T+A DAC 200

[OE333]

On 3/2/2023 at 7:29 PM, Tihon said:

@OE333 would you be so kind to tell a little bit about difference between DAC200 and it's more expensive brother SDV 3100 HV?

 

I mean, it's very obvious that SDV 3100 HV has more bells and whistles built in, but I'm interested in DAC for HQPlayer only. Is 3100 really much better than 200 when it comes to converters and/or analog stage? I've noticed different output voltages and higher SNR of DAC200, which is strange. Besides this, the information on the site does little to help understand the advantages of the 3100 series.

 

By the way, SDV 3100 HV has a built-in NAA (sad that there is no such thing for MP200 or MP 2500R). How do you think, is it the best possible solution for HQP, or external high-end USB source might improve on it?

 

In principle the architecture of DAC 200 and SDV3100 are very similar and the same technology is used - like separate PCM and DSD DACs, T+A's unique true-1-bit converter, galvanic isolation between digital and analog, all discrete analog stages, etc. etc.

The differences are in the details:

• The SDV 3100 HV includes additional audio sources (like streaming client, FM/DAB tuner, bluetooth audio). In case of the DAC200 you will need the MP200 for these functions.
• The SDV3100HV has a built in NAA for streaming audio signals in highest quality from HQ Player
• Massive case design of the HV series made from up to 4cm thick aluminum plates
• Much higher weight of the SDV3100HV means better mechanical stability and vibration damping
• The SDV's case is divided into 5 different compartments for digital power supply, analog pwr-supply, control section, digital signal processing section and analog section. All these sections are heavily shielded against each other to eliminate electromagnetic interference
• The T+A 1 bit converter in the SDV3100HV has double the number of converter elements giving even better noise performance
• Both devices use the T+A "high voltage" technology  for the analog stages but the supply voltage in the HV series is still higher than in the DAC200 making the stages even more linear.
• The analog output stages in the SDV3100HV run on a much higher bias current giving an extended class-A operation range
• The SDV has a better headphone amplifier
• SDV3100HV has a special T+A proprietary connection for PDT3100HV SACD player. This connection transfers audio and clocking signals in highest possible quality and with extremely low jitter.
• SDV3100HV has higher power mains power supplies with highly sophisticated mains filtering and separate analog and digital mains inlets to connect analog and digital parts with separate mains cords to the mains minimizing coupling of electromagnetic noise from digital to the analog side.

 

All the above differences make a difference and the sound quality of a SDV3100HV is still a bit higher than the already excellent quality of a DAC200.

 

SNR performance

The measured noise performance of the DAC200 on the first look seems to be 2 dB better than that of the SDV3100HV. This difference is a consequence of the higher bandwidth of 300 kHz in the SDV compared to 200 kHz in case of the DAC200.  If the bandwidth of the SDV3100HV would also be limited to 200kHz it would give the SDV3100HV an advantage in noise over the DAC200 but we wanted to have the higher bandwidth because this improves the dynamic response and gives a more linear phase at high frequencies.

But the measured differences between DAC200 and SDV3100HV are really extremely small !

 

 

[Tihon]

@OE333

 

Thank you! Being able to get first-hand answers is so cool! This is very interesting and now I seem to want the SDV 3100 even more 🤣 

 

What do you think about the built-in NAA and external USB sources? I understand that some kind of single-board computer (RPi?) is installed inside the 3100, which is connected to the DAC via USB. There are quite a few USB audio devices on the market today. It uses separate dedicated power supplies, high-quality clocks, and much more to improve the quality of the USB signal. On the other hand, an additional device means additional cables, connections and ground loops. Is there any point in using external USB transport in the case of SDV 3100 and HQP?

 

Also, I wonder why different sources are done so differently? For CD/SACD you have a PDT 3100 dedicated transport which is transmitting audio stream with clocking data via IPA Link (I assume it's some kind of propietary I2s interface). For streaming asynchronous USB is used, which is clocked by DAC. It's pretty much obvious that disc spinning mechanism needs separate enclosure, what I don't understand is the clock source part.

 

[1125FPS]

I just got my DAC 200 working; no direct AB vs. Spring 3 KTE.  But DAC 200 did play a well-recorded Hiromie track I heard last night on Spring 3 KTE.  Using HQ Player, Linux on custom computer (commercial SRP estimate $3k USD,) Roon on separate NUC for library function only.  

 

First impression is that just unpacked DAC 200 is in a different league from S3KTE.  More later.

