A new center channel for the Octagon - AVS Forum
Forum Jump: 
 
Thread Tools
Old 08-27-2012, 02:04 PM - Thread Starter
pnw
Senior Member
 
pnw's Avatar
 
Join Date: Dec 2003
Posts: 439
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 7 Post(s)
Liked: 36
Okay, I’m going a little psycho (acoustic). The entire premise for this project is our ear-brain integration time and how it interacts with the acoustic source and room reflections.


Background:

Over the past couple of years I’ve followed posts by respected industry professionals from companies like KEF, Snell, JBL, etc. I pay particular attention to their holistic views on speakers and rooms…the perception of sound in real rooms. Among other things, this obviously includes acoustic reflections from room surfaces.

It seems generally accepted that our ear-brain has a very short integration time at high frequencies…we sense first arrival from the speaker, largely oblivious to room reflections. As frequency decreases, integration time increases progressively until, in woofer territory, we fully sense contribution of the room…our ear-brain acts sorta like windowed MLS measurements. I have been unable to find quantitative documentation of this “sliding scale”, but the tendencies seem clear.

Floyd Toole indicates we can detect acoustic flaws as narrow as 1/20 octave. SoundEasy doesn’t have a selection for 1/20 but it does do 1/24 octave which, unless otherwise stated, I’ll use consistently in this thread. Also for consistency, only Hanning windows are shown. (Hanning and other tapered windows like BH and KB emphasize the leading edge of the impulse, trailing off to zero influence at the end of the time window; a rectangular window weights all input equally.)

All measurements shown here are from the 4M listening distance. Why 4M? Many speakers are designed with practically anechoic measurements (1M, gated MLS, GP, etc). This is appropriate when the listening environment is unknown or unpredictable. However, IMO, it is somewhat shortsighted to ignore a known environment, like our purpose-specific listening rooms. In other words, you will not see the issue discussed here in your system unless you look for it with LP measurements in your room.

EQ of the “old” speakers evolved through hundreds of measurements. The most pleasing results were obtained with “mental integration” of measurements at 2ms, 4, 8, 16, etc, doubling through 128ms. 2ms provides the reference up around 10kHz (including direct arrival only) with increasing room reflections added until 128ms provided the subwoofer reference. Here is an overlaid example to help illustrate the process.



Through the measurement/EQ process, floor reflection impact was obvious, even with short “listening” windows. For example, a 2ms Hanning window timed to exclude floor bounce yields clean first arrival frequency response. Including floor bounce by extending even the very forgiving tapered window to 3ms begins to corrupt FR. To illustrate rapid onset of the issue, here is one of the “old” speakers from 4M, at 2ms, then 3, 5, 10, 20, and finally 100ms. (BTW, this and all following MLS FR plots in this post are run at the same level but manually separated for clarity.) The longer windows allow other reflections to enter the picture, but floor bounce still dominates midrange corruption because it is the nearest surface with the strongest return. Also notice the apparent increase in bass levels as longer integration time and the room come onboard.



The Octagon ceiling is already heavily treated with fiberglass stuffed in the space between (fabric covered) joists, but ten 2’x2’ floor pillows were the only practical tool I had to combat floor bounce. Unfortunately the pillows are not tunable and positioning them for improvement usually met with degraded performance elsewhere. No pillows were used in any of the measurements in this thread.

pnw is offline  
Sponsored Links
Advertisement
 
Old 08-27-2012, 02:04 PM - Thread Starter
pnw
Senior Member
 
pnw's Avatar
 
Join Date: Dec 2003
Posts: 439
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 7 Post(s)
Liked: 36
The objective defined:

So, the objective for this project is increased vertical directivity to reduce midrange combing generated by floor bounce between the speaker and the 4M listening distance.



The concept:


The “old” speakers are, nominally, about 60x50 above crossover. 60 degrees horizontal is wide enough for full audience coverage, but narrow enough to just skim the near sidewalls. However, at 4M, 50 degrees vertical almost fully illuminates floor and ceiling reflection points.

Since this system is used only for seated listening, a much tighter vertical window seemed reasonable…hummm…how about 15-20 degrees vertical? Limited to a 32x38 baffle, a 60x20 aspect ratio seemed impossible without stacking horns…and stacked horns for a close LP seemed irrational. Another alternative would be a vertical line array, but vertical polars for standard lines vary dramatically across the frequency range…even worse, they usually have uncontrolled asymmetric horizontal directivity…and I sure didn’t want to go backward in that area.

Having used shunt capacitors to taper line length vs frequency for many years, I knew caps could be used to achieve a fairly consistent vertical window. Thoughts moved to combining a progressive line for vertical control with a waveguide for the horizontal.

Baffles 32” x 38” limited “real” vertical control to around 400Hz but using low order slopes from woofers to the WGs, would extend line length and control. A skosh of WG waist-banding could extend horizontal directivity into the region of broad overlap with the woofers for more uniform system response. Borrowing from pro-sound, a little diffraction would be sacrificed in favor of horizontal control to a lower frequency (darned tradeoffs!)


