You're sure about this? All I could find was the coax/duplex debut in '47 whereas Altec Lansing's was in '43.
Could be Altec beat them to it but the company was founded way before Altec. Sadly, Altec no longer makes the speakers that made them famous. GPA makes them but they are outrageously expensive! I’ve used 604 and 605s and they have great mids but no low end in anything smaller than a refrigerator sized cabinet. They are crazy efficient but can only take 30 watts. Tannoy 3139is a modern take on the coincidental driver with a large x max and low Fs. You can get 40hz out of about a two cf box.
Right, true multi-cells don't pattern 'flip' like the Altec 511/811 series and their ~copycats, ditto the fake multi-cell Altec/GPA used for decades on their coaxes.
Hmm, you stated for HIFI/HT apps, so with 'pattern flip' only defining which way you need to orient the horn for the desired polar response, which is normally well defined horizontal, narrow vertical, so the normal wide horn width is horizontal and if you need a very wide horizontal, then rotate the horn (or coax) 90 deg for 'fan' mode over a wide arc.
Round coax horns OTOH are the theoretical ideal, but can require more room treatment if big ones are used.
All my info is in two burned up HDs and they want way too much to pick through them to retrieve whatever they can, if anything useful.
That said, found these docs:
http://adireaudio.com/Files/HE12-1Plans.pdf
http://adireaudio.com/Files/HE12-1Plans.pdf
Bummer! In my haste I copied the wrong links:
http://web.archive.org/web/20031203150146/http://adireaudio.com/Files/HE12-1Plans.pdf
http://web.archive.org/web/20050530005201/http://adireaudio.com/Files/HE12-1Plans.pdf
http://web.archive.org/web/20031203150146/http://adireaudio.com/Files/HE12-1Plans.pdf
http://web.archive.org/web/20050530005201/http://adireaudio.com/Files/HE12-1Plans.pdf
Directivity is linked to wavelength and physical size of the transducer/construct. Wavelength is physical size of sound. 1khz interacts with 34cm sized objects similarly as 2khz with 17cm sized objects, or 100Hz with 3.4m sized.
On a coaxial driver both the woofer and tweeter would have about the same directivity when wavelength is about and longer than size of the woofer. As wavelength gets shorter, frequency goes up, wavelength gets closer to physical features in size which have more effect to the pattern for both the woofer and tweeter, one to the other.
It is quite simple to imagine situation, two separate physical objects, even though merged into one and coincident, have very different properties as transducers and have very different immediate environments. Simplified, the other is very big ring and hole in the middle, and the other is the hole, or perhaps there is this separate waveguide depending on which pro coax driver one is looking at.
Anyway, its just the way things are, at crossover two physical objects need to emit same size wavelengths and its not possible to have them share physically same space and time. I think multiple entry horns have least impact of ways to each other, in other words it seems best approximation of a point source.
Message is, try make best out of it dont get stuck, both would probably work fine but result different radiating pattern and SPL capability for the whole system. I suggest think which kind of directivity youd like and when things get too big and so on. Now you know what sized coax you need.
It is really not possible to provide a polar map for a mid driver since it all depends on the enclosure it is fitted in. And not necessary, because all cone drivers conform the the same basic rule: when the wavelength a driver is asked to produce is smaller than the circumference of said driver, it will start to beam in a predictable way. As soon as it is mounted on a baffle, the dimensions of the baffle come into play. You might wish to read up on baffle step.
So that's why this information is not provided in spec sheets.
Because the Tannoy is not a reference speaker. Thats the only reason for the NS10 - everyone uses it even if it is cr*p... 😉 btw - thats the definition of reference - widely used and accepted (and not "the best"...)
You have no use of a crappy speaker that is not a reference.
//
Ha ha! NS10, my most hated reference. For bad reasons of course!
Most peoples have no clues of what their use is in a studio: they serve has a 'light house' for free lances engineers which change rooms on a regular basis. Nothing more nothing less. Genelec 1030/31 was the end 90's early 00's equivalent in my part of the world.
Fredygump,
Why do you think either your w371 won't work? You documented your search, have an ( even partial) understanding of principle at work and had done a research about it*. It's even confirmed by M. Gunness ( the whole recipe for active cardio is depicted in one sentence in the paper linked previously.
It's more or less the same about your concerns with directivity matching of the 3 ways. It should works if implemented wisely. If you want full cardio until waveguide it's possible but will require prototype and experiment.