 

[OE333]

@Tihon

The idea behind the SDV3100HV is to make all kinds of audio sources available to the user (S/P-DIF, HDMI, USB, network, disc player etc) at the highest possible quality.

 

The SDV (and also the DAC200) incorporates an extremely precise clock which is used to re-synchronize all incoming audio data streams  - no matter of which input is used. This ensures a top of the notch jitter performance for all sources connected to the DAC.

 

Also a complete galvanic isolation is included in our DACs which guarantees that no digital noise from any source can invade into the analog section.

 

[1125FPS]

My friend has a 2nd system in which his TV HDMI output feeds his Bricasti M21 DAC w/renderer ($16k USD SRP.)  Alternately the TV optical output can feed his M21.  He also sometimes employs M-Connect to stream Qobuz to the M21 renderer.  In this application, how might users compare audio and musical performance of the M21 vs. DAC 200?  

[Tihon]

@OE333 oh, and I've almost forgotten another question. Do T+A dads have same output level for PCM and DSD? Some DACs are playing DSD at lower levels than PCM.

[1125FPS]

My friend has been updating my server system for much of the day and I can hardly wait to hear DAC 200 when he is done.  

[OE333]

On 3/5/2023 at 8:38 PM, Tihon said:

@OE333 oh, and I've almost forgotten another question. Do T+A dads have same output level for PCM and DSD? Some DACs are playing DSD at lower levels than PCM.

 

Precise answer: DSD is 0.1dB louder than PCM with standard FIR Oversampling.

 

But the PCM output level slightly depends on the chosen oversampling filter (FIR, Bezier, NOS).

Generally speaking the output level differences between the different input formats and oversampling algorithms are always less than +/- 0.2 dB.

 

 

 

[Miska]

When HQPlayer produces PCM or DSD output, the levels are quite predictable. Generally HQPlayer doesn't produce DSD data exceeding 0 dB level, and upsampled PCM outputs are not allowed to clip. But the DSD spec allows periodic peaks to reach +3.5 dB. So if you play some raw DSD from some SACD rip, your levels may be something different. Or from raw PCM source depending how inter-sample overs are handled in the DAC ( @OE333 )?

 

So for HQPlayer the PCM gain compensation for DAC 200 / HA 200 is close to 0 dB.

 

[Nkam]

On 3/5/2023 at 10:38 AM, OE333 said:

@Tihon

The idea behind the SDV3100HV is to make all kinds of audio sources available to the user (S/P-DIF, HDMI, USB, network, disc player etc) at the highest possible quality.

 

The SDV (and also the DAC200) incorporates an extremely precise clock which is used to re-synchronize all incoming audio data streams  - no matter of which input is used. This ensures a top of the notch jitter performance for all sources connected to the DAC.

 

Also a complete galvanic isolation is included in our DACs which guarantees that no digital noise from any source can invade into the analog section.

 

 

does the DAC 200 play true NOS in PCM mode? 
I’ve heard that it isn’t due to the nature of the ΣΔ chips.   BB I think. 
Are the first 6 bits NOS and the rest ΣΔ?

 

Asking if it is a good DAC to use with HQplayer in PCM.  

 

Thank you 

[linuxprophet]

@Nkam

I have used the HA200 with OVS set to NOS2 with HQP in Ubuntu LTS 22.04 to sample DSD1024, DSD512 and various DXD and PCM files up to 768k with no issues.

 

I would presume the DAC 200 should perform the same. Totally lost with the ΣΔ though 

[linuxprophet]

Unboxing generally, is a kind of a personal high. I don't know if it's the same for other esteemed members here.

 

For me personally, Unboxing experience for the T+A gear was special. I can't say if it was because of the packaging or the sheer feel of the aluminum kit therein. The only other comparison I had was probably unboxing the Oppo PM-1s when they came out a few years ago.

 

Then you connect it all, and spend hours tweaking, listening to various material, comparing with similar gear, and generally behaving like a 5 year old with a new Lego set.

 

Kudos to the design teams, the quality of the equipment and the ability to use tools like HQP with them.

 

 

[OE333]

5 hours ago, Nkam said:

does the DAC 200 play true NOS in PCM mode? 
I’ve heard that it isn’t due to the nature of the ΣΔ chips.   BB I think. 
Are the first 6 bits NOS and the rest ΣΔ?

 

Asking if it is a good DAC to use with HQplayer in PCM. 

