Good explanation Greg.
I would add a little bit.
The simulation programs that are available now do a few things in principle.
They take empirical research and use mathematical language to describe what has been measured. When the two meet up together and shakes hands and make friends it is an awesome experience. When the mathematical calculations and the real life empirical measurements consistently agree we have something worthy of trusting.
Hornresp and Unibox are the two programs that I have confidence in. Add in AkaBak I guess. I haven't used AkaBak in quite a few years. These programs are carefully constructed to spit out realistic simulations of what really happens in the interactions between drivers and the enclosures that we can design for them.
The second thing that a good simulator does is allow us to play the what if game. What if I change this or that? What will happen to the response? The group delay? Can I extract a little more efficiency when I modify this section or that? It allows many people to do work that very few people had the chops to accomplish not that long ago.
Simulation programs that are consistent in explaining what happens in real life applications are invaluable in shortening the design cycle of a complex product. And they are awesome at pushing the envelope of what can be accomplished.
Now I throw in an off topic pet peave of my own.
How is it that the proliferation of tapped tappered pipes being tossed around are called tapped horns?
Tom Danley re-popularised the concept. It is an old one. There are examples of identical enclosure setups that are almost 60 years old.
But Tom's work actually has greater output than an equally sized vented enclosure. Usually 5 to 7 db more than an equally sized vented enclosure.
Put that criteria in front of the so called tapped horns available and see how many, if any, meet such a comparison test.
I know that Greg has published designs that do this. And I have. Patrick Bateman has also posted at least one prototype that met that goal. Greater efficiency gain than a vented enclosure.
Call them what they are.
Transmission lines. A T-line most of the time slightly betters a vented box if it borrows from Paul Voight in combining a tapered section.
I would add a little bit.
The simulation programs that are available now do a few things in principle.
They take empirical research and use mathematical language to describe what has been measured. When the two meet up together and shakes hands and make friends it is an awesome experience. When the mathematical calculations and the real life empirical measurements consistently agree we have something worthy of trusting.
Hornresp and Unibox are the two programs that I have confidence in. Add in AkaBak I guess. I haven't used AkaBak in quite a few years. These programs are carefully constructed to spit out realistic simulations of what really happens in the interactions between drivers and the enclosures that we can design for them.
The second thing that a good simulator does is allow us to play the what if game. What if I change this or that? What will happen to the response? The group delay? Can I extract a little more efficiency when I modify this section or that? It allows many people to do work that very few people had the chops to accomplish not that long ago.
Simulation programs that are consistent in explaining what happens in real life applications are invaluable in shortening the design cycle of a complex product. And they are awesome at pushing the envelope of what can be accomplished.
Now I throw in an off topic pet peave of my own.
How is it that the proliferation of tapped tappered pipes being tossed around are called tapped horns?
Tom Danley re-popularised the concept. It is an old one. There are examples of identical enclosure setups that are almost 60 years old.
But Tom's work actually has greater output than an equally sized vented enclosure. Usually 5 to 7 db more than an equally sized vented enclosure.
Put that criteria in front of the so called tapped horns available and see how many, if any, meet such a comparison test.
I know that Greg has published designs that do this. And I have. Patrick Bateman has also posted at least one prototype that met that goal. Greater efficiency gain than a vented enclosure.
Call them what they are.
Transmission lines. A T-line most of the time slightly betters a vented box if it borrows from Paul Voight in combining a tapered section.
Well, not surprisingly you get an F in reading comprehension. I addressed the bolded part of your quote (and more) in post 89 when I said "What YOU choose to build should be a compromise considering a lot of different variables - in the case of small vs large sealed you should be looking at how much space you have available, response curves of the various size boxes (system q), how much power you have available and how much the driver can handle and a bunch of other stuff."
Your footnote makes mention of system q - it's just one of the variables and compromises. Like I said, excursion vs power handling is at least as important. If your system q is too high you might never be able to reach xmax before the driver coil melts and dsp can't help with that.
If you think you are talking about things that have not been considered by the general diyaudio population you are delusional, this is beginner stuff and it has been discussed at length.
As for the original question - if there are a bunch of simple simulators spitting out .707 qtc alignments for beginners, that particular .707 qtc is a good suggestion as GM noted. If you have no idea what driver is being used or how or why the person is using the sub .707 is a reasonable compromise between frequency response, power handling, size and all the other variables. But no one except the most inexperienced beginner should feel locked into a suggested default.
I think you guys ( Ben and Anthony ) need some quality time with a brew in hand and a good long chat. You would get along famously. I should know. I have had the pleasure of very pleasant conversation with the both of you. Both Ben and Anthony are very intelligent fellows with firm convictions.
Why for you cannot play nice?
