Hi,
I couldnt find any subwoofer implementations that do linear phase correction for a bass reflex enclosure. But, I do remember seeing one DIY project a few months backusing a 15" or 18" driver.
Please provide links of phase corrected ported subwoofer implementations if you come across.
Thanks and Regards,
WonderfulAudio
I couldnt find any subwoofer implementations that do linear phase correction for a bass reflex enclosure. But, I do remember seeing one DIY project a few months backusing a 15" or 18" driver.
Please provide links of phase corrected ported subwoofer implementations if you come across.
Thanks and Regards,
WonderfulAudio
I would like to know the theory behind it.
It can be only a correction in the time domain with FIR filters.
And in the end: does it make a difference while listening?
It can be only a correction in the time domain with FIR filters.
And in the end: does it make a difference while listening?
I've been phase correcting my sub builds with FIR for 5 years or so.
Here's a mag and phase trace of a dual bass reflex in a PPSL setup used as push-push.
The traces would be the same for a standard single 18" bas-reflex given same tuning.
There is a 27Hz BW 18dB/oct hpf and a 100Hz LR 72dB/oct lpf in place.
This was with 6144 taps at 48kHz.
As to how it sounds...
It almost sounds like there is less bass, until you realize it is simply cleaner, and you can turn it up to regain the previous bass level you were used to.
And then the entire spectrum, bass and above, sounds more powerful, tighter, clearer.
Best heard outdoors though, rooms can muck up anything...
Here's a mag and phase trace of a dual bass reflex in a PPSL setup used as push-push.
The traces would be the same for a standard single 18" bas-reflex given same tuning.
There is a 27Hz BW 18dB/oct hpf and a 100Hz LR 72dB/oct lpf in place.
This was with 6144 taps at 48kHz.
As to how it sounds...
It almost sounds like there is less bass, until you realize it is simply cleaner, and you can turn it up to regain the previous bass level you were used to.
And then the entire spectrum, bass and above, sounds more powerful, tighter, clearer.
Best heard outdoors though, rooms can muck up anything...
So you had to do it by yourself - correcting the phase of your sub with measurements and FIR programming. It looks good - the smooth falling response at 30 Hz.
As the thread starter asked for commercial subs with linear phase.
I build one way full range loudspeakers with IIR correction (usual Equalizers) - they are as minimal phase designs always correct in the time domain when the amplitude is linear. Only on the low frequencies I wonder if my bass reflex system would profit of FIR correction.
I do not know of loudspeakers with FIR correction I had a real chance to listen to.
It is interesting that a big manufacturer like RCF now promotes time corrected loudspeakers.
As the thread starter asked for commercial subs with linear phase.
I build one way full range loudspeakers with IIR correction (usual Equalizers) - they are as minimal phase designs always correct in the time domain when the amplitude is linear. Only on the low frequencies I wonder if my bass reflex system would profit of FIR correction.
I do not know of loudspeakers with FIR correction I had a real chance to listen to.
It is interesting that a big manufacturer like RCF now promotes time corrected loudspeakers.
Hi Freedom666, yep, had to do it myself.
I don't think we will see any commercial subs using FIR, as the latency is just too great trying to correct the very low frequencies.
I agree regarding RCF and so many other big prosound manufactures now promoting time corrected (phase corrected) loudspeakers.
It's more the norm now, than not.
I think virtually all of the correction is taking place from around 500Hz up, again due to latency kicking in below that.
I expect IIR will long rule sub land, other than for pure playback folks like me.
Hi bansuri, 64ms.
6144 taps @ 48kHz = 6144/48,000 =128ms.
Using FIR as linear phase centers the impulse response peak, for a delay of 1/2 the entire filter.
Sometimes, i play with moving impulse center around, like moving it all the way to the start for 0 delay, when wanting to just replicate IIR.
Hi Mark,
I wonder if this is enough resolution for the sub
I also wonder how you measure the impulse response, which is difficult for a vented cab.
Otherwise I agree with your
I expect IIR will long rule sub land, other than for pure playback folks like me
also a hassle to adjust the delay in each video app
I wonder if this is enough resolution for the sub
I also wonder how you measure the impulse response, which is difficult for a vented cab.
Otherwise I agree with your
I expect IIR will long rule sub land, other than for pure playback folks like me
also a hassle to adjust the delay in each video app
Great question; and no, I don't think the 6144 taps is enough resolution for that sub in linear phase mode.
As i'm realizing you are aware, resolution is the inverse of the delay.
Or equivalently, the sample rate divided by number of effective taps.
So with 3072 taps in play with linear phase, resolution = 48,000 / 3072 or 15.625Hz.
The rule of thumb i've been taught is that lowest frequency of control should be 3x resolution, or about 47Hz for what i posted.
I've acquired a new processor (a used Q-sys Core) that has 16,384 taps @ 48kHz.
That gives me 5.9Hz resolution, or 18 Hz utilization per 3x rule of thumb.
I've already seen it works that low for electrical filters.
Hoping to get outside soon, on the driveway, to measure the sub itself.
I've always just measure via transfer function, and let the math derive the subs' impulse response.
As i'm realizing you are aware, resolution is the inverse of the delay.
Or equivalently, the sample rate divided by number of effective taps.
So with 3072 taps in play with linear phase, resolution = 48,000 / 3072 or 15.625Hz.
The rule of thumb i've been taught is that lowest frequency of control should be 3x resolution, or about 47Hz for what i posted.
I've acquired a new processor (a used Q-sys Core) that has 16,384 taps @ 48kHz.
That gives me 5.9Hz resolution, or 18 Hz utilization per 3x rule of thumb.
I've already seen it works that low for electrical filters.
Hoping to get outside soon, on the driveway, to measure the sub itself.
I've always just measure via transfer function, and let the math derive the subs' impulse response.
Time coherent loudspeakers above 500 Hertz - sounds good. As for a long time the criteria of linear phase was not a major concern in loudspeaker design - we should be happy now.
In germany Mr. Pfleid, beginning in the 80ies with its loudspeakers made one way loudspeaker designs with a very good 20cm driver in an 8 litre enclosure closed box. He had a patent on the driver construction. He even had a patent on this loudspeaker being able to reproduce square waves correctly on a multiple of frequencies. He had to defend his patent but was able to do it. Now its loudspeaker history.
In bass reproduction time coherence is not a priority. Low distortion and driver membrane stability for good impulse response are important criteria.
In germany Mr. Pfleid, beginning in the 80ies with its loudspeakers made one way loudspeaker designs with a very good 20cm driver in an 8 litre enclosure closed box. He had a patent on the driver construction. He even had a patent on this loudspeaker being able to reproduce square waves correctly on a multiple of frequencies. He had to defend his patent but was able to do it. Now its loudspeaker history.
In bass reproduction time coherence is not a priority. Low distortion and driver membrane stability for good impulse response are important criteria.