Edward J M
Senior HTF Member
- Joined
- Sep 22, 2002
- Messages
- 2,031
Introduction:
$600 can buy a lot of subwoofer these days, particularly if you are shopping the internet. No two brands get mentioned more in this hotly contested and highly competitive class than SVS and Hsu.
And what better match-up than the PB1-ISD vs. STF-3? Both of these $600 subs feature a vented box design, similar enclosure size, a 12" high excursion driver, a fixed tune point with similar rated extension, and a digital BASH amp.
To keep the report from getting too bogged down, there is a lengthy and detailed section named "Post Fight Analysis" at the end of this review discussing the specifics on test equipment and testing methodology. If you are a tech head, you might find this section interesting, otherwise it will probably put you to sleep.
Without further delay, let's see how they stack up. The scores after each round are rated on a scale of 1-10, and there is a possible total of 90 points.
Each score is relative to the $600 subwoofer price class. A high score in a particular category means that you won't find much better performance or features for $600. Scores cannot possibly be applied on an absolute basis, as there is simply too much price variability in the world of subwoofers for that to be a valid approach.
Tale Of The Tape:
In this corner we have the SVS PB1-ISD in gray-stone with black trim, standing 18" wide x 21" high x 20" deep and weighing in at 70 pounds. This subwoofer features a long-throw, down-firing 12" woofer, 320 watt BASH amp with detachable power cord, integrated base-plate with semi-compliant rubber feet, and a 4" (I.D.) port which flares to 6". Also included are four stick-on flexible rubber feet for hard surface applications where the sub might otherwise wander or vibrate at high volumes.
In the opposite corner we have the Hsu STF-3 in menacing matte black, standing 15" wide x 22" high x 22" deep and weighing in at 65 pounds. This subwoofer features a long-throw, down-firing 12" woofer, 300 watt BASH amp with detachable power cord, tall hard plastic spike feet (as opposed to a base-plate), and dual 3" (I.D.) ports which flare to 4.375". Alternate feet for hard surface applications are not included.
The SVS R&D and manufacturing facility is located in Liberty, Ohio. The PB1-ISD was designed, engineered, and built in the United States. For more information, go to www.svsubwoofers.com
The Hsu R&D and general headquarters is located in Anaheim, California. The STF-3 was designed and engineered in the United States, and was built in China. For more information, go to www.hsuresearch.com
Photos of both products are provided below, with a DVD box and CD case included for size reference.
http://home.rochester.rr.com/themull...0&%20STF-3.JPG
Round 1: Fit & Finish, Design Features, and Floor Interface:
The PB1-ISD is currently back-ordered in black, so SVS is offering the colored models at no extra charge. We ordered the gray-stone color, which is handsomely trimmed out with black baseplate dowels. The website pictures don't do this color justice; it has quite a bit more black in the coloration, and is very attractive in person. The coating feels smooth and is reminiscent of a semi-gloss urethane finish. The top and bottom caps are nicely fitted to the cabinet walls, but a tiny cap seam can be seen (in the right light) and felt in the coating. The amp, vent, and baseplate dowel mating surfaces all looked great.
The STF-3 comes in a matte black vacuum-sealed vinyl coating that is fine textured with a hint of satin. The coating application was extremely uniform with no detectable surface defects, giving the STF-3 a handsome, monolithic appearance. Hsu opted for a visible external ridge on the top and bottom caps, and this is an attractive design cue. The amp, woofer, and vents were fitted nicely, and overall this is a first rate enclosure.
The SVS is almost square in dimension, and requires more floor space than the taller and narrower Hsu. If floor space is a premium, make sure to measure for either sub to ensure they can fit. These are both pretty big enclosures in person.
The integrated baseplate and optional compliant rubber feet on the SVS is a superior and more flexible solution than the single set of hard plastic spikes on the Hsu. The baseplate provides a consistent firing surface regardless of the floor treatment, and the user has a choice of feet for carpeting or hard surface. Beware, once you mount the compliant rubber feet on the SVS, they will stick and drag if you try to move it again. Just placing the feet under the baseplate without peeling off the adhesive strip is recommended while you are experimenting with placement.
The hard spiked feet on the Hsu vibrated and buzzed on the Armstrong laminate floor at higher volumes (something other owners have also reported on the Hsu website). If you are using the Hsu on a hard surface, I recommend some small black rubber furniture cups, available at any local home improvement store.
Round 1 Score: Both are finished very well, but the Hsu gets the nod with an almost perfect cabinet finish in every respect. The integrated baseplate and optional compliant rubber feet definitely score points for the SVS. The Hsu takes up less floor space than the SVS, and can be squeezed into a narrower space.
Fit and Finish: SVS 9.5 Hsu 10
Floor Interface Solution: SVS 9.5 Hsu 8
Dimensions and Floor Space: SVS 9 Hsu 10
Round 2: Amplifier Feature Set:
The 320 watt SVS amp features include auto-on, low level L/R inputs, high level L/R inputs with a fixed (100 Hz) 1st order high pass filter, a continuously variable 40-120 Hz 2nd order low pass filter control with a defeat switch, a continuously variable 0-180 phase control, a replaceable fuse, and a detachable power cord.
The 300 watt Hsu amp features include auto-on, low level L/R inputs, high level L/R inputs with a fixed (100 Hz) 1st order high pass filter, a continuously variable 30-90 Hz 2nd order low pass filter control with a defeat switch, a 0/180 phase switch, a replaceable fuse, and a detachable power cord.
Both the SVS and Hsu amps employ a non-user adjustable infrasonic (i.e., high pass) filter to protect the woofer from over-excursion below the tune point. And both amps probably feature some level of mild equalization to flatten, extend, or otherwise optimize the frequency response. This practice is common, and there is nothing inherently wrong with mild equalization to the extent the woofer can tolerate it well, and provided it doesn't overtax the amplifier.
Round 2 Amplifier Feature Set Score: The SVS wins this round with a continuously variable 0-180 phase control compared to the 0/180 switch on the Hsu. SVS 9 Hsu 8
Round 3: Near-Field Frequency Response Measurements:
Measuring near-field helps to eliminate room modes and room gain, and provides a much better indication of the true frequency response of the subwoofer. While not truly anechoic, near-field measurements are nevertheless a very valuable evaluation tool, and can reveal tuning, equalization, and extension differences far better than in-room testing at the listening position.
Near-field measurements were conducted under the same conditions and amplitude for each sub. Refer to the Post Fight Analysis section for details.
The near-field frequency response for each sub was nearly identical from 100-24 Hz. Below 24 Hz, the PB1-ISD begins to exhibit an advantage in extension, holding a 3-5 dB edge in the 22-18 Hz bandwidth.
Round 3 Near-Field FR Score: The SVS takes this round with better near-field extension below 24 Hz, measuring 3-5 dB from 22-18 Hz. SVS 9.5 Hsu 8.5
Provided below is a TrueRTA screen shot of the near-field frequency response curves for both subwoofers. Also included is a photo of the Hsu STF-3 being tested near-field.
Round 4: In-Room Frequency Response Measurements:
In-room frequency response was measured with the mic at the key listening location, in the center of the room, on a chair at head level. In-room FR sweeps will naturally include room modes and room gain, to the extent they are present and cannot be eliminated by manipulating the phase control and the low pass filter.
In-room measurements were conducted under the same conditions and amplitude for each sub. Refer to the Post Fight Analysis section for more details. The in-room FR sweeps were processed through the A/V receiver bass management circuit with an 80 Hz subwoofer crossover frequency and the L/R mains set to small. Measuring the in-room FR sweeps in this manner best reflects how well the subwoofer will interact with the loudspeakers and the room under actual listening conditions.
A moderate amplitude room null was present in the 35-55 Hz region. The continuously variable phase control on the PB1-ISD allowed slightly better control over this null than did the 0/180 phase switch on the STF-3. Other than that, the in-room field frequency response for each sub was nearly identical from 100-24 Hz.
Below 24 Hz, the SVS again showed an extension advantage over the Hsu, measuring about 3-5 dB in the 22-18 Hz region. A TrueRTA screen shot of the in-room frequency response curves for both subwoofers is provided below.
