It's the most sensitive spectra of human hearing BUT most designers appear to choose it for their mid-high pass. Why is this? It would seem to me that in an ideal world, a three way with a midrange driver covering 400-4khz would be the solution used most often but I more often still see three ways crossed to a tweeter lower......sometimes much lower.
Crossing from midrange to tweeter at 4kHz would require a small (3" or less) midrange speaker in order to maintain a smooth power response. A 5" midrange would beam way too much to cross that high, while a typical dome tweeter would not beam at all when it takes over. This mismatch in directivity causes the power response to droop in the midrange and then suddenly jump up when the tweeter takes over. Also the center to center spacing of the midrange and tweeter would be too large, compared to the wavelength of sound at 4kHz.
Last edited:
Why? Because it is the right thing to do. You can only cross over
in the range where driver feels comfortable if high quality is what
you seek.
In an ideal world that would be the case but there is no such midrange driver so we are using what looks like the least compromise possible.
I mostly agree, but I would claim the range is more like 700 - 7000. I avoid crossovers in that range, although right at the lower edge is usually the only option.
But is it the best compromise?.....
One one hand we have a slight discontinuity in directivity and power response and on the other, we have shifting phase and overlapp. From a voicing standpoint, I'm no longer sure I've been choosing correctly.
I can get pretty close C to C spacing and directivity match with something like the SS 10F......and even if I need power power handling, a vertical pair which would introduce two lobes at +/- 20 degrees.......instead of bouncing more energy off of the floor and ceiling. Sounds like a good compromise.
I've read your viewpoints on this extensively which has led me to where I am now in my thinking. Get that XO down to 1khz or less and the system just becomes something different......special IMO.
Now if like to pick your knowledge at some point as to why a 1.4-2" CD 2waveguide with a super tweeter isn't a trade you're willing to make. Is it the size of the horn?......the anomolies of a 2nd HF device and the associated combing?
This new crop of 2-3" 'mid woofers' with very low HD, not so bad power handling and efficiency are giving me reason for pause whenever I contemplate a design that's looking to do something different from what we've been doing. I look at commercial designs like ANYA and see a changing of the rules when we apply our current and past thinking.
Obviously, I'm talking about line source systems to a degree. The current thoughts for home use......too big, too expensive, needs eQ.....and on and on. But I pose WHY do they have to be floor to ceiling? Who actually sits below or above a vertical axis of 20 degrees at home? Do these folks get up and dance around the room, looking for nulls and or lobes? I know I don't. I sit down with a refreshing drink, a book or tablet and........relax. If a symmetrical row of 2" Fullrange drivers with response from 500-5khz gets me there flanking a central line of ribbons, I ask 'why not?'.
It looks that BMS 4592nd solves your problem. I listened it in Oris Swing where mid driver covers frequency range from 300Hz to 7000Hz. Doesn't get much better than that. It is a bit expensive though but solves xover problem in most sensitive region once and for all.
Yuichi 290 can be a part of the solution. With jbl 2446 it can play from 600Hz to 10000Hz. I would add a supertweeter at the top and a good woofer for low frequencies and that would be it.
Last edited:
I'm thinking a pair of SS 10F's would avoid the needs for the complexity of a four way. With a third order HP around 400hz, those little mids would drive most right out of the room. There's quite a few cheaper, smaller options from Tymphany just put, including a prelim of a little 3" with 89db at 2.83 for less than $25 US. With really low inductance and a BIG motor.
Does anyone have any thought on destructive interference from a line source of say four of these drivers running up to 4-5khz?
How would you do this without beaming or rising distortion in the lower mids?
hello, by keeping the volume lower and then when i want it louder i either switch to a higher crossover frequency, or i increase the loudness of the reverberant field with a time delay of 10-12 mS or both.
The issue isn't "why not use a larger driver?", it's "why?" What advantage is gained by using a larger CD? A lower crossover means a huge waveguide to get down that low - not practical, more power handling is certainly not required for home use, and another higher driver and waveguide just adds more size, more cost and more complexity. All of that for what?
If you can get this to work then great, but trust me it would not be easy, or even possible. There will be wide variations in polar response unless it is very wide, which is one of my "never do" design requirements. High directivity is a must IMO and it must be well controlled. If you can achieve this to the extent that I can with a waveguide and a woofer then I'd love to see the data. Its all in the details and the end result.
The ear becomes quite tolerant of aberrations below about 500 Hz because it processes sound differently in that region than it does above there. Our hearing acuity peaks in the 2-3 kHz region and is large from 700 - 7 kHz. We are quite sensitive to aberrations in this region and ALL crossovers are aberrations to be avoided - not added whimsically.
So yes, a system that had a crossover down at 500 Hz would be good, but a 36 inch wide waveguide would be required and then it would NOT match the directivity of the woofer at that point unless the woofer were some 20-25 inches. Things get out of hand fast when one tries to lower the crossover point. In my NS-15 it is about 800 Hz which works quite well in a reasonable size system.
Last edited:
There isn't really a problem, as per say, crossing over anywhere in the audio spectrum providing that the drivers are correctly integrated and phase aligned when doing so.
There have a been a few files floating around that demonstrate the audible differences between how a 4th order LW, at I think 1kHz, affects the sound vs an unedited file. Basically you'd be hard pressed to hear anything. So much for filter audibility smack bang in the wrong place. I don't think this is news though.
The main thing you have to watch out for is what's going to happen to the polar response by crossing wherever you might be contemplating.
On axis there are lots of driver combinations that would appear to work well when crossed at say 500Hz and 5khz, but off axis things rapidly turn pear-shaped. There is absolutely no point in crossing your drivers over with the mentality of this thread if you're going to compromise everything else by doing so.
There is some truth to the argument that if you're crap at designing crossovers you're going to butcher the sound less if you place the crossovers out of the ears most sensitive band, but this isn't exactly a good argument to use. Loudspeakers are already full of compromises anyway, don't let your xover skills be one of them.
Geddes use of wave-guides down low is an excellent way to get around a large number of issues, it's a shame it's not more widely applicable.
There have a been a few files floating around that demonstrate the audible differences between how a 4th order LW, at I think 1kHz, affects the sound vs an unedited file. Basically you'd be hard pressed to hear anything. So much for filter audibility smack bang in the wrong place. I don't think this is news though.
The main thing you have to watch out for is what's going to happen to the polar response by crossing wherever you might be contemplating.
On axis there are lots of driver combinations that would appear to work well when crossed at say 500Hz and 5khz, but off axis things rapidly turn pear-shaped. There is absolutely no point in crossing your drivers over with the mentality of this thread if you're going to compromise everything else by doing so.
There is some truth to the argument that if you're crap at designing crossovers you're going to butcher the sound less if you place the crossovers out of the ears most sensitive band, but this isn't exactly a good argument to use. Loudspeakers are already full of compromises anyway, don't let your xover skills be one of them.
Geddes use of wave-guides down low is an excellent way to get around a large number of issues, it's a shame it's not more widely applicable.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.