Perhaps there should be a separate thread started for the "Two 18inchers in vented box vs 15inch in tapped horn" discussion so this can be focused on the Collaborative Tapped horn project?
Just a thought. Peace!
Just a thought. Peace!
MaVo said:
Hey, keep it amicable and try reading what I've said all along, which is that:
*** ACCORDING TO HORNRESP SIMULATIONS ***
the tapped horn doesn't seem to offer any big advantage over other boxes when both have correctly optimised drivers.
Since most people who are designing and building speakers -- including tapped horns -- rely on simulations to get a design at least close before cutting any wood -- including those on this group -- I don't see your problem here.
I've said many times that the results don't seem to tie up with what is claimed for the tapped horn, and that -- given Tom's reputation -- this doesn't seem likely, and that something doesn't add up. And I want to know what it is. And I'm not going to stop looking until I find out.
I've built many speakers in the past and intend to continue doing so in the future, but I'm not going to spend hundreds of pounds and a lot of wood and time building something which doesn't deliver what is predicted. When I am convinced I go ahead -- I've spent over $700 on a single 18" driver (sight unseen and nobody else had one in the UK) before now (replacing a blown one which cost over $500 and was also the first in the UK) -- have you?
It's easy to say "why don't you just build it?", but the kind of performance I'm looking for doesn't come from sticking a $50 driver in a cheap MDF box -- and if the tapped horn doesn't deliver, the driver parameters needed (as the simulations show) are very different to those for a reflex box, which means I can't transplant the (probably rather expensive) drivers into one to re-use them.
What I do find really annoying is people who just take things as gospel without either thinking about them or investigating them -- such as "tapped horns are always better than...", or "tapped horn efficiency goes up more in a corner", or "reflex ports always suffer from choking no matter how big they are" -- any or all of these will be true *in some cases*, I'm trying to find out if they're always true or not, and under what circumstances this happens.
And you have to admit, Tom didn't *exactly* help his own case by comparing a 15" driver in a >300l tapped horn (fine) with a single 15" driver in a >300l reflex, and then concluding that the tapped horn is great and the reflex driver couldn't be built -- of course not, nobody in their right mind would try to get high efficiency out of a single 15" in a ported box this size (which is why nobody does it), the right solution from the point of BL/Sd/Mmd is 2 18" drivers which can of course be built quite easily (which is why many people do it).
So let's wait until we hear from Tom (about whether he thinks the results I posted were correct) and David (about whether he thinks the Hornresp results are correct) before jumping to any further conclusions.
Like I said, I'll be happy to be proved wrong 🙂
Cheers
Ian
ion said:
My first post was actually saying "Hey guys, this is the best tapped horn I can come up with -- can anyone do any better, 'cos this doesn't seem to be any better than a reflex box, which doesn't agree with what everyone says?"
Somehow this seems to have turned into a tapped horn vs. reflex war -- all I ever wanted was a better tapped horn which would beat a reflex !
So back to square one -- has anyone managed to squeeze better performance out of a tapped horn than the one I originally posted for the TC PA-5100?
Cheers
Ian
P.S. If the predicted excursion in Hornresp is too high then a much cheaper driver would do the same job, 18mm Xmax isn't needed after all 🙂
AkAbaK vs. Hornresp.
Quick note. Although not looking at the design goals being targeted here (more interested in 20-50 Hz for domestic use), I ran a quick TH (BMS 12S330 in about 200l, actually not a very good design so I won't publish details) in Hornresp and AkAbaK.
The results are not identical but the excursion and response are both within 20% (and are both larger in Hornresp so the SPL for a given excursion is rather similar).
The difference is slightly frequency dependent - 20% is for the mid band (~30Hz).
I'm looking to see where the differnce could arise, but am rather busy with other things, so no promises.
Anyway no big difference like those hinted between model and reality at earlier today.
Ken
Quick note. Although not looking at the design goals being targeted here (more interested in 20-50 Hz for domestic use), I ran a quick TH (BMS 12S330 in about 200l, actually not a very good design so I won't publish details) in Hornresp and AkAbaK.
The results are not identical but the excursion and response are both within 20% (and are both larger in Hornresp so the SPL for a given excursion is rather similar).
The difference is slightly frequency dependent - 20% is for the mid band (~30Hz).
I'm looking to see where the differnce could arise, but am rather busy with other things, so no promises.
