Multiple Small Subs - Geddes Approach

Hi Duke

From studies of automotive noise we know that at LFs the tactile sensation is required for listeners to sense a realistic LF effect. This is important < 100Hz and critical < 50 Hz. For example, in a car that had 20 dB LF noise reduction the audible effect was almost nothing owing to the fact that only airborne noise was attenuated.

As to the timing of the LF tactile sensation, I know of no studies in this regard. That said, I cannot see how delay at these LFs could be sensed unless it is many many ms. long. We have a very poor sense of timing of LF signals. Timing is a HF phenomena.

Your example is interesting - for explosions - but for music it might be less pertinent. I once watched a demo butt shaker movie at a cinema and it was awful. Could have been poorly setup, I don't know. I do know that if there are any LF resonances in the "chair" that you sit in while listening that these resonances will play havoc with the LF perception. It's a highly complex situation that has little been studied.
 
You know that I wrote the original paper on BP enclosures defining all order from 4th through 8th and I used them extensively for many years. But that was pre DSP and with LF EQ it just doesn't matter what the design is as it all gets EQ'd and the subs with the most flexibility work out best.

What do you think of a 6/8th order cabinet which has ‘perfectly’ aligned the harmonics 1/4 and 3/4 of a (4/4 total so to speak , frequency ) and with driver entry from two closed ends creating the 0.349 (TL version of a fulcrum point or 1/3 the qw pipe length) but with a folding such that it’s 90 cm in three segments off the front output and one off the motor side . combining the driver positions (30 inn90 and then two 90cm segments between .a Joint exit of the two (1/4 and 3/4 long (360 cm total and exit at 270) pipes before one more section of that 30 cm offset in the solo 90 cm motor side pipe.

if that’s all as it hopefully sounds. In the sim it’s quite intersingky ‘in synch’ ?? I’ve got -990 dB in the outskirts of 860,8600hz and anything between peaks that ends defining the ‘perfect circle’ I guess?

I’ve got a pattern in this (it’s simmed in horn response and not a very complicated one either) and it follows our own solar system and even more? As million Miles (placer or Simijsr) are hz. ? what the heck? So I looked deeper and deeper and it doesn’t seem to wanna disappoint. I found the trails of what would be just about anything in the absolute version of ? Physics as it would be naturally? It even pints your flawed planets and almost shines what it why if you combine the sim with a search thruogh a nasa or planetary model. This then gets you searching trigonometry to find angles as numbers or either in math or metric system / standard system. it might already sound silly , but this all stsrted from trying to clean up a TL with a TL… I failed. :D:D

But I don’t know if it was so bad? hopefully this makes sense as typed? ? I don’t wanna pig up you’re post here but I wanted to ask you what the heck this is? (I can give you all the info and the complete sim if interested ).

thanks and thanks for everything already learned from you . two subs or four saved my day!
 
Honestly the whole area of 29.972458,30,31.428,33.33,34.3(as cm). It’s this oasis of opportunity that shifts around a variety of very good, to wow! I’m gonna jump in to Dr G’s BP4-8 paper and see if I can get a clue or a better way to describe what I’m trying too . it’s the horn response that’s allowing the View of things in ways not likely ever before? Little details expose opportunities? It ones that wouldn’t exist or be considered? I couldn’t be the judge if that, but I can throw you a sim to play with. You’re better at descriptions too :)
 
What do you think of a 6/8th order cabinet which has ‘perfectly’ aligned the harmonics 1/4 and 3/4 of a (4/4 total so to speak , frequency )

There are two patents on similar designs, one by Bose but 3/4, 1/4 and another much earlier one with 1/3, 2/3. Look them up for a more in depth description. The Bose "wave cannon" was of this design. Not a market success.
 