Here again you can find info here as a number of members already have done hard work! 😉 Tmuikku experimented recently and i suggest you search about member Keyser's thread ( and posts!). He is one of the brains behind D&D and documented his preliminary search ( let's take a look at 8c if they are unknown to you).
About the PA gear: a driver is a driver. It is designed with an application in mind. The bad rep PA gear have comes from low budget with questionable 'acoustic design' choices operated by not knowledgeable people.
In practice there is no difference between what is used for 'ultimate' quality PA or Quality Control in studio ( monitoring). Of course if you can afford, have the room ( or not! Let's look at 4345 in 12square meter room in Japan!) you can enjoy this at home too!
Some of the highest quality drivers you could find in the 70's and 80's were developed for PA, by a division of consumer grade mass products brand by a bunch of guys which one of them built the defacto 'standard' for studio monitoring/ hifi.
http://www.reyaudio.com/history-e.html
Iow, there might be difference between PA drivers and 'HIFI' ones but in the end not much more between two 'hifi' drivers... it depend of your application, needs and target.
About bsc:
https://www.trueaudio.com/st_diff1.htm
* could you provide links to all what you based your W371 search on? Pdf, link to interview, articles, etc,etc,... i'm lazy! 😉
Most peoples have no clues of what their use is in a studio: they serve has a 'light house' for free lances engineers which change rooms on a regular basis. Nothing more nothing less. Genelec 1030/31 was the end 90's early 00's equivalent in my part of the world.
Fredygump,
Why do you think either your w371 won't work? You documented your search, have an ( even partial) understanding of principle at work and had done a research about it*. It's even confirmed by M. Gunness ( the whole recipe for active cardio is depicted in one sentence in the paper linked previously.
It's more or less the same about your concerns with directivity matching of the 3 ways. It should works if implemented wisely. If you want full cardio until waveguide it's possible but will require prototype and experiment.
Here again you can find info here as a number of members already have done hard work! 😉 Tmuikku experimented recently and i suggest you search about member Keyser's thread ( and posts!). He is one of the brains behind D&D and documented his preliminary search ( let's take a look at 8c if they are unknown to you).
About the PA gear: a driver is a driver. It is designed with an application in mind. The bad rep PA gear have comes from low budget with questionable 'acoustic design' choices operated by not knowledgeable people.
In practice there is no difference between what is used for 'ultimate' quality PA or Quality Control in studio ( monitoring). Of course if you can afford, have the room ( or not! Let's look at 4345 in 12square meter room in Japan!) you can enjoy this at home too!
Some of the highest quality drivers you could find in the 70's and 80's were developed for PA, by a division of consumer grade mass products brand by a bunch of guys which one of them built the defacto 'standard' for studio monitoring/ hifi.
http://www.reyaudio.com/history-e.html
Iow, there might be difference between PA drivers and 'HIFI' ones but in the end not much more between two 'hifi' drivers... it depend of your application, needs and target.
About bsc:
https://www.trueaudio.com/st_diff1.htm
* could you provide links to all what you based your W371 search on? Pdf, link to interview, articles, etc,etc,... i'm lazy! 😉
Last edited:
Fredygump,
For the behavior in directivity of 12" you have an example with the Labrat 12 datasheet: you have polar map and overall dimension availlable, you know driver reference so you have access to sd: perform a sim of an 'ideal' driver and look if it corelate ( there will be difference as the membrane shape isn't considered as other factors influencing directivity in real life, but close enough).
The link about 12c362 being almost constant directivity is taken from an article in german where independant measurements were performed. There is a link in the thread from were picture is taken ( but you'll have to use archive.org) .
That said you can feel what is all about from other built, there might be differences in scales but the trends should be the same ( this message and the following one) :
https://www.diyaudio.com/community/...er-uniq-adventures.377221/page-9#post-6879588
For the behavior in directivity of 12" you have an example with the Labrat 12 datasheet: you have polar map and overall dimension availlable, you know driver reference so you have access to sd: perform a sim of an 'ideal' driver and look if it corelate ( there will be difference as the membrane shape isn't considered as other factors influencing directivity in real life, but close enough).
The link about 12c362 being almost constant directivity is taken from an article in german where independant measurements were performed. There is a link in the thread from were picture is taken ( but you'll have to use archive.org) .
