Word of advice to future finish workers... Thinning is the key to simple success of staining and top coating.
I learned this the hard way. Mineral spirits or Turpentine are go to's for oil based products but I did not learn about water based products until yesterday. I thought that I could simply thin the water based poly with water but I did not do enough research, so when it was time to apply, I followed the rules and only added 10-15% water to the poly. The problem, this is no where near thin enough for self leveling performance. I ended up removing the majority of that application. Come to find out, I could have cut the water based poly 50-50 with denatured alcohol, this would have rendered a viscosity that I found acceptable. You cannot cut the water poly 50/50 with water fyi. SO Now I am preparing to possible reapply another application of oil based color and then after some time, I can try again the water based poly top coat.
Last edited:
Fact is, that it can also be done with more subs, see Toole et all.
To-may-to to-mah-to, similar thing just practical implementation is slightly different.
The more you add, the more you can average.
Obviously at a certain point they don't add much anymore, so around 3-5 is often fine.
Most just depends on the size, shape of the room and position of like the listening position(s) etc.
For just practical reasons, I always recommend having two symmetrical stereo front subs.
That gives you a bit of leverage to cross a bit higher when needed.
When the midwoofers can handle enough low-end that is obviously not needed.
Below roughly 100-120Hz there is no directional information anymore, so there is physically no way of sensing the location of subwoofers.
If you can, there are some higher order secondary "resonances/issues" going on.
What about harmonic distortion for example? Because of our frequency dependent hearing sensitivity, it's possible to hear the harmonics louder even if they are lower in level than the fundamental. This is especially true for bass frequencies, where the HD is typically not low (of course this depends on the whole system) and the harmonics lies in the midrange.
I think the low pass set at 98hz, lr24db IIRC could be 12...I used one corrective peak filter at 38hz, approximately +10db 10Q, just a guess, I was trying to match the shape of the dip...The thing is, when I see a smooth phase transition, I usually achieve better EQing.
Listening position ~6.5ft away, left channel
Strangely enough when I removed the top baffle of the ppsl FR problems increased, below, filtering is just the low pass.
Listening position 3ft away, left channel
Below is after the addition of the peak filter I spoke of earlier at 38hz, I spent minimal time trying to design a crossover for a picture, but I think it is generally accepted that this could be a nice final FR with some TLC? Opinions?
Listening position 3ft away, left channel
HD distortion (not IMD) is pretty hard to hear at those lower frequencies.
We have to generalize things here.
So let's create a bit of a scenario with a thought experiment.
Say we are using a closed subwoofer with a LR 4th order LP filter at 100Hz (meaning 100Hz will be at -6dB).
So around 70Hz or so, we can expect harmonics at 140Hz as well as 210Hz.
I will leave anything higher out, because those harmonics are generally not significant.
(if they are, get an higher quality sub, because there is obviously something wrong)
If you look at a typical HD distortion curve, you'll see that a lot of (sub)woofers perform extremely well between roughly 80-300Hz or so.
Often less than 0.5% for 2nd order and often half of that for the 3rd order.
Which is -46dB lower than the actual SPL.
Especially taken the fact that our hearing is already not very sensitive at that frequency range, I don't see how that is audible?
It could be a different story for intermodulation distortion.
But at lower frequencies the distortion is mostly dominated by HD distortion from the non-linear compliance and a tiny bit of BL (see Klippel).
From a practical point of view it makes even more sense to just have two front subs in a symmetrical setup.
haha, well it depends.
If people can't fit anything bigger, there is no other option than going for displacement.
If people can't fit anything bigger, there is no other option than going for displacement.
two 18's sealed at 1meter seems like the perfect amount for the conservative, judging one channel... With one more 18" on each channel, it would put me in the next league for headroom. The dip at 38hz is possibly due to a 1/6 cancelation from the distance of the corner. With a rear placed 18" I could bring it in much lower mimicking a ported design, but leaving out the bad stuff, like the harmonics. I was thinking to bring it at a point around 50hz where 1/4WL summing begins.
@b_force If you had a 8ohm and a 4ohm 18" (same make n model) planned to be a pair... with DSP running FR is there any worry? of a mismatched signature?
@b_force If you had a 8ohm and a 4ohm 18" (same make n model) planned to be a pair... with DSP running FR is there any worry? of a mismatched signature?
Last edited:
It was pretty easy. First I used the manufacturer that I had used for decades - B&C - because of their high performance, reliability and consistency. I wanted the woofer as efficient as possible tro best match the waveguide - low efficiency woofers require the waveguide to be padded, which is not real bad issue, but I like high efficiency in general. Then I looked for a woofer that was well controlled at its upper edge. While the NDX15-100 is not ideal in this regard it is still better than most. In the end, the singular issue with the NS-15 was the bleed through of the woofer into the midrange because of the HF peak in the woofer. I moved the crossover as low as I could to help alleviate this issue, but I can still see it in the final result. I never even looked at the TS parameters nor the THD numbers - they simply did not matter.
This is, of course, a serious confounding factor in the localization of subs at LFs. The subs need to be very clean or the results are tainted. Same with resonances.
They become less significant by the numbers, but far more significant as far as perception is concerned. That's why I wouldn't completely discount them. Think of our hearing at LFs as a sort of HP filter up to about 500 Hz., then it flattens out.
I was mostly talking about just the level of them.
At that point they get swamped by the (background) noise.
I am missing context here, but just as plain closed subwoofers and you have enough amplifier power and voltage, sure why not.
Don't know about difference in T/S therefor cabinet size etc.
That 20-100Hz frequency response is +/-7dB, ~14dB variation.
In my opinion, a final frequency response would be within +/-1.5dB, ~3dB variation would be "nice".
I'm relatively easily pleased 😉
If by "planned to be a pair", you mean both drivers run in parallel, the 4 ohm driver would draw approximately +3dB more power than the 8 ohm, overall would deliver ~3dB more SPL.
In addition to that 3dB difference, the frequency response between the two may fall within +/-2dB, so the response could vary by 5 dB between the two.
If the two different drivers used individual DSP and amplification, they could be equalized to the same response.
Well, there are B&C woofers with higher efficiency, smoother top end, lower Mms and lower Le than the 15NBX100 have, e.g. the 15PLB76, 15NW76 or even the 15NA100.
If you don't care about the T/S parameters or distortion, the only advantage of the 15NBX100 is the higher power handling (which is rarely needed in home environment IMO) over the few, above mentioned drivers.
Although those three drivers have higher sensitivity than the 15NBX100, they also have reduced response ~100 Hz by as much as -6dB.
Easy to attenuate the mid band of the HF driver, but can't bring up the low end of the woofer by +6dB with a passive crossover.
There's always something gotta give..
Nice, but extremely difficult even for several subs with DSP.
I could certainly have made any one of those drivers work, but I don't think that they all were available when I was still building speakers. Your right about power handling in a home. I have far far far more SPL capability than I would ever use. And Camplo is after even more - I never have understood that. There is something "macho" about more horsepower though! I love it in cars, maybe HT as well?
My real point is that it is the system design that matters, almost any drivers can work - drivers are a commodity - especially now with DSP. The low end of reproduction is room/system design integration far more that any particular driver. And the high end is all about polar control - the driver hardly matters. - waveguide is critical.
Are we talking about mid-range performance or low-end subwoofer performance?
Because in the low-end freq range it's mostly cone excursion, Sd and power are the dominant factors.
In that case we can let go of a bit of sensitivity/efficiency.
Even more so when there is plenty of power available.
The general idea with most PA drivers is just to get as much as BL as possible.
This seems to be the focus of B&C in particular, as being shared by Bennett Prescott in their videos.