Obviously, moving either sub around the room would have provided a totally different in-room FR curve. However the front left corner has proven to be a good sub location, and furthermore the goal of this comparison test was to illustrate the differences between the two products under identical conditions, with end user optimization of the feature set controls being the only valid variable.
Round 4 In-Room FR Score: The SVS wins this round with a slightly smoother mid bass response (primarily resulting from fine tuning the CV phase control), and noticeably stronger in-room deep extension, measuring 3-5 dB in the 22-18 Hz region. SVS 9.5 Hsu 8.5
Round 5: Total Harmonic Distortion (THD) Testing:
Harmonic distortion occurs when harmonics (multiples) the fundamental signal are produced due to nonlinear behavior of the electrical, magnetic, or mechanical mechanism of the driver.
THD at various frequencies and amplitudes is one of the most important benchmarks of subwoofer performance. A subwoofer with low THD at all frequencies within its normal operating range will sound clean and distinct, while high THD will sound muddy and unclear. At the lowest frequencies, a high THD reading means the listener will feel less of the true fundamental note, and hear more of the false harmonics. It can mean the difference between a clean, spine-tingling organ note which truly pressurizes the room, versus a muddy blur of notes which lack true foundation.
For THD testing, the mic height was fixed at 18" and was placed along the horizontal centerline of the subwoofer, exactly 3 feet away from the depth midpoint. While the depth of the cabinet was only 2" different between the two models, this placement method helps to eliminate cabinet size as a variable, and helps to keep the mic at the same distance from the woofer axis and the port(s) of each model.
THD measurements were conducted at 18, 20, 22, 25, 30, 35, 40, 50, and 60 Hz. Unless otherwise noted, THD was limited to 10%, since this represents the approximate linear operating limits of the driver. This test methodology is employed by Tom Nousaine.
The total sound pressure level specified at each test frequency includes the fundamental group, all distortion harmonics, and the noise floor of the test rig. At 10% THD, the contribution from the harmonics and the noise floor only amounts to a small fraction of a dB; nearly all of the sound pressure results from the fundamental group.
The average SPL over various bandwidths is also provided below. Since the dB scale is log10, it cannot be directly averaged, and doing so is a violation of mathematical laws. In order to correctly calculate an average SPL in dB, the values must be converted to linear units, then arithmetically averaged, and the resultant must then be re-converted to the dB scale.
Regardless, Tom Nousaine directly averages his dB values, and truthfully this method of data presentation (although mathematically incorrect) probably better reflects the real world differences we perceive in subwoofer output in the lowest octaves. Furthermore, many enthusiasts rely on Nousaine's data for comparative purposes, so both methods of calculation are provided below.
SVS PB1-ISD:
18 Hz: 84.3 dB
20 Hz: 94.6 dB
22 Hz: 103.3 dB
25 Hz: 110.3 dB
30 Hz: 119.7 dB
35 Hz: 118.2 dB
40 Hz: 121.7 dB (4.28% THD; amplifier limited)
50 Hz: 115.5 dB (9.14% THD; amplifier limited)
60 Hz: 111.4 dB
Average SPL 18-40 Hz: 116.6 dB (107.4 dB Nousaine Method)
Average SPL 20-40 Hz: 117.3 dB (111.3 dB Nousaine Method)
Average SPL 22-40 Hz: 118.1 dB (114.6 dB Nousaine Method)
Average SPL 25-40 Hz: 119.0 dB (117.5 dB Nousaine Method)
Average SPL 18-60 Hz: 116.4 dB (108.8 dB Nousaine Method)
Average SPL 20-60 Hz: 116.7 dB (111.8 dB Nousaine Method)
Average SPL 22-60 Hz: 117.2 dB (114.3 dB Nousaine Method)
Average SPL 25-60 Hz: 117.9 dB (116.1 dB Nousaine Method)
Hsu STF-3:
18 Hz: 79.2 dB
20 Hz: 87.7 dB
22 Hz: 100.6 dB
25 Hz: 108.7 dB
30 Hz: 118.1 dB (9.36% THD; amplifier limited)
35 Hz: 116.3 dB
40 Hz: 120.1 dB
50 Hz: 114.8 dB (3.89% THD; amplifier limited)
60 Hz: 115.4 dB
Average SPL 18-40 Hz: 114.9 dB (104.4 dB Nousaine Method)
Average SPL 20-40 Hz: 115.6 dB (108.6 dB Nousaine Method)
Average SPL 22-40 Hz: 116.4 dB (112.8 dB Nousaine Method)
Average SPL 25-40 Hz: 117.3 dB (115.8 dB Nousaine Method)
Average SPL 18-60 Hz: 115.0 dB (106.8 dB Nousaine Method)
Average SPL 20-60 Hz: 115.5 dB (110.2 dB Nousaine Method)
Average SPL 22-60 Hz: 116.1 dB (113.4 dB Nousaine Method)
Average SPL 25-60 Hz: 116.7 dB (115.6 dB Nousaine Method)
At certain test frequencies, the 3rd order distortion harmonic from the STF-3 was louder than the 2nd order distortion harmonic. Subjectively, odd-order harmonic distortion is considered more objectionable to the listener than even order harmonic distortion. Even-order harmonics sound similar to the fundamental, but at a higher octave. Odd-order harmonics sound dissonant because they are related to the fundamental only in a mathematical sense, and not a musical one.
At 20 Hz, the SVS holds a nearly 7 dB advantage in clean output over the STF-3. Since the dB scale is log10, a 6 dB increase represents a doubling of the sound amplitude. At 20 Hz, it would therefore require two co-located STF-3's to match the clean output capability of one PB1-ISD.
In the 22-25 Hz region, the SVS holds a 2-3 dB advantage. In the 25-50 Hz bandwidth, the SVS holds a small advantage, and the Hsu took the 60 Hz test frequency by 4 dB.
Clearly, the frequencies in the 20-30 Hz region will be the limiting factors for both subwoofers. In real world use in a moderate size room, the maximum audibly undistorted output for these subwoofers will be in the 110-113 dB region at the listening position, and this coincides well with the SPL peaks measured during the DVD testing section.
A complete description of the THD measurement methodology, equipment, and calibration techniques is discussed in the Post Fight Analysis section.
A TrueRTA screen shot of the 20 Hz and 40 Hz distortion spectrum generated by each subwoofer is provided below.
Round 5 THD Score: At all frequencies except 60 Hz, the SVS holds an advantage in distortion-limited output. At 20 Hz, it would require two co-located STF-3's to match the clean output of one PB1-ISD. In the 22-25 Hz bandwidth, the SVS still holds a significant clean output advantage. Above 25 Hz, the two subwoofers match up very closely. SVS 9.5 Hsu 8.5
Round 6: Blind Listening - Music
It is very important when comparing the performance of two subs on music and DVD, that they be calibrated to the same exact level. A calibration difference of even 2 dB can significantly alter the reviewer's impressions on back-to-back comparisons. Both subs were exactly level matched over the frequency bandwidth where the two FR curves overlapped the longest (essentially the 21-50 Hz region). A complete description of the level matching methodology used for music and DVD is discussed in the Post Fight Analysis section.
The test subject for blind listening is a musician. He has been involved in hi-fi, music, and playing instruments for the better part of 25 years. He knows good sound when he hears it, and I trust his opinion.
I swapped out the subs (only took a few seconds since we used the same interchangeable power cord and sub cable), and took notes on his observations while he listened without knowing which model was playing. Provided below is a listing of the music selection, his subjective impressions, and the preferred subwoofer revealed in parentheses.
1) Kamkiriad (Donald Fagan, Reprise Records) - "Springtime".
Subwoofer A: This sub is nice and lean, with a good tactile response on the deepest sections. Subwoofer B: At times this sub sounds a bit thick and wooly in the mid bass strings. Preference: Subwoofer A (SVS)
2) Kamkiriad (Donald Fagan, Reprise Records) - "Snowbound".
Subwoofer A: I really like the fret work and the finger slides, and it sounds nice and airy with good presence around the strings. Subwoofer B: Well balanced, but the finger slides were a little less pronounced in comparison. Preference: Subwoofer A (Hsu)
3) Kamkiriad (Donald Fagan, Reprise Records) - "Trans Island Skyway".