That said you can feel what is all about from other built, there might be differences in scales but the trends should be the same ( this message and the following one) :
https://www.diyaudio.com/community/...er-uniq-adventures.377221/page-9#post-6879588
For my
To be precise, I do not know for sure if it will work in terms of cardiod bass when it is close to a wall. This article discusses the issues: https://www.prosoundweb.com/counterintuitive-cardioid-subwoofers-can-be-a-bit-claustrophobic/
This issue would be adequately addressed by instructing customers to place the cabinets a minimum distance from walls. I should see if I can find more documentation about the system to see if they specify this.
The info you shared about passive cardiod designs highlighted the importance of the same signal playing from the front and rear. They even said that the distortion level should be the same, so that the distortion cancels too. So if this is true, then applying different EQ to the 2 drivers would hurt the cardiod effect. But Genelec discussed doing this.
I am definitely at a point where I need to start doing my own tests! But I am still waiting for that DSP unit to arrive!
And even if placing near a wall is a problem, I can try to build an absorber specifically to address this.
To be precise, I do not know for sure if it will work in terms of cardiod bass when it is close to a wall. This article discusses the issues: https://www.prosoundweb.com/counterintuitive-cardioid-subwoofers-can-be-a-bit-claustrophobic/
This issue would be adequately addressed by instructing customers to place the cabinets a minimum distance from walls. I should see if I can find more documentation about the system to see if they specify this.
The info you shared about passive cardiod designs highlighted the importance of the same signal playing from the front and rear. They even said that the distortion level should be the same, so that the distortion cancels too. So if this is true, then applying different EQ to the 2 drivers would hurt the cardiod effect. But Genelec discussed doing this.
I am definitely at a point where I need to start doing my own tests! But I am still waiting for that DSP unit to arrive!
And even if placing near a wall is a problem, I can try to build an absorber specifically to address this.
To the question of the directivity maps I posted before, is it safe to say that the horn beams at the same frequency as the driver it is installed in? We talk about how the horn uses the woofer as the wave guide, but I was not sure to what degree this works.
But if so, then the pattern is omnidirectional until it approaches the driver's beaming frequency, and the horn is essentialy beaming at the same frequency?
I wondered if the HF pattern might deviate from the 12" driver pattern. But it seems it does not?
Then that would be the answer for D1sco!
But if so, then the pattern is omnidirectional until it approaches the driver's beaming frequency, and the horn is essentialy beaming at the same frequency?
I wondered if the HF pattern might deviate from the 12" driver pattern. But it seems it does not?
Then that would be the answer for D1sco!
This is what to expect from a coax which have been highly optimised ( it's a Kef, either sp1753 either the one used on ls50, i've lost track on Mr Stitcha's thread).
Anyway look at pictures and make them scroll, you'll see the three different freq of xover and the effect. The last picture is the coax mid only, it gives you hint that the 'beaming' is taking in charge by it only. In this ( small width) box the overall response is close to what you could expect of a good 12" ( except the xover point would be located at 1khz/1,5khz and this is why i'm excited by it beside it's cons ( low efficiency, xover in a range too close to 3,5khz,...).
Compare to data in labrat 12 datasheet ( polar map).
https://www.diyaudio.com/community/...r-uniq-adventures.377221/page-13#post-7012661
Anyway look at pictures and make them scroll, you'll see the three different freq of xover and the effect. The last picture is the coax mid only, it gives you hint that the 'beaming' is taking in charge by it only. In this ( small width) box the overall response is close to what you could expect of a good 12" ( except the xover point would be located at 1khz/1,5khz and this is why i'm excited by it beside it's cons ( low efficiency, xover in a range too close to 3,5khz,...).
Compare to data in labrat 12 datasheet ( polar map).
https://www.diyaudio.com/community/...r-uniq-adventures.377221/page-13#post-7012661
More 'food for thought', 604-8K w/Mantaray (CD) WG, polars: https://www.lansingheritage.org/html/altec/specs/components/604-8k.htm
Posted this a few times when relevant (attached) to help 'work through' combining co/tri-ax polars, whatever.
Posted this a few times when relevant (attached) to help 'work through' combining co/tri-ax polars, whatever.
Yanno, I've never actually seen an example or diagram of midbass woofers actually working as waveguides, so I never 100% put faith in any given driver to behave as it's sold. I definitely should be accounting for this, but good pointing it out. I'm talking about Constant Directivity, like you are, and my concerns that a wide coaxial design will likely have a few big directivity spikes but ultimately the same nulls before the crossover region. It's tough to keep the shape of the beam perfect in a wider design. As I'm sure you know, the added reflections color the sound in-room. Which definitely brings me to the next point...
