There are always group of people who are interested on the objective measurements by DAC. We have in house measurement systems to check all our recording and playback equipments. You can read DAC202U and DAC2 in house measurements by ourselves. The measurement is done via a digital loop through**.
Reference Digital Test Tones -> DAC (Test unit) -> reference ADC -> Measurement equipment.
From the above signal flow, one can see the reference ADC (analogue to digital converter) can also be a factor. If the DAC output distortion is lowered than the reference ADC, the measurement will only show the limitation of reference ADC, rather than actual DAC performances.
Here are some reference measurements of Weiss flagship Medea+ DAC.
It is impossible to measure the (THD+N) quality of Medea+ with the DAC – ADC approach. One has to do it as shown in the above thd-fft-96.gif file. There the 1kHz signal has been suppressed in the analog domain via the AP system (as it is done in a THD measurement, you see the dip around 1kHz caused by the notch filter which notches out the 1kHz tone). This signal then is A/D converted and a FFT is done. Because the 1kHz signal is suppressed, the dynamic range of the signal is much less, i.e. just the harmonics and noise and thus the A/D converter does not have to process a huge dynamic range. The AP then recalculates the proper scaling as if the 1kHz tone was still there. So the 3rd harmonic in the graph is more than 120dB below full scale.
Same with THD vs level, there you can add the input level (horizontal axiss) and the measured THD+N to get the THD+N relative to full scale. As an example at -60dBFS the THD+N is -70dB, so the THD+N relative full scale is -130dB. This THD+N figure,we cannot use and DAC-ADC method to measurement.
The linearity graph is incredibly good.
For people who buy a Medea+ will not need to consider these measurements. There are also other factors (output impedances, output stage topology etc) which do not shown on above basic measurements. The Medea+ uses 8pc of OP1-BP gain stages, which are close to perfect on many analogue applications. While I read some audiophile presses saying our hifi industry is lacking of real new technologies, I would raise most of the audiophiles today are looking for something else rather than real technologies.