Not sure if I still have any of those plastic couple of litre enclosures any more but it wouldn't be hard to knock something up on the CNC to test it out.
Not sure I really like the design of that tweeter, while it looks kind of space aged, I would be concerned about cabinet refraction.
If we are going to be outside the box, might as well be right out of the box... Looks a little Mayan.
Last edited:
Foam pads seem to be just for this, they've also noticed the vertical gets so narrow they had to do something about it. The foam pads probably help as they are implemented so maybe it is not too severe. They would shade shorter wavelengths more than longer ones making the ribbon acoustically shorter at high frequencies. I would still make some quick test fixture with two tweeters, at different heights and try figure out if it matters subjectively. You could hold them on your hands in front of you and play with tilt and height and listen if its of any concern. At least factor in ability to tilt the center speaker some, simple math to figure out how much is needed to get "the beam" at ear height at listening position. Anyway, if system costs >10k then I'd would want it to be very comfortable to listen to so problems with sweetspot feel unacceptable in that sense and worth checking out before committing.
Huge screens of today are real problem for center channel audio I think. If I remember correctly recent interview in Erin's Audio Corner there was mention that its difficult problem to get the center sound to proper height, although the conclusion was picture helps immensely and tricks brain to hear the sound coming from the talking mouth even though the speaker is much below or above physically.
Last edited:
To me the Wilson system looks to be made by marketing folks 😀 Anyway, the kimbal thingy is kind of ability to "tilt back" the beam if needed. point to ears. The tweeter is probably ~1" in vertical (as well as horizontal) dimension so the vertical beams shorter than ~1" wavelengths, meaning that similar problems with beaming as with the long ~5" ribbon happen roughly at 5 times higher frequency and not as much of a problem.
Last edited:
Edge diffraction is serious here, the ribbon is only 14mm wide so it does not beam on horizontal axis, like at all. This means the sound radiates to baffle edge up to very high frequency, basically all the way from baffle width to tweeter width wavelengths. 50cm wide box with 1.4cm tweeter (and any other drivers radiating omni) on it would diffract all the way from ~700Hz to ~20kHz.
I would try to avoid the wide design altogether and would instead make as slim as possible structure for the ribbon and mid with as big round overs as feels comfortable aesthetically and to manufacture, woofers somewhere else than on the front beside. Rule of thumb is to minimize flat baffle area around the transducer. Put the woofers on the sides of the box for example if depth is not problem, if they don't fit stacked vertically on the front. All these might make crossover tougher to make though. Anyway, it would be worth it to address diffraction if at all possible. Any round over on any box benefits.
If I had the tweeters and small mids I would probably make identical tweeter + mid boxes for all the speakers and then varying sized separate bass boxes for them. The tweeter mid box would be just minimal box with >1" radius roundovers (on the horizontal axis, vertical edges) starting immediately outside the drivers frame. Stack em to some bass boxes. On the center the drivers could be just side by side on flat wide box below this midhigh box.
Last edited:
While this is one option you have, a flat baffle isn't causing a problem by itself.
Yes the flat baffle doesn't diffract, but as baffle gets wider ever longer wavelengths are now supported by it and in half space radiation, showing edge diffraction ever lower frequency. Edge is the problem, extra baffle area just makes the edge "exposed" for lower frequency. Basically diffraction happens above bafflestep frequency, below which the box is acoustically small and sound doesn't interact with it as much.
To counteract the diffraction at the edge of a wide baffle (with extra flat area) a larger round over would be needed at the edge. Smaller radius roundover will do if the baffle is small because the problems are higher in frequency, at shorter wavelengths.
Basically what this is that the more the enclosure approximates a sphere the less diffraction there is. If you imagine the baffle to be surface on a sphere, how big the sphere is? Baffle can't be smaller than the transducers on it, but it doesn't have to be any bigger. If baffle needs to be bigger for some reason then it would be better to employ bigger transducers instead, big woofer and waveguide the tweeter, to control the sound to lower frequency and reduce radiation towards the edge. Basically keeping the flat baffle area minimized.
To counteract the diffraction at the edge of a wide baffle (with extra flat area) a larger round over would be needed at the edge. Smaller radius roundover will do if the baffle is small because the problems are higher in frequency, at shorter wavelengths.
Basically what this is that the more the enclosure approximates a sphere the less diffraction there is. If you imagine the baffle to be surface on a sphere, how big the sphere is? Baffle can't be smaller than the transducers on it, but it doesn't have to be any bigger. If baffle needs to be bigger for some reason then it would be better to employ bigger transducers instead, big woofer and waveguide the tweeter, to control the sound to lower frequency and reduce radiation towards the edge. Basically keeping the flat baffle area minimized.
HAh, I imagined this one but didn't mind posting illustration 😀 Add two more woofers to make it middle finger giving "in your face traditional center channel"! yeah its better diffraction wise assuming all the blue area is made round over / slant, beside the tweeter especially.
Thinking all this again, I'd guess that a center channel would be more intelligible if the speaker didn't radiate towards front wall and in that sense it would make sense to make wide center channel to reduce radiation to front wall so perhaps the traditional type is fine as long as there is (generous) round overs on the sides. If at all possible try keep the woofers below tweeter at least, or on sides of the box behind past baffle edge corner. I have no idea which one is more intelligible, less diffraction or less front wall radiation. Preferably both so you could employ acoustic treatment behind a low diffraction speaker, to reduce reflections from the front wall.
