But if our sighted impressions will screw up the sound so much, why expect to perceive the "sound we liked" during the blind test when listening sighted?
Can we detect "poor, better, good, great" sound in sighted conditions, or not? If not...again...what does it matter if we detected them in the blind tests?
We aren't listening blinded at home.
One must be more brutal here: It is not that when we are listening sighted, we hear this or that according to what our eyes see and the sound just stays that way forever. Sound is still sound, what our brain however makes out of it changes from situation.
Our vision is dominant over our hearing, biologically, and influences how we rate our sonic perceptions. When you buy the speaker with an eccentric frequency response, but it had these diamond membrane woofers and the high gloss finish, sighted, they sounded good, because your brain was imagining how they would decorate your livelihood and make you appear respected and a gentleman of good taste. Another day, you find asking yourself if they make you appear like a prick, your eyes look upon them negatively and they do not suite the room as good as previously thought. Now you hear this etching sound and it creeps to your mind, nagging. Why did I buy them?
The sound is a hard fact, as is the interdependence of vision and hearing. But it is not that we can simply hear what we want because we derive it from visual imagination. The sound is there, and as soon as our established mental image of the speaker collapses, we can perceive it again with our hearing.
I am the proud owner of a chipboard monkey coffin with a 10-inch PA driver, compression driver and waveguide. A very simple speaker. It is driven by a DSP amplifier. I am limited by gated indoor measurements, and due to the low resolution at lower frequencies, there is always some doubt if I had actually set up everything correctly. This is going for about two years now. And I have made changes to the filters throughout the time.
Recently, I read a post by
@kimmosto how waveguides would create a closed in sound. And yes: I was looking at the 90x45 waveguide and I could hear it: no spaciousness, very contained. I knew, and I could see it, it was this speaker construction. I had clearly invested in the wrong approach. The lack of openness, I reasoned, is also because it is lacking a dedicated midrange. And waveguides after all are not as useful. However there was an issue with the waveguide causing this: How could I then enjoy so much the Genelec 8030A at the broadcasting office? They have a waveguide. Obviously it must be the dome tweeter I thought.
After some more time living in painful doubts, I thought: So you had slightly reduced the airy frequency range from 3k to 8k. What if below that, some energy was missing? I added a 0.3 dB boost around 1.7 kHz with a low Q PEQ, where there might be some energy missing because of the crossover at 1.25 kHz. It sounded so gorgeous! It just opened up the whole presentation.
And so I once again had to discard an assumption that I had created from written knowledge, projected onto the speaker as visual object. This is so hard to learn, and it has happened so many times already. Bass, anechoic bass vs. room gain, low mids, obviously the 1k region, 3 to 8k presence region, lastly the airy bit above. I had found etching sound to be exceptionally clear, boomy setups to be full, anemic setups to be precise, closed-in setups to be balanced. All the time, the ears where slower than my pre-established concept of what they would be hearing, and they were patient and accepting my ignorance for weeks over weeks. Until, finally, my brain gave room for their perceptions and I heard something wrong and needed to find out what it was.