The irony of the kick drum at the beginning of the song above to this conversation is, "wild", as the kids would say.
On the subject of test tracks, much of the Hugh Padgham-produced Phil Collins material has pretty tight, 'attacky' drums sounds, and played/programmed by a brilliant drummer!
It can be quite visceral on a good PA, but usually flat as a steamrollered pancake on most tiny, compressed home systems...
It can be quite visceral on a good PA, but usually flat as a steamrollered pancake on most tiny, compressed home systems...
I have heard kick from PA "heads". I presume that they play down to 50Hz or so. But to note is that the effect was without low end reinforcement. I wonder if it might be that, in the case of full-range systems (i.e., with LF reinforcement), excessive group delay in the low end might destroy the effect. Any thoughts?
True, verified.
Though the designs for some PA systems are ruining what you need for "sense of presence", IE on very good recordings you need low GD even lower in frequency, sense of space from large but slow volume of air movement recorded by the microphones. Many PA setups have excessive GD near FS because of tuning, they simply ignore anything happening below 40-50-60hz because it is not cost effective, you need so much larger designs to produce the low frequency content and to put it bluntly most people are sheep who will never notice it until they are taught the difference.
Last edited:
Art, I wanted to let this lay because it appears to be of interest only to you and I.
But I disagree with you, and think the view you stated paints a one-sided picture of how measurements work.
I know for certain that frequency response measurement does not have to automatically come first.
(And when it does, nit-picking here, it is an inverse-FFT that parses it to back to the time domain.)
I routinely take impulse measurements first, that then compute the frequency domain from them, using a variety of standard noise signals, and sine sweeps/chirps.
Frequency response graphs are nearly always the first thing measurement software shows, frequency response being of primal interest,
But the order of displayed info doesn't assuredly reflect the order of data analysis/computation of the time domain vs frequency domain.
I think until we understand exactly how each measurement software program we are using works...how it calculates spectrum analysis, single and dual channel transfers and impulses ....we don't know much of anything about chicken vs egg.
The more i research this, the more convinced I get the frequency domain and the time domain, are as inseparable in analysis, .....as the chicken and egg are in nature.
Understanding this science is not just academic exercise for me. I've aged past academics for academics sake LoL
Trying to find real-world better tuning results, standard procedure has become to take impulse responses, instead of transfer functions.
Frequency domain info is the same either way.....but i find the time domain info of impulse is superior.
Why should better time domain info matter?
If I tune a speaker via transfer function to achieve flat mag and phase, I know I'll get a good looking impulse response.
Along with beautiful step, waterfalls, spectro's, yada yada... (the ole identity property in play 😉
But despite those perfect mag and phase graphs, there has been variance in how the same speakers sound, that were tuned on repeated tries in the same conditions, to give the same flat mag and phase, and same IFT derived excellent impulse.
What I've found (though instruction from the author of Crosslite+), is to look beyond the acoustic impulse, and into the electrical filter impulse.
It's utterly remarkable how different the impulse responses can be of FIR (and IIR) filters, that give the same mag and phase traces (and overall speaker impulse)
I've found the cleaner I can make the electrical impulses, the cleaner the acoustic impulses become....all the while mag and phase stay the same good ole flat.
The sound is cleaner/clearer...it's like the noise floor is reduced. Between notes is quieter, notes are a little more vibrant. For lack of better words.
FWIW...
A frequency response could be considered a "one-sided transfer function" of the frequency domain with no reference to the time domain, the transient response to an impulsive signal.
My view merely disagreed with your statement that "transient response = frequency response" is a "fact".
No disagreement, the hardware and software used determine how you measure the frequency response and/or transient response.
Right, two speakers can have identical (equal) frequency response, but far different (unequal) transient response to an impulsive signal.
A subwoofer that takes more time to react to an impulse, and rings long after, won't have the same "kick" as one that has better transient response.
Keep up the good work!
Art
Surely this suggests that it is being used outside it's (low passed) capabilities.
If we take two drivers, one flat to 200Hz, the other flat to 1000Hz and low pass steeply at, say 80Hz, why should the speaker with a higher bandwidth react differently to an impulse?
If you are considering an unfiltered signal, then I agree, but that's not quite the real world...
I also absolutely agree that any driver or enclosure which rings long after the signal has ceased will necessarily blur signals in the time domain. To my ear and in my experience, 'tight bass' (Sorry to use that term, couldn't think of an alternative!) can be correlated to low energy storage in the subwoofer.
