I think we're getting there. Thanks for the perseverance. Cheers for that.
Why would people want to use their ears to identify shortcomings instead of measuring?
Not everyone has measuring equipment and software. (I do have a mic that I use to help me postion the speakers and listening spot and to measure the response in the final position to generate EQ filters; i did make a few gated measurements when I briefly had a play with horns)
Unfortunately lot of equipment hasn't been measured.
Many speaker measurements show only basic FR and impedance plots.
Some measured shortcomings are audible and others are not or not significant.
To better correlate listening with measurements.
As it once happened sometimes a particular dispersion pattern whilst deemed ideal won't work well in your room.
To upgrade with more intention instead of accidentally.
In reply to your last question, I agree that one will partly use benchmarks and prior experiences but mostly one must try spot potential problems by listening to different fit-for-purpose recordings.
Fair enough. I'm think we're coming at it from different angles, as I only use speakers that I've designed and therefore have quite thoroughly characterised objectively. I've also gone through a lot of the studies into audibility thresholds of all the various forms of distortion, so I believe I have a pretty decent ability to look at data and determine whether issues will be audible.
And having designed a lot of speakers and also having done some basic mixing/mastering work, I know that my ears (and especially my ears/eyes/brain working deviously in concert) are really not something I'd want to rely on, except as a last resort.
For a classic example demonstrating why this is so, a few times while mixing I've sat there adjusting the EQ or some other setting on an effect unit, sometimes for 15 minuntes or longer, taking it for granted the whole time that I'm finessing what I hear coming out of the speakers, only to realise half an hour later that the effect unit wasn't connected to the track I thought I was working on.
Another example comes up in testing for audibility of various forms of distortion, e.g. nonlinear distortion, group delay etc, with myself as the subject. I do this by creating an impulse response or some other filter containing X degree of distortion in whatever parameter I'm testing, and then using an ABX comparator, much like I suggested earlier today on the other thread. Typically, I find that when I first run a music signal through the filter, it seems absolutely obvious that the distortion is audible. Switch back and forth blind a number of times, and usually what I thought I heard has disappeared. Or it doesn't seem to have, so then I conduct an ABX test on myself. Then, only in a small number of cases do I actually pass the test.
And my chances of success are always way higher when I focus on a tiny snippet, usually 2-3 seconds, and switch very quickly. This is a highly unnatural listening condition, but it's often the only way of having any hope of discerning a difference. The idea that I'd be able to hear things in long-term listening that I miss under these circumstances seems pretty outrageous to me.
And well, you know, I'm a pretty rational person with a lot of listening/mixing experience who's also done some systematic ear training, so I know that I'm reasonably sensitive to a lot of these things compared to the average, and pretty self-critical when it comes to my hearing abilities. Yet I still seem to repeatedly be able to fool myself into thinking I can hear things that, upon closer scrutiny, it turns out I can't.