Anyway no big difference like those hinted between model and reality at earlier today.
Ken
That matches the findigs a few 🙂)) pages before. I can´t find the posts right now, but akabak and hornresp were pretty close in most accounts. Since then, many more were modeled, so far I haven´t encountered great differences in SPL or cone movement when modeling THs with both of them while using the same parameters.
iand said:
So we've had enough.... this leaves you three options that I can see.
1) Show Tom some measured proof,
2) Apologise to Sabbelbacke (and others), or
3) take up the fork and knife and start banging on the table for support, or ye'll be walking the plank.... arrr!
Sabbelbacke said:
So there's definitely something strange going on, then -- if AkaBak and Hornresp results for tapped horns are close to each other (presumably for excursion as well as sensitivity and frequency response) why don't the simulated numbers agree with those in the tapped horn white paper or the measured results in the TH-115 data sheet?
Not trying to attack anybody here, just trying to find out *why* there's such a big discrepancy -- if the simulations are wrong then anyone using these figures to design a tapped horn is likely to reach the wrong conclusions (as I could well have been doing 🙂
Cheers
Ian
iand said:
See #1284 " the excursion and response are both within 20% (and are both larger in Hornresp so the SPL for a given excursion is rather similar)"
That was, indeed, the point.
Ken
kstrain said:
Of course, which makes the discrepancy between [my Hornresp simulations] and [Tom's predicted excursion and frequency response (in the white paper) and measured frequency response] for the TH-115 even more puzzling.
Given that they're independently written it seems unlikely (but not impossible, see below) that both Hornresp and AkaBak would be wrong in the same way, so the fact that they agree reasonably well would suggest that neither is far out -- or at least, not by almost 2x (excursion) or 3dB (efficiency).
One possible explanation is that the model I used for the TH-115 isn't a good fit to the real thing -- but I'm pretty sure the 15TBX100 driver is correct, and no amount of tweaking to the design that I could find gave results any closer to the measured SPL or Tom's calculated excursion.
Another possibility is there's some real-life effect that Tom's simulations take into account and neither Hornresp or AkaBak does, in which case it would be really useful to know about this -- and I'm sure David would incorporate whatever it was into Hornresp pretty damn quick, it only took him a week or so to put the combined response feature in (open-back horn) after I suggested it.
Tom, can you shed any light on this?
Cheers
Ian
iand said:
Hi Ian,
As indicated in the Help file, Hornresp calculates diaphragm displacement assuming linear behaviour at all input voltage levels. No allowance is made for low frequency high power amplitude compression.
Another thing not specifically stated in the Help file, but hopefully users will have appreciated, is that the model assumes a steady-state, single-frequency, sinusoidal forcing function.
The mean-to-peak diaphragm displacement in millimetres is therefore calculated as follows:
D = Sqrt(2) * V / (2 * Pi * f) * 1000
Where:
V = rms velocity of the diaphragm piston in metres/sec
f = frequency in hertz
Multiplying by Sqrt(2) converts rms to peak, multiplying by 1000 converts metres to millimetres.
This is the standard equation for displacement, given sinusoidal motion. It is also the expression used by AkAbak, AJHorn and most other loudspeaker simulation programs - the only difference being that AJHorn chooses to quote rms rather peak values, possibly in an attempt to account for the fact that practical input signals are rarely pure sinusoids.
By exporting an *.aks script file from Hornresp and running it in AkAbak, it can be readily shown that the displacement results calculated by the two programs are identical. The results calculated by AJHorn are also the same as those generated by Hornresp when the AJHorn results are multiplied by Sqrt(2).
Hope this helps.
Kind regards,
David
Hi David
Thanks for the confirmation that Hornresp, AJHorn and AkaBak agree, this is obviously crucial to any simulation vs. real life discussion.
I think everyone realises that all these simulations are using small-signal models, on the other hand if we're operating within the linear Xmax of a driver this is a reasonable assumption -- if there was sufficient compression to alter the excursion results significantly (more than say 10%) the resulting harmonic distortion would be huge, and this is evidently not the case.
Have you also cross-checked the Hornresp results for ported/sealed boxes against other simulation programs to see if they agree with each other?
Do you (or does anyone else?) have any results comparing measured excursion in a horn with that predicted by Hornresp?
Cheers
Ian
Thanks for the confirmation that Hornresp, AJHorn and AkaBak agree, this is obviously crucial to any simulation vs. real life discussion.