@fredygump how are you going to measure this anyways? I think it's one of the reasons people ask why you're doing your own design vs. copying an existing, proven one. Without being able to measure directivity and model appropriately, it could take a long time to get everything perfect.
BTW, that beamwidth measurement seems to be just the tweeter. It seems to imply that the driver loses constant directivity around 2k with either model. I would question the woofer moreso in either case.
I wasn't too impressed when Adason said I should do a "proven" design, and you choose to ask the same question in the same words. This question just reveals your biases. It really makes me wonder why you want to participate in this conversation at all!
You are correct that the beam width measurement is only the horn, because only the horn is playing at 2khz and above. The interesting part is that even though the cone is not making any of this sound, the pattern narrows dramatically at the beaming frequency of the 6" cone, the same as would happen if the cone were generating the sound.
The only logical explanation I can find is that it doesn't matter if the sound is generated by the horn, or if it is generated by the cone itself.
You opened by saying, "I've never actually seen an example or a diagram of midbass woofers actually working as waveguides....", but I think you were looking at the example you say you haven't seen when you wrote that.
I found this polar map comparision of several well known speakers. I think it's great because it includes the Genelec 8040A, speakers I used to own for a few years. So it gives me a nice reference when trying to understanding what the polar maps mean in the real world. (imgur.com/F5rFf35)
This appears to be good news, because I thought the 8040A was quite good, but here it clearly has issues in the vertical. These are the issues that the coax does not have.
This appears to be good news, because I thought the 8040A was quite good, but here it clearly has issues in the vertical. These are the issues that the coax does not have.
And I missed these before, but on the website usspeaker.com includes polar maps for B&C drivers. It is interesting. I looked at several, and there is always a close similarity between the two. But not exact. The horn is slightly wider than the woofer beam on all of them. And it takes a few moments to examine them, because the range is different on both images. The top one is 500 to 5,000 hz, and the bottom is 1khz to 20khz. (This image is for the B & C 12CLX64)
I doubt it proves anything, but just more data.
I doubt it proves anything, but just more data.
Unfortunately I find myself a bit behind the curve on KEF designs! Once realized how good studio monitor speakers can be, I sold my amp and ignored passive speakers. I had seen KEF speakers, and I saw they were different than others. But I didn't even realize they were coaxial.
Their design is interesting. I just found a paper about the design of the Genelec 8300 speakers, and I instantly saw principles from the Genelec design in the KEF speaker. For example, the paper explains how there can be issues at the edge of the wave guide, so the entire front baffle should be the wave guide. And KEF is accomplishing this with the baffle being a continuous curved surface.
I will take your suggestion and read some of the KEF white papers as you suggested a couple times!
If you allow me Fredygump, don't be touchy please, D1sco's concerns are valid and don't be offended people's suggest a prooven design: there is always high risk of fails in designing a loudspeaker. It cost money, time,... can be frustrating. Nobody wants to end up with OCD like some audiophiles have.
Even the great Joseph D'Apollito failed on a design which was published.
Anyway,
D1sco ok your comment about narrowing makes more sense to me with this clarifications.
I think there is a misunderstanding: the constant directivity can only be achieved through the use of the waveguide itself.
The woofer parts cannot have a constant directivity by definition: we use it's beaming range to achieve directivity match with the waveguide which then can achieve a constant behavior over a wide range.
When 'constant directivity' is claimed for a reference it can only be valid for the waveguided part of the spectrum. But i've not met a lot of reference doing so.
The fact the Bms have a range which behave as if it had constant directivity is a nice discovering ( independent measure) and the brand don't advertise about it ( but in no way i think it is an accident/random).
I'm not this concerned about the membrane profile of coax. about the directivity behavior, more about the diffraction which can occurs.
I once asked Earl Geddes if an OS profile could serve as a basis for a coax woofer shape, the answer was negative for a number of reasons. That said in the same thread Earl explained he once tested a 10" or 12" B&C, and that the outcome was surprisingly good naked, without a box. Of course no bass. He didn't pursued the idea as there was low commercial potential.
About measurements, well it ask a bit of diy but a manual rotating table is not this difficult to build. There is example thread atm here ( Fluid is working on a motorised one driven by Arta for automatic measurement).
Issue with room: either outside measurement, gating technique or you have acces to a basketball hall/gym...
Anyway if you want to go past the pleasure of build something (achieve a set of verified performance) measurements are mandatory.