I did have a look at the site before posting, but did not find the maximum SPL information, as it is somewhat confusingly on the "Isolation" page, rather than "Sound Quality". They do not mention what frequency the maximum SPL is tested at, though. If they test it at 20 Hz, then 3% THD is not that bad, it is fairly common for dynamic headphones to have relatively high distortion on loud bass.
For those who are interested, I show the results of some tests I did with my DT770 Pro and DT880 Pro (both 250 Ω), using an in-ear microphone created from a measurement microphone capsule. While this method is not suitable for accurate (or even usable) measurements over the entire audio range, it works well enough below about 1-2 kHz, and in this particular case we do not care about the treble anyway:
Top left: DT770 Pro 250 Ω without a good seal (rolled off sub-bass with higher distortion)
Top right: DT770 Pro 250 Ω with better seal
Bottom left: DT880 Pro 250 Ω without a good seal (wearing glasses) and from 100 Ω source impedance (all the others are 10 Ω)
Bottom right: DT880 Pro 250 Ω under the same conditions as the top right graph
Blue: frequency response at -10 dBV (~92 dB SPL for the DT770, ~90 dB for DT880)
Green: frequency response at 0 dBV (~102 and ~100 dB SPL, respectively)
Yellow: THD at -10 dBV (up to 5th order harmonics are included, the THD is referenced to the total RMS level of the signal); -20 dB is 10%, -40 dB is 1%, -60 dB is 0.1%, etc.
Red: THD at 0 dBV
- the SPL levels are simply calculated from the assumed sensitivity of the headphones (96 and 94 dB SPL at 1 mW and 1 kHz), based on information I could gather from various sources, and therefore may not be perfectly accurate
- because of the simple measurement setup, the results are only reliable up to 1-2 kHz, and basically useless above 4-5 kHz; the lowest bass frequencies are also slightly rolled off by the microphone (other than that, it has flat response in the usable range)
- bass performance is significantly affected by how good the seal is; bass roll-off also makes the distortion look worse by attenuating the fundamental more than the THD products
- microphone placement has a dramatic effect on the measured high frequency response (which you should ignore anyway)
- the right driver of the DT880 distorted the bass more than the left one, but still slightly less than Tyll Hertsens' DT880-250
Perhaps I should also have done a THD vs. level test at some low frequency, for example increasing the voltage up to 2 Vrms at 40 Hz. I also have IMD tests at 55 Hz + 2 kHz, but the result is similar to what would be expected from the low frequency THD (which is not good, i.e. the distorting bass can "rattle" the mids and treble).
By the way, if loudness and bass are the most important to you, why not just get Beats by Dr. Dre Pro ? Tons of bass, and no less than 118 dB (!) maximum usable SPL.
Seriously, those are actually decent for the market they target.
Edited by stv014 - 12/14/12 at 2:24am