I think everyone realises that all these simulations are using small-signal models, on the other hand if we're operating within the linear Xmax of a driver this is a reasonable assumption -- if there was sufficient compression to alter the excursion results significantly (more than say 10%) the resulting harmonic distortion would be huge, and this is evidently not the case.
Have you also cross-checked the Hornresp results for ported/sealed boxes against other simulation programs to see if they agree with each other?
Do you (or does anyone else?) have any results comparing measured excursion in a horn with that predicted by Hornresp?
Cheers
Ian
bass reflex v. tapped horn comparison
Hi Ian,
I am one of those who are interested in the workings of the tapped horn, and I feel that even though your needling has produced a few interesting responses, in general it is a sidetrack into a comparison of tapped horn and bass reflex design benefits, and belongs into another thread, also, the following seems a little disingenuous:
quote: "diyAudio Forums > Top >Loudspeakers >Subwoofers >Collaborative Tapped horn project
iand:
Post #1283: ...Somehow this seems to have turned into a tapped horn vs. reflex war -- all I ever wanted was a better tapped horn which would beat a reflex..."
- particularly in light of the fact, that you announced in another thread that:
quote: "diyAudio Forums > Top >Loudspeakers >Subwoofers >Who makes the lowest distortion pro-sound subwoofers?
inad:
Post #15: ...Tapped horns aren't necessarily more efficient, lower distortion or have higher maximum output than a reflex of the same size -- I'm just going to post an interesting comparison to the "Collaborative tapped horn project" thread...
Post #17: ...*if* you compare one driver in a tapped horn with one driver in a reflex, which is what Tom Danley did in his white paper when he concluded that a tapped horn was much superior. But he was really comparing it to the wrong thing... ...if you compare one (expensive) driver in a tapped horn with two (cheaper) drivers in a reflex you end up with the same box size, efficiency, maximum output, cone travel, and cost..."
In other words it looks to me, that you already had made up your mind that Tom Danley's approach was wrong and you are now just out to prove your point.
Anyway:
I second MaVo's motions in Post #1258 and 1279, and especially ion's in Post #1281: this much work deserves it's own thread.🙂
Hi Ian,
I am one of those who are interested in the workings of the tapped horn, and I feel that even though your needling has produced a few interesting responses, in general it is a sidetrack into a comparison of tapped horn and bass reflex design benefits, and belongs into another thread, also, the following seems a little disingenuous:
quote: "diyAudio Forums > Top >Loudspeakers >Subwoofers >Collaborative Tapped horn project
iand:
Post #1283: ...Somehow this seems to have turned into a tapped horn vs. reflex war -- all I ever wanted was a better tapped horn which would beat a reflex..."
- particularly in light of the fact, that you announced in another thread that:
quote: "diyAudio Forums > Top >Loudspeakers >Subwoofers >Who makes the lowest distortion pro-sound subwoofers?
inad:
Post #15: ...Tapped horns aren't necessarily more efficient, lower distortion or have higher maximum output than a reflex of the same size -- I'm just going to post an interesting comparison to the "Collaborative tapped horn project" thread...
Post #17: ...*if* you compare one driver in a tapped horn with one driver in a reflex, which is what Tom Danley did in his white paper when he concluded that a tapped horn was much superior. But he was really comparing it to the wrong thing... ...if you compare one (expensive) driver in a tapped horn with two (cheaper) drivers in a reflex you end up with the same box size, efficiency, maximum output, cone travel, and cost..."
In other words it looks to me, that you already had made up your mind that Tom Danley's approach was wrong and you are now just out to prove your point.
Anyway:
I second MaVo's motions in Post #1258 and 1279, and especially ion's in Post #1281: this much work deserves it's own thread.🙂
I have to agree, 15" driver in a tapped horn VS 2 18" drivers in a BR discussion needs its own thread.
As to the excursion dilemma. I cannot accurately measure it, but I certainly can tell when its moving 44mm p-p and when its moving 10mm p-p. It never reaches near full excursion at any frequency above 20Hz. Only when it gets down to 10Hz does it start to reach xmax.
I know the capabilities aren't whats important, but rather the measured vs the calculated. Thats what I'm saying, akabak predicted over 20mm excursion at 20Hz with a 1400W signal, I am not seeing half of this.
