$2,000-3,000 Tower speaker comparison: Need help picking the contenders

[NewHTbuyer]

I'm very interested as well. For me it's either a pair of Ascend Towers (with the NRT tweeter = $2k) or the SVS = $2k.

I'm thinking I like the form factor of the Ascend Towers a lot - 7.5" wide, just like my favorite speaker (figure it out). And I generally like soft dome tweeters more than metal tweeters. And it's 30 lbs lighter, which is a plus in my book.

Still, SVS is a fine company and the Ultra tower looks to be a LOT of speaker for the money.

So, very interested in your comments. And try the Ascends with the NRT tweeter!

I had a couple others suggest trying the NrT towers, but I spoke with Dave at Ascend and we mutually agreed not to go that route at this time.

From a design perspective I think the Ascend are better loudspeakers. (if you can, push for the tweet upgrade...)



Why?

Soft dome tweeters (especially silk) tend to breakup earlier in frequency than metal.

It's a common misconception that metal sounds metallic and silk sounds silky (whatever that means...)

Soft domes tend to flex by nature (and design), much more, & quicker than metal but they are much more damped so it doesn't cause irritation (or huge FR peaks). If the designer knows what he is doing, IMO, a stiff dome (read metal) will be superior 99.9998% of the time as they don't flex very much at all which is great for detail (RAAL anyone??). I have seen VERY few metal tweeters that breakup within the audible band.

Why do you think the Ascends are better from a design perspective? Also, in my brief listening, the SVS have not seemed too hot or forward.

I just finished getting them all set up. On Monday I plan to tweak positioning etc. to try to get them dialed in. So, I don't want to really post many listening impressions yet.

I do have a question. Over on AVS one member mentioned that he thought the crossover from the midrange to the dual 8" woofers, which is at 160Hz, is too low, especially if you use the towers with subs and crossover at 80Hz. It would seem to me that is not a design flaw as long as the midrange can play down to 160hz and the crossover is done right. Any thoughts from the experts? I do agree that for HT, the extra low-end on the Ultra towers might be a waste.

Here are some pics:

 

[NewHTbuyer]

More pics:

 

[monkish54]

Why do you think the Ascends are better from a design perspective?

They cross very large midrange drivers to a tweeter on a 180 degree waveguide (baffle) at a fairly high frequency. There will be a large directivity mismatch between the tweeter and midrange. This will most likely cause less accurate imaging in most rooms.

The Ascend (the RAAL one especially) has a pretty good polar response and likely has better imaging in larger rooms..

Over on AVS one member mentioned that he thought the crossover from the midrange to the dual 8" woofers, which is at 160Hz, is too low, especially if you use the towers with subs and crossover at 80Hz.

I agree with him...

It would seem to me that is not a design flaw as long as the midrange can play down to 160hz and the crossover is done right. Any thoughts from the experts?

Two 8" woofers will have less distortion (in the form of VC heat, thermal power compression, etc) than two midrange drivers. This is why the Revel Salon 2 uses three 8" woofers as apposed to one 14". The heat is spread across 3 high power handling 8" drivers as apposed to one very high power handling 14" driver.

Also, why not have more bass produced by the larger surface area?

I'm sure they did it for a good reason..perhaps the midrange have better HD at 160hz? Perhaps the woofers can't play very high with minimal distortion? This is only a 2K speaker with many drivers and large ones at that; Not to mention the complex XO...cost has to be cut somewhere...

 

[Dennis Murphy]

I don't know how high the midranges are crossed to the tweeter, so I can't comment on that. But there shouldn't be a real issue unless it's over 3 kHz. As for the lower crossover point, you have to cross low if the woofers are on the side. Otherwise they will beam to the sides and you won't get a proper blend in the front listening position. So 160 Hz isn't a design flaw--it's a necessity. Is this an all-passive system, or are the woofers actively driven? It's quite difficult to achieve a 160 Hz crossover point passively, but if they've pulled it off, more power to them.

 

[NewHTbuyer]

I don't know how high the midranges are crossed to the tweeter, so I can't comment on that. But there shouldn't be a real issue unless it's over 3 kHz. As for the lower crossover point, you have to cross low if the woofers are on the side. Otherwise they will beam to the sides and you won't get a proper blend in the front listening position. So 160 Hz isn't a design flaw--it's a necessity. Is this an all-passive system, or are the woofers actively driven? It's quite difficult to achieve a 160 Hz crossover point passively, but if they've pulled it off, more power to them.

The tweeter to mid xover is 2kHz. Then they have a bottom midrange taper frequency at 950 Hz and then the xover to the woofers at 160. Mark Mason at RMAF explained why they did not do a MTM for the two mids and the tweeter, but it was mostly over my head. They are all passive.