Yeah, it seems as soon as you join the exits of a would be ‘bose wave cannon’ and offset the driver entries you arrive at an entirely different scenario and acoustic center inside. 0-30cm, then 90 cm from each closed end driver and then a fold, with 0-90 twice more between those. The exit is at 270cm/90cm of the 360 total. the driver entry’s serperated by 150 cm (half of 300cm) and the entire simulation and ultimately resulting 0/90/90/90(270total) and 0/90cm total design which is (if placed on a unit circle and radians) working the (substitute ‘30’ for 33.33cm or 31.43cm or .029.972458cm and add one last interval as a combined ‘series’ exit.

The SPL response and pressure phase plots are ridiculously ‘perfect’ if not a continuous repetition of circles (phase rotating) that only pauses for a hemisphere (180) on a location and subsequently others (hard to see at these freqs without working through it manually ) of (86400 seconds in a day(hz as 860,and 8600…) which is also the mean orbital radius of Saturn (log scale kms )minus the suns own radius. And the diameter of the sun 864000 kms which is its self the metric equivalent of 22/7. Which is 6.96xxx Miles in radius and 13.92 miles in diameter (big bang number as years) and includes the 11 year /22 year /33 year 44 year season or magnetic equinox cycle of the own equatorial twist. (Polar too in a weird dual motion?) .

Work this sounds kinda silly. Further investigation reveals how rmserious(real) this is . And even leap year as 365.25 days and the 1440/1461 cycle (1433 in a speaker duct as described) is part of the biggest earth cycle 25920 which is the axis tilt swap (precession of the equinox cycle Season’s flip hemispheres and suddenly we are looking at global warning in further complexity of a cycle outside of the earth bound spins but trapped in the solar spins and individuals as a whole. Jupiter and Saturn rule this area in ways that complicate everything, but the big bug picture is oddly represented right here in this little speaker ducting model and odd as it sounds , it only needs a more qualified person or credentialed and well known respected person in the industry to take a peak.

I could easily be mistaken. But my ‘I’ and ‘t’ are dotted and crossed. If I wrong here then it’s an honest mistake and eating peoples time was accidental. either way I can’t get any further into this. I can’t find an error? Only the hz in which everything is defined by a location or a spore without (velocity )vector? As a lie or fake ‘constant’ such as 29972458m/sec in 1496000000m. 7500000hz is the speed of light ? Or 300 (C) in 150(AU) is 510 seconds(time of flight in parts of 60) at 75 (Hz).

Acoustical waves and gravity waves are in the audible spectrum. Sept 15, 2016 this was recorded ? this was amazing. it was an impedance chirp off a DATS or anything…the sound of two black holes colliding was heard? Recorded and tweaked to be simple .

I don’t have those ends tied up in this typing.. but there’s way more
 
There are two patents on similar designs, one by Bose but 3/4, 1/4 and another much earlier one with 1/3, 2/3. Look them up for a more in depth description. The Bose "wave cannon" was of this design. Not a market success.

Yes sir. The Bose was a deterrent for a long time for that reason. but it also pointed out flaws that would later represent necessary changes in length and phase (pressure ) it seems? Reflection as sharp Imediate folding at hexagon like math(6 in 360 or 240, intervals? 90 / with offset driver entry points on each end at 30 and 330 from 0 and 360? working the c = 2pi r with quarter wave as ‘r’ or similar which plays out in the circle model? Math of a sphere in a rectangle so to speak ?


Parts of .2619, 349, .5238, .785, 1.0472, and 2.09, 3.14, 4.188(volume of a sphere as ‘r’ =1). But that equation is 5.236 if ‘r’ is 90(11/7)

4/3 pi r^3 = V of a sphere. But that 4.188 (r=1) and (0.5238)/0.2619 or 5.238/2.618 (r= 360,270,(90)or (30) or (150 or 15)in pi numbers (22/7) is 5.5/7, 11/7,22/7,33/7,44/7) And the rectangle as a cube is .5236 hypotenuse. And there’s 4 affective rectangles here or hexagon triangle shaped parts as cone arcs ? 90,180,270,360 in length . From 0. On either end . 2.0944 is the center of this . And 4.188 is the ends? v= 1.333(3.143) (1cm ) ^3 is 4.188cm3. And that’s part of the sim too. CSA in cm2 is critical . and TS parameters will dictate the choices used? Seems there’s a few 12 abd 18” drivers that fall into this ambiguous area and work well? 18ds115-4 by B&C . But also a cheap Chinese car audio driver . An odd unicorn I guess? It’s not perfect , but it models well prior to any ‘real world’ changes in inductance, etc (reality vs sim ). Still not too shabby for a lab rat or experimenting with numbers or qw anomalies? whatever weird science is officially? :D


