Why horn-loaded loudspeakers are subject to design tradeoffs

[djk]

FMD is proportional to the bandwidth in octaves squared.

The narrower the bandwidth the better.

Edited August 1, 2014 by djk

[DrWho]

You're the one claiming THD plots are misleading compared to IMD.....so you tell us what behaviors aren't captured there.

[Chris A]

It's clear to me that a "paradigm shift" is required in order to see the effects that we are discussing. It is important to note that "Doppler distortion" is a well known effect, but the real issues that we are dealing with are "audible distortion at listening levels".

For the moment, I'm going to focus only on bass bin distortion--the reasons being that this is where most of the distortion is found in any loudspeaker at any level of SPL output, and that all the distortion frequencies become audible due to modulation effects--including but not limited harmonic distortion (which turns into IMD if the loudspeaker bass bin plays anything other than single sine wave frequencies).

If your loudspeakers cannot play music back at anywhere close to concert levels without audible distortion overtaking the listening experience (as it does for my listening with any direct radiating loudspeaker bass bin I've ever heard) then what is happening is that your ears are getting used to hearing that distortion, and you get used to evaluating your loudspeakers relative to other loudspeakers, and not live performance. This statement may be a bit shocking at first, but I believe that it is one that those owning fully horn-loaded loudspeakers really are trying to express in discussions on this forum.

Here are two references that I'd hope that the readers here will take a little time to digest:

http://www.artalabs.hr/AppNotes/AP7-EstimationOfLineardisplacement-IEC62458-EngRev01.pdf - look on pages 8 and 9. These show that with higher displacement of driver cones found in direct radiators, the higher the distortion. And the red and blue traces in each graph need to be very much less than 10% if the listener is to suspend disbelief that what they are listening to is anything other than that close to a live performance.

Any time you boast about a "modern driver" having "crazy low distortion", in order to compare its distortion with horns, you must use driver cone displacements in the direct radiator that are at least 5 times that of a horn loudspeaker using that same driver with a horn for equal SPL output, for apples-to-apples comparison purposes.

Now we are getting closer to what I would call "useful arguments" about distortion in bass bins by type of enclosure. You cannot only look at total harmonic distortion (THD) at some arbitrary level of cone displacement (Xmax) and compare that with using that same driver in a good horn assembly, because the horn-loaded assembly can do the same job at 1/5 the cone displacement. The lower the distortion (any type), the better the bass bin sounds, and this is strongly related to multiple distortion types occurring simultaneously from several different sources.

The second reference is a tutorial that Wolfgang Klippel put together, and I have to admit that I'm still digesting it. It's possible that it is the best single article on loudspeaker or driver performance that I've ever read:

http://www.klippel.de/uploads/media/Loudspeaker_Nonlinearities%E2%80%93Causes_Parameters_Symptoms_01.pdf

More to come...

Edited August 2, 2014 by Chris A

[DrWho]

I only compare distortion at the same SPL.

No paradigm shifts required.

[DrWho]

To make this interesting, let's see the distortion numbers from whatever you feel is the lowest distortion bass horn.

[Chris A]

To make this interesting, let's see the distortion numbers from whatever you feel is the lowest distortion bass horn.

http://www.hometheatershack.com/forums/subwoofer-tests-archived/970-intermodulation-distortion-measurement-results.html

Apparently the only horn-loaded subwoofer in the bunch is the "DIY PA horn"...all the others are direct radiators.

EDIT: It looks like the cutoff frequency of this "DIY PA Horn" is more like that of a Khorn or Jubilee bass bin, so if they are pushing the response of this unit down to 20 Hz, the THD and other modulation distortion products of this horn-loaded bass bin is probably much higher than most for-purpose horn loaded subwoofers with longer horn lengths (e.g., my TH subs have a horn length of 21 feet [6.4 metres] while the Khorn bass bin is more like 8 feet [2.4 metres]).

