Rerurn to Romy the Cat's Site

Horn-Loaded Speakers
Topic: MF trick

Page 1 of 1 (18 items)


Posted by koh123 on 05-20-2021

I am going to show two different simulations for a 105Hz tractrix horn with the same 8”driver (B&C 8PE21).  The parameter values for these two simulations are quite different.   However the simulations predict the same great flat freq response within the range I am interested in (150-800Hz).  The diaphragm displacements are different as to be expected from s1. The first simulations values are close to InlowSound.com 135Hz mid-bass.  And the second simulations values are close to Aries Cerat model Contendo Reference.  I am at a loss as to how different the two sets of values are but the freq results are close to the same.  I have the room to build either one.  Is there something I am missing?  Phase? Time alignment? Which is actually the better one and why?

Inlow Sound 135Hz mid-bass


    

Aries Cerat model Contendo Reference>>


    

Posted by rowuk on 05-22-2021
Hornresp only calculates pressure not tone. That means you can have perfectly flat response and not be able to integrate it with anything.In addition, no single channel can be viewed out of context. They will both change depending on what is below and above and the efforts that the owner went through to get the result..There is no such thing as "better" or "worse". There are only people that "think" and those that "try stuff out". If you research this site, you will see a basic maxim: Random process means random result.

Posted by Romy the Cat on 05-22-2021
 koh123 wrote:

I am going to show two different simulations for a 105Hz tractrix horn with the same 8”driver (B&C 8PE21).  The parameter values for these two simulations are quite different.   However the simulations predict the same great flat freq response within the range I am interested in (150-800Hz).  The diaphragm displacements are different as to be expected from s1. The first simulations values are close to InlowSound.com 135Hz mid-bass.  And the second simulations values are close to Aries Cerat model Contendo Reference.  I am at a loss as to how different the two sets of values are but the freq results are close to the same.  I have the room to build either one.  Is there something I am missing?  Phase? Time alignment? Which is actually the better one and why?

Koh, I tend do not answer this type of questions.  Would it be any difference between Inlow and Cerat? Yes, probably it would. Would one of them be “better” this is a very loaded questions and since it was asked in nuance-less way then I would ignore you as it has no meaning. BTW, if you are looking at 105Hz tractrix horn then you might also consider slower opening profiles. The 100Hz is the horn rate range what the thins begun to be auditable…

Posted by Romy the Cat on 05-23-2021

I was cooking a dinner today for my kids and they asked me if any taste difference between ice-tea in Donkin and Starbucks. In my head popup an association with this thread.  The same with the horns. I do not know the difference, but it certainly exists. Now, an interesting question: as we identified the difference then will the difference be manifest itself in the final sound? Let pretend that it will and let pretend that the author of the system speaks perfect audio language and was able to build a complete installation just around the difference between those two types of the horns. What this type of the system would say about design objectives? When we build our playbacks, we are trying to manufacture sound and to navigate our sensation as the result of our exposure to the sound. We do not build all playback as topological consequences. I admit, we could but in my view, it is a dead end, guess how I know…
 
So let's reformulate the question thus those two versions of tratrix horns have any psycho acoustic differences is that worth to capitalize. Or even let me too complicated, if those differences exist then how to make those difference transparent in the context of entire installation. I very much not trying to do the things more complicated but to answering this question in practical terms, considering everything I said above, I feel that it should not be differences worth any practical attention. while we designing our installation with hundred cycles horn, we are dealing with so many other aspects that actually have many tangible impact to sound that if a person care about versus of two tratrix then he have addressed everything else. it is my sincerely believe that if person have addressed anything else then he would not care about tratrix differences.