The drivers:

A compression driver would likely require a separate “horn within a horn” with a mouth terminating in sharp edges…problematic for diffraction. I decided to use the natural directivity of a tall tweeter, with foam to trim tweeter directivity as needed. Since I’d already constructed several large ribbon drivers, a ribbon was the first consideration…but a light 5 micron ribbon in a WG probably can’t co-exist with eighteen 15” subs on the same baffle! Next on the list was an AMT, specifically the Aurum Cantus 30130.

For the mids, Zaph’s testing of the Vifa TC9FD18-08 caught my eye….very low native distortion in the anticipated frequency range and broad bandwidth; providing flexibility in both upper and crossover regions. Small size with a square frame meant eight could be stacked in a close-spaced midrange line for fine taper gradients. High Q, the Vifa likes a large box which, in this case, is no problem.

Anticipating a lower XO frequency, the woofer arrays were changed from twelve (progressive) Dayton RS180’s to four SEAS W26FX. The W26s are pushed to the corners of the 32”x33” lower baffles for maximum directivity. Volume displacement of multiple 180s & W26s is nearly the same -and I had another use for the 180s- so woofers from the Raptor project were re-purposed…re-purposed for the 5th time since I’ve owned them!

pnw is offline  
Old 08-27-2012, 02:05 PM - Thread Starter
pnw
Senior Member
 
pnw's Avatar
 
Join Date: Dec 2003
Posts: 439
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 7 Post(s)
Liked: 36
Construction:

I’ve finally realized the last change to the Octagon will come at some unknown point in the far distant future. Until I hit the wall with improvements, faster progress is made by building just one speaker at a time. Also, I find critical evaluation far more effective with a single channel…stereo and multichannel are so enjoyable they can cover sins that might otherwise be missed. Rationalization aside, I didn’t know how to model a line in a horn so I wasn’t certain the concept would perform as planned…and I didn’t want to waste a lot of time and money!

Basic construction is a 2x32x38” frame for mounting the WG to the 2” baffle wall. Within the frame, the elliptical section flare is 1/8” hardboard formed over a plywood buck with a station for each driver (think boatbuilding).








The hardboard flare is backed with 1.5” plywood strips glued on edge.



All mid and tweeter acoustic centers were aligned relative to the 4 meter sweet spot, providing flexibility in the final choice of mid-tweet XO slopes. Admittedly, perfect alignment is only correct at one point in space (my seat;) but a 3.93 meter radius inside a WG, on a sloped 38” baffle, most of which is above listening height, was a fun challenge. Individual driver mounting plates were taped together; packing tape stretching enough to make the subtle curve in the buck while glue cured between plates. Driver cutouts were drilled after the plates were fixed to the completed flare.


To provide a near continuous radiating line with the mids, the 30130 is physically shortened by removing the faceplate; clamping the driver in a sub-chamber directly to the WG. Removing the faceplate also smooths the diaphragm to WG interface.




The low-radius portion of the flare is contained within a large midrange enclosure, only the higher curvature/strength portion is exposed to subwoofer backwaves within the IB wall. The sealed mid enclosure is well braced 3/4” ply with wool felt suspended behind the drivers; fill to the sides of the enclosure is simple fiberglass home insulation.




WG end plates are ¾” plywood reinforced with curved strakes inside the flare.




To suppress vertical reflections within the WG, the dual purpose strakes support ½” thick wool felt pads spaced off the parallel surfaces.




Man this thing is heavy!




Temporarily installed for crossover work.




Crossovers are 2nd order at 380 and 3k. High-pass and low-pass for the Vifa mids are both passive. Passive high-pass for the Aurum Cantus tweeter is cascaded with the mid HP; providing another protective knee below crossover as well as reduced IM distortion. Low-pass for the W26’s is digital so those drivers can be delayed to match WG arrival time.

pnw is offline  
Old 08-27-2012, 02:05 PM - Thread Starter
pnw
Senior Member
 
pnw's Avatar
 
Join Date: Dec 2003
Posts: 439
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 7 Post(s)
Liked: 36
Results:

Here is a comparison of new (1st in plot pairs) vs old (2nd in pairs). Moving the IB wall outside for vertical polars is a little impractical so in a “duh” moment I decided to measure for the desired result…speaker/room interaction at listening height, at the listening dist of 4M.

The first pair of plots is 2ms…both new and old produce decent first arrival response. The next pair of plots are the same impulses gated at 20ms…midrange through low treble is the target for improvement. The final pair are 100ms windows, illustrating the difference over a wide range of arrival time.




Still a decent horizontal listening window. Rather than fixed seating, we use easily moved chairs in the Octagon so here is my “sweet spot” vs an overlay of 1 seat, 2 seats, 3 seats, and 4 seats wide (~50 degree window). A bit too much waistbanding, but good enough.




New center distortion sanity check (sweep, not 1/24 MLS):




Subjectively, how does the new center sound???

