• 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

wgscott

Active Member
Joined
Mar 1, 2016
Messages
179
Likes
37
Interesting. Thanks again for posting. I learned (at least) three things (assuming I understand the measurements correctly):

(1) I paid $900 more than I really needed too. Phool me once...
(2) I am fully justified in turning the volume up to 11.
(3) The one difference I thought I reliably heard (upsampling to DSD made it sound a bit worse) seems to be measurable.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,451
Likes
36,880
Ok, here is some more measurements before I rush to ship the unit back.

Now look at what happens when I upsample to DSD:

View attachment 9461

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!

Did you have an ultrasonic filter for the output when doing DSD? Would be interesting if that made all the artifacts go away.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,303
Likes
233,697
Location
Seattle Area
Did you have an ultrasonic filter for the output when doing DSD? Would be interesting if that made all the artifacts go away.
Too late to experiment as the box is in the post office. It does have selectable ultrasonic filter for DSD. I don't unfortunately remember the default value I left it at.

Just looked and Exasound E32 has no support for upsampling. So I will test with HQplayer next and see what it does for DSD conversion.
 

wgscott

Active Member
Joined
Mar 1, 2016
Messages
179
Likes
37
DSD Cut-off Filters 50kHz, 150kHz (Switchable)

(Those are the only two options)
 

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,563
Likes
10,120
Location
North-East
Now look at what happens when I upsample to DSD:

View attachment 9461

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!

The DSD output looks like there are only about 4 voltage levels being used to represent the sine wave (once you filter out the ultrasonic components). In other words, just 2 bits worth of information. This may not be that surprising at -90dB if the input is only 16 bits (or if the internal DSD processing is only limited to 16-17 bits).
 

Frank Dernie

Master Contributor
Forum Donor
Joined
Mar 24, 2016
Messages
6,441
Likes
15,766
Location
Oxfordshire
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):

View attachment 9459

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:

View attachment 9460

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

Now look at what happens when I upsample to DSD:

View attachment 9461

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!

I am particularly interested that the DSD upsampling increases level.
An acquaintance of mine has digitised all his LPs and after some experimentation he went for DSD rather than pcm, he said they sound better and sent me a couple of samples. The DSD sounded better to me too, but louder. I wondered if it only sounded better because it was louder but was not confident to be able to measure the loudness difference so I could level match accurately enough.
I have also had a demonstration of a dCS sreamer which can upsample to DSD and it sounded better than pcm into a Devialet amp, which immediately converts DSD to pcm...
Maybe DSD upsampling always adds a bit of level??
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,303
Likes
233,697
Location
Seattle Area
The DSD output looks like there are only about 4 voltage levels being used to represent the sine wave (once you filter out the ultrasonic components). In other words, just 2 bits worth of information. This may not be that surprising at -90dB if the input is only 16 bits (or if the internal DSD processing is only limited to 16-17 bits).
This test uses 24-bit undithered samples. There is another test stereophile does that (the squarewave) which uses 16 bits. That aside, you are right that it looks that way.
 
Last edited:
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,303
Likes
233,697
Location
Seattle Area
Maybe DSD upsampling always adds a bit of level??
It is a challenge with signal processing algorithms to make sure at the end levels are not changed. But as much as possible, this should be a goal to achieve. Let me do more testing and see if we can find if this is generally true.
 

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,563
Likes
10,120
Location
North-East
I am particularly interested that the DSD upsampling increases level.
An acquaintance of mine has digitised all his LPs and after some experimentation he went for DSD rather than pcm, he said they sound better and sent me a couple of samples. The DSD sounded better to me too, but louder. I wondered if it only sounded better because it was louder but was not confident to be able to measure the loudness difference so I could level match accurately enough.
I have also had a demonstration of a dCS sreamer which can upsample to DSD and it sounded better than pcm into a Devialet amp, which immediately converts DSD to pcm...
Maybe DSD upsampling always adds a bit of level??

Generally, DSD is lower by 6dB compared to PCM. Some DSD to PCM converters automatically apply this 6dB gain. Maybe there's built-in compensation in NT-503 to bring DSD playback closer to PCM levels?
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,303
Likes
233,697
Location
Seattle Area
Generally, DSD is lower by 6dB compared to PCM. Some DSD to PCM converters automatically apply this 6dB gain. Maybe there's built-in compensation in NT-503 to bring DSD playback closer to PCM levels?
My Roon converter had that 6 db adjustment by default and I found that it was not correct. I think I set it to 4 db.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,451
Likes
36,880
My Roon converter had that 6 db adjustment by default and I found that it was not correct. I think I set it to 4 db.
I don't have any DSD capable converters. I convert in software. Over the last couple years various needledrops and other files put into DSD from the same source seemed to vary from the PCM version by 4.5 to 5 db. A couple times people sent me PCM copies of PCM and DSD playback and those also varied by 4. 5 to 5 db.

I can't say I quite understand why or even if it is true, but some people who should know say DSD acts a bit like tape saturation at high signal levels. If that effect is very large it would be like a very slight compression, which combined with actual loudness differences would explain how some claim PCM to DSD for playback sounds better than PCM. Would also indicate however much some may like the sound that DSD is not a transparent medium. Can anyone here say if the saturation like effects are part and parcel of DSD?
 

wgscott

Active Member
Joined
Mar 1, 2016
Messages
179
Likes
37
The DSD output looks like there are only about 4 voltage levels being used to represent the sine wave (once you filter out the ultrasonic components). In other words, just 2 bits worth of information. This may not be that surprising at -90dB if the input is only 16 bits (or if the internal DSD processing is only limited to 16-17 bits).

I confess to having only a very superficial understanding of how the tests are done. The volume adjustment (which for the most part is attenuation) on the Teac NT503 is analogue, so it shouldn't decimate the signal in the way you describe. Is the signal being attenuated in some other way for the measurement?
 

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,563
Likes
10,120
Location
North-East
I confess to having only a very superficial understanding of how the tests are done. The volume adjustment (which for the most part is attenuation) on the Teac NT503 is analogue, so it shouldn't decimate the signal in the way you describe. Is the signal being attenuated in some other way for the measurement?

As I understand it, Amir fed a -90dB digital signal to the input of the DAC and measured the analog output. If that input was only 16 bits, then the signal would be just the least-significant bit flipping between 0 and 1. That would normally result in just two distinct analog output levels. But, Amir fed it 24 bits input, so it would appear that some bits were lost in processing as only four distinct levels appeared at the output. This makes it equivalent to about 17 bits DA conversion, if my math is right.
 

A.wayne

Active Member
Joined
Mar 21, 2016
Messages
175
Likes
20
Interesting. Thanks again for posting. I learned (at least) three things (assuming I understand the measurements correctly):

(1) I paid $900 more than I really needed too. Phool me once...
(2) I am fully justified in turning the volume up to 11.
(3) The one difference I thought I reliably heard (upsampling to DSD made it sound a bit worse) seems to be measurable.


Wait, So now up sampling is bad, sounds worse .... :)
 
Top Bottom