I understand that. And when for example the turnover is say 500Hz, then 500Hz is +3dB above the level of 1Khz and any flat part of the curve. So to check accuracy of an EQ you normalise the 1Khz output measurement and go to 500Hz which should be +3dB higher.
But if the turnover is specified at 1,000Hz then it isn't possible to get it +3dB relative to 1kHz because it's one of the same thing, so is it then permissible to measure from an otherwise flat part of the curve at say 5Khz ? And is this documented anywhere ? I have, I think, all text books on the subject but can't find a definitive reference. I am looking for both friendly advice and also a definitive reference please.
Other experts, feel free to chip in too ?
Dave (BTW Gary sorry for the spelling mistake)