Because of pattern control and therefore, correlation between on-axis and off-axis SPL. A vertically small waveguide will have lost pattern control above crossover frequency for most typical crossovers. While the horizontal pattern is regarded as most important for imaging, there would still remain a problem: When the direct axis sound is flat, the off-axis would have too much energy due to the widening pattern that happens on the vertical axis. And vice versa, a continuous off-axis curve would necessitate a dip of the on-axis frequency response. There can be solutions with using asymmetric crossovers, using cancellation around crossover, but you could easily start with a competent waveguide in the first hand.
My first adventure into the 3D CAD world:
Learning curve of FreeCAD is quite steep, but I just might start to get it. Enclosure is parameterized. The spreadsheet controls everything. Should allow me to switch between high/deep Calpamos, 12" versions, ...
A friend of mine works for a company that sells woodworking machines. I can use their demo CNC machine (for a six-pack or two I guess).
Learning curve of FreeCAD is quite steep, but I just might start to get it. Enclosure is parameterized. The spreadsheet controls everything. Should allow me to switch between high/deep Calpamos, 12" versions, ...
A friend of mine works for a company that sells woodworking machines. I can use their demo CNC machine (for a six-pack or two I guess).
https://www.pispeakers.com/contents.html
Have you looked at these 2 ways. They use easy to source parts and are available as DIY with parts kits or you can simply join the forum and ask for the free plan and round up the pats yourself. I built a pair of the 3 model with the upgraded drivers and crossover.
Have you looked at these 2 ways. They use easy to source parts and are available as DIY with parts kits or you can simply join the forum and ask for the free plan and round up the pats yourself. I built a pair of the 3 model with the upgraded drivers and crossover.
They don't look that well spread, the height is very close to double the width and depth. What GM says above is correct from an internal perspective.
Diffraction can also be affected strongly from dimensions that are similar to half dimensions being repeated.
The golden ratio of having dimensions 1.618 : 1 can be beneficial in breaking this up if it becomes practical.
If you have not considered it....I prefer a slot over a round port.....I mean if I make a rectangular slot.....the width of the cab, more or less. I'll have 3/4" walls blocking reverb into the tunnel...and no plastic ports to deal with. I just never enjoyed them in the past.... Most Pa boxes use slots for permanency I believe, or maybe the reduction of one more separate part. I didn't know if you'd considered it or not, so now ya know.....
Yeah, if all three dims are divisible by a whole number or itself, avoid it like the plague. Re the common golden ratio(s), there's been some advancements first seen (by me) in Don & Carolyn Davis's 'Sound System Engineering' and here's a bit more recent paper: https://www.researchgate.net/publication/28578741_Room_sizing_and_optimization_at_low_frequencies
There is arguments to be made either way, the biggest problem with reduced vertical size is addressed by sheeple. By reducing the vertical coverage angle with a smaller sized vertical dimension the narrower directivity is only maintained over a more limited bandwidth. In pro sound this can be a real problem and the effect is known as waistbanding. 90 x 60 is a reasonable compromise, in anything more than 1:1.2 ratio the effect can be seen. Trying to increase the size and maintain the narrow directivity is not so easy to accomplish in practice.
The horn has to be crossed over to the woofer and having a horn that is very similar in size to the woofer in both dimensions allows the best directivity match between the two in both planes. Gedlee Summa and JBL M2 take this approach. The wider vertical angle does put more sound towards the floor and ceiling where it isn't wanted but it does so with a spectrum that is more similar to the direct sound.
Which of the available options is best is argued by those that favour either side, if in doubt going with Earl Geddes seems reasonable to me
I've written a quick and dirty script that optimizes the ratio in order to have peaks as far away as possible from each other.
Here is the golden ratio:
And here is the optimum that was found (1.088):
Although the lower peaks are more evenly spread for the golden ratio, the 'optimized' variant does better in the region f > 900Hz.
Are the lower frequencies more important? Do I need to worry if I have enough damping as @GM says?