Why for you cannot play nice?
This is categorically untrue, misleading at best. Post up a sim of any tapped horn that you think applies and I'll show you how it doesn't.
Now it is true that tapped horns don't suffer nearly as much from port compression as ported boxes, but that is also true of the transmission lines you claim are not tapped horns so I know that's not what you are getting at.
THs are a bit like 6th order BPs. You can define for efficiency, or low-frequency extension, or a bit of both. And of course there are trade-offs to consider, box size being one of them.
I've found myself settling on THs that provide about 3dB of increased efficiency in the passband and greater low frequency extension than vented systems, similar to what I used to do with my BP designs.
Hello Anthony.
The simple test I perform when I come across any tapped horn design is already stated.
Take an equal volume vented enclosure and test it's output against the so called tapped horn.
It's pretty simple. I have been saying this for quite a few years.
Agreed about port compression. But that again is assuming that people are still sticking to the concept of undersized ports.
A shelf vent properly sized takes care of that. There is lots of information on properly sizing ports if you look up the research papers.
I have said many times that I don't like the sound of the tapped horns. And I don't like the sound of a band pass enclosure either.
But that being said means I have heard them. And I have built them. They have a purpose. And in the reproduction of some types of music they are well received.
Where the reproduction of the original event that was recorded is the goal I stay far away from a tapped horn or a bandpass.
In the 80's I made quite a few one note wonders for the car audio crowd. Hated them then. And still do. But for kick drum a way larger than life they can't be beat.
The simple test I perform when I come across any tapped horn design is already stated.
Take an equal volume vented enclosure and test it's output against the so called tapped horn.
It's pretty simple. I have been saying this for quite a few years.
Agreed about port compression. But that again is assuming that people are still sticking to the concept of undersized ports.
A shelf vent properly sized takes care of that. There is lots of information on properly sizing ports if you look up the research papers.
I have said many times that I don't like the sound of the tapped horns. And I don't like the sound of a band pass enclosure either.
But that being said means I have heard them. And I have built them. They have a purpose. And in the reproduction of some types of music they are well received.
Where the reproduction of the original event that was recorded is the goal I stay far away from a tapped horn or a bandpass.
In the 80's I made quite a few one note wonders for the car audio crowd. Hated them then. And still do. But for kick drum a way larger than life they can't be beat.
Last edited:
Hi there Brian.
You have hit on pretty much what I have discovered. Most tapered tapped horns posted gain about three db.
If you have access to one of Earl Geddes papers that he did with Mark Gander on band pass enclosures you will find that the maximum gain in a bandpass enclosure is 7 db. The narrower the frequency range generated ( passband ) the greater the possible gain.
It is a similar thing with a proper high gain tapped horn.
Knowing how to choose the driver is a very important thing to start with.
I see way to many that have weak motors and are used anyway.
On the more prosound threads there are some really good examples of how to choose a driver and develop an enclosure.
You have hit on pretty much what I have discovered. Most tapered tapped horns posted gain about three db.
If you have access to one of Earl Geddes papers that he did with Mark Gander on band pass enclosures you will find that the maximum gain in a bandpass enclosure is 7 db. The narrower the frequency range generated ( passband ) the greater the possible gain.
It is a similar thing with a proper high gain tapped horn.
Knowing how to choose the driver is a very important thing to start with.
I see way to many that have weak motors and are used anyway.
On the more prosound threads there are some really good examples of how to choose a driver and develop an enclosure.
This is why I wanted you to post a sim.
If you look at Danley's designs they are all WAY larger than a "max flat" ported design with the same driver(s). And that's where most of the extra sensitivity comes from. Usually tapped horns with high sensitivity compared to ported boxes are in the neighbourhood of 4x larger than a "max flat" ported box tuned to the same frequency.
You are very correct in that a lot of tapped horns you see have only 3 db more sensitivity than a ported box. But that's the designer's choice - they actively chose small size over large sensitivity gain. If size is no object it's fairly easy to get 30 db more sensitivity. Leach's ideal horn math produces horns with about that much extra sensitivity.
Another reason I wanted you to post a sim is to show that the tapped horn's sensitivity gain only matters at a fixed power level - for example 1w comparisons. The sensitivity gain also comes with higher excursion so if you compare at xmax instead of at 1w the difference isn't much, in some cases the ported box will actually win even if the tapped horn is more sensitive.
I'm aware of papers and resources to properly size ports. A shelf port isn't that different than a round port wrt velocity issues and port compression, it's the cross sectional area of the port wrt velocity concerns that matters. If the port is properly sized to allow no more than 10 m/s at xmax there will be little port compression - regardless of whether the port is a shelf port or round port. But as we all know, ported boxes are almost never designed to have 10 m/s velocity at xmax. So the tapped horn always wins even if a sim says there is no difference in max spl.