Something like that is working here? I’m probably screwing up the explanation as it’s not my strongest quality:) tying or math as words? only angles or hypotenuses, arcs on cone vs edge on hexagon . And offset 30 degree hexagon is then pressure nodes for a system if 240 in 360? I wish I was a Platonic solid or a greek mathematician right about now?? Points on the exterior or nodes as a wooden cube on a lathe turned down to a circle… but that is when I noticed a bowl as a waveguide shape that needed to expand . and nodes went with that . But pipe was easier in the brain. And in reality it served a purpose to?
 
Last edited:
Regardless of any particular similarities in the sphere/trigo/ rectangle/ radians /circumference maths in the enclosure another very interesting thing I noted was the ability to add a variety of pipe sections or conical flare pieces to the vent opening via cap screws and threaded inserts . Since it’s so ‘in phase’ so to speak it’s tolerant (simulation) to a great variety of those to use depending on wants /needs? I’m kind of assuming that hints at room behavior being consistent , not a gigantic burst of harmonic junk typical of a tapped horn vent or randomly folded TL/taped/ compounded, etc?

Car audio was a great expirmental thing. Exploiting anything quickly. these are aweful when not thought out and folded at nodes/anti. some really weird issues in previous versions that had to be fixed (sorta) with mass loading restrictions or materials at exit damping output into finished returns .. so this ‘new’ version really stands out to me and the new ‘PH’ sim in horn response allows for some specifics that would never be visible otherwise.
 
The Bose "wave cannon" was of this design. Not a market success.

Not sure why that's the case, because Bose is usually successful with their products. I suspect that it had a lot to do with its form factor.

The Paraflex designs follow a similar concept, but they fold the whole thing up into a box that's likely quite easier to store, stack and transport. I'm particularly intrigued by the "Type A" ones, as they promise a wider bandwidth and better suppression of out-of-band "noise". At least on paper. Hopefully later this year I get to try out an actual build to see how they really perform.
 
This application was car audio experiment to explore/exploit any boundary , so ignore everything except the smoothness and lack of anything akward or flawed by a phase issue of two or ( way more than two specific) paths or chambers engaged here. these typically have obvious response glitches in the region of the 85-100 hz 3 x1/4 and zigzags thereafter if not stuffed and bandaided. this didn’t require anything an inspired the pursuit of another to be used as multiples(not in a car, in home or other) in the hopes it might suite the Gedes approach (multiples) and still playing with the fun of (it’s not sealed, it’s probably not easy, but easy isn’t curious. Sealed might be inevitable but meanwhile it’s not. gotta look to know the brown grass before green.?) .

I never realized there was so much good BP4/6/8 info on the ‘Gedes’ website. It was waveguide things I grabbed to help me understand. this was food for more thoughts!!
 

Attachments

  • image.jpg
    image.jpg
    1 MB · Views: 300
Last edited:
I never realized there was so much good BP4/6/8 info on the ‘Gedes’ website. It was waveguide things I grabbed to help me understand. this was food for more thoughts!!

BP's rule!

FLH = BP4 with flared port.
TH = BP6S, where the whole enclosure is the flared port.
Ported FLH = BP6P.
Paraflex = BP6S with BP8 impedance response.
ROAR = BP6S.
Karlson = BP6S.
Skhorn = BP6P.
Devastator = BP6P.
MAUL = BP6P.
 
BP's rule!

FLH = BP4 with flared port.
TH = BP6S, where the whole enclosure is the flared port.
Ported FLH = BP6P.
Paraflex = BP6S with BP8 impedance response.
ROAR = BP6S.
Karlson = BP6S.
Skhorn = BP6P.
Devastator = BP6P.
MAUL = BP6P.

