@shigzeo
You got me wrong. I know that and never said that it would. All i said was that I prefer devices that are able to give high levels of damping (preferably -85 dB and more) and control the volume in 0.5 or 1 dB per step. I don't know Plenue D's attenuation step size in dB and also don't know whether it keeps consistent over the whole range of 100 attenuation steps, simply because the D was too loud for me with the lowest possible volume setting for listening quietly, even with masters that are rather quiet. That's also why I didn't measure its attenuation step size in dB but sold it shortly after it arrived, which was quite sad as it was one of the quietest DAPs with low outpu impedance I've come across (just very little hiss and not as quiet as the DX90, however still very quiet).
And it's quite obvious that the numbers are just a scale and not total numbers in dB which wouldn't even be possible to display without knowing the headphone's sensitivity and impedance. Nonetheless such high RMS dB levels aren't even possible in the earth's atmosphere.
Read my post again and you'll see that I just stated that I prefer fine-grained volume attenuation/dampening where every step is equal, no matter if in the low or high attenuation range. And didn't relate Plenue D's scale to anything other than it is - a scale with 100 steps, not an info about the actual loudness.
You just got me wrong. Having been a highly regarded member in the German audio community before I went over to Head-Fi, given advice in many thousand threads, raising the importance of metrological volume matching when comparing gear and that the subjective differences disappear most of the time or are very small when the volume is correctly matched within less than 0.5 dB, raising the importance of why multi-driver in-ears require a very low impedance output to sound "correct" and having performed many RMAA tests to measure the output impedance/frequency response deviation/detect caps in the signal path, I am aware of all of that - you just got me wrong.