If you look at Danley's designs they are all WAY larger than a "max flat" ported design with the same driver(s). And that's where most of the extra sensitivity comes from. Usually tapped horns with high sensitivity compared to ported boxes are in the neighbourhood of 4x larger than a "max flat" ported box tuned to the same frequency.
You are very correct in that a lot of tapped horns you see have only 3 db more sensitivity than a ported box. But that's the designer's choice - they actively chose small size over large sensitivity gain. If size is no object it's fairly easy to get 30 db more sensitivity. Leach's ideal horn math produces horns with about that much extra sensitivity.
Another reason I wanted you to post a sim is to show that the tapped horn's sensitivity gain only matters at a fixed power level - for example 1w comparisons. The sensitivity gain also comes with higher excursion so if you compare at xmax instead of at 1w the difference isn't much, in some cases the ported box will actually win even if the tapped horn is more sensitive.
I'm aware of papers and resources to properly size ports. A shelf port isn't that different than a round port wrt velocity issues and port compression, it's the cross sectional area of the port wrt velocity concerns that matters. If the port is properly sized to allow no more than 10 m/s at xmax there will be little port compression - regardless of whether the port is a shelf port or round port. But as we all know, ported boxes are almost never designed to have 10 m/s velocity at xmax. So the tapped horn always wins even if a sim says there is no difference in max spl.
Last edited:
I have a client that has asked me to design a serious tapped horn. But that my friend is not getting posted.
If you look over on AVS forum I did a tapped horn posted publicly that had 7db of gain.
Greg did there as well.
And Greg has done a few more.
Patrick has one that I can remember using the MCM 8 inch driver if I remember correctly. I think I razzed him about being a capable wood butcher.
I'll see if I can find the tapped horn design I did. I lost a whole mess of data around that time. One of those viruses that takes out everything. Every start up gave me a screen of nothing but good old hash. Like you used to see when there was no TV station being broadcast.
Yes 24 hour TV has not always been the norm.
If you look over on AVS forum I did a tapped horn posted publicly that had 7db of gain.
Greg did there as well.
And Greg has done a few more.
Patrick has one that I can remember using the MCM 8 inch driver if I remember correctly. I think I razzed him about being a capable wood butcher.
I'll see if I can find the tapped horn design I did. I lost a whole mess of data around that time. One of those viruses that takes out everything. Every start up gave me a screen of nothing but good old hash. Like you used to see when there was no TV station being broadcast.
Yes 24 hour TV has not always been the norm.
Excellent illustration and from "the horse's mouth" no less.
But I suspect my conclusion in looking at those swell pictures is rather different than David's. I'd say anybody knows the Q=1.4 will sound the best (assuming the bump is deep low enough). Or maybe Q=1.
Something that looks flat to your eyeball isn't necessarily something that sounds right as a subwoofer. And that is a message for the sim-lovers.
Anybody want to dispute that?
There's no magic (or even reasonable compromise) in ".7". It is just a mathematically convenient number. (Hint: it has something to do with the square-root of 2.)
Ben
Last edited:
Positive flare = tapped horn.
Straight flare = tapped quarter wave pipe.
Negative flare = tapped tapered quarter wave tube.
All 3 enclosures are model in Hornresp with the the TH function. That's probably where the generalization of the term Tapped Horn is coming from.
Any simulation is only as good as it's practical application.
Add to that the knowledge of and the ability of the person making the application.
Take Davids example of high Q.
Perfect for outside use where you have a great deal of sound pressure level losses.
Would sound rather boomy inside though.
Respectfully, a couple of dbs extra south of maybe 40 Hz (and we are talking about good subs here) would typically sound real nice, not boomy and not sound one-note'ish.
Ben
You quoted me saying ".707 is a reasonable compromise between frequency response, power handling, size and all the other variables."
I was talking about qtc so I'm not sure why you showed a qts comparison with that quote.
However this does illustrate a good point - contrary to what Ben thinks there is no "whomp up" at fs unless driver qts is quite high - a value of 1 or higher. It was in the church organ thread where Ben went on about the "whomp up" and would not believe my illustrations of qts effects, would not believe weltersys when he said there was no "whomp up" and where he asked for a decisive ruling on the matter despite plenty of evidence.
The more you type the more you reveal you don't know what you are talking about. This shows a comparison of driver qts, not system q aka qtc. You can tell by the way the midband efficiency (the higher frequencies) drops with higher q. Oh, and the graphs are also labelled qts.
Yes I'd like to dispute that but it's hard to know where to start, there's so many problems with this.