From a theoretical standpoint, what would be advantages of using five inch mids instead of the 6.5 inch? I can assume that since the xover to the side firing woofers is so low, a smaller mid would not play down to 160 Hz effectively, but that is just a guess.

What are xovers in the Phil 2? Also, the Kef Blade has side firing woofers, I just looked and they cross the tweeter to mid at 3kHz (but it is coaxial). They cross to the four side firing woofers at 350 Hz.

Interesting

 

[Irvrobinson]

Two 8" woofers will have less distortion (in the form of VC heat, thermal power compression, etc) than two midrange drivers. This is why the Revel Salon 2 uses three 8" woofers as apposed to one 14". The heat is spread across 3 high power handling 8" drivers as apposed to one very high power handling 14" driver.

Also, why not have more bass produced by the larger surface area?

I'm sure they did it for a good reason..perhaps the midrange have better HD at 160hz? Perhaps the woofers can't play very high with minimal distortion? This is only a 2K speaker with many drivers and large ones at that; Not to mention the complex XO...cost has to be cut somewhere...

The Salon 2 design decision is in the brochure, and you're half right. Voecks also wanted a narrower cabinet. (Though I suppose he could have gone narrow and deep, a la NHT. Whatever.) Here's the quote:

In both instances the multiple woofer arrays
in Salon2 and Studio2 offer a fundamental
advantage over the more traditional “big” woofer
approach without sacrificing a decibel of bass
foundation. The smaller cone diameters permit a
narrower cabinet waistline, critical for managing
diffraction artifacts. The result is pristinely focused
imaging and as wide a soundstage as the recording
will permit. Bass extension is likewise preserved by
virtue of the robust collective output of the
multiple woofers. For example, the three 8-inch
woofers of a single Salon2 displace almost the
equivalent diaphragm area of a single 14-inch
woofer. Similarly the total area of the twin 8-inch
woofers of a single Studio2 is actually greater than
an 11-inch woofer. Who said you can’t have your
cake and eat it too?

 

[Dennis Murphy]

The tweeter to mid xover is 2kHz. Then they have a bottom midrange taper frequency at 950 Hz and then the xover to the woofers at 160. Mark Mason at RMAF explained why they did not do a MTM for the two mids and the tweeter, but it was mostly over my head. They are all passive.

From a theoretical standpoint, what would be advantages of using five inch mids instead of the 6.5 inch? I can assume that since the xover to the side firing woofers is so low, a smaller mid would not play down to 160 Hz effectively, but that is just a guess.

What are xovers in the Phil 2? Also, the Kef Blade has side firing woofers, I just looked and they cross the tweeter to mid at 3kHz (but it is coaxial). They cross to the four side firing woofers at 350 Hz.

Interesting

A 2k crossover to the tweeter is actually quite low--so there shouldn't be any issue. They have combined a 2.5 design and an MTM configuration. The lower midrange provides baffles step compensation and then rolls off, leaving the upper midrange to work with the tweeter at the crossover point. If you're crossing at 160 Hz, there's no point in using 6.5" woofers as the mids--5" drivers can usually go down to 55-60 Hz, so 160 Hz would be absolutely no problem, and the smaller drivers have better dispersion at higher frequencies. It seems like a well thought out design--no need for anyone to bash it from an engineering standpoint. The Phil's are crossed to the ribbon tweet at 2.9 kHz. The planar midrange is much narrower than a conventional 5" midrange, and therefore has better horizontal dispersion and can be crossed higher. The cross at the low end is at 600 Hz. The very low-mass planar driver can't reproduce lower frequencies well, and so the 8" woofer handles the load. The flip side is that the lower mass planar is in theory better suited to higher frequencies than a conventional driver. But the test is always in the listening, not the theory.

 

[LS35a]

Tower between 2 & 3K, has anyone mentioned this?


SoundStage! Hi-Fi | SoundStageHiFi.com | SoundStage! Hi-Fi | SoundStageHiFi.com

 

[TLS Guy]

I don't know how high the midranges are crossed to the tweeter, so I can't comment on that. But there shouldn't be a real issue unless it's over 3 kHz. As for the lower crossover point, you have to cross low if the woofers are on the side. Otherwise they will beam to the sides and you won't get a proper blend in the front listening position. So 160 Hz isn't a design flaw--it's a necessity. Is this an all-passive system, or are the woofers actively driven? It's quite difficult to achieve a 160 Hz crossover point passively, but if they've pulled it off, more power to them.

Dennis, you and I know you never really do pull that off. You end up with huge inductors and caps. Unless the inductors are very expensive you end up with high insertion resistance which really changes system Q and upsets tuning. You also end up with a an amp buster load and big amp stresses.

