- Jul 17, 2013
Noise level is relevant. Let's talk about THD and THD+N -- that's total harmonic distortion, and total harmonic distortion plus noise, these being two different kinds of measurements.
With THD+N, noise is part of the measurement. AP defines THD+N as "the sum of all the harmonic distortion products in a measurement, as opposed to selective harmonic distortion, which measures discrete harmonic products separately. THD+N adds noise to this figure." Further:
You can see, then, if the system in question is measuring THD+N, then noise should be considered. From the Audio Measurement Handbook by Bob Metzler:
That books has been around for many years, so let's discuss THD measurement (as opposed to THD+N). For a THD measurement (not THD+N) in Audio Precision's APx, what's being measured are only the bins that contain harmonic distortion products. If that's the method being used, does that mean exogenous noise has no effect? No, that's not what it means. Especially in an acoustical measurement, noise can still affect a THD measurement. Even if measuring only the bins that contain harmonic distortion products, external energy that reaches the mics and gets into those bins can be reflected in the THD calculation. This can be more evident, for example, if the FFT length is short and the bins are wide.
For passive headphones, we go with THD. InnerFidelity measures THD+N. (As for anyone else, I can't account for the different software various people are using and how distortion measurements are determined across the many different apps used.) I do believe Tyll mentioned in a recent post that InnerFidelity will soon be switching to Listen, Inc.'s SoundCheck measurement system, which is an excellent system, especially for electro-acoustic test. When he makes the switch, perhaps he'll continue to stick with THD+N to represent the distortion of passive headphones, but I am guessing that perhaps THD will come to the fore at InnerFidelity for that purpose after the system changeover.
Again, noise can definitely affect THD, and especially THD+N, measurements.
(Perhaps we can run a special set of measurements with the HD800 and/or HD800S to illustrate this, when time permits.)
We are talking about comparative differences between the HD800 and HD800S as measured on the same setups. So I don't see how noise floor is the answer.