With smoother midrange response, pink noise sounds richer, more dense/full/robust/complete. Somewhat unexpectedly, I’m hearing subtle notes and detail previously buried in the notches. Imaging depth is also improved, probably by revealing the smaller spatial clues present in some recordings. So…the “psycho” project provides actual practical benefits…yeah!!!

Is the new center perfect? No, but it shows vertical bounce can be addressed without 2’ thick absorbers on horizontal room surfaces! I’d still like to track down the little dip at 14k…possibly too-dense wool felt in the tweeter sub-chamber, or maybe a horn reflection, dunno yet. Also, dial back a bit of waistbanding if I can do it without damaging other parameters.

WMTMW anyone?

pnw is offline  
Old 08-27-2012, 06:44 PM
AVS Special Member
 
maxmercy's Avatar
 
Join Date: Nov 2008
Posts: 2,375
Mentioned: 2 Post(s)
Tagged: 1 Thread(s)
Quoted: 43 Post(s)
Liked: 147
Great stuff.

Are the L/R channels going to get the same treatment?

JSS
maxmercy is offline  
Old 08-27-2012, 07:30 PM
AVS Special Member
 
baniels's Avatar
 
Join Date: Oct 2006
Location: 52556
Posts: 1,028
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 39 Post(s)
Liked: 49
I'm going to read this a second time. Thanks for sharing the details.
baniels is offline  
Old 08-27-2012, 09:54 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 16,852
Mentioned: 26 Post(s)
Tagged: 0 Thread(s)
Quoted: 565 Post(s)
Liked: 1170
nice experiment, construction, results, and write-up.

all the ripple knocked out was by minimizing floor and ceiling bounce?

what do you think about those w26's relative to what you had before?

Listen. It's All Good.
LTD02 is offline  
Old 08-28-2012, 08:33 AM - Thread Starter
pnw
Senior Member
 
pnw's Avatar
 
Join Date: Dec 2003
Posts: 439
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 7 Post(s)
Liked: 36
Quote:
Originally Posted by maxmercy View Post

Great stuff.
Are the L/R channels going to get the same treatment?
JSS

Now that I know it works as planned, yes. Several other things I need to do first so it will be a few months.



Quote:
Originally Posted by baniels View Post

I'm going to read this a second time. Thanks for sharing the details.

Thanks...if this gets just a couple other people thinking I'll have accomplished my goal.smile.gif



Quote:
Originally Posted by LTD02 View Post

nice experiment, construction, results, and write-up.
all the ripple knocked out was by minimizing floor and ceiling bounce?
what do you think about those w26's relative to what you had before?

Yes, in a normal room the new center would reduce the ripple caused by both floor and ceiling bounce. Floor bounce was my problem and that's the ripple the new center knocked down in these measurements. (Ceiling bounce was taken out in the original room design....acoustically it is an absorbent "cloud"...fiberglass stuffed between 16" deep trusses and "sealed" with 3 layers of stretched cloth.)

In this application, 4 W26s are virtually the same as 12 RS180s. No significant difference in response, distortion, or sound. Since I needed 12 RS180s for a friends HT, I just used the W26s on hand rather than buy 12 more 180s. Expensive drivers aren't really so expensive if they are good enough to use over and over.

pnw is offline  
Old 08-28-2012, 10:51 AM
AVS Special Member
 
Smigro's Avatar
 
Join Date: Feb 2009
Location: Montgomery County, PA
Posts: 1,891
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 42 Post(s)
Liked: 110
I just started my first sub woofer build and that has me thinking. All I have to say about this is WOW. Not sure I even understood what I just read, but hopefully at some point I will. biggrin.gif
Smigro is offline  
Old 08-28-2012, 06:23 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 16,852
Mentioned: 26 Post(s)
Tagged: 0 Thread(s)
Quoted: 565 Post(s)
Liked: 1170
"Floor bounce was my problem and that's the ripple the new center knocked down in these measurements."

can you explain how that works? i thought that a cancellation occurs when one wave arrives 1/2 a wavelength out of phase with another.

just approximating from your setup, i place the speaker at 4 feet high and the listening distance at 14 feet. the on axis distance is 14 feet and the path from the speaker to the floor half way in front of you then bouncing back up to you is ~16.1 feet by simple trigonometry. that means your first floor bounce cancellation minimum will occur at a frequency where the half wavelength is 2.1 feet, which is 266hz.

going up from there you don't run into another cancellation until 532hz. continuing up the frequency spectrum, the next cancellation occurs at 1064hz and so on doubling every time.

however in your measured data base case, there are many more cancellations than that.

Listen. It's All Good.
LTD02 is offline  
Old 08-28-2012, 06:45 PM
AVS Special Member
 
maxmercy's Avatar
 
Join Date: Nov 2008
Posts: 2,375
Mentioned: 2 Post(s)
Tagged: 1 Thread(s)
Quoted: 43 Post(s)
Liked: 147
My guess is the octagon's old LCR driver arrays provided for multiple floor bounce distance opportunities, hence the waviness of the response...

