[chris3377]

I know this should go in the general audio forum, but since you guys always seem to know, or know someone who knows, I am posting this here, but you can change it. In all my reading and research I have never come across 2 things that are bugging me.
1. What is the smallest amount of time an average human ear can hear between 2 sounds? I have read the whole time delay feature on receivers is a joke, and since one manufacturer put that feature on, now everyone does it. I have read that the ear can pick up differences between left and right really well, but what about if both drivers are on the right side.
2. What is the dissipation rate of higher frequencies as they travel throguh the air? I know low frequencies can travel really far and high ones can't, but I want numbers.
  Now for the reason it's bugging me. I have been thinking of building a taller set of speakers, maybe a line source, that has multiple drivers and tweeters. During my research on ideas and designs I have seen many that have really ugly shapes. Even though I love the JM Lab Grand Utopia, it has one of the shapes I am talking about. They put the tweeter further back from your head and say it is because you want all the voice coils the same distance so one sound doesn't travel to your ear faster from one driver than another. So with the speed of sound, does the small difference between a tweeter and woofer's voice coils, say 2 inches, at 10 feet away make a difference? Then does the rate at which high frequencies fade away make it so you need the tweeter closer, negating the reason for making it further away in the first place. This whole idea of making the tweeter further away seems like voodoo science to me, that's why I ask. It's ight up there with people who claim that speaker wires need to be the same length. Thanks in advance for any help.

[EProvenzano]

I think the designers are actually trying to allign the drivers' acoustic centers.

On a flat baffle, the accoustic centre of a woofer is set deeper into the cabinet than the adjacent tweeter driver's centre.  I believe that's why the tweeters tend to be set back in a 'time aligned' system.

I'm no designer; this is from my reading.

[EProvenzano]

Here's a tread on the Ellis circle where Dave discusses some of the advantages of aligning the drivers in this fashion.

http://www.audiocircle.com/index.php?topic=20868&start=50

Danny, I hope you don't consider it poor taste, posting this in your circle.  If so, please feel free to delete.

Thank you.

[jon_010101]

Voodoo science?!    Not really...

Imagine 10kHz signal... period 1/10000 sec... wavelength=speed of sound * period, so ~340m/s*(1/10000s)=.034m=3.4 cm.  So, if the acoustic center of your tweeter is offset from other drivers by 2" you are shifting the phase at 10 kHz by ((2.54cm/in*2in)/3.4cm)*360deg=537deg!  And, of course, similar phase shifts are introduced throughout lower and more-relevant frequencies... but the effect is certainly small.  Still, significant misalignment would contribute to blurring of transient effects or a funky soundstage.  Of course, this simple analysis ignores other sources of phase problems... the crossover itself, the electromechanical properties of the speakers,  speaker placement and seating position, etc.  But if you can reduce phase shifts by driver placement, ummm, it definitely won't hurt (although it might not help much either).

Now you also mentioned the issue of dissipation and frictional attenuation through air?  ... this effect can be considered very small compared to many other variables, and certainly neglected for a living room.  The polar response of a tweeter (and of the speaker array) is much more significant in determining how much high freq sound reaches your ear.  This is what results in the "sweet spot".  Such geometric effects... regarding driver positioning, array configuration, room placement, cabinet baffle shape, etc. will easily dominate over frictional/diffusive/dissipative losses associated with acoustic propagation through air.

So, no worries!  

[_scotty_]

jon_010101 missed one of the primary problems that results from phaseshift.
  When phaseshift occurs as a result of a crossover something must be done
to place the drivers output in phase in the transition region or the drivers output will cancel out partialy or totally resulting in notches in the response curve at the crossover frequency.  If there is a substantial region of overlapping output between the two drivers multiple notches will occur and large lobing errors will be present resulting in poor imaging,room placement problems and a very small sweet spot where even slight motions of your head result in image shift.  
   Build your line source and enjoy the results and follow Danny's recommendations regarding driver placement on the baffle.
Scotty

[jon_010101]

jon_010101 missed one of the primary problems that results from phaseshift.
  When phaseshift occurs as a result of a crossover  ...

Ha, I did... and that is probably the most important problem.