The 0.1% THD figure is incorrect, and yes 0.1% is very important, as one can hear -120dB or 0.0001% for higher order harmonics. The published spec is 0.0005% at 3V RMS 1k, but this is conservative, using a passive notch filter I can measure 0.0001% THD 3v RMS - and its only 2nd and 3rd harmonic, no other distortion and no other noise floor modulation at all. The figure for 1v RMS will be much lower than the 0.0001% level. I don't make great claims about the measurements as there are more important things to talk about, and Hugo challenges the capability of test equipment.
On a technical level (and you can verify these points with measurements), if you wanted to give 3 reasons why people claim it has a unique sound then they would be:
1. Use of largest ever tap length digital filter - the 26,368 tap WTA filter. Now the interpolation filter has the job of recovering the bits in between the samples, and if you look at the math, if you use an infinite tap length filter, then you will perfectly recover the original bandwidth limited signal - it would be as if you had not sampled the signal at all! Now the problem with real life is that we can't use infinite tap length filters. Conventional DAC's use a few hundred taps, and this will give a good frequency domain response, but poor time domain response. The poor time domain response means that the transients (the starting and stopping of notes) is not reproduced accurately enough, so the brain can't tell when a note starts or finishes.
To verify that Hugo has the most sophisticated filter ever, measure the filter using 0dB random noise as a source. You will see the OP is very close to an ideal brick wall filter - much closer than any other digital product available today.
2. Lack of noise floor modulation. I won't go into the details technically why, as its too complicated, and this performance comes from many, many factors. Anyhow, if you measure Hugo's OP from 0dB to nothing, and use a passive notch filter, the noise floor is completely unchanging. This is very important subjectively - noise floor modulation makes the it sound hard and grainy if very bad; bright if moderate, and when it is completely removed, instruments have a much bigger timbre range. Hugo has un-measurable noise floor modulation, if you use a passive notch filter. You need to use a notch filter as the ADC's in test equipment have much more noise floor modulation than Hugo has.
3. Small signal fidelity. Low level signals are crucial to the perception of sound stage depth, and detail resolution. DAC's have big difficulties with small signals - they create distortion, and adjust the signal amplitude (effectively having different gain at different signal levels). This is measured with fundamental linearity tests. Now Hugo reproduces a -140dB signal perfectly - no amplitude error, no distortion. Indeed, you can go from -40dB to -140dB and see no distortion, no change in noise floor, with perfect fundamental linearity. You will not get this performance from other DAC topologies, and its never been done before at Hugo's price point.
If you would like more info on these points, have a look at my other posts. One of my posts has measurement FFT's of -40dB and -140dB.
A final point - we audiophiles are a fickle bunch, with lots of different passions and tastes - but everybody that has listened to it, the acclaim has been universal. Now that really is odd!