Subwoofer A: Tympanic strikes have a great balance of skin and kick and it digs really well on low E, staying clean. Subwoofer B: Tympanic strikes have good skin and a bit less kick but still clean. The guitar lines are a little harder to follow on the deepest sections. Preference: Subwoofer A (SVS)
4) Bela Fleck - "Flight Of The Cosmic Hippo"
Subwoofer A: The string reverb stays detailed and I can feel each pulse when it goes really deep. The kick drum is great and overall dynamics were super when things got hot in the middle. Subwoofer B: The string reverb was a just little muddier here - I can't feel each pulse as well on the deepest sections. The kick drum sounds just a tad tubby. The dynamics suffered in the middle (2:40 mark) where it sounded a bit strained compared to sub A. Preference: Subwoofer A (SVS)
5) SoundHound Classical Organ CD - "Track 4"
Subwoofer A: Gets low pretty well, but doesn't seem to convey amplitude changes with authority on the really deep stuff. Subwoofer B: This sub really wants you to wick it up; excellent tactile response and much better room pressure. Rattled more things in the room, for sure. Preference: Subwoofer B (SVS)
6) SoundHound Classical Organ CD - "Track 9"
Subwoofer A: Wow, did you feel it track that pressure change in the beginning (40 second mark)? This sub has some real authority way down low; man this CD is just nuts. Subwoofer B: Less tingle in the chair, less pressure in the air, and real trouble keeping up at the beginning when the pressure changed. Preference: Subwoofer A (SVS)
7) Underworld Soundtrack - Renholder "Now I Know"
Subwoofer A: Some decent air pressure but it's waffling a bit on the synth drum. Subwoofer B: Better...this sub is more iron fisted on the deep stuff with no waffling and better tactile response. Preference: Subwoofer B (SVS)
8) Al DiMeola - Kiss My Axe "The Embrace"
Subwoofer A: Good attack on both mini-kettle hits (2:30 and 3:30 marks) with a hint of flab. The initial hit is deep, with some decay evident. Subwoofer B: Excellent attack on the kettles; holy smokes this CD has major dynamics. Interesting - the initial hit is more tympanic and not as deep sounding, but the decay is more pronounced and deep. I didn't notice such deep decay with sub A.
(requests that both subs be played again in the same order on the same track - this time louder)
Sub A is a bit deeper on the initial hits but has some tub too and has less pronounced deep decay. Subwoofer B hits harder with no tub, and also has deeper decay. They are very close but Sub B gets the nod for better over dynamics. Preference: Subwoofer B (SVS).
Round 6 Blind Listening Music Score: The listener was asked to rank both subs on a scale of 1-10. He feels both are very good products that sound similar on most popular music. He noted the biggest differences on recordings with either very challenging dynamics or extremely deep content (or both). In these two areas, he gave the SVS the nod for being able to play louder and deeper while staying cleaner and providing a better tactile response. SVS 9 Hsu 8
Round 7: Home Theater Performance
Rather than go for sheer quantity on the DVD review, we decided to closely assess a few popular passages on some well known action and fantasy DVDs. The goal was to select difficult scenes which would expose any differences in low end extension, usable output, and dynamic range; all hallmarks of good HT performance.
As with the music comparison, both subs were exactly level matched over the frequency bandwidth where the two FR curves overlapped the longest. A complete description of the level matching methodology used for music and DVD is discussed in the Post Fight Analysis section.
Peak SPL monitoring was done with a Bruel & Kjaer (B&K) model 2205 sound level meter set on C-weighted Fast, mounted on a tripod at the listening position. The SPL values listed below are C-weighted; to obtain the approximate unweighted (i.e., true) SPL for a given passage, add about 4 dB.
1) Star Wars Episode I: Phantom Menace (DD-EX) Master Volume -5
Pod Explosion (1:01:39): SVS 105 dB; Hsu 104 dB
Pod Race Canyon Drop (1:03:34): SVS 110 dB; Hsu 111 dB
Pod Race Canyon Exits (time varies): SVS 105-107 dB; Hsu 105-106 dB
The subs stacked up well on the Pod Race. Both were dynamic, deep and strong. Occasionally the SVS would convey more infrasonic tactile information on the canyon exits, but neither sub was a slouch in this regard. The Hsu performed very well on the difficult Canyon Drop scene.
2) Jurassic Park III (DTS) Master Volume -13
Dino Fight: (0:26:40-0:27:25): SVS 105-107 dB; Hsu 105-106 dB
Both subs shook the room nicely, easily conveying the largely 25 Hz information in this scene.
3) Atlantis - The Lost Empire (DD-EX) Master Volume -8
Tsunami Intro (0:00:36): SVS 107 dB; Hsu 106 dB
Boiler Explosion (0:04:48): SVS 104 dB; Hsu 102 dB
Leviathan Moving (0:20:54): SVS 104 dB; Hsu 105 dB
Leviathan Escape Explosion (0:24:10): SVS 109 dB; Hsu 108 dB
Column Explosion (0:29:12): SVS 104 dB; Hsu 103 dB
Both subs handled Atlantis with aplomb, with only minor differences in character and sound pressure peaks. The rather deep boiler explosion was probably the only scene where the SVS seemed stronger.
4) Daredevil (DTS) Master Volume -12
Bar Fight Deep Somersault (0:23:14-0:23:17): SVS: 100 dB; Hsu 96 dB
Bar Fight Pool Cue Strikes (0:23:29): SVS 106 dB; Hsu 104 dB
Bar Fight Deep Content (0:24:04-0:24:07): SVS: 102-104 dB; Hsu 98-99 dB
This was the first DVD we tested where there was an obvious difference between the two subwoofers. This scene has considerable infrasonic content which the SVS played with noticeably more authority than the Hsu. The SVS pressurized the room well and tightly conveyed changes in amplitude. In comparison, the Hsu missed portions of the infrasonics in this scene and the dynamics were compressed.
5) Gladiator (DTS-ES 6.1 Discrete) Master Volume -12
Maximus and Tiger Falling (1:46:17): SVS 101 dB; Hsu 98 dB
There is a moderately loud but extremely deep aftershock when Maximus and the tiger fall to the ground. This is another scene where the SVS showed an advantage in deep extension and usable output in the 20 Hz region. The aftershock and room decay was visceral and palpable with the SVS; it was more of a dull thud with the Hsu, with no room decay.
6) LOTR-FOTR EE (DTS-ES 6.1 Discrete), Master Volume -12
What would a proper sub shoot-out be without the now infamous ring drop? Both subs well conveyed the weight and body of the ring drop, but the SVS captured a bit more of the infrasonic decay after the initial hit, and this was reflected in the SPL readings.
Sauron Ring Drop (0:03:58): SVS 108 dB; Hsu 106 dB
Round 7 Home Theater Score: Both subs handled HT challenges very well. On most titles and passages, it was hard to tell them apart, actually. In a moderate size room, both were able to cleanly handle playback levels at 10-15 dB below Reference Level, which is plenty loud for most listeners. On certain DVDs with extremely deep bass content, the SVS demonstrated an audible and measurable clean output advantage, conveying infrasonic tactile information better than the Hsu. SVS 9 Hsu 8.5
The Judges Decision - The Post Fight Score Card:
Fit and Finish: SVS 9.5 Hsu 10
Floor Interface Solution: SVS 9.5 Hsu 8
Dimensions and Floor Space: SVS 9 Hsu 10
Amplifier Feature Set: SVS 9 Hsu 8
Near-Field FR: SVS 9.5 Hsu 8.5
In-Room FR: SVS 9.5 Hsu 8.5
Total Harmonic Distortion: SVS 9.5 Hsu 8.5
Blind Listening Music: SVS 9 Hsu 8
Home Theater: SVS 9 Hsu 8.5
Out of a possible 90 points, the PB1-ISD earned 83.5 points for a 93% score, and the STF-3 earned 78 points for an 87% score. The SVS PB1-ISD emerges the winner in this $600 Subwoofer Shoot-Out.
Overall, these are both fine subwoofers, doing justice to both music and HT at playback levels loud enough to please all but the most ardent bass heads. In the end, the SVS scored a win primarily due to better deep extension below 24 Hz and considerably higher distortion-limited output in the 18-25 Hz region, resulting in superior deep dynamics and tactile presentation on both music and HT.
