Sorry i can' resist, they are suitable for those who love that album :
Boum, Boum and Boum... and boum... BADA BOUM BOUM
Just take a look at the number of views, please.
YouTube
PS : the entire album is a boumboum nightmare.
OK, -24 dB/200 Hz then. All too low to get any sort of location information, so with the mains off you shouldn't be able to detect it unless it's the sub or something quite near. If the mains are on, then seems like all you can do is BW limit the system to the ~40-1800 Hz BW to hopefully find the harmonic that's causing it.
GM
GM
This was my also first assumption but, in a small room, you are close to the subs and terefore there is a part of the direct field radiation of the lobe that triggers mechanically all the objects (generation of a wide BW of harmonics)
With my large woofers the direct field is close enough to trigger the skin sensors, an array of subs needs more space to breath... or tiny woofers maybe.
I even posted in that thread! but thanks for the reminder, I will reread it and maybe take more from it with a fresh perspective.
The new idea is that DSP doesn't really beat the restrictions on minimum phase, the signal must be delayed to allow some kind of "look ahead".
Similarly, if we subtract out the delay of the horn driver then the woofer is effectively time advanced and similar tricks to DSP should be possible.
At the least it should be possible to deal with the time offset nicely.
And this seemed a reasonable thread to ask in, it's for a 2 way, 15" woofer and compression driver/horn rather similar to many speakers shown here.
And hopefully relevant to the OP's search.
Best wishes
David
It is actually easier to delay the woofer than the tweeter with analog methods since lowpass filters have almost constant group delay within their passband while the highpass filters have their highest group delay below their passband. An often used method is to use a higher order lowpass (often Bessel because of its almost constant group-delay) and lower Order highpass for speakers of this kind.
With mine I use a special crossover with flat group-delay and an additional Bessel lowpass whose cutoff is above the crossover frequency to delay the woofer signal.
Regards
Charles
With mine I use a special crossover with flat group-delay and an additional Bessel lowpass whose cutoff is above the crossover frequency to delay the woofer signal.
Regards
Charles
David, when you talk about linear phase, do you consider only the direct sound? Because obviously the time delay between drivers will be a function of a spatial coordinate. In my experience, what only works in the end is to take the real data (the more the better), fetch it to a good simulator (*) and play with it for a while. Quicky one gets the idea and achieve some compromise for the overall acoustic response - you really can't look at the one axis only. I usually start with some low order filters and see how that works (including all the EQ, impedance compensations, etc). There are situations I have to increase the order(s) to perform better. But I guess you will never know in advance, especially because in these designs you work with woofers right at the end of their passband and every driver will be different to a degree. So for the compression drivers on their lower end. To your particular question of getting linear phase, I don't know and I have never seen such analysis. For me it just works out somehow - and usually not that bad. Of course with DSP and FIR you can correct for the excess phase after all is done, but it will allways work properly for the direct sound only as the drivers are simply not coincident.
(*) I had to write my own just for this purpose. Now there's the VituixCAD, which is almost as good
(*) I had to write my own just for this purpose. Now there's the VituixCAD, which is almost as good
Last edited:
Hmm, missed the small room/~near-field scenario. Still, been there, done that more than once, so now wondering about the house/room construction, but again something near/at the sub is really the only way to locate it. I mean triggering all the objects isn't going to focus only one sub, but a pressure wave to the body sure will, so how are the woofers located in relation to where you're sitting [Lp]?
GM
Yeah, I've been 'scratching my head' over this XO/TD conundrum as Altec taught me that the default XO for a baffle mounted horn woofer is to use a 3/2 XO and 'gap' them as required based on degrees of offset. Always seemed to work well, though with no fancy test gear, no clue how well it measures.
GM
The woofers where located at the place where the measurements were less catastrophic, with the room layout and furniture it was one dipolar near the center slighly on the left and one closed in the right bottom, the CSD was relatively clean near the center of the room. It is fun because you can ajust the bass level by moving your seat to the center of the room.
The small (10in) peerless XLS are unlocalizable in the same room with their smaller direct sound field, but therefore unable to impose their direct field force against room ringings.
Hmm, no experience with dipole subs nor know the room layout/construction, so not enough to work with though 'sounds' like you're going to have to choose between what measures good Vs what sounds good.
Out of curiosity and still not having read Dr. Geddes' paper; have you tried one sub up a wall at an odd harmonic?
When I started playing with this way back when I assumed that since the room is 3D I needed at least one 'sub' in each plane at an odd harmonic and it worked well at least by ear to a fairly wide range of folks over time since I didn't have the $20+ K it cost back then to replicate the local Altec distributor's test gear racks and even then not sure they could do as much as REW, etc..
Maybe the 'standout' just needs to get up off the floor. Also, have had some luck tilting subs/speakers to make them 'disappear', so wondering if it works with dipoles.
GM
Out of curiosity and still not having read Dr. Geddes' paper; have you tried one sub up a wall at an odd harmonic?
When I started playing with this way back when I assumed that since the room is 3D I needed at least one 'sub' in each plane at an odd harmonic and it worked well at least by ear to a fairly wide range of folks over time since I didn't have the $20+ K it cost back then to replicate the local Altec distributor's test gear racks and even then not sure they could do as much as REW, etc..