Thank you 

 

 

This is indeed an intersting question.

 

First: the facts:

The DAC200 uses 4 PCM1795 converter chips for converting PCM signals to analog. The PCM1795 consists of a 8x oversampling filter followed by a ΣΔ DAC.

In case of the T+A devices the 8x oversampling inside the PCM1795 is completely bypassed and the digital signal is directly fed to the ΣΔ  modulator.

If one of the DAC200's PCM oversampling filters (FIR1, FIR2, Bezier1, Bezier2) is selected, the oversampling is done in an external DSP (using T+A's own oversampling algorithms) , the internal filters of the PCM1795 are never used.

 

If the oversampling is switched OFF and NOS1 or NOS2 mode is selected, the external DSP is also bypassed and the incoming PCM stream is directly routed to the ΣΔ  modulator stage of the DACs without any signal processing before the modulator stage.

 

So, as a first result we can state that in the DAC200 all PCM oversampling filters are bypassed in case of NOS mode.

 

The question now is: does the ΣΔ  modulator stage perform oversampling or not ?

 

This question is a bit philosophical and the answer depends on weather you regard the output signal of the ΣΔ  modulator as a highly oversampled 1 bit digital signal (in this case it performs a kind of oversampling), or if you regard the modulators output signal as an analog signal having an average value representing the analog output value - in this case the modulator is DAC delivering an analog output voltage which only needs some averaging (analog low-pass filtering) to get rid of the unwanted high frequency noise and to deliver the wanted analog signal average.

 

Some more details about this topic are given in this paper: https://www.beis.de/Elektronik/DeltaSigma/DeltaSigma.html

 

 

More than this consideration it might be interesting to look at the output signal of the DAC200 and find out if it behaves as would be expected from a NOS DAC.

 

 

1.) Outut signal from the I/V stage in NOS2 mode when the DAC200 is fed with a digital step signal @fs=44.1 kHz (blue trace):

 

 

 

The output signal after the I/V stage of the DAC. It rises instantly at the digital signal step, without any trace of interpolation. A NOS R2R converter would behave exactly the same way.

 

Please note: This signal is the signal before the final analog reconstruction filter. It is an internal signal and can not be measured at the output jacks of the DAC200.

 

----------------------------------------------------------------------------------------------------------

 

2.) The same signal after the analog reconstruction filter, as it appears at the DAC200 output:

 

 

The output signal after the analog output filter slews a bit slower than the signal in the first measurement. This slower slew corersponds to the 120 kHz cut-off frequency of the DAC200's analog output filter (in "WIDE" mode) - it is NOT a consequence of digital interpolation.

 

----------------------------------------------------------------------------------------------------------

 

For comparison:

This is the DAC200 output signal with the same digital step input signal with oversampling filter "FIR1":

 

 

 

This output waveform shows the much slower response due to the interpolation performed by the digital FIR oversampling filter.

 

 

 

The presented step measurements clearly show the Non Oversampling behaviour of the DAC200.

The output is exactly what would be expected from a NOS DAC and the presented output signals are indistinguishable from the output signals of a NOS  R2R DAC fed with the same digital input signal.

 

Imho the DAC200 is very well suited to operate with high quality external oversamplers such as HQ player, because inside of the DAC200 no oversampling/interpolation takes place and the quality of the externally oversampled signal is preserved in all detail.

 

 

[Nkam]

21 minutes ago, OE333 said:

 

 

This is indeed an intersting question.

 

First: the facts:

The DAC200 uses 4 PCM1795 converter chips for converting PCM signals to analog. The PCM1795 consists of a 8x oversampling filter followed by a ΣΔ DAC.

In case of the T+A devices the 8x oversampling inside the PCM1795 is completely bypassed and the digital signal is directly fed to the ΣΔ  modulator.

If one of the DAC200's PCM oversampling filters (FIR1, FIR2, Bezier1, Bezier2) is selected, the oversampling is done in an external DSP (using T+A's own oversampling algorithms) , the internal filters of the PCM1795 are never used.

 

If the oversampling is switched OFF and NOS1 or NOS2 mode is selected, the external DSP is also bypassed and the incoming PCM stream is directly routed to the ΣΔ  modulator stage of the DACs without any signal processing before the modulator stage.

 

So, as a first result we can state that in the DAC200 all PCM oversampling filters are bypassed in case of NOS mode.

 

The question now is: does the ΣΔ  modulator stage perform oversampling or not ?