Here is the golden ratio:
And here is the optimum that was found (1.088):
Although the lower peaks are more evenly spread for the golden ratio, the 'optimized' variant does better in the region f > 900Hz.
Are the lower frequencies more important? Do I need to worry if I have enough damping as @GM says?
That does the opposite of optimizing as far as I am concerned. As frequency rises the modes will get closer together naturally until there is so many you can't see them any more. It is why above 1k in room the speakers response is the dominant factor. Below that the modes spread out more and more gaining more power. Having three spaced closely together low in frequency will make the combination problematic and harder to absorb due to reinforcement.
Tom, About box internal modes: the higher modes (say around crossover frequency) dampen easier (more) with damping material than the lower ones, whose wavelength is longer. You should be able to minimize the lowest modes by placing the woofer in the middle of that dimension, at least this is what happens in hornresp models. Put the woofer middle height if at all possible. In reality your horn robs internal height some so perhaps difficult to hit the sweet spot first time. If problem at all.
Prototypes are great. Make a quick one, if there is no measurable and audible issues with the internal modes then there is no problems. If there is you have a chance to do something about it
ps.
Three way speakers can enjoy compromise free enclosures. Here with big two way spekaer, as you need the one woofer to play lows and highs (mids) you need to compromise between. In order to get extended lows you need big box which diffracts and contains standing waves, panel resonances and what not, problems with port schuffing / leaking mids. With three way these can be built outside bandwidth, separation of concerns with separate bass and mid transducer (and enclosure). You can do it with 2 ways as well but lot of attention to detail is needed, if you want exceptional results (audible or not). If you want good two way mains, make the system three way, use separate sub woofers and don't worry about the bass extension of the mains in which case you can optimize mains woofers for mid performance (small enclosure without ports). But, of course you can make what you wanna! Wanted to provide another perspective to things. Audio related compromises can be pushed to added cost and complexity, if there is possibility to take that path given time and cost constrains and there is need to squeeze every it of "audio quality". On the other hand big speakers are so much fun small problems don't mean nothing. Small speakers can't compete no matter how well optimized
Prototypes are great. Make a quick one, if there is no measurable and audible issues with the internal modes then there is no problems. If there is you have a chance to do something about it
ps.
Three way speakers can enjoy compromise free enclosures. Here with big two way spekaer, as you need the one woofer to play lows and highs (mids) you need to compromise between. In order to get extended lows you need big box which diffracts and contains standing waves, panel resonances and what not, problems with port schuffing / leaking mids. With three way these can be built outside bandwidth, separation of concerns with separate bass and mid transducer (and enclosure). You can do it with 2 ways as well but lot of attention to detail is needed, if you want exceptional results (audible or not). If you want good two way mains, make the system three way, use separate sub woofers and don't worry about the bass extension of the mains in which case you can optimize mains woofers for mid performance (small enclosure without ports). But, of course you can make what you wanna! Wanted to provide another perspective to things. Audio related compromises can be pushed to added cost and complexity, if there is possibility to take that path given time and cost constrains and there is need to squeeze every it of "audio quality". On the other hand big speakers are so much fun small problems don't mean nothing. Small speakers can't compete no matter how well optimized
Attachments
Last edited:
Not sure that's at all relevant for speakers with controlled (sort of anyway) directivity for a decent part of their BW. I'd never set a pair of speakers like these up the same as I would a typical wide dispersion unit. Toe the CDs in so they cross in front of the LP.
All of those are relatively easy to damp if you incorporate that into the design from the start. Only the lowest near 200 is at all difficult.
Agreed.
The 2380 flare is horrible; I gave away my last PA versions maybe 20ya. I'd also never use a 4" diaphragm Ti 2" exit above 8-10k as they all oil can to some extent.
This is not related to directivity, it is about low frequency management in the room.
And that's going to depend largely on the construction of the room and whether or not EQ is used.
I was basing my comment on this ^. In that case the LF imbalance will not be that much of an issue even if the room is in a nuclear bunker.