One more thing to consider: moving mass, energy storage.
A woofer might go up quite high, disregarding dispersion for now, yet it still pays to let the midrange or tweeter go as low as possible, because the smaller driver on the HP side will give you more detail than the heavier driver on the LP side. This improves impule response as well. You can find nice info on energy storage in the burst tests of SL on his webpage. He shows a direct comparison of a good tweeter to a good midrange. The tweeter stores easily 10 times less energy - and that gives you more transparency.
As a personal experience, my 10" woofers have a useable FR to >1kHz. I used to cross them to the mid at 300 Hz. When I switched to 200 Hz or 160 Hz XO point, I heard considerably more detail.
Of course the smaller driver must still fulfill excursion requirements at the chosen XO point (i.e. have low distortion at a relatively high excursion, for its build).
A woofer might go up quite high, disregarding dispersion for now, yet it still pays to let the midrange or tweeter go as low as possible, because the smaller driver on the HP side will give you more detail than the heavier driver on the LP side. This improves impule response as well. You can find nice info on energy storage in the burst tests of SL on his webpage. He shows a direct comparison of a good tweeter to a good midrange. The tweeter stores easily 10 times less energy - and that gives you more transparency.
As a personal experience, my 10" woofers have a useable FR to >1kHz. I used to cross them to the mid at 300 Hz. When I switched to 200 Hz or 160 Hz XO point, I heard considerably more detail.
Of course the smaller driver must still fulfill excursion requirements at the chosen XO point (i.e. have low distortion at a relatively high excursion, for its build).
That would be great Dave, I'd love to attend. Let me know when you've got something scheduled. Hopefully my Vifa/Scanspeak is up and running by then:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=50601&highlight=
Basement reno's have forced it to the back burner for the last 2 months.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=50601&highlight=
Basement reno's have forced it to the back burner for the last 2 months.
dwk123 said:
I'll one-up you on this...how about planer bass. Cover the left and right walls of the room with relatively small woofers spaced less than a meter or so apart. Now, if the stereo signal coming out of your source is L(t),R(t), then feed the signal L(t)-R(t-x), R(t)-L(t-x) to the woofers, where x is a time delay equal to width of the room in meters/344. Now you've got virtual anechoic stereo bass...completely taking the room out of the equation. You can then add electronic reverb to make the room acoustically as large as you want it. Remember the number one rule of acoustics: small rooms suck. I haven't built this system yet, but it's just a matter of time (anyone know a source of cheap woofers in Europe?)
John
454Casull said:
from the above link...
"so if a linear-phase filter has noticeable preringing (for instance, if it's a high-Q filter), it might not sound as good as its minimum phase equivalent."
The "if" in that sentence is critical. There is nothing inherent about linear phase filters that makes them ring any more than other types of filters. In the above quote they are speculating that pre-ringing sounds worse than post-ringing. It's perhaps best to avoid both.
Also keep in mind that if the objective is to have a linear phase crossover, you don't need to apply linear phase filters to the drivers. I am using a delay-derived linear phase crossover with a 60dB crossover slope that exhibits no ringing:
S.P. Lipshitz and J. Vanderkooy, "A Family of Linear-Phase Crossover Networks of High Slope Derived by Time Delay", J. Aud. Eng. Soc, vol 31, pp2-20 (1983 Jan/Feb.).
454Casull said:
You have to be careful and differentiate between a single linear-phase filter stage and a *system* realized with net linear phase behavior. A single linear phase filter stage must be symmetric, and thus will 'pre ring'. However, multiple linear phase stages that sum flat will essentially have their pre-ring behavior cancel out. If done properly with minimal driver separation etc, this can be quite successful.
However, I tend to agree with John - the most successful approaches don't necessarily use linear phase stages directly, but derive filter transfer functions so that the sum is linear phase.
hancock said:
DUDE! Welcome back.
Yeah, great - my wife will just love this idea! Hey, honey - good news! You don't have to worry about wallpaper anymore.....
I'll have to think about this idea a bit more. Seems to me you'll still have modal behavior in the other two dimensions, so it won't quite be anechoic. I'd also be concerned that since bass is frequently highly mono, the quantities L(t) - R(t-x) will be small, and the net energy will be low.
I suspect you're just doing this to stay one step ahead of me. I'm finally getting the Unity/co-ax horn thing down, and you leapfrog again.
dwk123 said:
I just want to clarify this one just to make sure there are no misundertsandings on this important point. A linear phase filter will only pre-ring if it has a steep slope, just like any filter rings when it has a steep slope. The moral of the story is don't use excessively steep slopes and that applies to both linear phase and minimum phase filters.
I hope that helps rather than confuses...John
dwk123 said:
The modes in the front-backdirection are there, they just won't be excited if all the energy is going left-right. I should clarify that this only works in a rectangular room and if there are any acoustically large objects in the room, you'll get reflections that will reduce the anechoic effect. Luckily in the bass range, things have to be pretty big to be acoustically large--bigger than 1/4 wavelength in dimension.
