I always understood the formula to be F = 344/longest dimension. However, I have recently seen Bill Fitzmaurice (who knows what he's on about!) saying that the gain begins at half wavelength point. As you point out, this seems to give really high room gain starting frequencies.
I'd also agree to an extent with Paul on the 12dB/oct rise, that is on the optimistic side.
Regarding the GD issue, are you saying that GD is related to cone excursion? Maybe I ought to re-check, but my own findings were that GD of sealed boxes is always orders of magnitude better than vented. Will WinISD with the LT feature and see.
I'd also agree to an extent with Paul on the 12dB/oct rise, that is on the optimistic side.
Regarding the GD issue, are you saying that GD is related to cone excursion? Maybe I ought to re-check, but my own findings were that GD of sealed boxes is always orders of magnitude better than vented. Will WinISD with the LT feature and see.
formula
If room gain starts when the longest dimension of the room equals half a wavelength the formula is 344/(length*2)
In other words if the wavelength is twice as long then the frequency is an octave lower.
If room gain starts when the longest dimension of the room equals half a wavelength the formula is 344/(length*2)
In other words if the wavelength is twice as long then the frequency is an octave lower.
No, I'm saying that GD changes with a sealed box has eq or a LT. I don't know if there is a direct link between cone excursion and GD but I doubt it. A sealed and vented box with the same response and GD will not have the same excursion.
In my experience (with modelling), a sealed box has much lower GD before eq. Such a sub might be used for music and would be considered "tight." I'm not sure if this is a characteristic of how it integrates with room gain, or if it's related to GD, or if it's some other characteristic of sealed boxes, or if it's simply that earlier but gradual rolloff gives an inherently "tight" sounding bass ... I suspect there is some kind of combination of the above working in this situation.
My guess is that your comparisons that you are thinking of did not have the same response. I suspect that when you compare a vented and sealed alignment both properly implemented and both with the same response, the most noticeable difference would be output. In the absense of vent noise and boominess caused by poor integration, I suspect it would be hard to tell the difference, or at least, not as easy as you would expect based on so many negative comments about vented bass.
Regarding how to integrate the LT with room gain, there are two basic ways to approach it. One is to forget room gain in designing your LT and deal with the room with BFD or similar. The other is to tweak your LT with a drooping response that room gain will bring up.
I'd probably go with the former approach, although it's a bit more involved. If you want to go with the latter, have a play around with WinISD, mess with the Q and the F3 point.
The problem with designing a sub for your room is that your room might change if you move. Or you might move your sub.
I'd probably go with the former approach, although it's a bit more involved. If you want to go with the latter, have a play around with WinISD, mess with the Q and the F3 point.
The problem with designing a sub for your room is that your room might change if you move. Or you might move your sub.
paulspencer said:
I agree with all of that.
paulspencer said:
If you have an unequalised speaker with Fc 20Hz and Qtc 0.7, annother speaker with Fc 40Hz and Qtc 1.0 and apply LT to get the same Fc and Qc as the first example, they will have identical amplitude, phase and group delay characteristics. Both will have better GD than a vented box using one of Small's alignments with the same F3. However, it may be possible if the sealed box has very high Qc that GD could be similar to a vented box using a 'good' alignment by Small. I've not checked at this outline situation as I don't run Qc that high.
I agree that with 'normal' EQ such as bass boost shelving controls, that the GD gets mucked up and can be comparable to an 'equivalent' vented box as you point out. However, the LT works in a totally different way to shelving EQ and as a consequence of the pole cancellation does not muck up the phase and GD.
Room gain thing:
Of course I know by now that all the effects will changea lot with placement. I have read a lot of articles around this. The room response calculator was also a great help!
Now I am rebuilding my stereo amp so the sub project is waiting for at least a week or so. I will start to build the electronics first (already started the power amp), the LT will be the last one. And of course the box itself.
The new formula works! According to that my room gain will start somewhere around 32 Hz - this is quite the same as the graph produced by the room response calculator!
I still don't care too much about GD ever since I have decided to build a closed box. Even if eq-ing makes the situation somewhat worse, it cannot be that bad. I've seen the graphs in winisd pro, quite nice.
Of course I know by now that all the effects will changea lot with placement. I have read a lot of articles around this. The room response calculator was also a great help!
Now I am rebuilding my stereo amp so the sub project is waiting for at least a week or so. I will start to build the electronics first (already started the power amp), the LT will be the last one. And of course the box itself.
The new formula works! According to that my room gain will start somewhere around 32 Hz - this is quite the same as the graph produced by the room response calculator!