Maybe the 'standout' just needs to get up off the floor
. Also, have had some luck tilting subs/speakers to make them 'disappear', so wondering if it works with dipoles.GM
Hmm, no experience with dipole subs nor know the room layout/construction, so not enough to work with though 'sounds' like you're going to have to choose between what measures good Vs what sounds good.
Out of curiosity and still not having read Dr. Geddes' paper; have you tried one sub up a wall at an odd harmonic?
When I started playing with this way back when I assumed that since the room is 3D I needed at least one 'sub' in each plane at an odd harmonic and it worked well at least by ear to a fairly wide range of folks over time since I didn't have the $20+ K it cost back then to replicate the local Altec distributor's test gear racks and even then not sure they could do as much as REW, etc..
Maybe the 'standout' just needs to get up off the floor
GM
An odd harmonic injection for my subs ! why not but i've dismantled all of them and choosen to face a wide directfield wavefront generated by two 26" dipolar flying saucepans.
The measurements of a small room full of reverbreations is too tricky, so the tricks like your "odd harmonics" is the only wav to do things correctly.
I've also written several over the years, but now I want to jump into Vituixcad and I have concerns. Can I ask you a little about it?(*) I had to write my own just for this purpose. Now there's the VituixCAD, which is almost as good
Can I expect to prepare a number of response plots, have the ability to weight them as I see fit, then get a power calculation?
Will it divide plots, and other such combinations and transformations?
The program offers many simulations. I do not want anything simmed except the network, just manual control. Does this sound reasonable?
I still have only a very limited experience myself, I know it mostly from the presented results of others. At the beginning it was a bit clumsy to use but it seems that it got a lot better since. Sorry, that's all I can say at the moment. I still mainly use my old software as it is so handy and effective to me, but I just can't maintain it anymore... VituixCAD definitely looks as the best (free) alternative now.
Last edited:
I think what you point out, that the delay of the woofer is already accounted for in a complementary crossover like LinkwitzRiley, is one of the least understood aspects of crossovers.
I guess folks see the group delay graph for the crossover going to the woofer, and think it is a substitute for offsetting physical time alignment.
Which of course, it isn't.
My strategy for overcoming physical offset is to first minimize the physical offset as much as possible physically, and second minimize the lobing from drivers sharing the crossover region.
I've found steep linear phase crossovers reduce the crossover range of sharing, and significantly help vertical polars (assuming drivers are aligned vertically).
On a two-way like this thread is about, I'd try to get my CD down to wherever I need to cross to the woofer, given pattern control and maintaining pistonic frequency response from the woofer.
At that crossover freq, I'd use something like a 16th order linear phase LR, and time delay for the physical offset.
Some may balk at this, claiming pre-ringing issues. But it's almost always hearsay balking I think, with little actual listening or testing. Maybe my ears aren't good enough to hear pre-ringing, but I can hear the improvements in the crossover region.
I've found if I stick to linear phase steep complimentary crossovers, that it's much easier to find acceptable horizontal and vertical polar response than using anything else.
Plus, the most important part, it sounds better ...slamming transients and a extra coat of windex on the sound, so to speak, for improved clarity.
I get where you are coming from not wanting a software solution.
The best software solution I've found is JRiver's convolution on their Linux build. Simply because it's Linux as opposed to....
The best hardware solution I've found is QSC's Q-SYS. A Core110f would be ideal. Core 110f - Q-SYS Cores - Products, Peripherals & Accessories - Q-SYS Platform - Products - Systems - QSC
I've seen them go for as low as $750 on ebay.
You could do 'whatever' with this....
Sorry i can' resist, they are suitable for those who love that album :
Boum, Boum and Boum... and boum... BADA BOUM BOUM
Just take a look at the number of views, please.
YouTube
PS : the entire album is a boumboum nightmare.
The cover reminds of this
It's the kind of music academic hipster girls like. Play this at an afterparty and you're ready to go... boum boum.
Last edited:
I'd say smooth blending helps, and this means a little overlap, as difficult as this might sometimes be.
Following the discussion about closed box versus bassreflex, I found this whitepaper, boldly titled: Why Bassreflex is not Suitable for a Subwoofer.
The author is a Professor in Mechatronics and member of the team behind this loudspeaker:
Interesting paper, thx.
I'm sure he's technically correct, but not so sure pragmatically correct.
And I see he thinks a 60 litres is a big box ! he doesn't know DIY fanaticism, huh?
I must say, I doubt it
For me, best bass i ever hear is outdoors....where there are no modes.
Ok, now how to match that indoors....
Two opposite approaches I think...
Build a room with the goal of minimizing the number and strength of modes.
Golden ratio room dimensions, modeled sub placement(s) (usually ends up in corners), significant investment in proper absortion/RT 60 dampening.
Or decide that can't be done, and maximize the number of modes with mutiple subs , properly placed to achieve a uniform dense modal distribution.
Minimize vs maximize.
Which is more likely to sound like outdoors?
My bet/ my experience is minimize.
I wonder what the best recording studios do ?