First, there is no driver I am aware of (probably no driver in the world) that will have a qts of 1.4 combined with flat response down to fs combined with an fs low enough to put that bump "deep low enough" as you say. If the driver did exist, the sensitivity would be horribly low and since high qts drivers are almost always very cheaply made it wouldn't have enough power handling to overcome the low sensitivity and provide any kind of decent output.
Second, a driver that is flat down to fs and having a bump at fs like the 1.4 qts graph will sound absolutely terrible in room as Kravchenko pointed out. It will boom terribly, sound "slow" and have all the characteristics that you attribute to resonant enclosures like tapped horns.
There's no problem having a bump at low frequencies, in fact most people will prefer that type of house curve, but having a flat 2 pi response all the way down to the bump will sound terrible in room.
Again, you couldn't be more wrong. The reasonable compromise in a .707 qtc is exactly as GM explains. It's a "max flat" alignment, providing the lowest possible f3, a reasonable enclosure size, a reasonable power vs excursion level and reasonable compromises in all other areas. This means the .7 qtc enclosure could be expected to work reasonably well in a variety of different environments for a variety of different purposes.
On the other hand, if you have a very high qtc (like 1.4) the box will be incredibly small but f3 will be very high and with very low qtc (like .5) the low frequency efficiency gets better but box size gets way out of hand quickly so the potential usefullness of the alignment is limited to applications that can handle a very large box size.
You might be right Ben.
My thinking is that in a corner, you no need any more reinforcement of the midbass.
And in a corner is where the vast majority of subs are placed.
If it would only be around 40 hertz I agree that I would enjoy a little lift in some music.
Just for fun, here's a sim of a high qts driver in a sealed box. I chose the Eminence Alpha 15A because it's a popular driver, it's a high quality driver (there aren't many to chose from in this qts range) and it has a decent sensitivity rating. Unfortunately xmax is very low (3.8 mm) but that's what you get when you combine high qts, high quality and high sensitivity.
Two alignments are shown, both sealed boxes. The blue line is a 200 liter box with qtc close to 2 and the red line is a 1000 liter box with qtc close to 1.6. As you can see you need a MASSIVE box to make the high q bump low enough in frequency to be near fs. Even so, with 41 hz fs the bump is still too high in frequency to be really useful. Ideally fs would be lower and qts would be higher so the bump would be narrower and more pronounced and centered at a lower frequency. But good luck finding a quality driver with those parameters.
Max spl at xmax is limited to about 105 db at the top of the passband. This type of result is possible with a 6 inch driver in 1 - 2 cubic feet, so this high qts driver system is overwhelmingly mediocre at best.
Maybe this will show that the type of system you seem to prefer doesn't actually exist because there's no available driver that can pull it off and the required box would be so large it wouldn't fit through a doorway.
Two alignments are shown, both sealed boxes. The blue line is a 200 liter box with qtc close to 2 and the red line is a 1000 liter box with qtc close to 1.6. As you can see you need a MASSIVE box to make the high q bump low enough in frequency to be near fs. Even so, with 41 hz fs the bump is still too high in frequency to be really useful. Ideally fs would be lower and qts would be higher so the bump would be narrower and more pronounced and centered at a lower frequency. But good luck finding a quality driver with those parameters.
Max spl at xmax is limited to about 105 db at the top of the passband. This type of result is possible with a 6 inch driver in 1 - 2 cubic feet, so this high qts driver system is overwhelmingly mediocre at best.
Maybe this will show that the type of system you seem to prefer doesn't actually exist because there's no available driver that can pull it off and the required box would be so large it wouldn't fit through a doorway.
An externally hosted image should be here but it was not working when we last tested it.
Wait a minute there Master Anthony.
Drivers you may know about yes. There are no drivers.
But in times past. This idea of a very high Qtc was not that rare. Look up Acoustic Research. Henry Kloss was a pretty smart guy. And he used his smarts to capitalize good sound out of a smallish enclosure in exactly the manner described by Ben.
Not every thing is new and improved. Some things have in fact regressed over the years.
Everybody likes a bit of a bump at low frequencies, I don't think that's in question. Often the room gain alone will provide that with "regular" qtc systems. BUT the bump has to be centered as low as the lowest notes in the music or it sounds stupid. If you have notes that are lower than the bump the music sound bloated in the higher notes and weak in the lower notes.
The problem is that it's nearly impossible to build this bump into a sealed box response as I just showed. The fs has to be very low to get the bump low enough to be useful and the qts has to be very high to make the bump narrow instead of a broad hump through the whole passband. And the box will inevitably be huge. There are no drivers with suitable parameters (including high enough xmax and power handling) to accomplish this with a bump at 30 hz or lower, which is where the low notes in MY music are.
- Status
- Not open for further replies.