JSS
maxmercy is offline  
Old 08-28-2012, 09:46 PM
AVS Special Member
 
mayhem13's Avatar
 
Join Date: Jan 2008
Location: nj
Posts: 3,959
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 13
Excellent exercise Paul.......and yes......you've got me thinking!......Hmmmnnmm;)
mayhem13 is offline  
Old 08-29-2012, 10:55 AM - Thread Starter
pnw
Senior Member
 
pnw's Avatar
 
Join Date: Dec 2003
Posts: 439
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 7 Post(s)
Liked: 36
Hi LT

Multiple things going on here.

XO of the old speaker is 780Hz. With low mounted woofers there is virtually no floor bounce below XO (though sidewalls are present plus, at very low frequency, reflections penetrate the ceiling treatment).

Distance measurements are to the driver AC (I measured 3.93M precisely to align ACs in the new WG) and my hypothesis is that bounce calculations should be from the WG mouth (baffle) or 3.42M. Mic height was about 38” or .965M.

I don't know how the floor "views" a large WG mouth, but plugging the center height of 1.245M into a floor bounce calculator http://tripp.com.au/avfcbm.htm gives us a cyclic rate of about 530Hz for the WG. First dip (1/2 wave) at 795Hz, first peak at 1060, dip at 1325, peak at 1590 and so on...the interval does not double between peaks or between dips. It may be helpful to think in terms of cyclic rate or events.

Clearly there are other lower level reflections -with differing phase- adding complexity to the picture but floor bounce is the primary reflection.

pnw is offline  
Old 08-30-2012, 02:39 AM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 16,852
Mentioned: 26 Post(s)
Tagged: 0 Thread(s)
Quoted: 565 Post(s)
Liked: 1170
ah, so it is additive. 0.5 wavelength for the first cancellation (based on the differential distance), 1.5 for the second, 2.5 for the third, and so on...that makes sense. to a first order approximation that does appear to match up with your measurements. i was just curious what was going on. thanks.

Listen. It's All Good.
LTD02 is offline  
Old 09-14-2012, 02:08 PM
AVS Special Member
 
coctostan's Avatar
 
Join Date: Oct 2005
Location: Indy
Posts: 1,960
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 129
Paul-

I had a random thought pop into my head...what if we combined some of Danley Synergy principles with your design? I'm thinking instead of having the mids at the throat, place two vertical array behind the horn walls with bandpass ports for entry into the horn. The tweeter portion would probably still be an AMT with foam shading. The mids would be shaded but shouldn't require any delay ala CBT due to the narrow targeted ~15-25deg vertical window. The advantages I see are that you would be able to fit more mids and have the advantage of running them bandpass instead of direct.

The main issue I'm thinking about is how the array of bandpass mids should probably be curved in their placement on the horn with the center mids closer the mouth and the outer mids closer to the throat. In fact, it looks like Danley is doing this on his Jericho JH2. Now that I'm looking at the JH2, this might be like a JH2 mini but with shading to create a very narrow vertical dispersion.

I just mocked something up and in a 40" x 36" x 24" deep horn, I could fit a Aurum Cantus 30130 and 12 Pyle PDRM5's (probably 16-20 of the Celestion 4" drivers if those are available). It should give a horizontal 75deg pattern down to about 300-400hz and a vertical 20deg down to maybe 500-600hz. Ultimately not all that different than what you did, but with absurdly low midrange distortion.

Anything I'm missing?

coctostan is offline  
Old 09-14-2012, 03:58 PM - Thread Starter
pnw
Senior Member
 
pnw's Avatar
 
Join Date: Dec 2003
Posts: 439
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 7 Post(s)
Liked: 36
I don't think you're missing much wink.gif

My original concept was to run the mids bandpass, but modeling showed physical construction was going to be extremely tricky...the front chamber had to be wicked small to get the necessary HF response from the mids...extended response necessary to use "soft" slopes to the AMT for a smoother directivity handoff. (Vertically, the AMT is actually too broad at the bottom of it's range.) I obviously chickened out on the incredibly small front chambers frown.gif Early on, being in awe of Danley, I considered running pairs of drivers on the horn walls but Vd was already very high, plus it added another risk of failure at HF.

I didn't understand your comment about a convex (rather than concave) mid layout...can you point me to a drawing of how Tom arranged the JH2 mids...any DSP manipulation going on?

To show you how much our thoughts parallel, I bought the PDRM5 for eval. One thing to ponder is that the Pyle and Celestion mids don't look pretty (IIRC) above about 1.5k. With smoother response I remember thinking the 5" Celestion might be a better candidate than the 4"...but that has been a while ago. I also bought a stack of various 2-3" closed back mid-tweets but never did anything with them.

