• 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 Teac NT-503 Networked DAC

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,303
Likes
233,662
Location
Seattle Area
Let me confess that I have had a man-crush on the look of the teac NT-503 for quite a while! :) Compared to a lot of utilitarian DACs that I test, the Teac looked so much nicer and professional in the pictures. Alas, at $1,000 retail price, I was reluctant to purchase it. Fortunately due to kindness of a friend of a member here, I received a unit for evaluation.

As noted above, the unit is quite handsome: https://smile.amazon.com/NT-503-Dua...F8&qid=1509320332&sr=1-1&keywords=teac+nt-503

main.jpg

In person the rack handles are a bit much and get in the way a bit so could do without them. There is a handsome white LED/OLED that is informative and pretty. Menu navigation is through the handle in the middle and easy to figure out. Outputs are controlled through that which means it shuts off one or the other (RCA or XLR) depending on which one you need to use. Not sure why this was necessary.

The unit is hefty and sits where you put it which I very much like.

Plenty of inputs is provided including S/PDIF in coax and optical plus USB and even an external clock sync! Format support is excellent , all the way up to DSD 256 and PCM up to 384 Khz.

I used the unit without any of the drivers, letting Windows 10 creator edition automatically detect it. Playback was through Roon which used WASAPI but not without turning off event driven mode. Better compatibility would have been nicer.

Measured Performance
For measurements I decided to compare it to my Exasound E32 (retail $3,500) and Topping D30 (retail $120) bookmarking it pretty well at the high and low end.

As usual, I started with my testing with j-test signal, first comparing it to more expensive Exasound E32:

J-test vs Exasound.png


Here, the Exasound shows lower noise floor, and absence of low frequency noise components and close-in sidebands. Of course we are talking very small values so not an audible concern. But from engineering point of view, the Exasound beats the NT-503 DAC.

Next I compared it to the Topping D30 which is my current champ of budget DACs:

J-test vs Topping D30.png


Here, I am powering the Topping D30 with my lab supply. Otherwise it would also have some more low-frequency noise, created by mains leakage of its switchmode supply. The D30 also has some close-in side-band jitter like the Teac NT-503. But it also has lower noise floor. Overall, despite its cost being 8X lower, I give the medal to Topping D30.

Next up is 1 Khz harmonic distortion test. Contrary to prior tests, I am using my Audio Precision analyzer as the source, creating a 1 Khz 0 dbFS output. First up is comparison to Exasound E32 DAC:

1 Khz.png


Once again we see that the Teac NT-503 has higher noise floor than the Exasound E32. Harmonic distortion is more or less comparable. This is using the left channel. For some reason, my Exasound E32 has substantially more distortion in the other channel! If I had used that, the Teac would win this race as far as harmonic distortion. The Teac distortion was similar in both channels as it should be.

Comparing to Topping D30 we get this:

1 Khz vs Topping D30.png


So again we see similar results with the Topping D30 DAC having lower noise floor. The Teach is free of 5th harmonic distortion however compared to D30.

In order to make my tests more comprehensive and comparable to what John Atkinson does in stereophile, I have added a test of a very low amplitude -- -90 dbFS 24-bit, undithered sine wave -- to my testing. In an ideal situation we would see a perfect sine wave. Anything else indicates noise or lack of linearity at extremely low amplitudes. Here is how the Teac NT-503 did using coax input:

24-bit -90 db.png


We see a facsimile of a sinewave, albeit with some noise and distortion. Likely reason is the low-frequency noise we saw in the J-test. There is also a shifting up and down level in the way the waveform goes up and then comes down (not shown). This could be AC mains leakage causing drift.

Let's compare that to Exasound E32:

24-bit -90 db.png

Ok, ok. So money does buy you more happiness when it comes to low level accuracy. :) While we still have the shifting up and down the sinewave is very accurately rendered and definitely clearer than Teac NT-503. Not shown but when testing the balanced output of Exasound E32, the performance suffered quite a bit here! Will post more about this later so for now, this performance is only available through its unbalanced/RCA output.

Here is our budget DAC, Topping D30:

24-bit -90 db.png

It is splitting hairs to say whether the D30 is better than NT-503. The Topping output level does not shift up and down so I would give the nod to it.

