[beowulf]

I was just reading about the IDS-25 on Roger Russell's site.  The IDS-25 is a Line Array that uses 25 identical, full range 3-1/2" (90 mm cone radiators).  No crossover, mid, tweeter or woofer drivers are used.  However, the IDS-25 are super expensive (as one can imagine using 50 drivers and an equalizer must cost).

But after reading about Russell's design, it got me to thinking how awesome it would sound to use a similar design using Louis' Omega RS5 drivers.  Louis, have you ever created or considered doing a Line Array with your own drivers?  Do you think the performance gain would justify the cost, etc. and is that may drivers really needed to get the same effect (i.e. could one use say 5 larger drivers to achieve the same effect or is the actual height what's important about the Line Array design in the first place)?

Thanks!

[gab]

Do you think the performance gain would justify the cost, etc. and is that may drivers really needed to get the same effect (i.e. could one use say 5 larger drivers to achieve the same effect or is the actual height what's important about the Line Array design in the first place)?

Thanks!

http://audioroundtable.com/misc/nflawp.pdf

gab

[JLM]

Line arrays have several challenges:

 - They are not point sources, so cannot image vertically
 - Arrival time from varying points result in phase smear (unless a complex delay system is used and then can only work correctly at a single listener position)
 - To avoid comb effect the driver spacing cannot exceed the highest frequency wavelength (13,200 inches/Hz)
 - The comb effect thus requires midrange and tweeter drivers with crossover
 - The number of drivers involved and the series/parallel wiring scheme needed to maintain workable impedance would require a large amount of wiring
 - The required number of drivers drives cost up or quality of drivers down (done correctly would require about 65 cone/dome drivers per speaker)

So the proper implementation of line arrays is either very costly or flawed and still can't image vertically.

[beowulf]

Hey gab, I meant to reply to this before ... thanks for posting that paper.  It's quite the read and lot's of great info in there!

[beowulf]

Line arrays have several challenges:

 - They are not point sources, so cannot image vertically
 - Arrival time from varying points result in phase smear (unless a complex delay system is used and then can only work correctly at a single listener position)
 - To avoid comb effect the driver spacing cannot exceed the highest frequency wavelength (13,200 inches/Hz)
 - The comb effect thus requires midrange and tweeter drivers with crossover
 - The number of drivers involved and the series/parallel wiring scheme needed to maintain workable impedance would require a large amount of wiring
 - The required number of drivers drives cost up or quality of drivers down (done correctly would require about 65 cone/dome drivers per speaker)

So the proper implementation of line arrays is either very costly or flawed and still can't image vertically.

Interesting, the IDS-25s do not have a crossover, although they do use equalization of some sort.  It makes a lot of sense though that by having so many drivers and attempting to keep the costs down would mean cheaper drivers.  Thanks for pointing out some of the cons to this design.

[JLM]

I owned a McIntosh decades ago, but frankly would never put on my resume that I had designed McIntosh speakers (expensive yet totally unremarkable).

The IDS-25 will have comb effect over 2,500 Hz and can't image vertically.  The "real" efficiency of the speaker (without the array effect) based on the manufacturer should be 82.5 dB/w/m.  Overall not impressed for the price (just like any McIntosh speaker I've heard).

Patients aren't necessarily good ideas.