The big challenge for the bandpass approach is, I think, to get tweeter directivity down low enough to work well with BP mids for a 1.5-2k max XO. Maybe a Danley style Paraline tweeter, or $tacked AMTs, or maybe a shaded stack of good small format hard domes.

pnw is offline  
Old 09-15-2012, 04:36 PM
AVS Special Member
 
coctostan's Avatar
 
Join Date: Oct 2005
Location: Indy
Posts: 1,960
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 129
I think I'm missing enough to where my idea won't work. What are you guessing your vertical beamwidth is in this speaker? I'd guess it is still pretty wide at the bottom of your AMT's range up until maybe 6khz where it would be under 75deg.

I wasn't actually aware of paralines until you mentioned them, but I've read the entire Internet about them over the past day. It is very intriguing. I think we might be able to use one to give us a good vertical beamwidth to a low frequency. It seems you can set them up to effectively shade similar to how a CBT works, but on a single small line.

Looking at the data for the VTC EL210 which uses two ~ 5" tall paralines to cover ~1200hz and up, they have a beamwidth of ~60 deg tapering to 30deg by 4khz and down to ~10deg around 7khz. I'm not sure what kind of vertical pattern control we can get from a wider bandwidth paraline with both compression driver and mid. Maybe a few paralines could be arrayed and possibly shaded/delayed.

Ever since looking into the CBT, I've wanted to figure out a way to take the positives of the CBT (primarily minimizing floor bounce effects) with the controlled horizontal directivity of a 2 or 3 way horn. The synergy horns are an obvious solution, but don't really handle floor bounce all that much better.

I'll probably chat with "Patrick Bateman" over at DIYAudio and pick his mind on the validity of what I'm thinking. He seems to have thought about paralines more than anyone short of Danley himself.
coctostan is offline  
Old 09-16-2012, 10:15 AM
Member
 
jeno's Avatar
 
Join Date: Apr 2009
Posts: 33
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 6 Post(s)
Liked: 17
Danley JH2



I think the 24 5" midranges are coupled to the horn through a paraline together with the 12 1" compression drivers.
The convex arranged holes I think are for the six 18" woofers.

Very interesting thread by the way. Inspiering and food for thought.

Pnw:

Have you measured the AMTs vertical dispersion up high?
And do you have any thoughts on if the diffrencies you experince from your previous horn also could have something to do with using an AMT and direct radiators over a compression driver?

Coctostan:

I have also been thinking the same direction as you. A combination of a synergy horn to maintain driver integration and horisontal dispersion, and CBT to take care of the floor bounce and vartical dispersion. I don't know how to do that yet, and I don't have room for it either for the time being, but some day...cool.gif

-Jan
jeno is offline  
Old 09-16-2012, 04:26 PM - Thread Starter
pnw
Senior Member
 
pnw's Avatar
 
Join Date: Dec 2003
Posts: 439
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 7 Post(s)
Liked: 36
Max,
The original height target was 20 degrees, but I'm not sure where it ended up. Mid beam height was tightened over the original line sims by ear (listening to pink noise while adjusting cap shading) plus I don't know exactly how the line interacts with the horn. You're right, the bottom of the AMT was tricky, but mostly solved with "soft knee" 2nd order slopes (high overlap with the central mids...same approach as the mid-bass XO).

John (Patrick) probably is a good source for info on the Paraline...see the "square pegs" thread if you haven't already tuned in. Also, I haven't looked for Tom's patent/application but it should be extremely informative as he tends to fully explain his thoughts.


Jan,
Vertical dispersion of the tall 30130 is very narrow in the top octave, particularly above 15k. Unfortunately, I can't hear 15k, so no comment on the audible consequence. The original plan was to use open-cell foam to shape the very top end but I found, at least at listening distance, foam degraded VHF output.

No substantial thoughts on attributing any difference in sound to the differences in drivers.

pnw is offline  
Old 09-16-2012, 09:37 PM
AVS Special Member
 
coctostan's Avatar
 
Join Date: Oct 2005
Location: Indy
Posts: 1,960
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 129
Jan-

Good catch on the JH2 using a paraline for HF and mids with horn wall holes being the woofers.

Paul-

I've read the entire Square Peg thread on DIYA. The thing I can't seem to determine is how to size the paraline. For instance, what is the downside to making a very tall paraline? I assume it will exhibit a narrower vertical dispersion, but it doesn't seem to behave like a typical line source (ie the AMT's where it narrows as you would expect from a source that tall). It appears you can also alter the width dimension relative to the height dimension and get a different vertical beamwidth. John (P. Bateman) posted some info on that along with a spreadsheet to calculate beamwidth. Basically, by narrowing the "eye" you are applying some delay to the upper and lower portions of the mouth relative to the middle. If we could also figure out how to also shade the signal at the top and bottom relative to the middle we could create a CBT effect with a single paraline.

Ultimately, minimizing floor bounce is the goal. Looking at post 4 of this thread the results are astounding. People spend huge amounts of money on uber drivers and ultimately, room interaction simply tears that apart. Mitigating room interaction like floor bounce should be a first priority.

