Kemper Profiler MK 2

[Sascha Franck]

It’s likely that something that has a deep null with a source probably sounds very close to it, but when comparing a few different profiles, they could all null to a similar amount, but in areas we perceive differently (either based on the source, the frequency range, the amplifier, the tone, the playback system, the users ears).

This! A nulltest could fail miserably, but in case it's in a less important frequency range, the sound might still be more accurate than another contender showing better nulltest results, in case the "offsets" happen in a more relevant frequency range.