Post Fight Analysis - Technical Notes On Equipment and Methodology:
Description Of Test Equipment
The test rig hardware consists of a Behringer ECM 8000 measurement microphone, powered by a Behringer UB1002 featuring patented Invisible Mic Preamp and Ultra Low Noise circuitry. The mic and preamp were both professionally calibrated by Kim Girardin.
Test rig software is TrueRTA Level 4 v3.0 high resolution real time audio spectrum analyzer by True Audio. This software is run with an Intel Pentium 4 processor with a 2.0 GHz clock speed and 512 MB of DDR SDRAM.
The soundcard is a Creative Technology Audigy 2 ZS with a powerful and sophisticated DSP, 24 bit DACs, and a 108 dB S/N ratio. This soundcard exhibits an excellent dynamic response to a digitally generated chirp, a very low noise floor, an extremely flat frequency response, and a very high dynamic range.
The software accepts correction factors for the mic and preamp, and also corrects for frequency response anomalies in the PC sound card. The overall corrected frequency response of the test rig is 10-25,000 Hz +/- 1 dB or better.
The TrueRTA software also accepts sound pressure level calibration, and the test rig was calibrated with a Bruel & Kjaer (B&K) Model 4230 Sound Level Calibrator. This factory certified calibrator generates a 94 dB tone at 1,000 Hz when the nose of the microphone is placed into the receiver cup.
I am confident the test results accurately reflect the differences in frequency response and distortion characteristics between these two different products. Provided the two subwoofers are still present at the test location, I am open to privately arranging for an independent party to observe the regeneration of this test data.
Initial Set-Up
The test room is approximately 12'x18'x8' with an attached foyer and open stairwell. The floor is a slab-on-grade concrete covered with Armstrong laminate. This floor is acoustically inert and does not transmit any vibration or resonate, making it the ideal platform for evaluating subwoofers.
I have acoustically mapped out several locations in the test room, and have determined that the front left corner (about 12 feet from the prime listening location) provides the best overall response characteristics, with only a moderate null in the 35-55 Hz bandwidth.
Both subs were tested and measured under identical conditions. They were placed in exactly the same location for all measurements and listening tests, with the rear of the sub 6" from the wall, and the side of the sub 2" from the wall.
In-Room Frequency Response
In-room frequency response was measured with the mic at the prime listening location, in the center of the room, on a chair at head level. In-room FR sweeps will naturally include room modes and room gain, to the extent they are present.
In-room FR sweeps were measured using the TrueRTA Quick Sweep feature. The in-room FR sweeps were processed through the A/V receiver bass management circuit with an 80 Hz subwoofer crossover frequency and the L/R mains set to small. Measuring the in-room FR sweeps in this manner best reflects how well the subwoofer will interact with the loudspeakers and the room under actual listening conditions.
Since the Quick Sweep function poses no threat to overheating the voice coils, the in-room FR sweeps were conducted at a realistic test volume, averaging about 100 dB over most of the curve, with room peaks being considerably higher. The best subwoofers will retain a linear frequency response as the volume rises, resisting the tendency to compress dynamics and roll-off the lowest octaves.
During initial set-up and dial-in, each subwoofer was first calibrated with Avia to about 85 dB on the subwoofer tone. The phase and low pass filter controls were then tested in all possible configurations in an effort to get the best possible FR from the sub. Normally the low pass filter is disabled if bass management is conducted at the pre/pro, but this control can be used with considerable success to manipulate the FR in the 60-100 Hz region if the user can accurately measure its effect at each test setting. The best settings for the SVS were with the phase set at 100 degrees, and the low pass filter enabled at its highest setting. The best results with the Hsu were obtained with the phase set to 180 and the low pass filter disabled.
Each final in-room FR curve represents careful level matching over the largest sections of the curves for an easy to visualize comparison. Each FR curve also represents an average (as calculated by TrueRTA) of 5 Quick Sweeps. Both subs were extremely consistent, providing near identical overlay results for each of the five sweeps.
Near-Field Frequency Response
All feature set controls were left at the same settings used for the in-room FR testing. Instead of being placed at the listening position, the mic was now placed directly in front of the subwoofer along the horizontal centerline, and exactly 3 feet away from the depth midpoint of each sub. Using this method helps eliminate cabinet size as a variable, and keeps the mic at the same distance from the woofer axis and the port(s). Mic height was fixed at 18" for all near-field FR testing.
Since the near-field testing was conducted at a much closer mic distance, the test volume for the Quick Sweeps was increased to about 110 dB, again to evaluate the ability of the sub to retain a linear frequency response at high volumes and resist the tendency to compress dynamics and roll-off the lowest octaves.
Near-field FR testing was conducted with a direct input from TrueRTA to the subwoofer versus processing the signal through the AVR bass management system (as was done with the in-room FR testing). This eliminates the possibility that the L/R front mains could affect the near-field FR test results.
Each final near-field FR curve represents careful level matching over the largest sections of the curves for an easy to visualize comparison. Each FR curve also represents an average (as calculated by TrueRTA) of 5 Quick Sweeps. Both subs were extremely consistent, providing near identical overlay results for each of the five sweeps.
THD Analysis
I am fortunate to have quasi-anechoic acoustics at the near-field frequency response and THD test location. This helps eliminate room modes from affecting the THD readings, and provides a truer reflection of the THD-limited output capabilities of the subwoofer. For example, if the frequency response at the THD test location showed a null at 25 Hz, the THD-limited output at that frequency would be artificially low (and the converse would be true for a 25 Hz peak at the test location). Unless the frequency response at the THD test location is verified to be free of peaks and nulls, the THD data will be of limited value, and valid comparisons to THD data sets generated by others will be difficult at best, if not impossible.
For THD testing, the mic height was fixed at 18" and was placed along the horizontal centerline of the subwoofer, exactly 3 feet away from the depth midpoint. While the depth of the cabinet was only 2" different between the two models, this placement method helps to eliminate cabinet size as a variable, and helps to keep the mic at the same distance from the woofer axis and the port(s) of each model.
THD testing was conducted with a direct input from TrueRTA to the subwoofer versus processing the signal through the AVR bass management system (as was done with the in-room FR testing). This eliminates the possibility that the L/R front mains could contribute distortion to the test results. During the THD testing, the low pass filter on each sub was disabled. The phase control for the SVS was left at 100 degrees, and the Hsu was set to 180 degrees.
THD measurements were conducted at 18, 20, 22, 25, 30, 35, 40, 50, and 60 Hz. Unless otherwise noted in the test report, THD was limited to 10%, since this represents the approximate linear operating limits of the driver, where suspension stiffness rises to approximately 4X the resting position, and the motor strength (BL) falls to about 70%.
THD was calculated using the 2nd through 10th order harmonics of the fundamental. The total sound pressure level specified at each test frequency includes the fundamental group, all distortion harmonics, and the noise floor of the test rig. At 10% THD, the contribution from the harmonics and the noise floor only amounts to a small fraction of a dB; nearly all of the sound pressure results from the fundamental group.
Level Matching Methodology On Music and DVD Comparisons
A Bruel & Kjaer (B&K) model 2205 sound level meter was used for initial level calibration, and was also used to measure SPL bass peaks during DVD playback. To ensure the meter was functioning optimally, a fresh battery was installed, and it was then calibrated with the aforementioned B&K Model 4230 Sound Level Calibrator.
To ensure both subs were calibrated to the same level for music and DVD comparisons, the SVS was initially calibrated with Avia to an average of 82-83 dB, as measured at the listening position with the B&K sound meter set to C-weighted Slow. Without altering the gain setting on the SVS amp, TrueRTA was then used to run a Quick Sweep on the SVS to generate an in-room FR curve at the listening position.
The Hsu was then swapped into place and TrueRTA was again used to generate multiple in-room FR curves. The Hsu gain control was slowly adjusted after each FR sweep until the FR curve exactly overlapped that of the SVS in the 24-100 Hz region.
If the FR curves are similar (as is the case with the PB1 and the STF-3), this method of level matching is considerably more accurate than attempting to match levels using just a sound meter with a fluctuating rumble tone on a calibration disc.
