1. Speakers are inaccurate. All of them, period. Some more, some less. There is no such thing as a speaker with a flat response, just those with differing degrees of deviation from flat.
2. Speakers are already full of filters (which is what an eq is). The box is a filter. A crossover is at least two filters. The drivers themselves are filters.
3. The frequency response deviations are associated with phase deviations that are often corrected with the very same filter set that corrects their frequency response.
3. The best sounding large concert systems have lots of electronic filters applied to correct the system itself. This is totally aside from room issues. Bad sounding concert systems probably also have a bunch of filters applied, they're just not done properly.
The purist attitude isn't supported by the reality, which is far from pure. I'm not talking about random "fiddling" with tone controls, I'm talking about specific corrections for specific inaccuracies. In some cases, for example my cheap Soundblaster computer speakers, the inaccuracy is obvious and an adjustment can be made by ear with the media player's eq. In other cases a bit of actual measurement is needed to determine the exact filter parameters (on a parametric eq) needed to bring the system closer to flat. And by measurement I don't mean a crude, time-blind RTA but something that takes time into consideration so you can sift out room issues from speaker issues and deal with them appropriately. Just as it is true that you can't truly fix room resonance issues with filters, you can't truly fix speaker frequency nonlinearity with room treatment.
Okay, there are speaker systems so close to flat that correction is not really required, but I bet on this forum that applies to a minority of monitoring setups and casual listening systems. You guys are basically saying you'd rather hear inaccurate sound from your speakers than be "impure" and apply a corrective filter. That's not rational.