I still don't care too much about GD ever since I have decided to build a closed box. Even if eq-ing makes the situation somewhat worse, it cannot be that bad. I've seen the graphs in winisd pro, quite nice.
M127,
I have found that the FRD RRRC is very helpful and that it seems to match what I think I'm hearing in my room. A few things that you might like to try out:
* play around with the values for the room boundaries - large windows will absorb bass more than walls due to flexing glass, concrete or masonry will absorb very little, plasterboard will absorb somewhat; you might try to simulate the effect of bass traps in the room by increasing the amount of absorbance - you will notice the modes becomes much less severe
* try alternative placement - both listening position and speaker position
* try "direct field" placement - the sub as close as possible to your seating location (ie. 1m), the result is that the modes are less pronouned as the ratio of sound directly radiated is increased - behind your seating position or to the side
* try it for your main speakers as well as your sub - you have to put up with what you get for the most part with mains, but with subs you have more flexibility to put them where they get a smoother response; you then have to balance this with what works with your room layout and deal with localisation issues
* you can also experiment with 2 subs to see if you get smoother response
I have found that the FRD RRRC is very helpful and that it seems to match what I think I'm hearing in my room. A few things that you might like to try out:
* play around with the values for the room boundaries - large windows will absorb bass more than walls due to flexing glass, concrete or masonry will absorb very little, plasterboard will absorb somewhat; you might try to simulate the effect of bass traps in the room by increasing the amount of absorbance - you will notice the modes becomes much less severe
* try alternative placement - both listening position and speaker position
* try "direct field" placement - the sub as close as possible to your seating location (ie. 1m), the result is that the modes are less pronouned as the ratio of sound directly radiated is increased - behind your seating position or to the side
* try it for your main speakers as well as your sub - you have to put up with what you get for the most part with mains, but with subs you have more flexibility to put them where they get a smoother response; you then have to balance this with what works with your room layout and deal with localisation issues
* you can also experiment with 2 subs to see if you get smoother response
just a short one now:
I won't build two subs, don't have the place, don't wanna spend that much. Though I know life is easier if you have 2 of them.
Btw how do you make sure that eqs will not make other things worse while making a flat response? (Other things like delays, phase, etc.) I'm thinkin' about complex designs like the BFD.
I won't build two subs, don't have the place, don't wanna spend that much. Though I know life is easier if you have 2 of them.
Btw how do you make sure that eqs will not make other things worse while making a flat response? (Other things like delays, phase, etc.) I'm thinkin' about complex designs like the BFD.
Performing the disappearing act ...
The room can make so much mess that it's more important to get it reasonably flat. Good question, I'm not sure of the answer, but I don't think it's critical.
The difference I have noticed with a system set up with BFD is the sub does the disappearing act. Disable BFD and suddenly the sub draws attention to itself and the room.
The room can make so much mess that it's more important to get it reasonably flat. Good question, I'm not sure of the answer, but I don't think it's critical.
The difference I have noticed with a system set up with BFD is the sub does the disappearing act. Disable BFD and suddenly the sub draws attention to itself and the room.
You mean the quality is so good that you can't hear the sub itself (apart from the other loudspeakers) when the BFD is on?
If I buy a BFD, is it correct not to use a separate low pass filter but fix that also in the BFD?
I haven't checked, is the BFD available in Europe?
If I buy a BFD, is it correct not to use a separate low pass filter but fix that also in the BFD?
I haven't checked, is the BFD available in Europe?
Without BFD a sub usually will drop out in certain parts, and at others it will be overemphasised. Often room modes suck out the midbass, but then the low stuff is boosted by room gain. This draws attention to the fact that there is a sub. Calibrate the sub and the mains, and eq out the room modes and your attention won't be drawn to the subs, they will just give more authority, depth and power to your mains.
BFD is not a substitute for the low pass filter, or a rumble filter. Still use them.
Check at Behringer.com to see if it's available in Europe.
Do a google search for "behringer feedback destroyer" and you will probably find online retailers. Look at local PA places, ebay, etc. Again, check the website for dealers.
BFD is not a substitute for the low pass filter, or a rumble filter. Still use them.
Check at Behringer.com to see if it's available in Europe.
Do a google search for "behringer feedback destroyer" and you will probably find online retailers. Look at local PA places, ebay, etc. Again, check the website for dealers.
Hi Paul,
I have purchased an XLS10 on ebay at 87.- EUR. It is quite OK because it would cost more than 200 EUR (!) here in Hungary. I haven't seen any of the 12" model though. So I have a driver now.