$600 can buy a lot of subwoofer these days, particularly if you are shopping the internet. No two brands get mentioned more in this hotly contested and highly competitive class than SVS and Hsu.
And what better match-up than the PB1-ISD vs. STF-3? Both of these $600 subs feature a vented box design, similar enclosure size, a 12" high excursion driver, a fixed tune point with similar rated extension, and a digital BASH amp.
To keep the report from getting too bogged down, there is a lengthy and detailed section named "Post Fight Analysis" at the end of this review discussing the specifics on test equipment and testing methodology. If you are a tech head, you might find this section interesting, otherwise it will probably put you to sleep.
Without further delay, let's see how they stack up. The scores after each round are rated on a scale of 1-10, and there is a possible total of 90 points.
Each score is relative to the $600 subwoofer price class. A high score in a particular category means that you won't find much better performance or features for $600. Scores cannot possibly be applied on an absolute basis, as there is simply too much price variability in the world of subwoofers for that to be a valid approach.
Tale Of The Tape:
In this corner we have the SVS PB1-ISD in gray-stone with black trim, standing 18" wide x 21" high x 20" deep and weighing in at 70 pounds. This subwoofer features a long-throw, down-firing 12" woofer, 320 watt BASH amp with detachable power cord, integrated base-plate with semi-compliant rubber feet, and a 4" (I.D.) port which flares to 6". Also included are four stick-on flexible rubber feet for hard surface applications where the sub might otherwise wander or vibrate at high volumes.
In the opposite corner we have the Hsu STF-3 in menacing matte black, standing 15" wide x 22" high x 22" deep and weighing in at 65 pounds. This subwoofer features a long-throw, down-firing 12" woofer, 300 watt BASH amp with detachable power cord, tall hard plastic spike feet (as opposed to a base-plate), and dual 3" (I.D.) ports which flare to 4.375". Alternate feet for hard surface applications are not included.
The SVS R&D and manufacturing facility is located in Liberty, Ohio. The PB1-ISD was designed, engineered, and built in the United States. For more information, go to www.svsubwoofers.com
The Hsu R&D and general headquarters is located in Anaheim, California. The STF-3 was designed and engineered in the United States, and was built in China. For more information, go to www.hsuresearch.com
Photos of both products are provided below, with a DVD box and CD case included for size reference.
http://home.rochester.rr.com/themull...0&%20STF-3.JPG
Round 1: Fit & Finish, Design Features, and Floor Interface:
The PB1-ISD is currently back-ordered in black, so SVS is offering the colored models at no extra charge. We ordered the gray-stone color, which is handsomely trimmed out with black baseplate dowels. The website pictures don't do this color justice; it has quite a bit more black in the coloration, and is very attractive in person. The coating feels smooth and is reminiscent of a semi-gloss urethane finish. The top and bottom caps are nicely fitted to the cabinet walls, but a tiny cap seam can be seen (in the right light) and felt in the coating. The amp, vent, and baseplate dowel mating surfaces all looked great.
The STF-3 comes in a matte black vacuum-sealed vinyl coating that is fine textured with a hint of satin. The coating application was extremely uniform with no detectable surface defects, giving the STF-3 a handsome, monolithic appearance. Hsu opted for a visible external ridge on the top and bottom caps, and this is an attractive design cue. The amp, woofer, and vents were fitted nicely, and overall this is a first rate enclosure.
The SVS is almost square in dimension, and requires more floor space than the taller and narrower Hsu. If floor space is a premium, make sure to measure for either sub to ensure they can fit. These are both pretty big enclosures in person.
The integrated baseplate and optional compliant rubber feet on the SVS is a superior and more flexible solution than the single set of hard plastic spikes on the Hsu. The baseplate provides a consistent firing surface regardless of the floor treatment, and the user has a choice of feet for carpeting or hard surface. Beware, once you mount the compliant rubber feet on the SVS, they will stick and drag if you try to move it again. Just placing the feet under the baseplate without peeling off the adhesive strip is recommended while you are experimenting with placement.
The hard spiked feet on the Hsu vibrated and buzzed on the Armstrong laminate floor at higher volumes (something other owners have also reported on the Hsu website). If you are using the Hsu on a hard surface, I recommend some small black rubber furniture cups, available at any local home improvement store.
Round 1 Score: Both are finished very well, but the Hsu gets the nod with an almost perfect cabinet finish in every respect. The integrated baseplate and optional compliant rubber feet definitely score points for the SVS. The Hsu takes up less floor space than the SVS, and can be squeezed into a narrower space.
Fit and Finish: SVS 9.5 Hsu 10
Floor Interface Solution: SVS 9.5 Hsu 8
Dimensions and Floor Space: SVS 9 Hsu 10
Round 2: Amplifier Feature Set:
The 320 watt SVS amp features include auto-on, low level L/R inputs, high level L/R inputs with a fixed (100 Hz) 1st order high pass filter, a continuously variable 40-120 Hz 2nd order low pass filter control with a defeat switch, a continuously variable 0-180 phase control, a replaceable fuse, and a detachable power cord.
The 300 watt Hsu amp features include auto-on, low level L/R inputs, high level L/R inputs with a fixed (100 Hz) 1st order high pass filter, a continuously variable 30-90 Hz 2nd order low pass filter control with a defeat switch, a 0/180 phase switch, a replaceable fuse, and a detachable power cord.
Both the SVS and Hsu amps employ a non-user adjustable infrasonic (i.e., high pass) filter to protect the woofer from over-excursion below the tune point. And both amps probably feature some level of mild equalization to flatten, extend, or otherwise optimize the frequency response. This practice is common, and there is nothing inherently wrong with mild equalization to the extent the woofer can tolerate it well, and provided it doesn't overtax the amplifier.
Round 2 Amplifier Feature Set Score: The SVS wins this round with a continuously variable 0-180 phase control compared to the 0/180 switch on the Hsu. SVS 9 Hsu 8
Round 3: Near-Field Frequency Response Measurements:
Measuring near-field helps to eliminate room modes and room gain, and provides a much better indication of the true frequency response of the subwoofer. While not truly anechoic, near-field measurements are nevertheless a very valuable evaluation tool, and can reveal tuning, equalization, and extension differences far better than in-room testing at the listening position.
Near-field measurements were conducted under the same conditions and amplitude for each sub. Refer to the Post Fight Analysis section for details.
The near-field frequency response for each sub was nearly identical from 100-24 Hz. Below 24 Hz, the PB1-ISD begins to exhibit an advantage in extension, holding a 3-5 dB edge in the 22-18 Hz bandwidth.
Round 3 Near-Field FR Score: The SVS takes this round with better near-field extension below 24 Hz, measuring 3-5 dB from 22-18 Hz. SVS 9.5 Hsu 8.5
Provided below is a TrueRTA screen shot of the near-field frequency response curves for both subwoofers. Also included is a photo of the Hsu STF-3 being tested near-field.
Round 4: In-Room Frequency Response Measurements:
In-room frequency response was measured with the mic at the key listening location, in the center of the room, on a chair at head level. In-room FR sweeps will naturally include room modes and room gain, to the extent they are present and cannot be eliminated by manipulating the phase control and the low pass filter.
In-room measurements were conducted under the same conditions and amplitude for each sub. Refer to the Post Fight Analysis section for more details. The in-room FR sweeps were processed through the A/V receiver bass management circuit with an 80 Hz subwoofer crossover frequency and the L/R mains set to small. Measuring the in-room FR sweeps in this manner best reflects how well the subwoofer will interact with the loudspeakers and the room under actual listening conditions.
A moderate amplitude room null was present in the 35-55 Hz region. The continuously variable phase control on the PB1-ISD allowed slightly better control over this null than did the 0/180 phase switch on the STF-3. Other than that, the in-room field frequency response for each sub was nearly identical from 100-24 Hz.
Below 24 Hz, the SVS again showed an extension advantage over the Hsu, measuring about 3-5 dB in the 22-18 Hz region. A TrueRTA screen shot of the in-room frequency response curves for both subwoofers is provided below.
