ear".

 
As I see it, there are only three possible explanations here:
 

• During the uDAC-2 development, NuForce did not perform even basic testing which would have included such things as the dynamic range. So they were unaware it was clipping 0 dBFS signals until I emailed them. Perhaps they were designing mostly by ear or lacked the right instruments for proper testing. This would explain why they had to go make measurements before they could respond if my uDAC-2 was within spec or not.
• NuForce did perform testing at some point during the development, and there was no obvious clipping, but something changed in the design (or perhaps in manufacturing) after the testing and the units they're shipping now clip and have the problem. And perhaps they don't have sufficient quality control procedures in place to catch such a problem. So again, they were unaware of the problem until they received my email. This is what I expected might have happened when I first discovered the issue.
• NuForce discovered (or intentionally created as they argue) the clipping during the design phase and decided it was a good design choice to improve the dynamic range by only about 1 dB from approximately 93 dB to 94 dB by having the DAC clip with TEN TO FIFTY TIMES more distortion on a 0 dBFS signal. And then, somehow, the VP of R&D and the CEO both forgot about making such a critical and highly unusual design decision by the time they received my email. If they had remembered, wouldn't it have been much easier to just tell me my uDAC-2 was performing as designed rather than than trade 15 more emails, go out and make measurements, etc?

 
I'll let those reading this decide which of the above three choices they consider most likely. And perhaps NuForce will offer up some other explanation. I'm not trying to be difficult  or unreasonable, I'm just trying to be rational and present the situation from my perspective as an engineer familiar with design, manufacturing, quality control and testing of audio equipment. Other engineers, like Wildfire, apparently agree with me. I welcome other feedback--especially if someone here sees where I've made a mistake in my analysis.
 
 
 
 
 

 

[Bmac]

Well done, OP. I only wish there was more of this type of information available out there. Other than Stereophile's measurements and a few sites like Audioholics there aren't a lot of reputable sources for independent measurements of audio gear. Thanks and keep up the good work!

 

[Xan7hos]

nwavguy, been lurkin around your blog and I must say your reviews have been a fresh of breath air around these parts. head-fi has been and is plagued with so much misinformation and uninformed consumers, so I have to say again your articles are definitely much appreciated and I definitely agree with your points made about nuforce. I understand that EVERY product has an intended customer base, and a product should be designed and sold/marketed around such a base. In regards to the uDAC-2, it's definitely considered an "entry level" item, and I'm sure that many people who bought the item will no doubt enjoy it. But your findings definitely demonstrate a disconnect between what NuForce has claimed and what they have delivered, which is always a no-no to me (even though I'm sure its more common than I'd like to believe, not just with NuForce but other co. as wells)

 

[attenuated 3db]

I need to re-read the blog of "OP," but I still see no outright deception on the part of NuForce for what is admittedly an entry-level product among a broad range of products:

Excerpted from: http://www.headfonia.com/the-udac-2/


" ... As with the first version, probably one of the strongest factor behind uDAC’s strong sales performance is the first class build quality that we rarely see, even on many higher priced products. If you look at the closest competitors to the uDAC in terms of price, the HRT MS2, the Audinst HUD-MX1, the Styleaudio UD-1, and even up to the $300 price bracket products, none of them simply have the beautiful craftmanship that the uDAC has. Not only does it look prettier and more sophisticated than the HRT MS2′s DIY-like build, but the uDAC also has a better quality finishing than some $1,000 DACs. Being super tiny also help with the overall positive impressions. The uDAC is definitely one of the most iconic looking entry level audio products in the market today.

The original uDAC was priced at $99, and we all wish that the uDAC 2 has the same pricetag as the original. It’s funny that psychology affects our judgment a big deal here, as suddenly $130 seems a steep price to pay for the uDAC 2. However, the improvements in the DAC section is very real, and given how much the uDAC brand name has grown, I think that $130 is ultimately still a fair price for such the upgrades in sound.

Overall, the uDAC 2 makes for a nice simplistic system alongside a netbook. But if you’re planning to build a decent headphone system with a separate headphone amplifier, I’d still recommend you to get something better like the HRT Music Streamer as it gives a more mature DAC sound than the uDACs."