Hi, as I understand ringing after signal cease is just system Q. the woofer would return to rest with Q amount of cycles of resonant frequency. For example, if you simulate a closed bass box and Q is 0.5 and resonant frequency is 50Hz, it takes 0.5 cycle of 50Hz so 10ms to settle from what ever position the cone is at, the settling happens at the resonant frequency, not some arbitrary high frequency what made the cone move / halt. This would produce some amount of sound at resonant frequency, small boom when you stop signal suddenly, very small since displacement might have been very small, compared to how much displacement was needed to get equally loud sound at the resonant frequency.
Now, unhook the woofer from amplifier and start drumming it, and it booms like a drum, a wicked long resonance! What the heck!? The system Q changed, because electrical damping was removed by making infinite impedance between driver terminals, and Q shot up. But you'd see this immediately as high peak in frequency response if you measured the freq response with high output impedance amplifier. This stuff can be read mostly, if not completely, from frequency response.
Last edited:
The filtered signal that goes to the subwoofer driver physically has ringing. Simple mathematical fact - a band limited signal cannot be time limited, and the converse is true. It’s not the subwoofers’s JOB to clean up the impulse response, which has a lot of high frequency content. Even a gated kick drum has a lot of higher frequency content ADDED to get that fast decay. The listener hears what comes out of the sub AND the tops - and to fully and properly reproduce the edges requires the summed output of BOTH. How well that decay is matched depends on the overall response flatness or at least smoothness, and how closely in time the subwoofer’s output and the tops match.
That's why I don't have low tuned BP4's in my cars. I TRY to stay in the 60-70hz range for midbass attack and low frequency extension. A 60hz BP4 should get a 30-90hz response in the average car. I also use an 120hz lowpass filter to bring up the midbass.
SH50 = BP4 enclosure.
TH need a rising response instead of flat response. Flat responses look good ground plane, but when you get to 1.0 and 0.5pi, that low end gain kicks in on TH's. I had my home theater TH firing into the corner initially. As soon as I flipped the enclosure 180*, the midbass kicked in.
You guys need to check out the Devastator thread on avsforum.com. Most of the BP6P designs enhance midbass frequencies.
look(listen/measure) for the 'blur' and interesting mess within that area further with EQ or steep filters?
What do you think?
they make it 'louder', but do they 'enhance it'?
The paraflex is ‘louder’ there too, but it’s all messed up because the low tuned resonator gets closer to a full wave length (cancelation) before the added resonator length on the other side ‘matches’ up and it’s audibly weird @ and above 3 * Fb.
Attachments
I always thought it is the front facing reflex port that causes kick in chest effect. None of my sealed woofers do it.
See air cannon:
See air cannon:
Hehe yes, I was very amazed when I saw air cannon for the first time. But looking at that you can easily start figuring up explanations. The air cannon fires donut shaped vortexes, that whirl into themselves, that's why they travel slower. That kind of output would destroy sound quality of a reflex port for sure, so you'd image loudspeakers are designed to avoid this. That's why we now have flared ports with optimized flow and carefully calculated port and cabinet size ratios. But some ports are simple tubes that are even less perfect. With correct volume and circumstances, they should be able to fire fast air pressure waves that kick chest. Actually, isn't that the reflex effect? Probably sealed woofers can too, but to generate same sort of pressure wave they need some wild travel.
When I put my hand in front of a woofer, no touching, I normally won't feel the music in my fingers. But when I put my hand in front of a port, no touching, I can feel the resonant frequency beats even at low volumes.
I'm to busy/lazy to test with precise methods to say for sure.
When I put my hand in front of a woofer, no touching, I normally won't feel the music in my fingers. But when I put my hand in front of a port, no touching, I can feel the resonant frequency beats even at low volumes.
I'm to busy/lazy to test with precise methods to say for sure.
Oh I can feel the air moving in front of small speaker. Well fed 4" can be felt, and by 6" it is very obvious. It deends on the distance I guess. Near by the air indeed can add to the experience, but after 3-4m, it is out of question. Though the aerodynamics is interesting. The port certainly can blow 1-2m into distance, but it does not suck the air in in the same way. Sucking is omnidirectional.
No, the "reflex effect" is the slug of air contained in the port moving in phase with the driver around the box Helmholtz resonance frequency (Fb), the resulting waves travel at the speed of sound.
The "windy" air pressure waves created when the slug of air spills out of the port travel at a small fraction of the speed of sound, and won't be in phase with the driver or port output past a few centimeters.
Out of phase output reduces "kick".