You may recall this thread--the test procedure is briefly mentioned here.

Edited August 4, 2014 by Chris A

[DrWho]

In the range of 60-100Hz, that horn is roughly 4% THD with a 115dB sweep. Solution size is 25" x 25" x 38".

The B&C 21SW152-4 in a 24" x 35" x 31" enclosure tuned to 25Hz is roughly 1% THD with a 115dB sweep.

Putting the 21SW152-4 into a tapped horn (24" x 36" x 36") actually isn't buying any distortion improvement with the same 115dB sweep (also 1% THD):

(The CEA2010 short-term average over the range of 20-80Hz only picks up another 3dB of output)

There are plenty more drivers here if you'd like to make more comparisons:

http://www.data-bass.com/data?page=systems&type=0

I recommend sorting by the CEA2010 short-term average over the range of 20-80Hz (the headings are sortable).

Maybe you'd like to blame the XLS12 driver? That's why I'm providing the vented versus tapped horn comparison here. The XLS12 driver is actually quite horrible:

Due to the adiabatic curve of air, I think there is a point of diminishing returns with the horn:

As we reduce the distortion from the driver, the benefit of the horn is going to go down (as we see in the examples presented here). Hypothetically, if the linearity of the driver gets low enough, then the non-linearity of the higher pressures in the horn are can actually dominate the distortion of the system. I don't have the Klippel measurements of the B&C driver handy, but it was touted as one of the best drivers in the world in Voice Coil magazine. The Bl / CMS linearity and symmetry was almost flat over most of the operating region.

In other words, I think the benefit of horn loading is greatest when you're working with cheap drivers - which is very common practice across the horn loaded community. I also don't necessarily believe that throwing in a better driver just solves everything about those horns either. It could be that the driver is just masking the artifacts from the horn, and then removing the driver artifacts could leave the horn distortion as the next biggest issue. This is almost a daily exercise for me in the electronics world....lower the noise floor and/or the distortion, and new artifacts start revealing themselves that were previously masked.

Btw, Keele has an interesting article about direct radiator versus horns:

http://www.xlrtechs.com/dbkeele.com/PDF/Keele%20(1976-05%20AES%20Preprint)%20-%20Efficiency,%20Horns%20vs%20DR.pdf

He concluded that for a given volume, the systems will at least be comparable - if not the vented system being able to go louder. I think Danley and the real world have demonstrated that the tapped horn has been shown to exceed the vented cabinet, but it's still a real close race.

[DrWho]

Here is some information on trying to compensate for the non-linearity of air:

http://doc.utwente.nl/58981/1/Schurer94modeling.pdf

The new IPAL series from B&C is kinda approaching the same issue from a different angle:

http://www.bcspeakers.com/products/lf-driver/21-0/21ipal

http://www.powersoft-audio.com/en/technologies/dpc-differential-pressure-control

Basically adding a feedforward control system to a crazy GBW driver to have even lower distortion. You can completely offset Doppler Effect with this kind of approach too - something even a horn doesn't eliminate.

[Chris A]

Interesting THD data that I'll look at more closely. However, I was reporting on IMD, you reported on THD. Also, the problem with THD is that each harmonic is not created equal - higher order harmonics were weighted more strongly in the IMD tests that were reported in the HTS thread on IMD measurements that I referenced above (as well as here). I don't believe that the THD figures reflect this weighting, so I'm not sure that we have "apples-apples" here.

Perhaps we can both look for more references on bass bin IMD measurements (...as I found one already...)? I don't believe that looking only at THD is going to adequately address the subject of IMD in bass bins.

In the mean time, I'll continue the discussion on IMD by discussing the "derating" of higher output capability loudspeakers for home use in order to further reduce IMD (among other improvements in performance), and also a discussion of HF drivers (and horns, as the case may be) and FMD effects. It's always been interesting to me why so many audiophiles revel in the "full-range driver" world--including but not limited to Nelson Pass, who developed his First Watt amplifiers and other electronics almost exclusively for this market.

