What affects the size of speakers' effective phantom center area?

LOUT

For example; you have a pair of speakers playing and the central parts of the music are perfectly centered when you're seated in a roughly centralized
spot, but when you stand up and walk over toward one side this "phantom center" effect breaks. 

At around 0:14 into this video

there's a simple polarity test using mid-focused noise and flipping phase in/out between the speakers. In-phase, while sitting in the "sweet spot" it sounds nicely centered while the out of phase noise sounds seperated L/R and distant or hollow. However, if I lean a couple feet left or right, then the out of phase signal actually sounds more centered while the in-phase noise sounds a bit more noticeably separate L/R. 

Is this normal speaker behavior (the out-of-phase signal ending up sounding more like the in-phase one when outside of the sweet spot)?
What things can a DIY speaker builder do to increase the size of the sweet spot where centered music and effects sound..centered?

LOUT

Part of the reason this bothers me is because flipping the polarity of one tweeter so the speakers' woofers stay in-phase while the tweeters are now out ruins the narrow "sweet spot" but seems to give a better phantom center effect to the spots on either side which happen to be where the actual seats are. Right now, wiring the tweeters out of phase with each other seems to share the "centered" effect to a couple people while wiring in-phase makes a narrower sweetspot that's really only good for one person...or two really skinny persons that are okay with cuddling. 

jr@mac

Focus on symmetry. Gain and frequency response in particular. Your system should match as close as possible in those areas from left to right. 1db over an octave or two will have a noticable effect on imaging. Driver variance, component variance can vary as much as 5-10%, that will impact it. Auddysey can correct for quite a bit of that, give it a try. 

If the source is not mastered to that end, it won't happen.

LOUT

For a bit of a sanity check I tried switching over to a pair of full-range single driver boxes with the bafflestep crossover removed. This can't fix driver variance, but it should remove at least some of the potential pitfalls of component variance and woofer+tweeter timing/phase. The center imaging and the oddity with in-phase and out-of-phase signals when listening from exact center or a foot+ to either side still happens and it appears to act exactly the same. I've tried moving my seating position forward and backward and it doesn't seem to help a lot. I'm guessing it might be something to do with the room and speaker positions or me going crazy.

But I'm happy it doesn't seem to be tied to the driver spacing/timing or the crossover. 

Kornbread

Lay a cheap laser pin along the side of the speaker to help align them.   

 

Have you had your hearing checked lately?  Not trying to be a dick, but images that should be dead center favor my left side.  I've changed a ton of things trying to find out what's wrong until noticing a recent hearing test administered at work.  It appears my right ear has lost some sensitivity in the upper range while the left remains normal ... for my age.  Since being unable to find anything else wrong I'm thinking it may be me.   

Tom_S

I have the same problem - singers & instruments sometimes move off the center as they go up and down the scale. I've known for quite a while my ears are noticeably different. But it could be worse. My boss is 5 years younger than I am and he had to get hearing aides two years ago. 

LOUT

Ha! No offense taken, that's probably a good point for center drifting dependant on frequency (assuming there isn't something wonky with the tweeter or crossover). In my case the problem isn't a drifting center nor an off-center....center, it's just a complaint that the seating position where the center image is at its most "pin-point" happens to be too small for multiple people to enjoy at once. On the positive side, for music and TV the center image is still there, it's just less precise. Moving my head over into the sweet spot makes the center image so "tangible" that it draws my eyes in an attempt to see where the sound is coming from. 

I'm doubly fortunate that I don't seem to have an issue with the "center" drifting for different frequencies, and I don't think my hearing is noticeably unbalanced L/R. I can only differentiate up to 10K for pinknoise VS fullrange pink and only hear around 14K for sinewave, but I'm not sure how much that affects this. 

kenrhodes

You may want to consider the room itself as part of the problem.