OP
- Thread Starter
- #61
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Review and Measurements of Audio-gd NFB-27.38 DAC and Headphone Amplifier
- Thread starter amirm
- Start date
-
- Tags
- audio-gd nfb-27.38
Good luck making that stable enough to drive anything but a resistor!
60+ dB of gain is readily achievable if challenging. Most applications don't need that much in one stage, and of course many designs use cascaded gain stages in lieu of multiple feedback and/or feedforward loops to achieve stable high gain.
Regarding TIM (Otala's classic paper), yes, that is a typical example (mine too) to bolster the "too much feedback is bad" argument, and is one of the (or maybe just "my") classical examples of ears leading technologists to figure out what was going on. The catch is you can't treat it in isolation; TIM depends upon loop gain and bandwidth, dynamic range of the various gain stages, headroom and overload characteristics, etc. Sufficient loop bandwidth and headroom will control the problem of clipping, as will better overload recovery circuits. We have learned a bit in the past 40 years, and technology has advanced somewhat, fortunately.
Edit: And, what Rene said.
Regarding TIM (Otala's classic paper), yes, that is a typical example (mine too) to bolster the "too much feedback is bad" argument, and is one of the (or maybe just "my") classical examples of ears leading technologists to figure out what was going on. The catch is you can't treat it in isolation; TIM depends upon loop gain and bandwidth, dynamic range of the various gain stages, headroom and overload characteristics, etc. Sufficient loop bandwidth and headroom will control the problem of clipping, as will better overload recovery circuits. We have learned a bit in the past 40 years, and technology has advanced somewhat, fortunately.
Edit: And, what Rene said.
I was thinking 60dB of feedback in a power amplifier driving undefined reactive loads. Isn't that what Otala was concerned about?
Many opamps today achieve 80-100dB of gain in a single stage, albeit with a bandwidth of <10Hz.
Radar amps do it with several GHz of bandwidth centered at 10 GHz and above and have to deal with some ugly loads at times. Not sure I can get any more off-topic than that... 
And yes that is what Otala was mainly concerned about IIRC, though most power amps have gain ~30 dB rather than 60 dB. Wideband high-voltage and high-power transistors were hard to find in the 1970's, little easier to find these days.
And yes that is what Otala was mainly concerned about IIRC, though most power amps have gain ~30 dB rather than 60 dB. Wideband high-voltage and high-power transistors were hard to find in the 1970's, little easier to find these days.
They have a point regarding using the appropriate drivers. In the past you refuse to do so with other products saying it needs to just work with win 10. However that doesnt excuse the measurement which I assume i using a coax spdif.
OP
- Thread Starter
- #67
Why? They say it is UAC2 compliant which means it should be plug-and-play with Windows 10. If it is not, they need to look into why that is the case. Installing drivers is not the answer.
Also, the drivers are sent from a third-party, using .rar format. There is no app to open such a format on Windows 10 by default. So you have to go and find yet another program to install (many of which come with crapware) just to open the darn package. They need to provide this in .zip format as a minimum, and own the package on their own website.
Ultimately I did not ding them on this other than noting the problem. I am accepting on faith that others can make it work.
junki
Member
- Joined
- Mar 24, 2018
- Messages
- 55
- Likes
- 43
This is all true, but I'm surprised this community doesn't already know this? Everyone in the hobby knows this. Sorry, I'm new here; I just appreciate seeing measurements for things I haven't seen measurements for.
Anyway, measurements are important, but they don't tell you how something sounds. They will tell you how well circuits are designed and soldered, and they'll definitely tell you whether something is defective. There are some metrics, like seeing even low order dissipating harmonic distortion, that'll imply some tubey euphorics in the sound, but most measurement metrics are diagnostic.
The only way to figure out "what source chain will pair well with this headphone" is by auditioning it. DAC and AMPs, measuring well enough to be "well made", will sound different from each other.
--------
Example:
This is the noise and harmonic distortion comparison between the SDAC and the OL DAC. What should be one's take aways here?
If you say: The SDAC is better cuz it has lower noise, I'll laugh at you.
The truth is both noise floors are so low that it doesn't matter. In other words, when it comes to noise, both PASS. What ends up mattering just a bit more is that the OL DAC has only lower order harmonic distortion,whereas the SDAC has more higher order distortion up to 16 KHz. I say "just a bit" because this level of distortion also is too low to hear any difference at all.
But if you must compare them with regards to distortion and noise specifically, my money is the OL DAC being the winner. The SDAC wins in the other measurements relatively speaking, but both PASS on all measurements and you'll only be able to tell which one is the better sounding DAC to your ears against your equipment chain by auditioning them.
Last edited:
Let's get the record straight. The quote above is not Amirm's. It is from a post on Reddit by Max of Germany that Amirm included in post #41, here.
It is much more likely that headphone cups as well as their strange impedance characteristics are noticeable than well spec'd signal chain items.
It is much more likely that headphone cups as well as their strange impedance characteristics are noticeable than well spec'd signal chain items.
Last edited:
I don't disagree. You are saying this is how it should be. But it's not how it is. The company provides the appropriate drivers and you are not using them .
It's not the end of the world to use the drivers. Yes you can unzip a rar file. I've used 3rd party programs like Winrar for more than a decade. No, it's not going to break your computer. I think this is silly. You can choose not to use the drivers specified for the dac but people will question you including the manufacturer.
OP
- Thread Starter
- #71
Please read my tutorial on how these measurements are performed: https://www.audiosciencereview.com/forum/index.php?threads/understanding-audio-measurements.2351/
As I explain there, the frequency domain graphs represent significant amount of "FFT gain" so what you see on the graph, is far "better" than reality. I also comment on audibility issues you were asking about.