 

This question is a bit philosophical and the answer depends on weather you regard the output signal of the ΣΔ  modulator as a highly oversampled 1 bit digital signal (in this case it performs a kind of oversampling), or if you regard the modulators output signal as an analog signal having an average value representing the analog output value - in this case the modulator is DAC delivering an analog output voltage which only needs some averaging (analog low-pass filtering) to get rid of the unwanted high frequency noise and to deliver the wanted analog signal average.

 

Some more details about this topic are given in this paper: https://www.beis.de/Elektronik/DeltaSigma/DeltaSigma.html

 

 

More than this consideration it might be interesting to look at the output signal of the DAC200 and find out if it behaves as would be expected from a NOS DAC.

 

 

1.) Outut signal from the I/V stage in NOS2 mode when the DAC200 is fed with a digital step signal @fs=44.1 kHz (blue trace):

 

 

 

The output signal after the I/V stage of the DAC. It rises instantly at the digital signal step, without any trace of interpolation. A NOS R2R converter would behave exactly the same way.

 

Please note: This signal is the signal before the final analog reconstruction filter. It is an internal signal and can not be measured at the output jacks of the DAC200.

 

----------------------------------------------------------------------------------------------------------

 

2.) The same signal after the analog reconstruction filter, as it appears at the DAC200 output:

 

 

The output signal after the analog output filter slews a bit slower than the signal in the first measurement. This slower slew corersponds to the 120 kHz cut-off frequency of the DAC200's analog output filter (in "WIDE" mode) - it is NOT a consequence of interpolation.

 

----------------------------------------------------------------------------------------------------------

 

For comparison:

This is the DAC200 output signal with the same digital step input signal with oversampling filter "FIR1":

 

 

 

This output waveform shows the much slower response due to the interpolation performed by the digital FIR oversampling filter.

 

 

 

The presented step measurements clearly show the Non Oversampling behaviour of the DAC200.

The output is exactly what would be expected from a NOS DAC and the presented output signals are indistinguishable from the output signals of a NOS  R2R DAC fed with the same digital input signal.

 

Imho the DAC200 is very well suited to operate with high quality external oversamplers such as HQ player, because inside of the DAC200 no oversampling/interpolation takes place and the quality of the externally oversampled signal is preserved in all detail.

 

 

 

wow, thanks for the detailed response to my question.  
I’ll have to spend some hours digesting it as my technical knowledge is very limited.  
 

here is another question from me if you are kind enough to give me your opinion. 
 

The link to the thread is here.    
I just don’t know why, but every op amp analog output DAC I’ve tried doesn’t sound as good as discreet.  
why is that?
 

https://audiophilestyle.com/forums/topic/67752-someone-please-explain-is-a-discreet-analog-output-stage-better-or-op-amps/

[OE333]

Just now, Nkam said:

I just don’t know why, but every op amp analog output DAC I’ve tried doesn’t sound as good as discreet.  
why is that?
 

https://audiophilestyle.com/forums/topic/67752-someone-please-explain-is-a-discreet-analog-output-stage-better-or-op-amps/

 

To say it with the words of  Ludwig Mies van der Rohe: Less is More

 

or

 

in an integrated OP-Amp there is no choice - all components have to made from silicon. In a discrete amplifier the designer can choose the best possible type of component for each part of his circuit He can also tailor the circuit exactly to the specific needs and requirements of the device he is designing.

Very often this results in less complex circuits with fewer but higher quality parts and less chances for errors and imperfections.

So in many cases an experienced analog designer can tweak a discrete circuit to higher performance levels than a universal-to-use  off-the-shelf integrated standard part can deliver.

[Nkam]

14 minutes ago, OE333 said:

 

To say it with the words of  Ludwig Mies van der Rohe: Less is More

 

or

 

in an integrated OP-Amp there is no choice - all components have to made from silicon. In a discrete amplifier the designer can choose the best possible type of component for each part of his circuit He can also tailor the circuit exactly to the specific needs and requirements of the device he is designing.

Very often this results in less complex circuits with fewer but higher quality parts and less chances for errors and imperfections.

So in many cases an experienced analog designer can tweak a discrete circuit to higher performance levels than a universal-to-use  off-the-shelf integrated standard part can deliver.

 

that makes total sense to me. 
however there are those who will say that todays op amps ‘ performance’/measurements will be superior to a discreet design.  
 

what do you say to them?   Because my ears also always have preferred discreet as well.  