Summarizing, the Teac NT-503 doesn't bring anything objectively to the table than the Topping D30 which retails for 1/8 its price. The Topping is also feature rich but lacks XLR output or a display. And headphone output.

Listening Tests: Headphone
For this test, I compared the Exasound E32 headphone output to that Teac NT-503. AB testing is difficult because I had to plug and unplug my headphones and the Teac mutes and unmutes its headphone output every time you do that. So we are talking 8-10 second switching which is not ideal at all.

Keeping that constrain in mind, I preferred the output of the Teac NT-503 to that of Exasound E32 when testing my Sony MDRv6 headphones (DC resistance: 78 ohms). The NT-503 had better impact in bass and more separation of instruments. This could of course be biased so take it with some grain of rice. Until I create a test set up that allows me instant switching, this is all I have to offer.

Listening Tests: Unbalanced Line Out:
For this test, I used my Stax SRM-007t. I used this because it has dual inputs and allows me to make very fast AB switching. I then used Roon and ganged the NT-503 and Exasound and then NT-503 and Topping into zones. The outcome was synchronized playback on both devices at the same time. Boy do I love Roon! :)

Testing first against the Exasound, I could not detect any difference. Both the Exasound E32 and Teac NT-503 produced delightful sound. Gone was what I was hearing with headphones. Alas, the Exasound drivers are quite fragile and just after a few clips, it refused to initialize no matter how many times I restarted Roon, Exasound or plugging and unplugging the USB cable. :( The Teac was also flakey until I put it in non-event driven mode in which it became much more reliable So all in all, this testing was cut short but I stand behind the outcome.

Testing against the Topping D30 resulted in tiny preference in favor of Topping D30. Alas, its output is also 1 or 2 db higher and I think that is what I was hearing. Longer term I will do more level matched testing. Outside of that tiny difference, again both were delightful to listen to. Indeed I am doing that as I type this and would be happy with either device playing my reference playlist.

Summary
The Teac is a beautifully made networked DAC. I had no use for its networking (because it doesn't support Roon), but outside of that, other than some tiny issues, its measured performance is without concern. Alas, it loses to both cheaper and more expensive DACs so no advantage is provided there.

Its headphone output though seems excellent, driving my high impedance Sony headphone with ease and authority. To that end, I think it makes an excellent benchtop DAC next to your computer -- something neither the Topping D30 (which has no headphone output) or the Exasound (not as nice of a headphone amp) can do.

If you don't need headphone output, or balanced out, my suggestion is to save money and by the Topping D30. This is a little DAC that could! :)

As always, appreciate any feedback, questions, concern, correction and wisecracks.
 
Last edited:

wgscott

Active Member
Joined
Mar 1, 2016
Messages
179
Likes
37
Thanks for posting the measurements.

My primary complaint is with the networking function, and DLNA support in particular. The NT-503 deviates from the protocol. I have yet to find any controller that permits genuine gapless playback, and many of them do not work with it at all. Teac's own supplied iOS controller comes the closest for gapless, but it is kind of a pig to use, and shockingly buggy.

As a USB or Coax DAC, it seems to function well enough in my system, where it clearly is not the weakest link in the chain. It provides for a variety of upsampling regimens and filters, including DSD (which to my ear sounds inferior). But it is nice to have the choice for experimenting.

I am not sure why one would use an external clock, but Teac is about to make one available that costs about 1.5X what the DAC itself does. Maybe for multi-channel or video syncing it would be useful?
 

wgscott

Active Member
Joined
Mar 1, 2016
Messages
179
Likes
37
In order to make my tests more comprehensive and comparable to what John Atkinson does in stereophile, I have added a test of a very low amplitude -- -90 dbFS 24-bit, undithered sine wave -- to my testing. In an ideal situation we would see a perfect sine wave. Anything else indicates noise or lack of linearity at extremely low amplitudes. Here is how the Teac NT-503 did using coax input:

This has me wondering if the behavior would vary as a function of what filter is used (and if so, what filter option was used?) Also, does it have to use the coax input? I have wondered whether USB is better than coax on this thing (I cannot hear a difference, but I use a USB to coax converter, which might mask any differences.)
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,303
Likes
233,662
Location
Seattle Area
I'd like to see the NT-503 and Mytek Brooklyn compared on the same plot. That could be interesting. :)
Unfortunately, I need to have them both on my bench at the same time to do that. Seeing how I returned the Mytek already, that is not possible. Here is the J-Test as I saved before:

J-test.png


The Mytek Brooklyn has cleaner, distortion free J-Test output so in that sense it is superior to Teac NT-503.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,303
Likes
233,662
Location
Seattle Area
This has me wondering if the behavior would vary as a function of what filter is used (and if so, what filter option was used?) Also, does it have to use the coax input? I have wondered whether USB is better than coax on this thing (I cannot hear a difference, but I use a USB to coax converter, which might mask any differences.)
I can test with different filters. As to USB, I currently do not have the same test signal on the computer. I will create one in the future but for now, I had to use the Audio Precision Analyzer to test.

That said, I did create an approximation of it and its results were pretty close (over USB) to what AP generated. I will test that later tonight also and post.
 

astr0b0y

Active Member
Joined
Aug 12, 2016
Messages
273
Likes
201
Location
Melbourne Australia
Is it expected that the results would stay the same regardless of bit depth and sample rate of the input? For example, if you were able to do the same tests feeding DSD256 sample rate, would the same levels of noise and jitter been observed?
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,303
Likes
233,662
Location
Seattle Area
Is it expected that the results would stay the same regardless of bit depth and sample rate of the input? For example, if you were able to do the same tests feeding DSD256 sample rate, would the same levels of noise and jitter been observed?
DSD output would be different. I am working on developing DSD tests. Unfortunately I am on a short fuse on this DAC and need to return it. In the future I will include more DSD tests once I develop the right test content.
 

astr0b0y

Active Member
Joined
Aug 12, 2016
Messages
273
Likes
201
Location
Melbourne Australia
Cool, interested to see what kind of differences there may be given that hi-res support in DACs and up-sampling capabilities of Roon etc seem to be in high usage.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,442
Likes
36,872
Where is the velvet-ness graph? The Teac has Velvet sound. How could you skip that measurement. :p
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,303
Likes
233,662
Location
Seattle Area
Where is the velvet-ness graph? The Teac has Velvet sound. How could you skip that measurement. :p
Plan to use it as a pillow tonight. Answer tomorrow morning. :D
 

A.wayne

Active Member
Joined
Mar 21, 2016
Messages
175
Likes
20
Cool, interested to see what kind of differences there may be given that hi-res support in DACs and up-sampling capabilities of Roon etc seem to be in high usage.


Upsampling IME , kills the sonics and is a big no no ....
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,814
Likes
16,456
Location
Monument, CO
Almost all DACs upsample. Those that don't introduce aliasing artifacts.

Note it would be images, not aliasing, but I am not sure how many DACs actually upsample unless requested (e.g. user or program directs it to interpolate or otherwise upsample the incoming data). Note that the majority of audio DACs are delta-sigma designs, which are oversampling, but that is not the same as upsampling.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,303
Likes
233,662
Location
Seattle Area
Ok, here is some more measurements before I rush to ship the unit back.

First, the question of USB input vs S/PDIF. Please note that my USB test file may be different because it is produced differently. Still, as you see, there is really no difference in using USB and S/PDIF when it comes to low level accuracy: (all graphs are for balanced XLR output rather than RCA posted before for completeness):

Balanced AP vs USB.png


The pair of graphs by the way, are for left and right channels.

I then turned on upsampling to 8X, still using USB input. This is what I got:

Balanced USB native vs 8X upsample.png


Now I am showing one channel only. Again, seems to make no difference.

Now look at what happens when I upsample to DSD:

Balanced USB native vs DSD upsample.png


Hell broke loose! :) Low level accuracy has degraded substantially (in red). Hard to see it as a sine wave anymore. In addition, the waveform is shifted down which indicates negative DC offset (constant negative voltage).

Also, the levels are higher which in a listening test without control, may make people like DSD better even though objectively the waveform is degraded fair bit.

No time for further investigation. Have to see if my Exasound has DSD upsampling and does the same thing. Food for thought!
 
Top Bottom