So what I'm getting at is the THs output isn't xmax limited nearly as much as your simulations suggest.
As to the excursion dilemma. I cannot accurately measure it, but I certainly can tell when its moving 44mm p-p and when its moving 10mm p-p. It never reaches near full excursion at any frequency above 20Hz. Only when it gets down to 10Hz does it start to reach xmax.
I know the capabilities aren't whats important, but rather the measured vs the calculated. Thats what I'm saying, akabak predicted over 20mm excursion at 20Hz with a 1400W signal, I am not seeing half of this.
So what I'm getting at is the THs output isn't xmax limited nearly as much as your simulations suggest.
Re: bass reflex v. tapped horn comparison
I hadn't made up my mind at the time -- what I said was that I'd got some results which seemed to show that they were equivalent.
And for two separate cases (the "challenge" and the TH-115) this still seems to be the case , nobody including Tom has produced any evidence to the contrary -- and if you read his own white paper it actually says that the two are equivalent in SPL.
If this is true -- still not proved! -- the tapped horn is another solution to how to build a good speaker, but not necessarily the great leap forward that many people seem to believe.
Of course it's still possible that the tapped horn does make the driver easier to build, but this isn't the case for either of these.
I expect most people on this group are trying out tapped horns because they *do* think there's a big advantage; surely if this isn't always the case this is a pretty significant finding about tapped horns, especially if the combined wisdom of the group is trying to build optimised ones?
I'm always sceptical about anything which claims to be a great leap forward, on the principle that if something looks too good to be true it usually is 🙂
Cheers
Ian
tb46 said:
I hadn't made up my mind at the time -- what I said was that I'd got some results which seemed to show that they were equivalent.
And for two separate cases (the "challenge" and the TH-115) this still seems to be the case , nobody including Tom has produced any evidence to the contrary -- and if you read his own white paper it actually says that the two are equivalent in SPL.
If this is true -- still not proved! -- the tapped horn is another solution to how to build a good speaker, but not necessarily the great leap forward that many people seem to believe.
Of course it's still possible that the tapped horn does make the driver easier to build, but this isn't the case for either of these.
I expect most people on this group are trying out tapped horns because they *do* think there's a big advantage; surely if this isn't always the case this is a pretty significant finding about tapped horns, especially if the combined wisdom of the group is trying to build optimised ones?
I'm always sceptical about anything which claims to be a great leap forward, on the principle that if something looks too good to be true it usually is 🙂
Cheers
Ian
Ian
You still don't seem to get it that your reflex sims are far from real. I strongly encourage you to take a reflex box you have already built, and modeled, put it in a large flat area and measure it, near X-max or it's thermal limit, right across it’s usable frequency range.. You won't get anything near your model. If it comes within 6dB I'd be very surprised. Stop fooling yourself with inaccurate predictions. GO AND MEASURE SOME SPEAKERS!
Even more telling, you could go and buy one of the cheap Eminence 15” drivers mentioned earlier in this thread and build the tapped horn as modeled, then measure it. A comparison between this box and any thing else in its price/size class would teach you a few valuable home truths.
Until you build and measure, you’re just kidding yourself with numbers.
Cheers
William Cowan
You still don't seem to get it that your reflex sims are far from real. I strongly encourage you to take a reflex box you have already built, and modeled, put it in a large flat area and measure it, near X-max or it's thermal limit, right across it’s usable frequency range.. You won't get anything near your model. If it comes within 6dB I'd be very surprised. Stop fooling yourself with inaccurate predictions. GO AND MEASURE SOME SPEAKERS!
Even more telling, you could go and buy one of the cheap Eminence 15” drivers mentioned earlier in this thread and build the tapped horn as modeled, then measure it. A comparison between this box and any thing else in its price/size class would teach you a few valuable home truths.
Until you build and measure, you’re just kidding yourself with numbers.
Cheers
William Cowan
Just some food for thought: When messing around in Hornresp with the 6.5" Tang Band W6-1139SI Erik (Volvotreter) recommended (Post #1085), I noticed that a grouping of 8 makes for a very high output tapped horn of about the dimensions Ian (iand) is looking for. With the current PartsExpress price of $29.– ea. that seems almost like a reasonable thing to try.
Attachments
Hi Ian,
Yes. For ported and sealed boxes, Hornresp and AkAbak diaphragm displacement results are identical. This is not unexpected, given that the two programs are essentially using the same model.