OP
- Thread Starter
- #72
It can be when it comes from random sources overseas. Drivers can also cause significant instability as compared to being class compliant and using OS manufacturer's driver.
What do we do if the next OS version comes out with slightly different driver interface and their drivers no longer work?
What value do we get from having to install drivers? None that I can see.
There is a reason Microsoft warns you when you try to install unsigned drivers. It is not something good that you want to do unless you really need to do so. In this case, there are so many choices of DACs that I don't see a need to put up with such incompatibility.
I gave a thumbs down to the Exasound E32 that I bought for a few thousand for this very reason. So don't ask me to go easy on this. I won't.
"NFB, when used incorrectly especially to ludicrous extremes, is bad. Therefore all NFB is bad."
"Novel is always better than mainstream"
"Cheap is always crap."
"Expensive is always good."
-Audiophillia logic in a nutshell.
Stop making so many reasonable arguments amirm, my brain can't take it any much longer! How dare consumers demand some sort of standards when it comes to product support for a $1600 product, darn these entitled millennials.
Last edited:
I'm more or less in the same camp, according to the measurements all the "recommended" equipment should sound the same. Alas, they do not. Still, the information presented here is invaluable, I'd rather fine tune the sound knowing I'm trying the components that measure well, rather than those that do not, especially if they are marketed as the best in its class.
Jimster480
Major Contributor
honestly this is not a surprise to me. When I talked to kingwa from AudioGD about ordering their products.... he was very much in tune with audiophile Hype.
I asked him technical questions which he couldn't answer, he then told me that there is a "language barrier". So I then had my chinese friend type of my question in chinese and I sent it to him and he said the same thing in chinese (that he couldn't explain it).
Which actually meant that he didnt know the differences.... I was asking him about the 9038Pro chip being "true balanced" with just one DAC chip vs using dual chips in a mono mode... and he just kept saying that you can just do balanced with one chip if you like. And I asked him the technical difference then vs a single ended and he couldn't tell me what it was.
I asked him what the difference is in terms of components from the NFB-11 to the R2R-11 and the R2R-2, etc considering that the "dac only" unit was twice the price of the DAC+amp "budget" unit.
Once again he couldn't explain the actual technical difference or advantage to spending more than twice the money on only one component.
He said that r2r is for "older audiophiles because they like the sound" and that if I am a "new" audiophile that I will like the Sabre based one instead.
I opted to not spend any money on their products as I didn't get a good feeling from someone who is touting technical terms but cannot explain anything technical.
I asked him technical questions which he couldn't answer, he then told me that there is a "language barrier". So I then had my chinese friend type of my question in chinese and I sent it to him and he said the same thing in chinese (that he couldn't explain it).
Which actually meant that he didnt know the differences.... I was asking him about the 9038Pro chip being "true balanced" with just one DAC chip vs using dual chips in a mono mode... and he just kept saying that you can just do balanced with one chip if you like. And I asked him the technical difference then vs a single ended and he couldn't tell me what it was.
I asked him what the difference is in terms of components from the NFB-11 to the R2R-11 and the R2R-2, etc considering that the "dac only" unit was twice the price of the DAC+amp "budget" unit.
Once again he couldn't explain the actual technical difference or advantage to spending more than twice the money on only one component.
He said that r2r is for "older audiophiles because they like the sound" and that if I am a "new" audiophile that I will like the Sabre based one instead.
I opted to not spend any money on their products as I didn't get a good feeling from someone who is touting technical terms but cannot explain anything technical.
^ That is a bit unfair. The guy made a lot of gear so he must know something, unless of course he is just a businessman with engineers working for him which may well be the case as I can't just see a single person being able to design such an extensive product line in solo. On the other hand we do have an example of experts who tout technical technical terms and can explain everything technical without an accent, yet their gear still measures less than meh.
Actually, in the vast majority of cases, they do. If there are reasons for two pieces of equipment to sound different the measurements will reveal them.
With sighted testing, you sit there listening and in the main hear just what you think you should hear.
Put the devises under bias controlled blind listening conditions and all those things you think you hear will magically disappear..
My two Cents: The link between measurements and perceived SQ is there IME. Amps with higher THD+N tend to sound fatter, more sibliants and more airy in the top than with low THD. Many like the fat and spacious sound however. When playing complex music higher distortion amps seems to clutter the sound and implode the soundstage, in my experience.
I've compared my Bricasti DAC (which should have good data) with a number of other DAC's. No DAC has subjectively had less sibliants than the Bricasti. The other DAC's tend to sound more in your face and with less depth. My guess is this has to do with low distortion and low Jitter.
Many audiophiles might prefer the more upfront sound of for example an Yggdrasil, however I prefer to build my system on the base tof the technically soundest possible products and take it from there.
I've found Win10 Fall Creators Edition to be a bit fickle with USB Class 2 Audio. When I purchased my Sabaj Da3, naturally, their web site was down for 3 days and I was unable to download the drivers for it. Win10 partially recognized the XMOS XU208 that many DACs use, but not completely. So, for example, it worked in Foobar 2000, but not Kodi.
I'll give both XMOS and Sabaj credit - once the website came back I was able to download properly signed drivers for Win10 and it has worked like a champ. While trying to find alternative drivers I found out the XMOS has program to provide signed drivers for manufacturers who use their products. I don't remember the specifics, but for low volume it was a free service and for higher volumes not especially expensive.
Similar threads
- Poll
- Poll
- Poll