Obviously, moving either sub around the room would have provided a totally different in-room FR curve. However the front left corner has proven to be a good sub location, and furthermore the goal of this comparison test was to illustrate the differences between the two products under identical conditions, with end user optimization of the feature set controls being the only valid variable.
Round 4 In-Room FR Score: The SVS wins this round with a slightly smoother mid bass response (primarily resulting from fine tuning the CV phase control), and noticeably stronger in-room deep extension, measuring 3-5 dB in the 22-18 Hz region. SVS 9.5 Hsu 8.5
Round 5: Total Harmonic Distortion (THD) Testing:
Harmonic distortion occurs when harmonics (multiples) the fundamental signal are produced due to nonlinear behavior of the electrical, magnetic, or mechanical mechanism of the driver.
THD at various frequencies and amplitudes is one of the most important benchmarks of subwoofer performance. A subwoofer with low THD at all frequencies within its normal operating range will sound clean and distinct, while high THD will sound muddy and unclear. At the lowest frequencies, a high THD reading means the listener will feel less of the true fundamental note, and hear more of the false harmonics. It can mean the difference between a clean, spine-tingling organ note which truly pressurizes the room, versus a muddy blur of notes which lack true foundation.
For THD testing, the mic height was fixed at 18" and was placed along the horizontal centerline of the subwoofer, exactly 3 feet away from the depth midpoint. While the depth of the cabinet was only 2" different between the two models, this placement method helps to eliminate cabinet size as a variable, and helps to keep the mic at the same distance from the woofer axis and the port(s) of each model.
THD measurements were conducted at 18, 20, 22, 25, 30, 35, 40, 50, and 60 Hz. Unless otherwise noted, THD was limited to 10%, since this represents the approximate linear operating limits of the driver. This test methodology is employed by Tom Nousaine.
The total sound pressure level specified at each test frequency includes the fundamental group, all distortion harmonics, and the noise floor of the test rig. At 10% THD, the contribution from the harmonics and the noise floor only amounts to a small fraction of a dB; nearly all of the sound pressure results from the fundamental group.
The average SPL over various bandwidths is also provided below. Since the dB scale is log10, it cannot be directly averaged, and doing so is a violation of mathematical laws. In order to correctly calculate an average SPL in dB, the values must be converted to linear units, then arithmetically averaged, and the resultant must then be re-converted to the dB scale.
Regardless, Tom Nousaine directly averages his dB values, and truthfully this method of data presentation (although mathematically incorrect) probably better reflects the real world differences we perceive in subwoofer output in the lowest octaves. Furthermore, many enthusiasts rely on Nousaine's data for comparative purposes, so both methods of calculation are provided below.
SVS PB1-ISD:
18 Hz: 84.3 dB
20 Hz: 94.6 dB
22 Hz: 103.3 dB
25 Hz: 110.3 dB
30 Hz: 119.7 dB
35 Hz: 118.2 dB
40 Hz: 121.7 dB (4.28% THD; amplifier limited)
50 Hz: 115.5 dB (9.14% THD; amplifier limited)
60 Hz: 111.4 dB
Average SPL 18-40 Hz: 116.6 dB (107.4 dB Nousaine Method)
Average SPL 20-40 Hz: 117.3 dB (111.3 dB Nousaine Method)
Average SPL 22-40 Hz: 118.1 dB (114.6 dB Nousaine Method)
Average SPL 25-40 Hz: 119.0 dB (117.5 dB Nousaine Method)
Average SPL 18-60 Hz: 116.4 dB (108.8 dB Nousaine Method)
Average SPL 20-60 Hz: 116.7 dB (111.8 dB Nousaine Method)
Average SPL 22-60 Hz: 117.2 dB (114.3 dB Nousaine Method)
Average SPL 25-60 Hz: 117.9 dB (116.1 dB Nousaine Method)
Hsu STF-3:
18 Hz: 79.2 dB
20 Hz: 87.7 dB
22 Hz: 100.6 dB
25 Hz: 108.7 dB
30 Hz: 118.1 dB (9.36% THD; amplifier limited)
35 Hz: 116.3 dB
40 Hz: 120.1 dB
50 Hz: 114.8 dB (3.89% THD; amplifier limited)
60 Hz: 115.4 dB
Average SPL 18-40 Hz: 114.9 dB (104.4 dB Nousaine Method)
Average SPL 20-40 Hz: 115.6 dB (108.6 dB Nousaine Method)
Average SPL 22-40 Hz: 116.4 dB (112.8 dB Nousaine Method)
Average SPL 25-40 Hz: 117.3 dB (115.8 dB Nousaine Method)
Average SPL 18-60 Hz: 115.0 dB (106.8 dB Nousaine Method)
Average SPL 20-60 Hz: 115.5 dB (110.2 dB Nousaine Method)
Average SPL 22-60 Hz: 116.1 dB (113.4 dB Nousaine Method)
Average SPL 25-60 Hz: 116.7 dB (115.6 dB Nousaine Method)
At certain test frequencies, the 3rd order distortion harmonic from the STF-3 was louder than the 2nd order distortion harmonic. Subjectively, odd-order harmonic distortion is considered more objectionable to the listener than even order harmonic distortion. Even-order harmonics sound similar to the fundamental, but at a higher octave. Odd-order harmonics sound dissonant because they are related to the fundamental only in a mathematical sense, and not a musical one.
At 20 Hz, the SVS holds a nearly 7 dB advantage in clean output over the STF-3. Since the dB scale is log10, a 6 dB increase represents a doubling of the sound amplitude. At 20 Hz, it would therefore require two co-located STF-3's to match the clean output capability of one PB1-ISD.
In the 22-25 Hz region, the SVS holds a 2-3 dB advantage. In the 25-50 Hz bandwidth, the SVS holds a small advantage, and the Hsu took the 60 Hz test frequency by 4 dB.
Clearly, the frequencies in the 20-30 Hz region will be the limiting factors for both subwoofers. In real world use in a moderate size room, the maximum audibly undistorted output for these subwoofers will be in the 110-113 dB region at the listening position, and this coincides well with the SPL peaks measured during the DVD testing section.
A complete description of the THD measurement methodology, equipment, and calibration techniques is discussed in the Post Fight Analysis section.
A TrueRTA screen shot of the 20 Hz and 40 Hz distortion spectrum generated by each subwoofer is provided below.
Round 5 THD Score: At all frequencies except 60 Hz, the SVS holds an advantage in distortion-limited output. At 20 Hz, it would require two co-located STF-3's to match the clean output of one PB1-ISD. In the 22-25 Hz bandwidth, the SVS still holds a significant clean output advantage. Above 25 Hz, the two subwoofers match up very closely. SVS 9.5 Hsu 8.5
Round 6: Blind Listening - Music
It is very important when comparing the performance of two subs on music and DVD, that they be calibrated to the same exact level. A calibration difference of even 2 dB can significantly alter the reviewer's impressions on back-to-back comparisons. Both subs were exactly level matched over the frequency bandwidth where the two FR curves overlapped the longest (essentially the 21-50 Hz region). A complete description of the level matching methodology used for music and DVD is discussed in the Post Fight Analysis section.
The test subject for blind listening is a musician. He has been involved in hi-fi, music, and playing instruments for the better part of 25 years. He knows good sound when he hears it, and I trust his opinion.
I swapped out the subs (only took a few seconds since we used the same interchangeable power cord and sub cable), and took notes on his observations while he listened without knowing which model was playing. Provided below is a listing of the music selection, his subjective impressions, and the preferred subwoofer revealed in parentheses.
1) Kamkiriad (Donald Fagan, Reprise Records) - "Springtime".
Subwoofer A: This sub is nice and lean, with a good tactile response on the deepest sections. Subwoofer B: At times this sub sounds a bit thick and wooly in the mid bass strings. Preference: Subwoofer A (SVS)
2) Kamkiriad (Donald Fagan, Reprise Records) - "Snowbound".
Subwoofer A: I really like the fret work and the finger slides, and it sounds nice and airy with good presence around the strings. Subwoofer B: Well balanced, but the finger slides were a little less pronounced in comparison. Preference: Subwoofer A (Hsu)
3) Kamkiriad (Donald Fagan, Reprise Records) - "Trans Island Skyway".
