Hi,
I was planning on a bass reflex speaker. To find an optimum BR speaker one would typically load in a spreadsheet and decide on the allignment and there will be an answer.
I was just thinking to myself, there are many ways you can optimize differently if DSP involved. How would you optimize for such a speaker?
I personally would optimize it for the -6db point rather than the -3db point. Then I would probably push up the bass by another 3db at the 6 db point.
How would you do it?
Oon
I was planning on a bass reflex speaker. To find an optimum BR speaker one would typically load in a spreadsheet and decide on the allignment and there will be an answer.
I was just thinking to myself, there are many ways you can optimize differently if DSP involved. How would you optimize for such a speaker?
I personally would optimize it for the -6db point rather than the -3db point. Then I would probably push up the bass by another 3db at the 6 db point.
How would you do it?
Oon
I would optimize it for the 0db point.
If you calculate a vented box with rising response at box fs, a simple HP filter at a slightly higher than fs frequency can flatten the response, so a naturally "boomy" design will become more neutral.
Most vented designs go into large excursion rather quickly below tuning frequency, so it's a very bad idea to boost output below port tuning.
If you calculate a vented box with rising response at box fs, a simple HP filter at a slightly higher than fs frequency can flatten the response, so a naturally "boomy" design will become more neutral.
Most vented designs go into large excursion rather quickly below tuning frequency, so it's a very bad idea to boost output below port tuning.
When it comes to bass reflex with DSP, here's what I'd do:
- Pick the tuning frequency according to the desired low-frequency cutoff
- Set the cabinet size such that you can get to Xmax above the port tuning frequency when maximum power input occurs (whether that's amplifier or driver limitations).
- EQ to the desired response shape.
That will get you maximum SPL down to your chosen LF cutoff from the smallest possible box.
Chris
- Pick the tuning frequency according to the desired low-frequency cutoff
- Set the cabinet size such that you can get to Xmax above the port tuning frequency when maximum power input occurs (whether that's amplifier or driver limitations).
- EQ to the desired response shape.
That will get you maximum SPL down to your chosen LF cutoff from the smallest possible box.
Chris
Altough this thread is a bit old: I have been building loudspeakers dsp assisted for years and still do. I only take care about the resonance frequency of the port and that port size or passive radiator are big enough for undistorted reproduction.
Then - by measurement - I adjust the wanted transfer function ot the loudspeaker via dsp. In some rooms a slow falling frequency response down to the port resonance can help against boomy bass due to room interaction.
If the room has good sonic qualities the loudspeaker could be tuned to be linear down to port resonance.
Then - by measurement - I adjust the wanted transfer function ot the loudspeaker via dsp. In some rooms a slow falling frequency response down to the port resonance can help against boomy bass due to room interaction.
If the room has good sonic qualities the loudspeaker could be tuned to be linear down to port resonance.
Sixth order alignments go back a long way. And are quite simple too, even with analog active crossovers. I'd recommend them. You have (a bit limited) options of various classic alignments up to lowering the Fres with more than 25% by means of a Q=1.93 second order highpass around one opamp.
Not any problem to do this with dsp of course. But that is brute force approach.
It is clearly audible with acoustical benefit to put a filter below the port resonance. With my Behringer Inuke NU3000dsp I can choose up to a 48db Linkwitz Riley Filter. The bass sounds immediately better defined and cone excursion is reduced. And chuffing of the reflex port is minimized as signals below port resonance give a big contribution to that phenomenon.
Tuning frequency varies with output level and temperature by several Hz; it is practically futile thinking in terms of single Hz accuracy of tuning of filtering.
Well, that was supposed to be a simulation, after which one has to find ways for making it work in the real world.
I wouldn't worry about the accuracy of the cutoff frequency a lot, as a little to this side or that would only slightly increase the cone excursion. Besides, Xmax is usually much lower than Xdamage, so it should be OK in most cases. A little margin is a good practice though.
The main funda is that the vented box has nothing (good) to offer below the tuning frequency, so it's best to remove anything that happens below that.
Cheers.
I wouldn't worry about the accuracy of the cutoff frequency a lot, as a little to this side or that would only slightly increase the cone excursion. Besides, Xmax is usually much lower than Xdamage, so it should be OK in most cases. A little margin is a good practice though.
The main funda is that the vented box has nothing (good) to offer below the tuning frequency, so it's best to remove anything that happens below that.
Cheers.
Last edited:
Fading away of bass signal below port resonance
I read that below the tuning of the bass reflex port still some Hertz the reflex function is still working, but 5 Hertz below port resonance the loading is ceasing away.
I observed this with my subwoofer with a port tuned to 25 Hertz. I watched the diaphragm moving with loud single bass signals of 25 to 20 Hertz (25, 24, 23, 22, 21, 20).
Yet the cone movement is obviously reduced at 25 Hertz but at 22 Hertz reflex function begins to cease.
So I have put the filtering below port resonance at 20 Hertz.
I read that below the tuning of the bass reflex port still some Hertz the reflex function is still working, but 5 Hertz below port resonance the loading is ceasing away.
I observed this with my subwoofer with a port tuned to 25 Hertz. I watched the diaphragm moving with loud single bass signals of 25 to 20 Hertz (25, 24, 23, 22, 21, 20).
Yet the cone movement is obviously reduced at 25 Hertz but at 22 Hertz reflex function begins to cease.
So I have put the filtering below port resonance at 20 Hertz.
I started this post more than a year ago. Learned new things since then. I would choose a port tuning frequency a bit lower than the usual alignment so it is about -6db before it rolls off. I then boost it about 6db at that point and I would cut via a high pass filter at about 5 to 10Hz below port tuning frequency as suggested by Freedom666.
This way I can push the low frequency response downwards and increase power handling. Bad news is you loose 6 db of power at the mids....
Oon
This way I can push the low frequency response downwards and increase power handling. Bad news is you loose 6 db of power at the mids....
Oon