Posted by Jorge on 05-27-2021
We have two models with different upper bass horns, one is indeed a 140hz horn (77 cm diameter) and the original, bigger horn at 110 hz (92 cm diameter) Both tactrix both made from the same materials (MDF) and finished in the same automotive finish.
The difference in sound is substantial, but also de difference in size, though it doesnt seem like much, once you check out the diameter and see them side by side you can see how different they are.  The 140 hz horn can play in a hotel room and still have an acceptable sound, the 110 hz is too big for that.  The minimal room size for a 140 hz horn I would say is between 13-14 ft  X 20 ft,  The 110 hz at least 18 ft on the smaller side, though Romy made some magic in his small apartment system in downtown Boston.
With a Fane Studio 8M driver on both, the measured sound differences are not huge, the 140 hz coupled to the floor will go down, room dependant, to around 115 hz, while the 110hz coupled to the floor will go down to under 100hz, again measured response.
The sound of the 140 hz horn is more open on top with a very natural midrange, it will blend very easily with almost any driver on top,  imagine it as the best high end system you have ever heard, but still a playback.
The sound of the 110 hz horn is closer to a live presentation, it is more dynamic, the response of the horn seems signifcantly lower to the point where it becomes very listenable even without a channel under it making the bass.  It relies more on the following driver to make a perfect midrange and integration takes more care, it seems effortless on the bass and more forgiving to the bass channel, since it is already doing most of the hard work.
A final system with the 110 hz horn will sound better, more lively and effortless, the system with the 140 hz is still fantastic and still alive but size does matter, and where you will really notice the differences will be when you play a big orchestra and get better definition of instruments in that range.
Both systems can play with the same bass modules we designed based on a tapped horn with response optimized for 110 hz horn, playing almost flat down to 25 hz.  They blend in very well with both front horns, but the bigger dynamic response of the bigger upper bass horn is still noticeable.  Maybe if we redesign the bass modules optimized for the 140 hz horn the differences will become smaller, but I am not be too confident this will happen, so much I have not gone forward with such redesign.
(PS: keep in mind The throat on the 140 hz horn is 5" and the one on the 110 hz horn is 4")

Posted by koh123 on 06-13-2021
Jorge,
Thanks for that information.  I am just about over COVID-19.  It really wasn't too bad, but it did stop me from doing just about everything.   I do have some follow up questions but I am still not thinking too straight yet.   I hope to be posting again in a couple of days.   Thanks so far  -KEN

Posted by koh123 on 07-02-2021
     My COVID-19 brain fog has finally lifted. I have been using HORNRESP as the gold standard for horn design.  Upon reflection I forgot that we ultimately listen to the sound produced and not the beautiful computer simulated SPL profile.  So I went back to old school thinking and re-generated the mid-bass tractrix horn with the mouth circumference being one wavelength in size at the cut off frequency.  The resulting SPL profile was still the same as before.  Now that's the SPL profile, not the actual sound.  I'm now thinking that HORNRESP is great for designing horns that have to be smaller than optimal size.  However I have a dedicated listening room large enough for a mid-bass horn of your proportions so why not (18x30 with a 12 ft ceiling)           My bass horn is a 8x5x2 ft spiral exponential horn(InlowSound.com) which yields an effective length of 17 feet.          So I am now rethinking the dimensions for my midrange tractrix horn.  I am now thinking of using the same logic used for the mid-bass horn. That is, 900Hz at cut-off yields a horn mouth diameter of 5" and the JBL 2446 yields a throat diameter of 2" and then use HORNRESP calculate the rest.  What are your thoughts on this?   What are you using above your 110Hz mid-bass horn?

Posted by rowuk on 07-02-2021
My thought is that you should ask John. He has the comparison, time and time over again.Horn response calculates pressure, but that has nothing to do with sound quality. Flooding a too small horn with too low frequencies is the sign of a beginner to horns, or someone that is deaf.Remember that you really do want physical alignment. That makes mixing exponential and tractrix horns not so easy.

Posted by koh123 on 07-02-2021
John Inlow uses exponential and tractrix profile horns in his current system.  So I don't know what you are trying to say negatively when it comes to "mixing" exponential and tractrix horns.  Second,  after reading Jorge from Austin TX response I was rethinking my mid-bass horn size.  I am thinking about using a larger than John Inlows sized mid-bass horn.  So once again I don't know what you are trying to say when you say "Flooding a too small horn with too low frequencies is the sign of a beginner to horns, or someone that is deaf"  Are you saying that John is using too small a horn and is therefore deaf?  If you don't have anything positive to add or feel that insulting someone who is trying to get advice is a good thing please stay away from this thread.

Posted by Paul S on 07-03-2021
The thing about horns is, "developing" them consumes a lot in terms of time and materials, and they are expensive to materialize. SPL modeling programs can be great for getting in the ballpark, but YMMV when it comes to "level of demands" from a hi-fi system, or when to stop "development" of a horn or the system it is part of. Plenty of people seem happy to "model" a speaker system and then build accordingly, then they listen to Techno with pleasure. This site has up to now been geared to classical Music, and I think there is a general notion here that performances vary in terms of quality, and that speakers that make this plain are OK, aesthetic disappointments notwithstanding. Drivers and horns that are "swamped" have a hard time rendering "serious" music credibly. Who decides what's "swamped", or what's ultimately off kilter? You do.