I gave up that road long ago. If you really need to crossover that low a semi active or fully active speaker is required. I think for passive designs a crossover point of 350 Hz is pretty much as low as you should go. I do will not do passive designs lower than that and have not for a long time. The trade offs are just not worth it sonically or economically.

The next issue is that unless the speaker is truly a full range one, and I can tell you having done it, that is a huge undertaking, that the speaker is going to be used with a sub. Then you have a situation were you have crossover points an octave or less apart. That is never a good thing, as it makes for an almost impossible perfect integration. Good designs, use the highest bandwidth drivers possible in the mid range and get the crossover points as far apart as possible, three octaves preferably.

So I agree totally with the AVS poster.

 

[Dennis Murphy]

TLS-type Guy: I said a passive cross at 160 Hz was very difficult (and I personally have never been able to pull it off). But if SVS hasn't ended up with a weird impedance profile or other problems in the midbass, then it's really not up to us to judge. I would have to see measurements and listen for myself. I agree that the lower crossover point they've chosen isn't well suited to use with a sub. But that probably wasn't the design intent. In my prior post, I thought SVS was using 5" midranges, not 6.5". But the low 2 kHz cross minimizes any problems from the larger diameter. I've done many 2-way designs with 6.5" drivers, and they seem to have met a market test.

 

[Irvrobinson]

TLS-type Guy: I said a passive cross at 160 Hz was very difficult (and I personally have never been able to pull it off). But if SVS hasn't ended up with a weird impedance profile or other problems in the midbass, then it's really not up to us to judge. I would have to see measurements and listen for myself.

Though I'm not a speaker design expert by any means, I have to agree, and I've argued this point with TLS Guy before. The Salon2 uses a 150Hz passive crossover and measures and sounds great. The Legacy Audio Focus 20/20 had a passive crossover at 80Hz, and while it wasn't the measurement champ that the Salon2 is, it didn't measure all that bad either. When it comes to technical solutions, saying it shouldn't be done or can't be done a certain way is dangerous ground to be treading on. In field after field so-called inferior solutions sometimes win the day with superior execution. Chevy's pushrod, two-valve head, port injection LS7 V8 beats Porsche's latest 911 S 9A1 engine in horsepower per pound of engine weight, by a lot. Intel's x86 CPUs are performance leaders (talk about execution over architecture...). Some companies made compact cassettes sound better than they ever had any right to sound (cassettes make pushrod engines look like a great idea...). FM Stereo radio actually works. Purists get their butts kicked a lot by engineers with determination and funding.

 

[AcuDefTechGuy]

The Salon2 uses a 150Hz passive crossover and measures and sounds great.

I think a few people might agree with you.

I believe Sound & Vision Magazine measured the FR of the Salon2 to be +/-1.3dB from 28Hz-18kHz. I think that's sort of maybe decent.

Maybe Harman International/ Revel does have some okay engineers.

 

[Dennis Murphy]

The issue is not really frequency response. It's impedance and choice of components. If you want to stick with air core inductors and poly caps, you will break the bank on the crossover even if you figure out a way to keep impedance up. If you can go with nonpolar electrolytic caps, things become more manageable. I don't use NPE's because I would take massive tons of abuse on the forums, and there may be something to the contention that they can drift out of spec over the long haul. But truth to tell, they're a very sensible solution in situations like this. (I don't know for a fact that SVS uses NPE's--I would just be surprised if they didn't.)

 

[Irvrobinson]

The issue is not really frequency response. It's impedance and choice of components. If you want to stick with air core inductors and poly caps, you will break the bank on the crossover even if you figure out a way to keep impedance up. If you can go with nonpolar electrolytic caps, things become more manageable. I don't use NPE's because I would take massive tons of abuse on the forums, and there may be something to the contention that they can drift out of spec over the long haul. But truth to tell, they're a very sensible solution in situations like this. (I don't know for a fact that SVS uses NPE's--I would just be surprised if they didn't.)

This is a very good point. The Legacy Audio Focus line used to throw concerns about impedance and phase angle out the window in the 20-100Hz frequency range. I still remember the look on my friend's face when he measured my old pair. Of course, it was just a good excuse to buy Krell monoblocks. One does have to wonder what is going on in the Salon2 crossover, because it has pretty reasonable impedance measurements, certainly no worse than other speakers with more modest crossovers.

I fully understand your reluctance to include components that would cause you to be panned in the forums. You have a rather unique target market at this time, and lots of them have what they think are educated opinions about how you should be designing speakers.

 

[monkish54]

As for the lower crossover point, you have to cross low if the woofers are on the side. Otherwise they will beam to the sides and you won't get a proper blend in the front listening position.