Thinking all this again, I'd guess that a center channel would be more intelligible if the speaker didn't radiate towards front wall and in that sense it would make sense to make wide center channel to reduce radiation to front wall so perhaps the traditional type is fine as long as there is (generous) round overs on the sides. If at all possible try keep the woofers below tweeter at least, or on sides of the box behind past baffle edge corner. I have no idea which one is more intelligible, less diffraction or less front wall radiation. Preferably both so you could employ acoustic treatment behind a low diffraction speaker, to reduce reflections from the front wall.
Last edited:
Alright there you have it. Just shave all the flat area off as much as feasible although a centimeter here or there is not that much better/worse especially if it means problems with structural strength.
I might be able to shave 5-10mm off the tweeter area. Because these are the RAAL 140-15 AM dipoles, they have a much wider face plate than the RAAL 140-15 AM monopoles I will use for the centre and main.
I could take about 170mm off the bottom if I put the lower woofer close to the floor, and lose one of the mids a bit more space. Problem is this current one the tweeter is at ear height so that is way above what I could get away with on a big screen.
Going to have to compromise on the height a bit somewhere.
I could take about 170mm off the bottom if I put the lower woofer close to the floor, and lose one of the mids a bit more space. Problem is this current one the tweeter is at ear height so that is way above what I could get away with on a big screen.
Going to have to compromise on the height a bit somewhere.
Which is better because of less room reflections?
I have an answer to every one of your points.
If reduced room reflections is the goal it is better to employ bigger transducers / waveguides / arrays, than to widen the baffle around small transducers no matter how I've looked at it from audio related issue perspective. Not considering aesthetics or possible size / cost limits one might have. There is also possibility to go in-wall, or very big baffles approximating in-wall situation, in which case the baffle edge diffraction wouldn't be no different to any other diffraction that happens with other objects in the room. Diffraction can also be distributed by using tall and narrow baffle for example, or offsetting transducer so that distance to any point at the edge varies but this variation should be considerable for good measure. If if varies between 5-10cm its not as distributed as if it was 5-80cm like it could be on a tall and narrow baffle.
One can employ what ever means for what ever reasons, there is just more diffraction controlling directivity on reasonable sized speakers without round over and whether that matters is up to the designer. It is possible to have directivity control without penalty in diffraction, by using minimal flat baffle around bigger transducers / waveguides / arrays. Its very hard to get completely rid of diffraction as there is the other drivers on a multiway speaker, usually on the same baffle, causing diffraction even if the edge was optimal without any diffraction what so ever.
Tried to cover it all here for you AllenB. You could come forward with your thoughts so I can make the repertoire more accurate.
One can employ what ever means for what ever reasons, there is just more diffraction controlling directivity on reasonable sized speakers without round over and whether that matters is up to the designer. It is possible to have directivity control without penalty in diffraction, by using minimal flat baffle around bigger transducers / waveguides / arrays. Its very hard to get completely rid of diffraction as there is the other drivers on a multiway speaker, usually on the same baffle, causing diffraction even if the edge was optimal without any diffraction what so ever.
Tried to cover it all here for you AllenB. You could come forward with your thoughts so I can make the repertoire more accurate.
Last edited:
This is about as compact as I think I am going to be able to get it. 500mm high, and close to that wide. I think this ticks as many boxes as can be ticked.
Cool 🙂 If that is tough on aesthetics or too difficult to build you could use a cylinder shape instead, or something that is close to half moon, to get the reduced radiation toward front wall while enjoying reduced diffraction. Simplest to manufacture would be just two minimal shoe boxes stacked, one for the mid/high and one for the woofers. Minimal for the mid/high but the woofer box could be bigger no problem, I don't know what the crossover needs to be and what not. If woofers play to 200Hz, which is about 1.7m long, the woofer box could be quite big without much problems. You might even get away with same woofer box for LR speakers in upright position and laying on its side for the center.
Last edited:
I think for the mains I would go with a similar design, but with the two woofer vertical. I have a massive sub yet to build that will take care of the low low.
It's an 18" TC Sounds LMS Ultra 5400 with two VMP 18" passive radiators. Not sure if I will use the TC Sounds Pro 5100 I am currently using for OB bass as boxed woofers / sub woofers.
Depends on where I wind up going with my active system. I don't think I would have any difficulty building the centre speaker on my CNC. It is plenty big enough to handle the sheet size required.
It's an 18" TC Sounds LMS Ultra 5400 with two VMP 18" passive radiators. Not sure if I will use the TC Sounds Pro 5100 I am currently using for OB bass as boxed woofers / sub woofers.
Depends on where I wind up going with my active system. I don't think I would have any difficulty building the centre speaker on my CNC. It is plenty big enough to handle the sheet size required.
Attachments
Last edited:
Nice! After you have landed with the system design best favor you could do to yourself is to make some quick prototype boxes and see if you like the sound and looks and fit and all that. As it is quite an investment in cost and time to manufacture multiple speakers it can potentially pay of in multiples if you manage to avoid some issues and be able to pick better options. Anyway, have fun with the project!🙂