Best regards,Paul S

Posted by rowuk on 07-04-2021
Koh,if time alignment is critical (in my world it is), the sources of the sound (the driver diaphragms) need to be aligned. A very long exponential horn and a short tractrix are not „impossible“ to align, but what reflective surfaces ftom the exp. horn mouth are in the way for what comes out of the tractrix horn?Simply draw a front and side view of your potential installation and think about soundwaves, the coverage of each horn and integration of all the pieces - sonic and cosmetic. Look at some successful horns like Cessaro, Kornhent, or Romys system well documented here. Think about why the designers did what they did.If you take the time to think about what I posted, you will not be insulted. If this is one of those „pride of ownership“ threads, I am the least of your problems. No one here will spoon feed anyone as it does not lead to any proper results.My recommendation to ask John Inlow stands. You can ask him why he does what he does and he will be able to tell you - with no BS.Horns are decade devices and trying to extend LF for a given horn results in honk - sometimes you can deal with it, but often not. John can tell you about this too. John knows me as „Robin“ so you don‘t have to hide that fact if you do communicate.

Posted by Romy the Cat on 07-04-2021
 koh123 wrote:
Upon reflection I forgot that we ultimately listen to the sound produced and not the beautiful computer simulated SPL profile.  So I went back to old school thinking and re-generated the mid-bass tractrix horn with the mouth circumference being one wavelength in size at the cut off frequency.  The resulting SPL profile was still the same as before.  Now that's the SPL profile, not the actual sound. 

Yes, it is correct. The horns generally and sub 100Hz horn in particular reposing very strange to SPL profile. You might perfectly have two horns with identical RTA scans but with very different sounds. I can give you a laundry list of parameters that HORNRESP or any other prediction some would not even consider. As far as I concern the HORNRESP help you to get only your lower cut off, so any simplistic spreadsheet that compare mouth diameter, throat diameter and length of the horn that would encode the horn profile. With proper implementation any profile and any horn will work, the key is to make all the rest properly. I generally advocate idea horns with 35-40” mouth diameter as it gives some flexibility for architect ring your hone stock for sitting listening position. Still, I feel that energetic characteristics of horn loading( amplification, drivers, electrical damping via amp and mechanical damping via back chamber) has even more prominences then pure SPL response.

Posted by koh123 on 07-05-2021
     If there is one thing that I have learned in life, and heard many times,  is that making a mistake is not a bad thing if you learn from it. In addition the lesson learned is just as valuable if you learned it from someone elses mistakes.  And the beauty of learning from someone elses mistakes is that you didn't have to put any work into making that mistake. That is why I am asking these questions on this website. 
 
    I realize that making these horns is going to consume a lot of time and materials.  I already know what I want, a 4-way horn loaded active XO system.  I know were I want my crossover points to be (2-3 octaves each horn).  I know that bass horns should be hyperbolic or exponential, midbass tractrix, midrange tractrix, tweeters exponential. I'm aware of the problems with the physical arrangement of horns and time alignment and therefore are going to use DSP. The mistake I did not want to make was picking the wrong dimensions for the midbass horn.  
 
      According to John Inlow his 4-sided 19" diameter horn he states "Drum snaps are crisp, but slap bass guitar is especially striking. There's an accentuation, a thrust if you will. I couldn't wipe the grin of my face if I tried. Effects like lightening (hard to get a real slap of thunder in older recordings) are much more believable" 
  
     And according to Jorge in this thread his round 36" diameter horn he states "sounds better, more lively and effortless, the system with the 140 hz is still fantastic and still alive but size does matter, and where you will really notice the differences will be when you play a big orchestra and get better definition of instruments in that range" 

    Well that brings up a red flag.  Both horns get great reviews from their builders and yet one is 1/2 the size of the other.  This is were I started this thread. 
  
     Romy has weight in with 36" diameter also, which seems reasonable because it is in the neighborhood of one wavelength in circumference at cut-off frequency.  I am now happy that I will not make that mistake.  Lets now talk midrange horn for a moment.  I am going to follow the JBL 2446 tractrix midrange with a EV T350 (1" compression driver) at about 4-5K XO what do you generally advocate the midrange mouth diameter to be?

Posted by Paul S on 07-05-2021
Are you saying you plan full range DSP? Don't think there would be much support for that here. The general idea for all drivers is to give them a light, "optimal" work load, both in terms of frequency response and "SPL". You want "headroom", and you want all your horns to stay free of self-induced problems, which include self noise and blocking other horns/drivers. It's a lot to think about and keep in mind as you develop each channel. Do you really think FR DSP will provide a shortcut? Again, I am well aware that this is how it usually done these days. And if you are pleased with results from this approach then you should be getting tips from those who have succeeded (in your own terms) at what it is you wish to achieve.


Best regards,
Paul S

Posted by rowuk on 07-05-2021
A DSP is a great tool for modelling crossovers. It lets us experiment with a lot of things with no additional costs. There is a sonic hit, but far lower than "guessing" at crossovers.

