Sealed vs Ported Loudspeakers: Which is Better?
sealed vs ported - courtesy of dummies.com
This is an age-old question: Which is better? There are certainly points on each side of the argument to consider. The debate of ported vs. sealed is intrinsically relevant and interesting, regardless of the arbitrary brand involved.
When AR (Acoustic Research of Cambridge MA) introduced its original AR-1 acoustic suspension system in 1954 (see figure 1), it represented a quantum leap forward in the advancement of the reproduction of accurate, extended, low-distortion bass from a home loudspeaker. Noted reviewer Julian Hirsch remarked to me that he “knew the industry would never be the same” when he heard the AR-1’s bass.
Fig 1 AR-1 of 1954
Using the air trapped inside a sealed cabinet as the restoring force to the woofer cone resulted in a clean, linear response that was a step forward over what existed at the time and is still extremely competitive, even today. The AR-1 (and its famous descendants like the AR-3, -3a, and –LST) had an honest, measurably verified frequency response that was smooth and level down to a minus 3dB point of 35 Hz, with less than 5% THD at that frequency at a 20-watt drive level! There are simply no current passive loudspeakers using just a single bass driver in an enclosure well under 2 cu.ft. that could match that performance, even today.
Not even the big Klipsch’s and Bozaks of the AR-1’s day could reach down to -3 dB at 35 Hz, with less than 5% THD, and they were 4-6-8 cu. ft. boxes. Sealed was a real advancement at the time. Without it, great bass from small boxes wouldn’t have existed, KLH, Advent and EPI wouldn’t have followed AR, and the whole stereo industry wouldn’t have exploded in sales like it did. The smaller box in the living room was icing on the cake (with stereo coming along in 1958, now you needed two speaker boxes), but that’s not what drove Ed Villchur to invent the AR-1. The advancement in bass performance drove AR, not “small box marketing,” as some revisionist historians have opined.
Since then, however, as pure research—led in a major way by Dr. Thiele—and computer technology have advanced through the ensuing years, better ported designs are now possible, to the point where acoustic suspension’s pure performance advantages, such as low distortion and linear response (lack of "boominess") are less apparent. In fact, the trade-offs in greater efficiency and lower 3dB down point often point in favor of a ported approach to system design.
What has changed over the last 50+ years to essentially eliminate this gap? Several factors:
1. Quantifiable, known research and documentation of driver, port, and enclosure parameters, combined with the arrival of fast, economical computing in the commercial corporate environment. Specifically, the work of Thiele and Small in the ‘70’s removed much of the previous uncertainty regarding vented speaker design, and when coupled with the newly available inexpensive computing power, it meant that a new level of speaker design accuracy was achievable. In the 40’s, 50’s and 60’s, designing a vented speaker involved a lot of cut ‘n try guesswork. Designers had to be happy with ‘close enough’ or ‘in the ballpark,’ because the precise calculations needed for optimum outcomes hadn’t yet been done. Sealed designs were more forgiving by nature, and didn’t sound offensive or low-fi when they were less than perfectly done. A poorly done vented system, in contrast, could sound like a bad acoustical joke.
There is a school of thought that points out that the T-S parameters are a bit of a simplification and therefore not fully accurate. A former KEF engineer states:
“The small signal behavior of the loudspeaker can indeed be calculated to better than 95% accuracy from a measurement of the impedance of the loudspeaker. To do this one needs a good model of the equivalent circuit of the loudspeaker system, than do an impedance fit to this model. From that, the equivalent circuit can show the current flow through the components that represent the vent mass and the driver mass, and hence calculate their acceleration. This is the equivalent to the far field response, ignoring cabinet diffraction.
One can then calculate the response from the impedance, measure the nearfield of vent plus driver, measure the inbox pressure, measure the farfield (in a well-calibrated chamber), and get all of these to agree to within 0.5dB down to 30Hz.
However, if it is then converted to T/S parameters, due to them ignoring some of the driver behavior, the result will not be as accurate.
Given that most people who measure the low frequency response do not have access to an anechoic chamber, this is actually a better way for them to get a reasonable idea of the LF performance.”