I also just re-read Keele's ground plane paper (http://www.xlrtechs.com/dbkeele.com/PDF2/Keele%20(2005-10%20AES%20Preprint)%20-%20CBT%20Paper%205.pdf) and it seems a ground plane line array is the best way to overcome floor bounce. I'm less interested in being CD in the vertical like the CBT, so I think a modified "Controlled BT" instead of constant would be fine. In other words, it doesn't have to have perfect legendre shading or ideal delay/curvature. I think that could be done with 3 paralines to keep budget within some reason.

Plan for modified CBT using 3 paralines:
All paralines with HF and mid covering 300hz and up. Mid/hf crossover is passive. Each paraline is 20" tall for a total speaker height of 5 feet. A single MiniDSP run 4 way with 4 amp channels divides the 3 paralines and a woofer section. A conical horn would be mated to the front of a size and shape needed to achieve the desired beamwidth.

Bottom paraline would be a normal symmetric (width is half of height) paraline and run at -0db and no delay.
Middle paraline would be symmetric on the bottom half and slightly taller than wide on the top half (think egg shape) to delay the top relative to the rest. This would be shaded slightly.
Top paraline would be the same as the middle paraline to delay the top relative the bottom. This would be shaded a bit more.

The idea would be to approximate a CBT curve and shading. IMO it is not necessary that it meet the exact CBT formulas. Instead the paralines would be adjusted to get the best directivity out of this compromise.

It might be possible to do this without a DSP. It would require building "throats" or extensions to each paraline exit in order to get the delay. Shading would be done via passive components. This would require more depth but there are ways it could be done without. Ultimately, I'd probably rather just use a few DSPs and amp channels.

Of course, it could be done with more paralines, but I'm not sure it would be worth the extra cost (more CD's, mids, DSP and amp channels). I think a ground plane near-CBT with controlled horizontal directivity is possible. Who wants to build it for me? smile.gif
coctostan is offline  
Old 09-17-2012, 07:01 AM
AVS Special Member
 
maxmercy's Avatar
 
Join Date: Nov 2008
Posts: 2,375
Mentioned: 2 Post(s)
Tagged: 1 Thread(s)
Quoted: 43 Post(s)
Liked: 147
Coctostan,

Each paraline is shaded like an individual CBT, greater output in middle, less at the ends (given the conventional 'eye' format), has to do with amount of paraline 'covered by' certain angles from source. Not sure you can shade a paraline so that it would fit into a larger CBT-like shading system with greatest shading on top. least on bottom. Will post more about it later, no time now.

JSS
maxmercy is offline  
Old 09-17-2012, 06:53 PM
AVS Special Member
 
flyng_fool's Avatar
 
Join Date: Aug 2009
Location: Frisco, TX
Posts: 3,860
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 182 Post(s)
Liked: 156
Paul, each time I read your build log of the Octagon I am re-flabbergasted. That has got to be one of the most incredible builds of all time. It is truly the most work I have ever seen anyone put into a a room. The only thing you didn't build is the stuff in the racks!

Television: Mitsubishi WD65737 DLP
Processor: Emotiva UMC-200
Amps: Carver AV 806x/Behringer EP4000
Mains: DCM TimeFrame 600 Center: AT 453C
Surrounds: AT 251.1 Sub: Danley DTS-10
Blu Ray: Panasonic DMP-BD655
flyng_fool is offline  
Old 09-18-2012, 06:30 AM
AVS Special Member
 
coctostan's Avatar
 
Join Date: Oct 2005
Location: Indy
Posts: 1,960
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 129
Quote:
Originally Posted by maxmercy View Post

Coctostan,
Each paraline is shaded like an individual CBT, greater output in middle, less at the ends (given the conventional 'eye' format), has to do with amount of paraline 'covered by' certain angles from source. Not sure you can shade a paraline so that it would fit into a larger CBT-like shading system with greatest shading on top. least on bottom. Will post more about it later, no time now.
JSS

I've been trying to figure out if a paraline shades the extremities relative to the center. I don't see how it would shade although you can certainly shape the wavefront of a paraline by making the width less than half the height. That shortens the path length as you get closer to the center.

Assuming the paraline produces equal dB across its wavefront, I've been able to come up with a delay pattern that approximates a CBT to maybe 90-95% of its curvature/delay. The design uses a combination of 3 curved wavefront paralines attached to "tilted throat extensions" and some electrical signal delay. The throat extensions provide linear delay at an angle from the bottom to the top of the paraline. I'm having trouble explaining it the concept clearly. I should sketch something up. It basically tilts the paraline back a specific slope prior to the exit slot. All three exit slots remain vertical and on plane. I'll attach my spreadsheet when I clean it up.

Of course, if the output is shaded across a single paraline, that will throw my plan off.
coctostan is offline  
Old 09-18-2012, 07:13 AM
Senior Member
 
JackNC's Avatar
 
Join Date: Aug 2011
Posts: 232
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 27 Post(s)
Liked: 18
did you find the posted explanation of the shaded amplitude lense? that is an extension to the paraline that does the shading...