Yes, This^^!!:D:D

But we must go through them (using horn response AND info outside of ‘acoustics’ to setup the architecture and polish the ‘horn’ /pipe into folded segments with respect to the pressure wave for its own harmonic(s), phase, node/antinode. The driver entry for both sides is a unique thing and creates a whole new level of ‘phase aligned’ . Its NOT the Wave cannon. Thankfully. It’s gotta bounce and start the perfect sequences that (from rarification and pressurization ) create superposition on the (11/7,33/7, 55/7, 156/7, genetic circle parts as quarter/ hemisphere ) in qw. Better described as 1/4,3/4,5/4 harmonic of a (4/4 as the 1/4,3/4 pipes compounded total affective)and then … strips the entire out of band of all potential energies for the passband area and only leaves the naked truth. The native 1/4 wave bare bone spikes with (-990dB in sim) between all 1/4 and between 5ths. All the way to 20000(in math and sim).


What’s distortion like when it’s only full of its own daughters or kinda like radioactive decay isotopes :)? I mean, the box is simple . All 90 cm (3 of them and a single ) and driver in at ~30cm on each side, two opposed to cancel even if order stuff and vibes and … two of these or 4… and an infra using the other architecture(ML version isn’t a big rear vented matching .

5 half lives and it’s stable . physics numbers in everything? Even uranium’s 235,236,238 AMU is falling apart at the meteor belts 235,000,000 km mark fro the nucleus’ that’s 696,000 in radius (sun) PI!! Assorts if PI . And those numbers are the sim. The same .:eek:

Paraflex has been fixed. it’s not stuffing. It’s the nodes.
 

Attachments

  • image.jpg
    image.jpg
    1,019.1 KB · Views: 251
So... Earl states that as long as we have or build a subwoofer with plenty of headroom.... most other specs do not matter much. Is Q included?
I looked at upgrading my 12" subwoofers to 15" - by going from XXLS 830982 to RSS390 HF or HO - simply because I have heard these drivers perform very good and playing nice and deep, at friends in different setups.


The HF should fit fine in my 95 liter closed box, that now holds the XXLS - but the Q is now 0,83 - given a box of 2,88 ft3 around 82 liters - stated by parts-express:
Dayton Audio - RSS390HF-4 15" Reference HF Subwoofer 4 Ohm
Would this "higher" Q value have any impact on sound quality, when playing purely as a subwoofer with frequencies below around 110hz or so?


The HF version should be the High Fidelity one.... apparently better for good sound quality - above the HO, HE and Ultra version. Does this even make sense for subwoofer use and the perceived sound quality?
 
Other specs are still important.

A woofer with a high Qt and high Vas can be corrected in a smaller enclosure, but compared to a woofer with low Qt and low Vas it needs a lot more power.

For this reason it is important to keep an eye on the power (or rather VA) graph with all active filters applied.

What you're looking for is low Qt (or Qes), low Vas, and low Fs.
Using the woofer below Fs (or 10-15% lower as absolute minimum) is not a good idea because of the extremely non-linear behavior of the compliance.
 
Other specs are still important.

A woofer with a high Qt and high Vas can be corrected in a smaller enclosure, but compared to a woofer with low Qt and low Vas it needs a lot more power.
Thank you. This is why I looked at the 12" HF version too. Maybe the smaller subwoofer with a better quality build, would fit my use better, when focusing on the best suitable closed box and optimum sound quality - without the box being huge.
Big drivers are fun... but I want best possible sound quality foremost.
I have 4 x 1000W amplifiers in 4 ohms and full DSP control for all drivers.
 
Well, in sense of distortion or non-linear behavior, bigger woofers usually distort less since their excursion is (a lot) less.

In the end it all depends on SPL and high low you want to go.
But how does Q translate to audio quality? Would a subwoofer with a Q of 0,9 "ring" audibly more - compared to one that has a Q of 0,5? Or does it mainly impact on how deep and loud you can practically play, with a given amount of amplifier power and EQ?