2. Computer-aided design programs. With today’s design programs, such as Finite Element Analysis (FEA) and 3D modeling programs such as Pro/Engineer, hundreds or thousands of driver/enclosure prototypes can be built in the virtual world of the computer before even one is actually built in the lab. A virtually unlimited number of iterations of magnet structure, cone shape, cone thickness, surround/spider compliance, voice coil length, coil winding, and pole plate variations can be tried and evaluated, before any physical parts are ever used. Different port lengths and geometries can be tried, vibrational analysis can be applied to ‘virtual’ cabinet walls to discover resonant modes and ascertain the structural integrity needed to optimize cabinet design within the cost target of the project. By the time a physical prototype is constructed, it is already most of the way there, with no big surprises.In contrast, in the good ‘ol days of the big bass reflex floorstander, after first trying differing port lengths/tunings, how many iterations and prototypes of the driver itself did they do? Two? Maybe three? Each time, hand-cutting a new pole piece, hand-winding a new voice coil, ordering and waiting for delivery of new magnets….
3. Advanced manufacturing processes and materials. Today’s most advanced loudspeaker manufacturing plants utilize very close-tolerance procedures with advanced adhesives and materials technology that were simply unavailable a generation ago.
It’s interesting to note a few basic factors that still fuel the sealed vs. ported issue, even today. At low frequencies, increasing SPL requirements demand ever-greater woofer excursion from a sealed system, which limits output and can increase distortion. In a properly tuned vented system, the port is providing the majority of low-frequency output, and the driver’s excursion—and hence its distortion—is very low.
The upshot of all this is that today’s speakers can be much better performers than yesterday’s if the company and/or designer so chooses. The blueprint for excellent performance is better known than it was in years past, and a skilled designer with clear, worthy goals working with today’s best tools can achieve astonishing performance at incredibly modest prices. It’s less a question of whether to take Sealed Avenue or Ported Road, and more a question of what you’ve chosen as your destination.
Unless you believe that the world’s finest, modern speaker companies are concerned primarily with speaker efficiency and the -3dB point, to the detriment of bass quality, the article does not really explain the reason that most of the world’s finest, modern speaker companies prefer ported designs.
It all comes down to one fact: cone excursion is vastly lower with ported designs than with sealed, acoustic suspension enclosures. This translates in a rather immediate way to reduced non-linear distortion in the deep bass. With the acoustic suspension approach, the strategy is to make the suspension very floppy, using the pressure differential across the cone to provide the force that opposes the motor force. The assumption is that the restoring force will be a linear function of cone displacement. But the obvious questions are with the linearity of the motor, and also the spider, which is part of the mechanical suspension. The voice coils are very long indeed, which leads to greater self-inductance, and also greater DC resistance and thus reduced sensitivity.
The alternative approach is to avoid the large cone excursion. This fundamental advantage gets lost in all the talk about -3B points and group delay. Group delay is accompanied by a time shift for output in the deep bass, relative to the rest of the spectrum, and becomes audible when, but only when, attenuation is severe. As long as the enclosure and port are tuned to adequately low frequency, such that you do not get the steep drop-off below the -3 dB point, it is not a problem at all. But there are other drawbacks. If the diameter of the port is not great enough, compression will occur, leading to reduced low-frequency output at higher volume levels. I.e., turn up the volume loud enough, and the bass disappears, and as it does so, the cone excursion increases. And this occurs regardless at frequencies below the -3 dB point, i.e., you still have high cone excursion and high distortion. There are other potential disadvantages that may or may not be significant, depending on who you ask: organ pipe effect and midrange leakage.
For medium-sized satellite speakers that rely on a subwoofer for deep bass, there is probably no significant difference between the two types, but it depends on the size of the satellite speakers, and also on whether equalization is used. With “home theater” speaker systems that use little tiny satellite speakers, the ported approach can be advantageous if used in conjunction with equalization, the reason being that you can use equalization to correct anomalies in the frequency response and extend the bass, while the port has the net, appreciable effect of reducing cone excursion, thereby reducing distortion. This same advantage, by the same reasoning, can apply to any speaker where porting is combined with equalization, whether the equalization is built in to the speaker as in the case of a powered monitor, or is accomplished by means of (ostensible) room-correction speaker setup in a modern home theater receiver.