Subwoofer A: Tympanic strikes have a great balance of skin and kick and it digs really well on low E, staying clean. Subwoofer B: Tympanic strikes have good skin and a bit less kick but still clean. The guitar lines are a little harder to follow on the deepest sections. Preference: Subwoofer A (SVS)
4) Bela Fleck - "Flight Of The Cosmic Hippo"
Subwoofer A: The string reverb stays detailed and I can feel each pulse when it goes really deep. The kick drum is great and overall dynamics were super when things got hot in the middle. Subwoofer B: The string reverb was a just little muddier here - I can't feel each pulse as well on the deepest sections. The kick drum sounds just a tad tubby. The dynamics suffered in the middle (2:40 mark) where it sounded a bit strained compared to sub A. Preference: Subwoofer A (SVS)
5) SoundHound Classical Organ CD - "Track 4"
Subwoofer A: Gets low pretty well, but doesn't seem to convey amplitude changes with authority on the really deep stuff. Subwoofer B: This sub really wants you to wick it up; excellent tactile response and much better room pressure. Rattled more things in the room, for sure. Preference: Subwoofer B (SVS)
6) SoundHound Classical Organ CD - "Track 9"
Subwoofer A: Wow, did you feel it track that pressure change in the beginning (40 second mark)? This sub has some real authority way down low; man this CD is just nuts. Subwoofer B: Less tingle in the chair, less pressure in the air, and real trouble keeping up at the beginning when the pressure changed. Preference: Subwoofer A (SVS)
7) Underworld Soundtrack - Renholder "Now I Know"
Subwoofer A: Some decent air pressure but it's waffling a bit on the synth drum. Subwoofer B: Better...this sub is more iron fisted on the deep stuff with no waffling and better tactile response. Preference: Subwoofer B (SVS)
8) Al DiMeola - Kiss My Axe "The Embrace"
Subwoofer A: Good attack on both mini-kettle hits (2:30 and 3:30 marks) with a hint of flab. The initial hit is deep, with some decay evident. Subwoofer B: Excellent attack on the kettles; holy smokes this CD has major dynamics. Interesting - the initial hit is more tympanic and not as deep sounding, but the decay is more pronounced and deep. I didn't notice such deep decay with sub A.
(requests that both subs be played again in the same order on the same track - this time louder)
Sub A is a bit deeper on the initial hits but has some tub too and has less pronounced deep decay. Subwoofer B hits harder with no tub, and also has deeper decay. They are very close but Sub B gets the nod for better over dynamics. Preference: Subwoofer B (SVS).
Round 6 Blind Listening Music Score: The listener was asked to rank both subs on a scale of 1-10. He feels both are very good products that sound similar on most popular music. He noted the biggest differences on recordings with either very challenging dynamics or extremely deep content (or both). In these two areas, he gave the SVS the nod for being able to play louder and deeper while staying cleaner and providing a better tactile response. SVS 9 Hsu 8
Round 7: Home Theater Performance
Rather than go for sheer quantity on the DVD review, we decided to closely assess a few popular passages on some well known action and fantasy DVDs. The goal was to select difficult scenes which would expose any differences in low end extension, usable output, and dynamic range; all hallmarks of good HT performance.
As with the music comparison, both subs were exactly level matched over the frequency bandwidth where the two FR curves overlapped the longest. A complete description of the level matching methodology used for music and DVD is discussed in the Post Fight Analysis section.
Peak SPL monitoring was done with a Bruel & Kjaer (B&K) model 2205 sound level meter set on C-weighted Fast, mounted on a tripod at the listening position. The SPL values listed below are C-weighted; to obtain the approximate unweighted (i.e., true) SPL for a given passage, add about 4 dB.
1) Star Wars Episode I: Phantom Menace (DD-EX) Master Volume -5
Pod Explosion (1:01:39): SVS 105 dB; Hsu 104 dB
Pod Race Canyon Drop (1:03:34): SVS 110 dB; Hsu 111 dB
Pod Race Canyon Exits (time varies): SVS 105-107 dB; Hsu 105-106 dB
The subs stacked up well on the Pod Race. Both were dynamic, deep and strong. Occasionally the SVS would convey more infrasonic tactile information on the canyon exits, but neither sub was a slouch in this regard. The Hsu performed very well on the difficult Canyon Drop scene.
2) Jurassic Park III (DTS) Master Volume -13
Dino Fight: (0:26:40-0:27:25): SVS 105-107 dB; Hsu 105-106 dB
Both subs shook the room nicely, easily conveying the largely 25 Hz information in this scene.
3) Atlantis - The Lost Empire (DD-EX) Master Volume -8
Tsunami Intro (0:00:36): SVS 107 dB; Hsu 106 dB
Boiler Explosion (0:04:48): SVS 104 dB; Hsu 102 dB
Leviathan Moving (0:20:54): SVS 104 dB; Hsu 105 dB
Leviathan Escape Explosion (0:24:10): SVS 109 dB; Hsu 108 dB
Column Explosion (0:29:12): SVS 104 dB; Hsu 103 dB
Both subs handled Atlantis with aplomb, with only minor differences in character and sound pressure peaks. The rather deep boiler explosion was probably the only scene where the SVS seemed stronger.
4) Daredevil (DTS) Master Volume -12
Bar Fight Deep Somersault (0:23:14-0:23:17): SVS: 100 dB; Hsu 96 dB
Bar Fight Pool Cue Strikes (0:23:29): SVS 106 dB; Hsu 104 dB
Bar Fight Deep Content (0:24:04-0:24:07): SVS: 102-104 dB; Hsu 98-99 dB
This was the first DVD we tested where there was an obvious difference between the two subwoofers. This scene has considerable infrasonic content which the SVS played with noticeably more authority than the Hsu. The SVS pressurized the room well and tightly conveyed changes in amplitude. In comparison, the Hsu missed portions of the infrasonics in this scene and the dynamics were compressed.
5) Gladiator (DTS-ES 6.1 Discrete) Master Volume -12
Maximus and Tiger Falling (1:46:17): SVS 101 dB; Hsu 98 dB
There is a moderately loud but extremely deep aftershock when Maximus and the tiger fall to the ground. This is another scene where the SVS showed an advantage in deep extension and usable output in the 20 Hz region. The aftershock and room decay was visceral and palpable with the SVS; it was more of a dull thud with the Hsu, with no room decay.
6) LOTR-FOTR EE (DTS-ES 6.1 Discrete), Master Volume -12
What would a proper sub shoot-out be without the now infamous ring drop? Both subs well conveyed the weight and body of the ring drop, but the SVS captured a bit more of the infrasonic decay after the initial hit, and this was reflected in the SPL readings.
Sauron Ring Drop (0:03:58): SVS 108 dB; Hsu 106 dB
Round 7 Home Theater Score: Both subs handled HT challenges very well. On most titles and passages, it was hard to tell them apart, actually. In a moderate size room, both were able to cleanly handle playback levels at 10-15 dB below Reference Level, which is plenty loud for most listeners. On certain DVDs with extremely deep bass content, the SVS demonstrated an audible and measurable clean output advantage, conveying infrasonic tactile information better than the Hsu. SVS 9 Hsu 8.5
The Judges Decision - The Post Fight Score Card:
Fit and Finish: SVS 9.5 Hsu 10
Floor Interface Solution: SVS 9.5 Hsu 8
Dimensions and Floor Space: SVS 9 Hsu 10
Amplifier Feature Set: SVS 9 Hsu 8
Near-Field FR: SVS 9.5 Hsu 8.5
In-Room FR: SVS 9.5 Hsu 8.5
Total Harmonic Distortion: SVS 9.5 Hsu 8.5
Blind Listening Music: SVS 9 Hsu 8
Home Theater: SVS 9 Hsu 8.5
Out of a possible 90 points, the PB1-ISD earned 83.5 points for a 93% score, and the STF-3 earned 78 points for an 87% score. The SVS PB1-ISD emerges the winner in this $600 Subwoofer Shoot-Out.
Overall, these are both fine subwoofers, doing justice to both music and HT at playback levels loud enough to please all but the most ardent bass heads. In the end, the SVS scored a win primarily due to better deep extension below 24 Hz and considerably higher distortion-limited output in the 18-25 Hz region, resulting in superior deep dynamics and tactile presentation on both music and HT.