BTW: the article that you linked from Don Keele is one that I've seen many times before - it's related to "size efficiency" of the enclosures. Unfortunately, that subject isn't of as much interest to me as it is to folks that are worried exclusively about "smaller loudspeakers".

Edited August 18, 2014 by Chris A

[DrWho]

I believe IMD and THD are directly related except for Doppler related distortions...at least I can't think of any other non-linearity that doesn't also show up in the THD.

IMD products are only generated when there is a non-linearity in the system. In other words, a linear system does not create new frequencies - which is actually the technical definition of a linear system, so I guess that's a bit circular. The point here is that IMD can't exist without the presence of THD because any non-linearity also shows up in the THD.

Ilkka made the following comment about the few IMD tests he made:

...these first results suggest that subwoofers with high output capabilities and low THD levels also have low IMD levels. One should carefully examine if IMD is even worth measuring with subwoofers due this relationship.

Read more: http://www.hometheatershack.com/forums/subwoofer-tests-archived/971-subwoofer-tests-explained.html#ixzz39SwFidcv

I can't say I've explored this exhaustively, but I'm yet to find anything that breaks this THD/IMD trend....definitely open-minded on the subject though - especially in light of that Jub LF versus 4x15" driver demo we both heard. That demo really rocked my understanding of things, and today I'm leaning towards the drivers themselves being different. I'd like to hear that performed again with modern drivers that are already starting out more linear (for both cabinets).

[Chief bonehead]

In the range of 60-100Hz, that horn is roughly 4% THD with a 115dB sweep. Solution size is 25" x 25" x 38".

DIY PA Horn_Distortion.png

The B&C 21SW152-4 in a 24" x 35" x 31" enclosure tuned to 25Hz is roughly 1% THD with a 115dB sweep.

21SW152_Distortion.png

Putting the 21SW152-4 into a tapped horn (24" x 36" x 36") actually isn't buying any distortion improvement with the same 115dB sweep (also 1% THD):

21SW152_TH_Distortion.png

(The CEA2010 short-term average over the range of 20-80Hz only picks up another 3dB of output)

WHAT??? Dangerous.... You are.

There are plenty more drivers here if you'd like to make more comparisons:

http://www.data-bass.com/data?page=systems&type=0

I recommend sorting by the CEA2010 short-term average over the range of 20-80Hz (the headings are sortable).

Maybe you'd like to blame the XLS12 driver? That's why I'm providing the vented versus tapped horn comparison here. The XLS12 driver is actually quite horrible:

diy peerless xls 12 ported 85l Distortion.png

Due to the adiabatic curve of air, I think there is a point of diminishing returns with the horn:

As we reduce the distortion from the driver, the benefit of the horn is going to go down (as we see in the examples presented here). Hypothetically, if the linearity of the driver gets low enough, then the non-linearity of the higher pressures in the horn are can actually dominate the distortion of the system. I don't have the Klippel measurements of the B&C driver handy, but it was touted as one of the best drivers in the world in Voice Coil magazine. The Bl / CMS linearity and symmetry was almost flat over most of the operating region.

In other words, I think the benefit of horn loading is greatest when you're working with cheap drivers - which is very common practice across the horn loaded community. I also don't necessarily believe that throwing in a better driver just solves everything about those horns either. It could be that the driver is just masking the artifacts from the horn, and then removing the driver artifacts could leave the horn distortion as the next biggest issue. This is almost a daily exercise for me in the electronics world....lower the noise floor and/or the distortion, and new artifacts start revealing themselves that were previously masked.

Btw, Keele has an interesting article about direct radiator versus horns:

http://www.xlrtechs.com/dbkeele.com/PDF/Keele%20(1976-05%20AES%20Preprint)%20-%20Efficiency,%20Horns%20vs%20DR.pdf

He concluded that for a given volume, the systems will at least be comparable - if not the vented system being able to go louder. I think Danley and the real world have demonstrated that the tapped horn has been shown to exceed the vented cabinet, but it's still a real close race.