Ported designs also make sense with very large speakers with very large woofers, since with these speakers, the benefit of reduced cone excursion still applies, as it invariably does. Most powered monitors are ported, because the driver itself does not displace sufficient volume to be capable of strong output in the deep bass. The required amount of excursion would be so great that high distortion would occur. With this type of speaker, you tune the enclosure and the port to a frequency well below the frequency that the alignment rules says that you should use, and then use equalization to correct the anomalies in the frequency response. This type of speaker is very smart from the additional perspective of the crossover, since by using separate amplifiers for each driver, the crossover is rendered immune from interaction with the complex impedance of the driver.
As a side story, in terms of sealed versus ported speakers, I first got into this hobby with acoustic suspension speakers. The first speakers I ever purchased were a pair of Cambridge Soundworks Model Six speakers. For those that are unaware, Edgar Villchur had a partner at AR when he released the AR-1, former student Henry Kloss. Kloss would go on to found KLH (he was the K), Advent, and Cambridge Soundworks. The CSW Model Six was/is (theoretically) loosely based on the old KLH Model Six. I couldn't qualify them as being great speakers by any stretch, but they were quite easy to listen to. On the other side of the coin though, I didn't find anything magical about acoustic suspension there either, and I've never caught myself pining for my old Model Sixes in spite of the fact that every speaker I own now is ported.
For any woofer optimal box sizes for a sealed and ported alignment are going to be very different. There are significant differences in driver selection and design for the two alignments.
I'll beg to differ on the 'lousy engineering' part - but you're certainly correct on the second half of your statement.
That is why we use a completely different variant of the Ultra 13.5" driver in our smaller SB13-Ultra subwoofer. That driver is optimized for operation in a sealed cabinet of that volume.
- We revised motor geometry (overhung) with a very high force factor (BL^2/Re). Extensive modeling and acoustic testing indicated that this is the preferred motor alignment for a sealed application in this size cabinet.
- We use a unique voice coil (bifilar wound, 3" diameter, aluminum, and with a longer winding height) optimized for a sealed application, resulting in higher power handling and excellent thermal management and heat dissipation.
- We also added a unique gap extension plate (nested in the top plate) for a more symmetrical force/displacement profile, resulting in lower distortion and increased linear stroke.
In the case of the PB13-Ultra, that variant of the 13.5" driver (which is completely different that the SB13-Ultra driver) is optimized for a large cabinet volume. As such, the performance remains very good, even in sealed mode. The PB13-Ultra driver is not terribly happy in a small sealed cabinet - but as you noted it performs very well in sealed mode in a large cabinet - and that's why we confidently offer this operating mode for the PB13-Ultra.
You'll note the very low Q knee and shallow roll-off profile. This minimizes phase change with respect to frequency, and the associated GD curve is very favorable. The PB13-Ultra sounds fantastic in sealed mode, and has plenty of output for mid-size rooms.
The sealed performance is respectable, but F3 is probably too high for HT enthusiasts who want to feel the explosions.
With respect to deep extension, the PB13U in sealed mode is only down about 7 dB @ 20 Hz, relative to the 40-80 Hz reference drive level. Due to the very shallow roll-off slope (something you won't often encounter in commercial sealed offerings), when used a mid-size enclosed room which exhibits a modicum of 'room gain' (i.e., the acoustic transfer function below the modal/pressure transition frequency of the listening space) the PB13U in sealed mode can/will exhibit remarkably deep in-room extension.
My personal reference HT room has rigidly constructed boundaries and is fairly tightly enclosed, and exhibits about 18 dB of room gain at 10-12 Hz. The PB13U in sealed mode in this particular room is flat to almost 10 Hz.
Further evidence of this can been seen in PBC's review of the SB13-Ultra (which has a very similar system Q and roll-off slope as the PB13U in sealed mode). Even in his mid-size and rather open/lossy listening space, the SB13-Ultra measured flat to 15 Hz in-room.
http://forums.audioholics.com/forums/subwoofers/81423-pbcs-svsound-sb13-ultra-user-review.html [forums.audioholics.com]
For any woofer optimal box sizes for a sealed and ported alignment are going to be very different. There are significant differences in driver selection and design for the two alignments.
Or a more balanced viewpoint is offering multiple tuning modes including the option to seal a ported design opens wider placement options and is in fact excellent engineering in doing so. The PB13-Ultra is this example one of engineered excellence as any owner or person that's actually heard one could vouch for.