Unfortunately I have not had the opportunity to conduct any predicted versus measured comparisons myself. The consistent feedback I receive though, is that measured displacements are somewhat less than predicted results.
As you say, it would be nice to know why 🙂.
Kind regards,
David
iand said:
Yes. For ported and sealed boxes, Hornresp and AkAbak diaphragm displacement results are identical. This is not unexpected, given that the two programs are essentially using the same model.
iand said:
Unfortunately I have not had the opportunity to conduct any predicted versus measured comparisons myself. The consistent feedback I receive though, is that measured displacements are somewhat less than predicted results.
As you say, it would be nice to know why 🙂.
Kind regards,
David
cowanaudio said:
Hi William
I've said *at least five times* that reflex boxes with too small ports and too high port velocity -- just like your 18LW1400 boxes with 30m/s in the ports -- do have big problems with port choking and nonlinearity, which should come as no surprise to anybody.
(incidentally, I'm still waiting for your reply to my post about port compression in the Seismic 8196 loaded box, as to how come if the port is choking so bad the diaphragm displacement is so small and the tuning frequency doesn't shift much with level...)
As I also said, to avoid this objection the "TH-115 reflex" has a huge 10" diameter port 20" long for each driver, which with only 9mm Xmax gives about 10m/s maximum port velocity, which is way lower than *any* other reflex box I've ever seen. If you're still saying that these will suffer significant port compression then I suggest that you're ignoring these facts 🙂
Anyway, let's stop the tapped horn vs. reflex debate for now because it's only generating too much "sturm und drang"...
Tapped horn simulation vs. measurement
----------------------------------------------------
I feel you're still missing the real point, which is that -- regardless of what realisation problems a design does or doesn't suffer from -- it can't ever do *better* than simulations (*if* these are accurate), only *worse* -- accurate simulations give an upper bound on performance.
So *if* Hornresp is correct (and I've repeatedly questioned this) neither type of box should be able to *exceed* the predictions for efficiency and maximum output, only fall short -- any losses and imperfections can only decrease SPL and increase excursion.
I've also pointed out that in spite of considerable trying I can't get results for either efficiency or excursion which agree with those in Tom's white paper on the TH-115 -- which are of course simulated, in spite of the fact you don't seem to believe in this. I'd also like to point out -- yet again! -- that he says that the tapped horn and reflex have similar efficiency when both are optimised, not that the tapped horn is higher 🙂
Hornresp parameters I'm using are:
S1=285 (3:1 CR) L=20/280/20cm (all exponential) S3=2400
-- exponential instead of conical gives less ripple because mouth area is bigger for required length, no other way to get low ripple close to measurements
-- added 1.3mH extra series inductance to match impedance with measured -- also reduces ripple
Efficiency
-----------
Here are SPL figures from Hornresp, the tapped horn white paper, the measured ones from the 2005 TH-115 data sheet (5/10/2005 test date), and the ones from the current data sheet (8/11/2005 test date) -- the bottom row is the trapezoidal mean from 40-100Hz:
Freq Hornresp paper 2005 2008
30 89.1 93 91 90
40 100.1 102.5 101 102
50 99.8 101.5 103.5 103
60 99.9 102 104 104
70 100.8 102.5 103 104
80 101.6 103 104 105.5
90 101.5 103.5 106 107
100 100.2 104 107 110
mean 100.6 102.6 104.1 104.9
The differences might not seem huge, but when you're trying to compare one design with another they are significant.
I can believe that the design I'm simulating in Hornresp is not optimum, which is why I've been asking if anyone can do better (with the 15TBX100, obviously) -- I'm not trying to clone/rip-off the TH-115, but it's the only real-life comparison we've got this much data on, and it's what Tom suggested.
So the 2dB difference between this result and the white paper could be down to this -- though I should point out that this is 1.6x increase in power efficiency which is not small.
If we assume that the white paper results are correct then the 2005 measured results are another 1.5dB better than this (another 1.3x factor) and the 2008 results are 2.3dB better (1.7x higher power efficiency).
Maybe all this is due to continuous improvement in the design by Tom, I don't know -- but in his own words "the tapped horn and reflex have similar efficiency", and we know the "optimum" reflex is about 101dB/W -- so where has the extra 3dB come from, this is double the power efficiency which is a *huge* difference!