[OE333]

10 hours ago, Nkam said:

what do you say to them? 

 

Listen !   😉

[Miska]

16 hours ago, OE333 said:

The question now is: does the ΣΔ  modulator stage perform oversampling or not ?

 

Well, it does, Sample And Hold aka Zero Order Hold. So there's no digital filter, the same sample is just copied N times. The DAC chip has selection for this, by default it is 8x. So eight copies are made of each sample at that stage.

 

[Nkam]

15 hours ago, Miska said:

 

Well, it does, Sample And Hold aka Zero Order Hold. So there's no digital filter, the same sample is just copied N times. The DAC chip has selection for this, by default it is 8x. So eight copies are made of each sample at that stage.

 

so the T+A DAC 200 has the same performance with HQplayer as a true NOS DAC does? 
in PCM that is.  

 

[1125FPS]

I finally performed AB test tonight, DAC 200 vs. Holo Audio's Spring 3 KTE, both using a custom HQ Player server with discrete NUC renderer.  I used solely DSD @ 128k.  Output impedances are almost identical.  Output V unbalanced: DAC 200 2.5V, S3 2.9V; I lowered the preamp level 1 dB for the S3.  This system is well sorted.  The SS amp has proprietary technology as does the multiple award-winning wave-guide main satellite speaker.  Bass below about 90 Hz comprises an award-winning distributed array.       

 

I used only one music program so far, Paul Schwartz, Suscepit Israel, orchestra, big chorus, a very good recording.  When the chain is capable the chorus portrays on risers, in a distinct arc 3-5 lines deep.  It took about half dozen or more comparisons to really get a handle on the differences.

 

First, I cannot overstate how good is the S3 and how great a value it is at $3100 USD.  My first impression was that I may have preferred the S3's instrument separation and huge magnitude of musicality; the S3 draws you into the music like a magnet and can quickly make you forget about audio.

 

But by the last comparison, especially focusing on the big crescendo near the end, the DAC 200's advantage was huge and crystal clear.  Both DACs detach well from the loudspeakers, but DAC 200 does so to a much greater degree.  Image density and sound scape size are in a completely different league.  The effect of being immersed in the sound scape is multiples greater.  Musical intensity is higher but there's no down-side to this effect.

 

I'll report more later with other music programs.  

 

I have not heard Holo Audio's May KTE DAC (about $5500 USD) but many if not all who heard May and my S3 (both with HQ Player) suggest the following: HQ Player provides the best performance; the 2 DACs perform much closer to each other than is indicated by the price difference and many call the May KTE a giant killer on an absolute scale.  If those things are true, then DAC 200 for $1700 more than the May KTE appears to be superb value.  There's nothing to criticize about the Chinese Holo's build quality but the fact that DAC 200 is made in Deutschland surely carries considerable cachet.  DAC 200 has considerably more features too; I love the meters. 

 

 

For the record, I used the same high-grade AC mains cable for both DACs, so mains power was off between auditions.  

 

  

[Nkam]

42 minutes ago, 1125FPS said:

I finally performed AB test tonight, DAC 200 vs. Holo Audio's Spring 3 KTE, both using a custom HQ Player server with discrete NUC renderer.  I used solely DSD @ 128k.  Output impedances are almost identical.  Output V unbalanced: DAC 200 2.5V, S3 2.9V; I lowered the preamp level 1 dB for the S3.  This system is well sorted.  The SS amp has proprietary technology as does the multiple award-winning wave-guide main satellite speaker.  Bass below about 90 Hz comprises an award-winning distributed array.       

 

I used only one music program so far, Paul Schwartz, Suscepit Israel, orchestra, big chorus, a very good recording.  When the chain is capable the chorus portrays on risers, in a distinct arc 3-5 lines deep.  It took about half dozen or more comparisons to really get a handle on the differences.

 

First, I cannot overstate how good is the S3 and how great a value it is at $3100 USD.  My first impression was that I may have preferred the S3's instrument separation and huge magnitude of musicality; the S3 draws you into the music like a magnet and can quickly make you forget about audio.

 

But by the last comparison, especially focusing on the big crescendo near the end, the DAC 200's advantage was huge and crystal clear.  Both DACs detach well from the loudspeakers, but DAC 200 does so to a much greater degree.  Image density and sound scape size are in a completely different league.  The effect of being immersed in the sound scape is multiples greater.  Musical intensity is higher but there's no down-side to this effect.

