Here's an interesting article championing sealed design:
SP Acoustics SP1M
Is it possible to have the inherent defects of vented designs (i.e. the opposite of the highlighted section above) and for it not to be fatiguing to listen to? The only downside to sealed designs seems to be cabinet volume, and for the dedicated DIY-er this isn't a fundamental issue. My sealed woofer cabinets are almost as big as the SP1M as it turns out.
Probably most can do a reasonable job of it at a subdued volume level - it's when asking for realistic SPLs that things become undone on many setups ...
We have an excellent entertainer locally who made a career of recreating the Sinatra sound, over many years - had the vocal skills to pull it off nicely. Went to see a live show of him, which had the full big band backing him ... disasterous!! The orchestra was PA'ed to the hilt, and badly in every sense. It was a battering ram of ferocious grit, and I clenched inwardly every time a crescendo hit - literally totally exhausted by the end of the show ... that's not fatigue, this is fatigue ...
Im familiar with the readings. I cannot see fatigue being concerned. Reading between the lines it only say that sealed has more usable low frequency, which is out of context for me.
The writer mentioned what is good with vented is the powerfull midbass. I mentioned the same thing and i related this to fatigue.
Sealed and classic TL needs certain energy level from amplifier to produce the wide range of lows. We have read in this thread where people relate fatigue with inefficiency. I use Dave Koz to test the midbass. With vented it is easier to make the music to sound like live. With sealed it is not because of the energy required to produce the midbass dominant with Dave Koz music.
I would agree with this. There are differences between sealed and vented but that down't mean that one will cause more fatigue than the other. In this thread many have been assuming that any measurable distortion must cause fatigue. (Where's the proof?)
This overstates the difference. Also "very precise time domain response" is nonsense. Both are precisely defined by the math of filtered systems.
Sealed is a 2nd order rolloff system (as are tuned lines). Vented is fourth order. The difference is roll off rate (obviously) and also transient response associated with a given extension frequency. It is true that if, say comparing 35 Hz cuttoff sealed vs. the same vented, that the transient response of the vented box will ring somewhat longer than a sealed box.
No one has ever shown that to be an issue of fatigue, or even necessarily an audible difference between the types. I have done testing in the past where I could electrically change the Q of a sealed system. Although the common assumption is that higher Q means an audibly poor transient response, I could hear the response differencee but not real as a time effect. Higher Q (within reason) sounded like more level at the corner frequency rather than some slow, turgid, long ringing mess. I think I was comparing Q of 1+ to Q of 0.5. The situation may change if you compare Q of 4 to Q of 0.5.
Now I don't need to argue this with audiophiles who will be convinced that higher Q must be an audibly degraded transient response ("just look at the impulse!"), but the fact remains that no one has proven that vented sounds inferior to sealed due to its higher rolloff rate. It is easy to take a purist stand, and much harder to prove significance in a blind test. I would suggest that all consider the much greater effects of the room(!) and also muse over the much reduced distortion that a vented box with its lower excursion, will have.
If you want to confuse yourself on the subject, you should also consider how response extension impacts transient response. If I have a given design (vented or sealed) and I magically extend its response down one Octave, its transient response will stretch to double the original time: longer ringing = worse transient response. 4 Octave = 16 times as long an impulse response. Extend it to near DC and the impulse response becomes infinitely long. It rings forever. So how do we improve our transient response? Does response extension go hand in hand with "slow bass"?
You also have to ask how to make an apples with apples comparison? Toole suggests that our judgement of response extension is roughly tied to the -10dB point of a system. Are we comparing systems with the same -10 point? Same cabinet volume? Same corner shape (note that a vented box can't be defined as a "Q" as it always has 2 Q values).
Certainly a sealed box can have a bumped corner and lingering impulse response, while a vented box can be designed with softer corners and a faster decay. Read Small's articles where he goes into transient response and effective reverberation time.
Careful with our generalizations, here.
David
In the case of the vented speaker, we are hearing both the port output, and the cone in parallel. As I understand it, at resonance, the cone is not moving much, so perhaps then, the result is a straightforward phase shift..? But what is the situation a few Hz above this, when we are hearing both acoustic outputs in parallel? Presumably we hear the sum of the two outputs and, in a steady state sinusoidal signal situation it would simply look like a phase shift. But in the case of transient signals, don't we get an output from the cone followed by a delayed output from the port, giving a result rather more complex than just a phase shift (the famous "time smear")? The sealed box on the other hand only has the one acoustic source, so its output can be characterised as a simple phase shift..?
Last edited:
Nope.
Even though we know physically that the sound is coming from 2 sources, the phase and frequency response of the combined system tells us everything. There is no time smear beyond that implicit in the measured response.
This is a basic tenant to the Fourier Transform: phase and amplitude response totally define the impulse response. Likewise, the impulse response totally defines the phase and amplitude response.
David
He's just trying to sell the speaker. Like speaker reviewers they have to say good things without selling themselves and you will know if they dont really mean it.
I guess so. Trying to work these things out intuitively is generally a mistake!
So, anyway, as a simple system it can be phase corrected electronically, should we so choose.
That still leaves the issue of high distortion from uncontrolled cone movement below the port resonance. Is that a real issue in your opinion?
Personally speaking I found that fatiguing speakers usually have problems above 1.5kHz or higher.
There are not many fundamentals up there so could it be that fatigue is somehow related to how well the harmonics fit the fundamentals ie too high in volume?
I find not many live instruments fatiguing but excess use of cymbals or trumpets can do it.
There are not many fundamentals up there so could it be that fatigue is somehow related to how well the harmonics fit the fundamentals ie too high in volume?