There is an awesome explanation at this link
http://www.authorstream.com/Presentation/BroMike-893215-s-a-l-t-treaty/
JackNC is offline  
Old 09-18-2012, 08:05 AM - Thread Starter
pnw
Senior Member
 
pnw's Avatar
 
Join Date: Dec 2003
Posts: 439
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 7 Post(s)
Liked: 36
The Paraline holds promise for an application like this but I don't understand the height (length) limitation. Turbulence in the narrow duct might become an issue over longer distances but what other limits exist? Any insight? (Single driver to keep it simple.)

pnw is offline  
Old 09-18-2012, 09:14 AM
AVS Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 5,101
Mentioned: 11 Post(s)
Tagged: 1 Thread(s)
Quoted: 572 Post(s)
Liked: 659
Paul,

Thanks for this thread. Very interesting.

My theater "area" just changed on me in the last few days. I now have about a 10' listening distance with essentially no sidewalls to worry about. My only reflections of concern are the laminate flooring, the ceiling tile (8'), and immediately behind my head. The speakers must be in-wall, therefore I have an enclosure depth of about 4" to work with.

So my plans of doing the usual bandwagon SEOS LCR are out the window. I'm trying to figure out what to do and have been scratching my head a lot. Would you, or anyone reading this, have any suggestions on what to do. Thankfully being an infinitly wide baffle will net me some smooth and sensitive frequency response. But I'd like to control vertical directivity. Oh, and I'm on a budget.

Was thinking MMTMM with Aura NS6 and a waveguide loaded SB29 in the middle (can cross around 1500hz fairly safely). I could possibly do an Eminence alpha8a with SEOS12/DNA-350 in wall. Height isn't a problem. Just cost and depth. Thoughts? Thanks.

Ryan
tuxedocivic is offline  
Old 09-18-2012, 09:27 AM
AVS Special Member
 
coctostan's Avatar
 
Join Date: Oct 2005
Location: Indy
Posts: 1,960
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 129
Quote:
Originally Posted by JackNC View Post

did you find the posted explanation of the shaded amplitude lense? that is an extension to the paraline that does the shading...
There is an awesome explanation at this link
http://www.authorstream.com/Presentation/BroMike-893215-s-a-l-t-treaty/

The GH looks interesting but I'm not sure how he is achieving the shading with this lens. If the paraline outputs an equal pressure waveshape I think my plan can work. BUT If a paraline inherently shades the output level away from center, then I don't think my plan holds water.

Do you have an idea of how we could use the GH shading technique?
Quote:
Originally Posted by pnw View Post

The Paraline holds promise for an application like this but I don't understand the height (length) limitation. Turbulence in the narrow duct might become an issue over longer distances but what other limits exist? Any insight? (Single driver to keep it simple.)

I'm not sure there is much limitation on the height aside from comb filtering similar to what any line source will have. Is that what you mean? Danely mentions using some rather tall paralines in some hifi stuff he has in his own collection. I can't recall the size, but I though it was a feet long.

I don't think turbulence is a factor. It doesn't seem to be an issue for Danley's JH2 which has 12 CDs and 24 mids pushing huge dB through a single paraline. Of course, I haven't done any actual testing so it is hard to say for sure. I put enough trust in Danley to at least pursue it. as a possible solution.

It basically transforms a point source, spherical wave shape to another shape, typically a line source (at least that is what I'm looking to do...you could do other shapes). If the paraline "folding" is keeps all pathlengths equal, the wavefront is flat. You can alter these lengths though and alter the shape of the wavefront through delay.

In your case, you could replace the AMT with a paraline fed by a compression driver and shape the wavefront to mitigate some of the HF comb filtering inherent to a line source like the AMT. I'm not sure it would be worth it in your case. You could also place your mids on a paraline. Ultimately I don't think it would do much for your specific setup.

Danley does it partly so that he can place multiple HF CDs on a single speaker to achieve absurd output levels from a relatively small source.

I'm mainly interested in it because it *might* allow me to create a CBT out of 3 CDs and pro mids and attach a conical horn section to the front to provide horizontal directivity control.
coctostan is offline  
Old 09-18-2012, 10:01 AM
AVS Special Member
 
coctostan's Avatar
 
Join Date: Oct 2005
Location: Indy
Posts: 1,960
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 129
Quote:
Originally Posted by tuxedocivic View Post

Paul,
Thanks for this thread. Very interesting.
My theater "area" just changed on me in the last few days. I now have about a 10' listening distance with essentially no sidewalls to worry about. My only reflections of concern are the laminate flooring, the ceiling tile (8'), and immediately behind my head. The speakers must be in-wall, therefore I have an enclosure depth of about 4" to work with.
So my plans of doing the usual bandwagon SEOS LCR are out the window. I'm trying to figure out what to do and have been scratching my head a lot. Would you, or anyone reading this, have any suggestions on what to do. Thankfully being an infinitly wide baffle will net me some smooth and sensitive frequency response. But I'd like to control vertical directivity. Oh, and I'm on a budget.
Was thinking MMTMM with Aura NS6 and a waveguide loaded SB29 in the middle (can cross around 1500hz fairly safely). I could possibly do an Eminence alpha8a with SEOS12/DNA-350 in wall. Height isn't a problem. Just cost and depth. Thoughts? Thanks.
Ryan