What seems to get the most traction here is using inherently well behaved drivers well within their comfort zone, thus not having to compensate with steep slopes or narrow filters.

Guessing at a mouth size for any horn depends a lot on the crossover frequency that you end up selecting - preferably after living with your mid/upperbass for a while (without anything on top). I would not even think about a HF solution until you are happy with the mid/upperbass and midrange. I ended up crossing over around 8K.
Of course we are all free to pile horns up as high as we want and then DSP them into our preferences, then there is not much to talk about...

Posted by Romy the Cat on 07-06-2021
 koh123 wrote:

     If there is one thing that I have learned in life, and heard many times,  is that making a mistake is not a bad thing if you learn from it. In addition the lesson learned is just as valuable if you learned it from someone elses mistakes.  And the beauty of learning from someone elses mistakes is that you didn't have to put any work into making that mistake.
I few things about mistakes. Also, I would generally agree with what you say but in context of horn audio very few people have luxury or perseverance to define things as mistakes. Typically we do what we feel it right and then live with consequences, defining it as success. Many of our project are one in a life and we in most cases have no options to take alternative way.   
 koh123 wrote:

I already know what I want, a 4-way horn loaded active XO system. 
You know that the number of horns is a very similar to the size of your boat. You always feel that if you have 1-2 feet more then you would be happier… 
 koh123 wrote:
I know were I want my crossover points to be (2-3 octaves each horn).  I know that bass horns should be hyperbolic or exponential, midbass tractrix, midrange tractrix, tweeters exponential. I'm aware of the problems with the physical arrangement of horns and time alignment and therefore are going to use DSP. The mistake I did not want to make was picking the wrong dimensions for the midbass horn.  
Generally, it is not accurate objectives as if you combine physical arrangement with DSP then you juts confused. If you do use DSP then you should not be worry about physical arrangement at all and your entire design objectives would be very different.
 koh123 wrote:
 According to John Inlow his 4-sided 19" diameter horn he states "Drum snaps are crisp, but slap bass guitar is especially striking. There's an accentuation, a thrust if you will. I couldn't wipe the grin of my face if I tried. Effects like lightening (hard to get a real slap of thunder in older recordings) are much more believable" 
I do not know what John you are talking about. The rate of MF hone mostly derives from the lower cut of off your MF driver and you desire to drive it too much atop or too much at bottom. I prefers to be at the save region of the bottom but it would depend from many many different factors.
 koh123 wrote:
 And according to Jorge in this thread his round 36" diameter horn he states "sounds better, more lively and effortless, the system with the 140 hz is still fantastic and still alive but size does matter, and where you will really notice the differences will be when you play a big orchestra and get better definition of instruments in that range" 
Jorge’s comment if not definitive, might be true only in context of his system as there is a lot of depends. Yes, rise matters but it will be all depends how you implement the midbass channel. If the top of your midbass is better, then bottom of you upperbas hone then all bets are off. The 36" is the elevation of your ears what you seat, plass-minus. You cam make 46” upperbass horn then MF will be firing to you from above in nearfield.
 koh123 wrote:
 Lets now talk midrange horn for a moment.  I am going to follow the JBL 2446 tractrix midrange with a EV T350 (1" compression driver) at about 4-5K XO what do you generally advocate the midrange mouth diameter to be?

Sorry, I do not know 2446 driver. The T350 is very interesting beast and you can get from it any sound you want, it all depends how you use it and what you want from it. To know it you need to know what your MF driver does atop of own range…

Posted by Romy the Cat on 07-06-2021
 rowuk wrote:
I would not even think about a HF solution until you are happy with the mid/upperbass and midrange.

That was very loaded statement but how a person with not a lot of listening experience to become happy with mid/upper bass and midrange. How become happy with o LF and tweeter. There is a solution: substitute quality with quantity.  It became discreet and literally countable. So, the objective is to make you MF to be able to demonstrate the max variety of differences, the differences of everything, tonal, dyadic, contrast, character, mood, and many others. The wider expressive performance of MF will be then it will be more properly designed. From certain perspective you should not listen how MF sound but rather to recognize what it does.

Posted by Paul S on 07-06-2021
Put the "differences" idea in the bank. We want "insight" and "involvement"; a "connection", as it were; not just the sound, per se, but the performances, and the performance as a whole. Also consider that Time is a factor; we may "get tired of" a given "perspective" over time. It's nice to have a system that is not stuck in one perspective. It is possible to "tune" a "good"  system by tweaking any number of "variable" points. Not at all sure how (or if) any of this works with DSP.


Best regards,
Paul S

Page 1 of 1 (18 items)