Post Fight Analysis - Technical Notes On Equipment and Methodology:
Description Of Test Equipment
The test rig hardware consists of a Behringer ECM 8000 measurement microphone, powered by a Behringer UB1002 featuring patented Invisible Mic Preamp and Ultra Low Noise circuitry. The mic and preamp were both professionally calibrated by Kim Girardin.
Test rig software is TrueRTA Level 4 v3.0 high resolution real time audio spectrum analyzer by True Audio. This software is run with an Intel Pentium 4 processor with a 2.0 GHz clock speed and 512 MB of DDR SDRAM.
The soundcard is a Creative Technology Audigy 2 ZS with a powerful and sophisticated DSP, 24 bit DACs, and a 108 dB S/N ratio. This soundcard exhibits an excellent dynamic response to a digitally generated chirp, a very low noise floor, an extremely flat frequency response, and a very high dynamic range.
The software accepts correction factors for the mic and preamp, and also corrects for frequency response anomalies in the PC sound card. The overall corrected frequency response of the test rig is 10-25,000 Hz +/- 1 dB or better.
The TrueRTA software also accepts sound pressure level calibration, and the test rig was calibrated with a Bruel & Kjaer (B&K) Model 4230 Sound Level Calibrator. This factory certified calibrator generates a 94 dB tone at 1,000 Hz when the nose of the microphone is placed into the receiver cup.
I am confident the test results accurately reflect the differences in frequency response and distortion characteristics between these two different products. Provided the two subwoofers are still present at the test location, I am open to privately arranging for an independent party to observe the regeneration of this test data.
Initial Set-Up
The test room is approximately 12'x18'x8' with an attached foyer and open stairwell. The floor is a slab-on-grade concrete covered with Armstrong laminate. This floor is acoustically inert and does not transmit any vibration or resonate, making it the ideal platform for evaluating subwoofers.
I have acoustically mapped out several locations in the test room, and have determined that the front left corner (about 12 feet from the prime listening location) provides the best overall response characteristics, with only a moderate null in the 35-55 Hz bandwidth.
Both subs were tested and measured under identical conditions. They were placed in exactly the same location for all measurements and listening tests, with the rear of the sub 6" from the wall, and the side of the sub 2" from the wall.
In-Room Frequency Response
In-room frequency response was measured with the mic at the prime listening location, in the center of the room, on a chair at head level. In-room FR sweeps will naturally include room modes and room gain, to the extent they are present.
In-room FR sweeps were measured using the TrueRTA Quick Sweep feature. The in-room FR sweeps were processed through the A/V receiver bass management circuit with an 80 Hz subwoofer crossover frequency and the L/R mains set to small. Measuring the in-room FR sweeps in this manner best reflects how well the subwoofer will interact with the loudspeakers and the room under actual listening conditions.
Since the Quick Sweep function poses no threat to overheating the voice coils, the in-room FR sweeps were conducted at a realistic test volume, averaging about 100 dB over most of the curve, with room peaks being considerably higher. The best subwoofers will retain a linear frequency response as the volume rises, resisting the tendency to compress dynamics and roll-off the lowest octaves.
During initial set-up and dial-in, each subwoofer was first calibrated with Avia to about 85 dB on the subwoofer tone. The phase and low pass filter controls were then tested in all possible configurations in an effort to get the best possible FR from the sub. Normally the low pass filter is disabled if bass management is conducted at the pre/pro, but this control can be used with considerable success to manipulate the FR in the 60-100 Hz region if the user can accurately measure its effect at each test setting. The best settings for the SVS were with the phase set at 100 degrees, and the low pass filter enabled at its highest setting. The best results with the Hsu were obtained with the phase set to 180 and the low pass filter disabled.
Each final in-room FR curve represents careful level matching over the largest sections of the curves for an easy to visualize comparison. Each FR curve also represents an average (as calculated by TrueRTA) of 5 Quick Sweeps. Both subs were extremely consistent, providing near identical overlay results for each of the five sweeps.
Near-Field Frequency Response
All feature set controls were left at the same settings used for the in-room FR testing. Instead of being placed at the listening position, the mic was now placed directly in front of the subwoofer along the horizontal centerline, and exactly 3 feet away from the depth midpoint of each sub. Using this method helps eliminate cabinet size as a variable, and keeps the mic at the same distance from the woofer axis and the port(s). Mic height was fixed at 18" for all near-field FR testing.
Since the near-field testing was conducted at a much closer mic distance, the test volume for the Quick Sweeps was increased to about 110 dB, again to evaluate the ability of the sub to retain a linear frequency response at high volumes and resist the tendency to compress dynamics and roll-off the lowest octaves.
Near-field FR testing was conducted with a direct input from TrueRTA to the subwoofer versus processing the signal through the AVR bass management system (as was done with the in-room FR testing). This eliminates the possibility that the L/R front mains could affect the near-field FR test results.
Each final near-field FR curve represents careful level matching over the largest sections of the curves for an easy to visualize comparison. Each FR curve also represents an average (as calculated by TrueRTA) of 5 Quick Sweeps. Both subs were extremely consistent, providing near identical overlay results for each of the five sweeps.
THD Analysis
I am fortunate to have quasi-anechoic acoustics at the near-field frequency response and THD test location. This helps eliminate room modes from affecting the THD readings, and provides a truer reflection of the THD-limited output capabilities of the subwoofer. For example, if the frequency response at the THD test location showed a null at 25 Hz, the THD-limited output at that frequency would be artificially low (and the converse would be true for a 25 Hz peak at the test location). Unless the frequency response at the THD test location is verified to be free of peaks and nulls, the THD data will be of limited value, and valid comparisons to THD data sets generated by others will be difficult at best, if not impossible.
For THD testing, the mic height was fixed at 18" and was placed along the horizontal centerline of the subwoofer, exactly 3 feet away from the depth midpoint. While the depth of the cabinet was only 2" different between the two models, this placement method helps to eliminate cabinet size as a variable, and helps to keep the mic at the same distance from the woofer axis and the port(s) of each model.
THD testing was conducted with a direct input from TrueRTA to the subwoofer versus processing the signal through the AVR bass management system (as was done with the in-room FR testing). This eliminates the possibility that the L/R front mains could contribute distortion to the test results. During the THD testing, the low pass filter on each sub was disabled. The phase control for the SVS was left at 100 degrees, and the Hsu was set to 180 degrees.
THD measurements were conducted at 18, 20, 22, 25, 30, 35, 40, 50, and 60 Hz. Unless otherwise noted in the test report, THD was limited to 10%, since this represents the approximate linear operating limits of the driver, where suspension stiffness rises to approximately 4X the resting position, and the motor strength (BL) falls to about 70%.
THD was calculated using the 2nd through 10th order harmonics of the fundamental. The total sound pressure level specified at each test frequency includes the fundamental group, all distortion harmonics, and the noise floor of the test rig. At 10% THD, the contribution from the harmonics and the noise floor only amounts to a small fraction of a dB; nearly all of the sound pressure results from the fundamental group.
Level Matching Methodology On Music and DVD Comparisons
A Bruel & Kjaer (B&K) model 2205 sound level meter was used for initial level calibration, and was also used to measure SPL bass peaks during DVD playback. To ensure the meter was functioning optimally, a fresh battery was installed, and it was then calibrated with the aforementioned B&K Model 4230 Sound Level Calibrator.
To ensure both subs were calibrated to the same level for music and DVD comparisons, the SVS was initially calibrated with Avia to an average of 82-83 dB, as measured at the listening position with the B&K sound meter set to C-weighted Slow. Without altering the gain setting on the SVS amp, TrueRTA was then used to run a Quick Sweep on the SVS to generate an in-room FR curve at the listening position.
The Hsu was then swapped into place and TrueRTA was again used to generate multiple in-room FR curves. The Hsu gain control was slowly adjusted after each FR sweep until the FR curve exactly overlapped that of the SVS in the 24-100 Hz region.
If the FR curves are similar (as is the case with the PB1 and the STF-3), this method of level matching is considerably more accurate than attempting to match levels using just a sound meter with a fluctuating rumble tone on a calibration disc.