(some of this is due to the rising efficiency with frequency of the tapped horn, the reflex is 101dB/w flat -- but the measured tapped horn is 104dB/W before this happens, which is still 3dB higher)
I'm absolutely *not* calling Tom a liar or doubting his measurements, Danley are by far the most honest provider of measured results that I know of -- but in that case I can only see 3 possible conclusions:
1. There's a much better tapped horn solution lurking in the same sized box (320l net) with the same driver (15TBX100) which I haven't found yet (but I'm still looking)
2. Hornresp is underestimating the efficiency of a tapped horn by up 3dB
3. Tom has some special "secret sauce" design tweaks which he's not telling about
==> Since I assume everyone here is interested in designing and building tapped horns which behave how they expect, I think it's kinda important to find out what the explanation is.
Excursion and maximum SPL
-----------------------------------
The tapped horn white paper shows a peak excursion of 6.1mm at 47Hz with 63V applied (500W/8 ohms); with the 15TBX100 mathematical Xmax of 9.5mm this means 1200W power handling (98V applied) before Xmax is exceeded. At 101.5dB/w this would give a maximum SPL of of 132.5dB, which agrees closely with the TH-115 data sheet.
Hornresp predicts a peak excursion of 14.8mm at 47Hz with 98V applied and an SPL of 130.9dB -- note that the impedance curve frequencies of maximum and minimum excursion are very similar to the TH-115, so the design can't be that far off.
This difference is even more worrying than the SPL one in many ways -- the SPL difference just means the output will be a bit higher than expected (which is good!), but the excursion difference has a radical effect on the design:
-- if Tom's figures are correct and Hornresp is wrong and you design a tapped horn using Hornresp, you'll end up using drivers with much higher Xmax than is actually needed -- this could make finding suitable drivers difficult or at the very least expensive (TC PA-5100)
-- if Tom's figures are wrong and Hornresp is correct then tapped horns don't handle as much power as you'd expect unless you use drivers with very large Xmax -- drivers with 9.5mm Xmax are easy to find, drivers with 15mm much less so.
==> Again, since I assume everyone here is interested in designing and building tapped horns which behave how they expect, I think it's kinda important to find out what the explanation is.
Why we (diyAudio people) need an explanation
------------------------------------------------------------
Tapped horns are pretty much impossible to build via trial and error because they are much more sensitive to a larger number of design parameters than a.n.other box, and it's almost impossible to guess whether a particular driver will work well in a tapped horn or not -- again, in contrast to other designs -- and what to do if it doesn't.
So if we want to come up with decent tapped horn designs, we all need a simulation tool which can predict the performance with reasonable accuracy -- but right now it seems there are considerable differences between simulation (all simulators) and real life with regard to efficiency, excursion and maximum SPL.
I suspect the only person who can shed further light on these inconsistencies is Tom, since he certainly knows far more about this then anyone else.
Cheers
Ian
tb46 said:
That does indeed look like a good design, though with lower efficiency (results are in 0.5*pi) and too much extension compared to my target.
If the real SPL is higher and the real excursion lower than the Hornresp simulations you could say that either:
1. this is a good thing (peak power handling would be even higher without exceeding Xmax -- if you've got a big enough amplifier), or
2. this is a bad thing because lower Xmax drivers might have given higher efficiency and therefore higher SPL for a given amplifier
Using large numbers of smaller drivers like the Tangband can be an excellent idea compared to one big driver if the cost and power handling are comparable -- it's often better to have several lower power drivers, ones rated at 1000W or so can physically handle these power levels but the resulting power compression can be higher than lower power drivers which don't get quite so hot.
If only we knew what "optimum" drivers for a tapped horn looked like for different cases (box size, cutoff, power handling) and could rely on simulation results to predict SPL and excursion, designing good tapped horns would be *much* easier... 🙂
Cheers
Ian
David McBean said:
Hi David
More than nice -- I'd have thought given your interest in exploring new things with Hornresp you'd be busting to find out why!
It does seem very strange that the different simulation programs seem to agree with each other, and that for reflex and closed boxes they agree with the real world (and presumably for conventional bass horns, since they've been used to design these for many years, including Tom's LABhorn), but they differ significantly from measurements for tapped horns.
After all, I expect the same basic equations which relate impedance, radiation resistance, excursion and SPL are used in all cases (is this true?), so why should tapped horns be so different?
Cheers
Ian