[Chief bonehead]

In the range of 60-100Hz, that horn is roughly 4% THD with a 115dB sweep. Solution size is 25" x 25" x 38".

DIY PA Horn_Distortion.png

The B&C 21SW152-4 in a 24" x 35" x 31" enclosure tuned to 25Hz is roughly 1% THD with a 115dB sweep.

21SW152_Distortion.png

Putting the 21SW152-4 into a tapped horn (24" x 36" x 36") actually isn't buying any distortion improvement with the same 115dB sweep (also 1% THD):

21SW152_TH_Distortion.png

(The CEA2010 short-term average over the range of 20-80Hz only picks up another 3dB of output)

There are plenty more drivers here if you'd like to make more comparisons:

http://www.data-bass.com/data?page=systems&type=0

I recommend sorting by the CEA2010 short-term average over the range of 20-80Hz (the headings are sortable).

Maybe you'd like to blame the XLS12 driver? That's why I'm providing the vented versus tapped horn comparison here. The XLS12 driver is actually quite horrible:

diy peerless xls 12 ported 85l Distortion.png

Due to the adiabatic curve of air, I think there is a point of diminishing returns with the horn:

As we reduce the distortion from the driver, the benefit of the horn is going to go down (as we see in the examples presented here). Hypothetically, if the linearity of the driver gets low enough, then the non-linearity of the higher pressures in the horn are can actually dominate the distortion of the system. I don't have the Klippel measurements of the B&C driver handy, but it was touted as one of the best drivers in the world in Voice Coil magazine. The Bl / CMS linearity and symmetry was almost flat over most of the operating region.

In other words, I think the benefit of horn loading is greatest when you're working with cheap drivers - which is very common practice across the horn loaded community. I also don't necessarily believe that throwing in a better driver just solves everything about those horns either. It could be that the driver is just masking the artifacts from the horn, and then removing the driver artifacts could leave the horn distortion as the next biggest issue. This is almost a daily exercise for me in the electronics world....lower the noise floor and/or the distortion, and new artifacts start revealing themselves that were previously masked.

Btw, Keele has an interesting article about direct radiator versus horns:

http://www.xlrtechs.com/dbkeele.com/PDF/Keele%20(1976-05%20AES%20Preprint)%20-%20Efficiency,%20Horns%20vs%20DR.pdf

He concluded that for a given volume, the systems will at least be comparable - if not the vented system being able to go louder. I think Danley and the real world have demonstrated that the tapped horn has been shown to exceed the vented cabinet, but it's still a real close race.

WHAT??? Dangerous....you are.....

[mustang guy]

Will anybody admit defeat?

Never!

The fact of the matter is that you can use a horn as a funnel, so they are better. Who can argue with that? Fact!

[DrWho]

Haha - I could see a K400 being used for an oil change....

[Chris A]

I have to say that the 21 inch diameter (53 cm diameter) driver that you linked to (http://www.data-bass.com/data?page=system&id=1&mset=32) is by far the best driver performance that I've ever seen - by a huge margin. I can see why you made your comments--I only wish that I'd investigated that woofer's harmonic distortion performance - which is about 10x better performance than any that I've seen before. [Note that they stopped measuring beyond the 4th harmonic, and I wish that they would have at least shown the higher harmonics--even though they said that they were negligible].

All this says to me something entirely different: how poorly designed almost all cone-type drivers have been--and for so long. All of the measurements and observations of PWK were based on real drivers, and now I see how poorly they've been designed from the standpoint of linearity, or stated conversely, with very high non-linear performance from the electromagnetic, thermal, and mechanical (compliance) perspectives.