I find not many live instruments fatiguing but excess use of cymbals or trumpets can do it.
Thats a tough one to generalize about. You will have more excursion for signals below cutoff but significantly less above. Which is better? You have to make assumptions about the general spectrum of music. I think with small speakers you may be worse off vented (in terms of total excursion). With larger speakers you are probably better off vented.
You can also pick allignments with more woofer stiffness (tends to more ripple) and they are less susceptable to subsonic over excursion.
David
Particularly larger speakers with more surface area (large driver or multiple drivers) and/or linear excursion.
IMO most of the sound of a ported enclosure that is generally deemed detrimental isn't so much the result of the freq. response (with added group delay) - in that the room itself is often causing irregularities in response that are usually worse, but rather impedance and phase angle as a load on any particular amplifier.
In fact, the speaker as a "load" may be causing fatigue with respect to *amplifier* performance.
So as a general case, a ported speaker presents a more difficult load than sealed?
I wasn't aware of that. The impedance difference between sealed and vented is that vented has two LF resonance peaks and sealed has one. Impedance magnitude of the peaks is typically about the same. Phase angle swing is about the same too.
Amps don't like the combination of high phase angle at low impedance as it tends to swing them out of their safe operating area. Still, most modern amps are pretty tough and can deliver low distortion into a wide range of impedance and phase.
A big jump to say that vented is generally fatiguing from this cause, in my opinion.
David
I disagree- the vented cabs tend to have lower impedance spikes, with the reactive notch from the vent tuning right in the middle of the impedance spike as it would be in a similar sized sealed system. In other words, the highest impedance part is cut out by the tuning. They get even lower if you damp the vents, and lower the Q of the tuning accordingly.
There are still some amps that don't like spikey impedance, tube amps and class D in particular, though with class D it's more in the HF that there's a concern, where any series notches or inductive rise from the drivers interact with the output filter. With tube amps with output transformers, the trafo can have a tough time keeping up with the impedance spike(s) (passing the properly multiplied impedance on to the tube in the circuit).
There are still some amps that don't like spikey impedance, tube amps and class D in particular, though with class D it's more in the HF that there's a concern, where any series notches or inductive rise from the drivers interact with the output filter. With tube amps with output transformers, the trafo can have a tough time keeping up with the impedance spike(s) (passing the properly multiplied impedance on to the tube in the circuit).
In part, but also depends on amplifier, as would type of ported enclosure. A standard BR would present a difficult load, whereas a MLTL would control this much better. Sealed aka acoustic suspension presents the easiest load. Again depends on the amplifier, valves having the greatest difficulty.
I personally look at this problem as it relates to RF design. These pressure/phase/dampening issues are akin to reflected energy (emf) in an antenna system. Reflected energy, causes VSWR, is driven back down the line ( be it a transmission line or coaxial) placing a massive load on the amplifier due to impedance mismatch. Rooms do the very same thing on the low end of the spectrum.
I am trying to figure out the time course on what you folks are calling "fatigue".
Is it the case that the speakers start out sounding good/fine and then become fatguing after how long (1's of minutes, 10's of minutes or 100's of minutes)? When it does finally occur, is this gradual or abrupt? Do "less fatigung" speakers take longer to become fatiguing? Is there a general consensus about the time course?
Is it the case that the speakers start out sounding good/fine and then become fatguing after how long (1's of minutes, 10's of minutes or 100's of minutes)? When it does finally occur, is this gradual or abrupt? Do "less fatigung" speakers take longer to become fatiguing? Is there a general consensus about the time course?
Not sure where you guys are getting that vented is a worse load.
You need to look at the equivilant circuit for a vented box. The woofer starts with the coil impedances in series with a parallel tank that represents the woofer mass and compliance. Going to vented puts a series resonant notch circuit around the previous tank. The notch is at the vent frequency and splits the original peak into two peaks. The two peaks are generally lower in height. The dip in the middle is still at least the resistance of the voice coil, so it never drops below DCR and isn't any lower than frequencies away from resonance already are. At the dip the load is resistive.
I don't see where any of this is a worse load. The minimum level is never any lower, the peaks tend to be a little less tall (probably reducing phase angle swing). Again, a "bad" load can only be one with way to low a resisitive component, or worse, a combination of high phase angle in combination with fairly low resistance. (lowest resistance is generally at a 0 degree phase angle) This fact was beat into our heads by Peter Walker of Quad who designed a lot of amplifiers and worried a lot about currentt protection.
Anyhow, anyone who is shying away from vented speakers for fear of the amplifier load they provide needs more important things to lose sleep over.
David
You need to look at the equivilant circuit for a vented box. The woofer starts with the coil impedances in series with a parallel tank that represents the woofer mass and compliance. Going to vented puts a series resonant notch circuit around the previous tank. The notch is at the vent frequency and splits the original peak into two peaks. The two peaks are generally lower in height. The dip in the middle is still at least the resistance of the voice coil, so it never drops below DCR and isn't any lower than frequencies away from resonance already are. At the dip the load is resistive.
I don't see where any of this is a worse load. The minimum level is never any lower, the peaks tend to be a little less tall (probably reducing phase angle swing). Again, a "bad" load can only be one with way to low a resisitive component, or worse, a combination of high phase angle in combination with fairly low resistance. (lowest resistance is generally at a 0 degree phase angle) This fact was beat into our heads by Peter Walker of Quad who designed a lot of amplifiers and worried a lot about currentt protection.
Anyhow, anyone who is shying away from vented speakers for fear of the amplifier load they provide needs more important things to lose sleep over.
David
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.