You're also right that this reduces the amount of sound energy in the room because there is no reflected sound. You'll only be hearing the sound coming directly from the speakers and this is 6-12dB lower than the reflected sound in a typical room. However, you'll put that energy back in by adding electronic reverb.
The whole point of this exercise is to maximize the feeling of bass envelopment you get from your system. Bass envelopment is something that concert hall designers spend a lot of time trying to maximize and unfortunately is something you will never get in a home listening room without going to a setup like I described--unless your room is the size of the Boston Symphony Hall, of course.
I'm on vacation right now, so I've got some time to explain this a bit more...
The brain localizes sound by comparing the relative pressure level and phase of sound in our two ears. At high frequencies this works great, since sound coming from the left is shadowed by your head and the volume of the sound will be lower in your right ear than your left. In addition the time delay between when the sound arrives in your left ear and when it arrives in your right ear is large compared to the frequency of the sound--the path length between your left ear and your right ear is large compared to the wavelength of the sound.
At lower frequencies, though, the shadowing effect of your head becomes insigificant (once the size of your head is smaller than a 1/4 wavelength) and the pathlength difference between your ears also becomes small relative to the wavelength of the sound, so it becomes increasing difficult to localize sound at lower frequencies. When your brain can't localize the bass sounds, you get that bass-in-your-head feeling.
If, however, you have sound coming from both the left and the right direction and they are not in phase, then you get cancellation effects and the sound in one ear will be louder than the sound in the other, so your brain will localize the sound to the left or the right (the apparent source of the sound will be from the left or right direction) and you get the bass out of your head. You only get this cancellation effect when the sound is coming from the left and right directions, not front/back, which is why concert hall designers try to maximize lateral bass energy.
Now here is the interesting bit...if the sound is just a pure tone--a single frequency, then the cancellation effect will result in the sound being louder in your left or your right ear and unless you move, the apparent source of the sound won't move. However, if the sound coming from the left direction is at a slightly different frequency than the sound coming from the right direction, then the apparent source will flip back and forth between your left and right ears.
The rate at which the apparent source switches back and forth depends on the frequency difference between the sound coming from the left and the right. If the frequency difference is 1Hz, the the apparent source will flip back and forth at a rate of 1Hz. This is slow enough that you brain can track the movement back and forth. If the oscillation of the apparent source gets above around 10-20Hz, your brain can't track the movement anymore and you get that sense of being enveloped by the sound.
Luckily music is not a pure tone, it is wideband and you do get envelopment. In fact one of the reasons vibrato sounds good is it increases the bandwidth of the sound and maximizes envelopment. To maximize the envelopment at bass frequencies in the home, you want to maximize bass coming in the left-right direction and minimize it in the front/back up/down directions. You also want the reverb from the left to be decorrelated with the reverb from the right. You want the sprectrum of the bass reverb to have tightly spaced peaks and valleys. Furthermore, you want the peaks in the sound coming from the left to correspond to the valleys in the sound coming from the right. This will give you maximum envelopment. You could never achieve this "optimimum" physically with a concert hall, but concert hall designers try to get as close as they can. With the home set-up I describe you actually could achieve this "optimum".
I guess the bottom line of all this is that there is a whole lot more to an ultimate electrodynamic speaker system than just should I use a 5" metal midrange. Certainly having speakers that have a flat on-axis response and low non-linear distortion is important and you can achieve that with mono-poles and horns/dipoles. However, if you want to really recreate the experience of being there, you need to worry not just about on-axis response, but things like directionality, power response, and surround sound. In my humble opinion, stereo mono-poles just won't get you to the ultimate.
Yes, it's all part of my evil plan...
John
Siegfried Linkwitz has a bit on the ringing of linear-phase filters.
http://www.linkwitzlab.com/frontiers.htm#I
http://www.linkwitzlab.com/frontiers.htm#I
Back to the original topic, nobody has really talked about the enclosure. No point having the best drivers & crossovers if you enclosure is flopping about.
You may want to consider materials such as corian, carbon fibre, concrete or epoxy, and the use of matrix bracing. Cost, weight & your skills are the criteria here.
Also consider diffraction effects of enclosure shape and preferably aim for curved edges.
One more thing, rather than a 12" driver, I think 2x8" or 2x10" is worth serious consideration. Better baffle rigidity & improved form factor.
Cheers
David
www.gattiweb.com
You may want to consider materials such as corian, carbon fibre, concrete or epoxy, and the use of matrix bracing. Cost, weight & your skills are the criteria here.
Also consider diffraction effects of enclosure shape and preferably aim for curved edges.
One more thing, rather than a 12" driver, I think 2x8" or 2x10" is worth serious consideration. Better baffle rigidity & improved form factor.
Cheers
David
www.gattiweb.com
Interesting idea. As I understand it, you're trying to cancel the sound bouncing off the opposite wall. I'd think you'd want to attenuate the signals you subtract somewhere between 5 and 10 dB because the reflected sound has to travel 3 times farther to reach your ear than the direct sound. Sound right?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.