Now the observations from the other side of the equation:

1. I'd love to have two or four of the 21 inch drivers to play with (since I've never heard anything like that-even in Hope, AR) -- except that I'm currently not going to spend $1400 for two woofers nowadays. That's a bit of an issue, but I understand how good these drivers are. I once sprung for TAD compression drivers (4002s) that made all the difference in the world: perhaps I might try a couple of these woofers in the future.
2. Now that I look at the passband of these woofers, I've got to say that I'm impressed with the Jub bass bins from a combination perspective of low distortion and extended passband (and the fact that you still have 2.5x mouth area as direct radiator for one 21 inch driver). These woofers used in a vented box really don't even have the passband of the TH horn that uses the same drivers - particularly in the bottom full octave (lower than 10 Hz for the TH Gjallarhorn vs. 20 Hz for the vented box). This is actually backward from the characteristics that I'm used to seeing in terms of passband of the different types of enclosures.
3. I don't believe that the basic observations on horn loading vs. direct radiating are wrong based even on these examples - but that the nonlinearities of virtually all available drivers have dominated the horns vs. direct radiator discussions, which from my perspective is still true since greater than 99.9% of all drivers nowadays have much worse nonlinear faults. Just look at the performance of all the direct radiating subs in the Home Theater Forums thread (which you also conveniently glossed over in your reply ). Virtually all of the commercially available direct radiating subs that are marketed nowadays are much worse than horn-loaded subs. It is only this one exception that I see thus far. That's a bit of a problem IMHO. My advice on this subject won't change in the near term until the marketplace starts to be populated much more often with much better drivers in the subs and bass bins.

Edited August 18, 2014 by Chris A

[DrWho]

BTW: the article that you linked from Don Keele is one that I've seen many times before - it's related to "size efficiency" of the enclosures. Unfortunately, that subject isn't of as much interest to me as it is to folks that are worried exclusively about "smaller loudspeakers".

I think the "space efficiency" matters a lot because bass horns are always undersized. If the size didn't matter, then why don't you have a full-sized straight 20Hz bass horn with constant directivity in your own setup? (or pick whatever arbitrary Fc for your particular application).

For whatever size enclosure is deemed acceptable for the bass horn, Keele's article shows that the EBP of an equal volume vented system will be comparable, and maxSPL is increased due to increased power handling. I would also posit that the linearity of the direct radiators is improved at the listening SPL because each driver is moving less. Can the linearity get better than that of the horn? I think it's possible based on the data I've seen, but then I've never seen Klipsch publish distortion sweeps. Perhaps our Chief Bonehead friend (haha, I love that name) is able to show us some better horn data?

I'm going to model the expected driver excursion of some ideal tractrix horns, and then see how many direct radiators it takes to match that excursion for a given SPL. If the horn is perfect, then the distortion should be identical (actually slightly in favor of the direct radiator due to the distributed heat, but I don't think that's significant here). It may take a few days before I'm able to do that though....writing all this from a phone is brutal.

[DrWho]

1. I'd love to have two or four of the 21 inch drivers to play with (since I've never heard anything like that-even in Hope, AR) -- except that I'm currently not going to spend $1400 for two woofers nowadays. That's a bit of an issue, but I understand how good these drivers are. I once sprung for TAD compression drivers (4002s) that made all the difference in the world: perhaps I might try a couple of these woofers in the future.
2. I don't believe that the basic observations on horn loading vs. direct radiating are wrong based even on these examples - but that the nonlinearities of virtually all available drivers have dominated the horns vs. direct radiator discussions, which from my perspective is still true since greater than 99.9% of all drivers nowadays have much worse nonlinear faults. Just look at the performance of all the direct radiating subs in the Home Theater Forums thread (which you also conveniently glossed over in your reply ). Virtually all of the commercially available direct radiating subs that are marketed nowadays are much worse than horn-loaded subs. It is only this one exception that I see thus far. That's a bit of a problem IMHO. My advice on this subject won't change in the near term until the marketplace starts to be populated much more often with much better drivers in the subs and bass bins.