With only 4" of depth, I'd suggest doing something pretty similar to PNW's but simply without the horn. I'd probably us 6-8 NS6's per speaker. That will allow you lower vertical directivity control. Of course you will need to shade these. For the tweeter portion, a paraline would actually work well if you don't want to fork over for a tallish ribbon/planar/AMT. Simply size the paraline to give you the directivity you want. Yorkville uses a paraline without a horn on the front. You will get rather wide beamwidth which you will likely get no matter what with 4" of depth. The good thing is that you should be able to get a narrow vertical beamwidth like PNW's speaker above and significantly mitigate floor and ceiling bounce.
coctostan is offline  
Old 09-18-2012, 11:33 AM - Thread Starter
pnw
Senior Member
 
pnw's Avatar
 
Join Date: Dec 2003
Posts: 439
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 7 Post(s)
Liked: 36
Thanks Flyer! It is a journey!


Ryan.
Some of my favorite speakers have been of the MMTMM variety. As Max suggests, extend the line as budget allows...since you don't need BSC, progressive cap shading will work well.

I quickly looked at the NS6 and it looks like they've traded XMAX for sensitivity...the long underhung coil gets out of the gap quickly so consider SPL requirements vs the number of drivers. By all accounts I've read, the SB29 is a good driver but finding a WG with good LF vertical control and short height for CTC mid spacing will be a challenge. I haven't followed developments, but some of the guys at PE have been working with domes in asymmetric WGs so you might see what they've been able to work up. I also favor asymmetric tweeters for MTMs. Paraline should work, but mostly uncharted territory for DIY.


Max,
"Feet long" will work...or did you mean "a foot long"? I'm not considering the Paraline as a replacement for the AMT, but the two L&R channels are up for rework in the coming months. Since I enjoy "stretching my envelope", the Paraline is on the list of things to try smile.gif

Floor and lateral is very well documented but, out of curiosity, have you seen anything regarding how the CBT handles front wall and lower frequency ceiling reflections?

pnw is offline  
Old 09-18-2012, 01:17 PM
AVS Special Member
 
coctostan's Avatar
 
Join Date: Oct 2005
Location: Indy
Posts: 1,960
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 129
Paul-

I believe it was a few feet long, but I'd have to find the quote. It was on diya. One possible solution would be three small paralines to fill your ~36" tall space. They would need to be mid/hi paralines. Something like this is what I'm thinking for the Paraline-array:



The middle paraline would be symmetrical and sized for a flat wavefront The top and bottom paralines would be more egg shaped to delay the upper-most and lower-most portions. You would basically have a miniature curved array and you could shade the top and bottom portions to taste. The advantage is that you would have more uniform vertical coverage because of the delay control, likely more headroom (3 compression drivers!!! not that you would go overboard on headroom) from the drivers and IMO, and a better wavefront shape to transition to the horizontal.

I don't know how the CBT handles front wall reflections. In my case, the Paraline-array would be baffle mounted similar to your setup (Octagon was my inspiration). Ceiling reflections are basically a function of the narrow, but constant vertical beamwidth. His data shows a 1.25m 45deg arc with a constant 30deg vertical beamwidth from 250hz and up. He does get major comb filtering above 12khz just like any line array (although I'm not sure the paraline will suffer from the same). The ceiling reflections should be minimal with a beamwidth that narrow.

The design I'm looking to do would mimic the CBT36 sold as a kit by PE. It should have a vertical beamwidth around 20-25deg from about 250hz and up. I can build a conical horn that is ~20" deep for the center and about 30" deep for my L/R. That should hold horizontal pattern pretty low too. smile.gif The nice thing is that the conical horn section should be easy to build and even try different profiles. I'll probably start conical, but try some tractrix and elliptical.



This is a screenshot of the CBT curve spreadsheet I slapped together last night. There are some loose ends to tie up and the parameters were just eyeballed so things could be off. Delay is represented in centimeters. Delay would be a combination of DSP electrical, "throat tilt" (which is a linear slope...the paraline will appear tilted back while the paraline exit is flat with the front) and paraline based (the "curve" of each paraline created by altering path lengths in the fold). I simplified the CBT to split into 3 sections instead of 5 like Keele's Legendre shading requires so directivity won't be as constant, but I'm not concerned about that. In a home, IMO vertical controlled is more important than constant. At my 10-16ft listening distances, a 20deg window is more than wide enough. Not much energy will hit my 8.5ft ceilings and the ground plane will take care of floor bounce. My L/R will be at ~40-45deg angles so there will be minimal side wall energy. I'm hoping for exceptionally consistent, and room independent measurements regardless of seat....then again I'm probably missing some ridiculously obvious issue.
coctostan is offline  
 
Thread Tools


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off