 

I'll report more later with other music programs.  

 

I have not heard Holo Audio's May KTE DAC (about $5500 USD) but many if not all who heard May and my S3 (both with HQ Player) suggest the following: HQ Player provides the best performance; the 2 DACs perform much closer to each other than is indicated by the price difference and many call the May KTE a giant killer on an absolute scale.  If those things are true, then DAC 200 for $1700 more than the May KTE appears to be superb value.  There's nothing to criticize about the Chinese Holo's build quality but the fact that DAC 200 is made in Deutschland surely carries considerable cachet.  DAC 200 has considerably more features too; I love the meters. 

 

 

For the record, I used the same high-grade AC mains cable for both DACs, so mains power was off between auditions.  

 

  

Agreed 

just looking at the components used on the T+A 200 , the fact that it is made in Germany where wages are consiiiiderably higher than China , taxes, overhead etc………..

yes personally the Holo Stuff should be much cheaper still.   I imagine their profit margins are much larger than T+A.   
 

 

[Nkam]

It’s wonderful partaking in these forums.   
 

chit chatting about stuff we love.  
 

but at the same time I can’t help thinking……

 

what is everyone listening through?

can we see a freq response graph of their room and what is going on RTDecay wise? 
what other equipment are they using?  Amps , speakers? 
 

 

Please no offense as I do not mean it in a bad way. 
 

i just know that when I started my first setup and first room.  I hired someone to teach me REW and my journey of setting up my speakers and room with what I already had , had begun.    
because I am not a pro , it took me around two years to make a HUGE a difference in the sound quality.  
measurably as well 

 

and I remember listening to some DACs back then or other gear , now knowing my speakers and everything else wasn’t setup nearly optimally.  
 

 

just pondering.  
 

Rant over. 
 

 

[Miska]

3 hours ago, Nkam said:

so the T+A DAC 200 has the same performance with HQplayer as a true NOS DAC does? 
in PCM that is.  

 

I never use mine in PCM mode. That would limit digital filter rate to max 16x instead of 256x or 512x I get through the DSD path. And I rather listen through my more advanced 7th order modulators than TI's simple 3rd order modulators.

 

Measured performance is also quite a bit better through the discrete DSD converter side. So I always run my T+A DACs in DSD mode.

 

[camott]

8 hours ago, 1125FPS said:

I have not heard Holo Audio's May KTE DAC (about $5500 USD) but many if not all who heard May and my S3 (both with HQ Player) suggest the following: HQ Player provides the best performance; the 2 DACs perform much closer to each other than is indicated by the price difference and many call the May KTE a giant killer on an absolute scale.  If those things are true, then DAC 200 for $1700 more than the May KTE appears to be superb value.  There's nothing to criticize about the Chinese Holo's build quality but the fact that DAC 200 is made in Deutschland surely carries considerable cachet.  DAC 200 has considerably more features too; I love the meters. 

 

I used to have the May KTE and just acquired a DAC 200. While it may have measured really well, I was never overly impressed with the May. Sounded flat - maybe I just don't like the R2R sound. The DAC 200 is in a different league in my opinion and a keeper. It has 99% of everything I love about my tubed Lampizator - which obviously measures poorly - plus a lot more. Very dynamic yet it is the least fatiguing DAC I have listened to - that's the biggest win for me. I use it purely with HQP upsampled to DSD256. The excellent line stage is icing on the cake. And yeah, I love the meters too.

[Nkam]

9 hours ago, camott said:

 

I used to have the May KTE and just acquired a DAC 200. While it may have measured really well, I was never overly impressed with the May. Sounded flat - maybe I just don't like the R2R sound. The DAC 200 is in a different league in my opinion and a keeper. It has 99% of everything I love about my tubed Lampizator - which obviously measures poorly - plus a lot more. Very dynamic yet it is the least fatiguing DAC I have listened to - that's the biggest win for me. I use it purely with HQP upsampled to DSD256. The excellent line stage is icing on the cake. And yeah, I love the meters too.

maybe out ears like up to and around 0.001% THD and when it gets to around 0.0001% the sound gets too flat.   
I don’t know.   The DACs with super super low THD always sounded a bit flat to me.    Even engineering giants like Benchmarks DAC3.  Sounded a bit flat to me . Although I own and adore their AHB2 and LA4.    Their DAC didn’t really fit my needs.   The low end of the DAC3 I must say had the cleanest most detailed sound of any DAC ive tried.