1) In the spirit of enjoying this hobby with others, we might be able to arrange for you to borrow a driver. Eventually I may be picking up a second driver at some point, so maybe you'd like to wait until then so you can play with two of them. Btw, I agree with your sentiments about what the Klipsch designs are able to accomplish - across the board actually.

3) I think the reason I've been pressing this issue so hard is entirely due to the idea of the fundamental physics at play. If as you say this driver is very different from the vast majority on the market, then that means some of the theory about horns and direct radiators is a bit over/under stated depending on how you want to look at it. The majority of audiophile systems are not horn loaded at all so I think we need to be a bit careful about overstating the importance of the majority. That said, I totally agree with your observation that the comparison is different today than when PWK published those articles....or really any of the horn literature for that matter.

Btw, I did want to add that the list of "good drivers" (like the B&C) has been growing. There are several more drivers showing up with amazing Klippel results - and some are much cheaper too. There are also some that are marketed to be awesome and part of the modern vein, but have exhibited other issues. That has really muddied up the waters for those that use only their ears.

[Chris A]

Now that I've had more time to review the various designs, I'd say that my basic observations of horn-loaded vs. direct radiator still hold, with the exception of the ability of some very high quality/high cost direct radiating (DR) woofers to provide good performance at exceptionally low THD levels. This is good news (but expensive news, just like my experiences with the TAD series of two inch diameter Beryllium diaphragm compression drivers)--TAD Be compression drivers make all the difference in the world, but they are very expensive. I dare say that they probably significantly outperform very expensive full range drivers when the TAD TD-4002s are used on the K-402 horn.

The issue that I see with the Data-Bass series of tested subwoofers is that these designs are really subwoofers and not bass bins. What that means is that you must pay particular attention to the bottom octave of performance for each design for comparing apples to apples.

For instance, the DIY Gjallarhorn with its single 18" driver is designed for the 10-40 Hz region, with its greatest design attribute its performance below 20 Hz. The BMS 18n862 Vented design just doesn't really compete in that octave. So the Gjallarhorn and the BMS 21" port sub are really apples and oranges.

So it goes for basically all the highest performance designs found at the Data-Bass site: apples, oranges, bananas, and strawberries--all are performing their intended design purpose with the design constraints placed on them (including size of the enclosure and the increasing degrees of insensitivity to cost, including high power amplifiers). If you design for lower frequency extension of the horn, you're not going to get the higher frequency performance relative to other designs (in general)--and that includes losing low THD performance.

Additionally, once you look at apples vs. apples in terms of designed FR, the results are clear - good horn-loaded bass bins are at least 15 dB(SPL) higher sensitivity than direct radiators using the same diameter and number of drivers. In general, if you're not trying to reproduce sub-20 Hz frequencies, then for like-quality drivers, horns have much lower IMD effects on average. If you are willing to pull out your pocketbook for a subwoofer (which I'm currently not willing to do) then you can get very clean 20+ Hz performance for LFE reproduction (i.e., short-term maximum SPLs at low frequencies). If, however, you are playing Bach Preludes and Fugues, you might want to rethink using direct radiating subs.

Tapped Horn (TH) subs: These seem to be the only examples of horn-loaded subs in the Data Bass list of tested systems. It's been my experience that TH subs are really good for one thing--and that is sub-20 Hz performance at very high sustained SPL in a very compact and relatively slim package, but covering no more than two octaves of passband (if that--I use just under one and a half octaves with my TH subs). Impulse response of TH subs is a weak point on the TH design, but fortunately, impulse response with a very narrow passband and low crossover frequencies doesn't seem to be as objectionable as when used in wider passband applications.