My plan is still the same, to build a closed box with Linkwitz Transform. My problem is that I tried to do some calculations with excel sheets and winisd pro, and it seems like the very low Qts of the driver can be a problem (at least there are some strange things which I'm sure are caused by that). The excel sheets (linkxfrm.xls and linktranadv29.13.xls) show that the 'k' factor in my case is negative, which it shouldn't be. I guess it has got something to do with the unequlized box parameters (f0, Q0) and my targets. They say if 'k' is negative then the LT won't work. What is the solution? Can you run your softwares and recommend a box to me?
I am also thinking about building a passive radiator instead, though I don't feel like paying the same price again for the pr, and I know that it would be much more difficult than a sealed box. The only advantage is that it seems like this is what xls was really made for.
I have purchased an XLS10 on ebay at 87.- EUR. It is quite OK because it would cost more than 200 EUR (!) here in Hungary. I haven't seen any of the 12" model though. So I have a driver now.
My plan is still the same, to build a closed box with Linkwitz Transform. My problem is that I tried to do some calculations with excel sheets and winisd pro, and it seems like the very low Qts of the driver can be a problem (at least there are some strange things which I'm sure are caused by that). The excel sheets (linkxfrm.xls and linktranadv29.13.xls) show that the 'k' factor in my case is negative, which it shouldn't be. I guess it has got something to do with the unequlized box parameters (f0, Q0) and my targets. They say if 'k' is negative then the LT won't work. What is the solution? Can you run your softwares and recommend a box to me?
I am also thinking about building a passive radiator instead, though I don't feel like paying the same price again for the pr, and I know that it would be much more difficult than a sealed box. The only advantage is that it seems like this is what xls was really made for.
I guess I tried everything last night, from 20 to 35 Hz, Q 0.5...0.7 and I couldn't get a positive 'k'.
Paul Spencer advised the xls range with a small sealed box and LT but now it seems like I cannot get the results I need, not even in theory with the caculations.
Maybe I'll buy the pr, build a really small box, and spend some months with adjusting... 🙂
Paul Spencer advised the xls range with a small sealed box and LT but now it seems like I cannot get the results I need, not even in theory with the caculations.
Maybe I'll buy the pr, build a really small box, and spend some months with adjusting... 🙂
I will have a look at it when I get the chance
What are the dimensions of your room and the type of construction?
What are the dimensions of your room and the type of construction?
OK, so here are the details.
My driver is a Peerless xls 10" (swr 269).
Fs= 18.9 Hz, Qts = 0.17, VAS=89.7 ltrs.
The room is 5.55 x 3.05 x 2.50 (metres).
I'd build any box with a reasonable size (= it has to be small 🙂 ). My first preference is a sealed box construction with Linkwitz-transform, but if the driver turns out to be not suitable for that then I'll buy the passive and build a pr box.
What is the 'k' anyway (in the LT calculations)? Why does it have to be positive and what happens if it's not?
My driver is a Peerless xls 10" (swr 269).
Fs= 18.9 Hz, Qts = 0.17, VAS=89.7 ltrs.
The room is 5.55 x 3.05 x 2.50 (metres).
I'd build any box with a reasonable size (= it has to be small 🙂 ). My first preference is a sealed box construction with Linkwitz-transform, but if the driver turns out to be not suitable for that then I'll buy the passive and build a pr box.
What is the 'k' anyway (in the LT calculations)? Why does it have to be positive and what happens if it's not?
I have done a quick design. It's a closed box of 20 litres net internal volume. The transformed response is Butterworth (Q 0.7071) -3 dB @ 22 Hz. I would have preferred to only go to 25 Hz but that results in the 'k' rule violation. At 22 Hz we are not too far out of the 12 dB gain boundary anyway.
Above is a screen grab of the output of my programs. Note that you need relatively large values of capacitors for C1 and C3.
Above is the schematic the program relates to.
'k' comes about because of the pole cancellation performed by the LT. At the end of the day it's a mathematical equation and there will always be some values which don't satisfy certain criteria.
An externally hosted image should be here but it was not working when we last tested it.
Above is a screen grab of the output of my programs. Note that you need relatively large values of capacitors for C1 and C3.
An externally hosted image should be here but it was not working when we last tested it.
Above is the schematic the program relates to.
'k' comes about because of the pole cancellation performed by the LT. At the end of the day it's a mathematical equation and there will always be some values which don't satisfy certain criteria.
- Status
- Not open for further replies.