If you don't highly value relatively low distortion sub-20 Hz performance above all other subwoofer performance attributes, then I wouldn't recommend a TH sub - but rather a conventional front-loaded subwoofer bass bin with a cutoff frequency above 20 Hz. The "Tuba" series of subs apparently fits that bill fairly well (except that their design performance apparently cannot be openly discussed in relation to other DIY sub designs due to the restriction on data by the IP owner--which is a really bad deal, IMHO).

Direct Radiating (DR) subs: My opinion on these type of subs is basically unchanged--unless you are willing to shell out the bucks (about $700+(US) for each driver) for the highest quality/low distortion drivers in DIY enclosures, I wouldn't recommend any of the commercially available bought subwoofers. There may be some professional DR subwoofers out there that use the highest quality drivers, and are therefore very low in THD and IMD, but I'm not currently aware of any of these. Almost without exception, DR subwoofers cover 20Hz and higher, but not really anything sub-20 Hz (unless the driver and cabinets are much larger and much, much more expensive than equivalent TH subwoofers).

Conventional Front-Loaded (FL) horn bass bins: Since conventional horns going lower than 40 Hz cutoff frequency (Fc), such as the Jubilee and Khorn bass bins--but not limited to these two designs--these are typically too large for home use and we generally do not talk about using this type of configuration for subwoofer designs. Instead, these are bass bin designs appropriate for the expanded passband of 30-400+ Hz--and are not subwoofers. Claude's quarter pie design fits this category.

Lower Fc front-loaded conventional horns are generally too large for home use, (i.e., we're basically talking about folded horn designs for cutoff frequencies below 200 Hz). What we're looking for here is broad passband up to the midrange driver/horn frequencies, such as a low of 250 Hz, but more likely 400-800 Hz.

Direct radiating bass bin designs of the highest quality drivers also appear to be able to supply this need while providing very low THD and very good impulse response, but at a cost in efficiency of the bass bin and the tendency of the bass bins to suffer from thermal compression distortion if used under high SPL conditions. I'm thinking that the Cornwall bass bin (and Cornscala design of Bob Crites & son) could really benefit from a much higher cost but higher quality direct radiator that might compete well with a Khorn or any of the direct-radiating bass bin Klipsch designs.

Edited August 18, 2014 by Chris A

[DrWho]

Btw, my original interest in the Othorn was its wide bandwidth - totally uncharacteristic of tapped horns. It is usable from 25Hz to 250Hz, which is an entire decade. That's better than the Khorn, and comparable to the Jubilee.

The Othorn violates the classic 2 octave limit of the tapped horn. The reason I built one was to see if there was still some "tapped horn artifact" still present that didn't show up in the measurements. As it turns out, the system is totally usable over the entire bandwidth and it doesn't sound like a tapped horn at all. I was actually planning to get rid of it after conducting the test, but I've never heard anything that compares to it, so I've kept it.

I have a pair of fancy 12" drivers destined for a tractrix'ish bass bin (I want to cover 80Hz to 1kHz)....once that happens I'll be sure to take a ton of THD/IMD measurements. It only takes a few minutes to whip up a vented enclosure to compare against too.

[Chris A]

Two things got my attention on the Othorn from the online database: its impulse response and harmonic distortion curves. I immediately knew it was a TH design.

However, what you're saying is that it doesn't sound like a tapped horn? Interesting.

I'm in process of measuring my subs for harmonic distortion using two-tone signals of varying frequencies which is now available with the V5 Beta 21 version of REW (released in May and since updated to V5 Beta 22 two days ago to fix a bug in the new IMD area).

I've learned recently that somewhere in my room or signal train that I've got a constant 90 Hz drone in my measurements, so I'm in process of hunting that down. I bet it's actually there acoustically, but I haven't isolated it yet. Since I'm taking measurements at lower SPL than 105 dB, etc., that drone and its harmonics affect the IMD measurements. [EDIT: it was the fan in my laptop computer... ]

Chris

Edited August 17, 2014 by Chris A