The wavelength of a 250hz wave is 54". What makes you think an 8" woofer would beam to the sides when the wavelength is so large compared to the diaphragm?

As earlier stated the KEF Blade crosses at 350hz.

[TABLE="class: tableStyle1, width: 540"]
<tbody>[TR]
[TD="colspan: 2"]Blade

The world’s first Single Apparent Source loudspeaker.[/TD]
[/TR]
[TR]
[TD="class: label"]Model[/TD]
[TD="class: detail"]Blade[/TD]
[/TR]
[TR]
[TD="class: label"]Design[/TD]
[TD="class: detail"]Three-way bass reflex, Single Apparent Source driver configuration[/TD]
[/TR]
[TR]
[TD="class: label"]Drive Units[/TD]
[TD="class: detail"]Uni-Q driver array:
MF: 125mm (5in.) Li-Mg-Al /LCP hybrid cone
HF: 25mm (1in.) Al dome
Bass units:
LF: 4 x 225mm (9in.) with force cancelling[/TD]
[/TR]
[TR]
[TD="class: label"]Frequency Response[/TD]
[TD="class: detail"]40Hz - 35kHz +/- 3dB
28Hz - 45kHZ +/- 6dB[/TD]
[/TR]
[TR]
[TD="class: label"]Harmonic Distortion[/TD]
[TD="class: detail"]2nd and 3rd harmonics (90dB/1m):
<0.5% 40Hz - 100kHz
<0.2% 200Hz - 10kHz[/TD]
[/TR]
[TR]
[TD="class: label"]Crossover Frequencies[/TD]
[TD="class: detail"]350Hz, 3kHz[/TD]
[/TR]
</tbody>[/TABLE]

I do like they use a 3.5..likely to compensate for the BS.

A 2k crossover to the tweeter is actually quite low--so there shouldn't be any issue.

Not low enough for similar directivity at XO. The wavelength of 2khz is 6.78". The midrange won't be beaming (-5db @ 60° or below), but it sure won't be in the high hundreds like the tweeter! It won't be a rough as the 800D, but there will be a mismatch...

The Salon 2 design decision is in the brochure, and you're half right. Voecks also wanted a narrower cabinet. (Though I suppose he could have gone narrow and deep, a la NHT. Whatever.) Here's the quote:

You're half right. Haha

Voecks: The Salon2 moves more air and has greater output, particularly in the bass. The Salon2's three 8" woofers have a combined area equivalent to a 14" woofer, but the heat generated is spread out among three voice-coils. This means that you won't get the heat buildup that leads to dynamic compression. (As voice-coils heat up, impedance goes up and leads to a mismatch in a speaker's filter network.) The Salon2 is more resistant to dynamic compression than the Studio2 because it has more drivers to dissipate the heat. The Salon2 also has a smaller midrange than the Studio2. This leads to a better match between tweeter and midrange drivers, helping control the Salon2's off-axis response.

 

[Dennis Murphy]

I may have to give you the one concerning beaming. The issue would be more of placement of sound--you can localize a source of sound above above 100 Hz, but I guess even there the woofers would image to the center of the cabinet. So maybe there was another reason they crossed so low--perhaps they wanted as much gap between the crossover points as possible, and/or maybe the woofers have breakup modes at a fairly low frequency, or other response problems that argue for a low crossover. But I ain't giving you the one about the upper cross. That would rule out some of the most popular speakers out there, and also some of mine (like the Salk HT2's). A perfectly symmetrical radiation pattern at the crossover point isn't necessary for a accurate sound production. There will be some dips or flares far off axis, but that's not what the ear focuses on. But let's not get into another discussion of the need for controlled directitivity. It's only one aspect of design, and a relatively minor one. It simply isn't fair to dismiss a criticize a design like the SVS because it doesn't meet every design goal you can think of for a loudspeaker.

 

[monkish54]

That would rule out some of the most popular speakers out there, and also some of mine (like the Salk HT2's).

And the SongTowers.

Not the SS8, 10, or 12 though.

A perfectly symmetrical radiation pattern at the crossover point isn't necessary for a accurate sound production.

I don't disagree that perfect symmetry is not not necessary.

There will be some dips or flares far off axis, but that's not what the ear focuses on.

I'm telling Harman on you!

It simply isn't fair to dismiss a criticize a design like the SVS because it doesn't meet every design goal you can think of for a loudspeaker.

And I will give you this one.

Although, I wouldn't say I dismiss the SVS, but that I like the design of the Ascend better because it does meet more of the design goals that I think are important.

 

[Dennis Murphy]

Oooops---I meant to redact that "dismiss" and replace it with "criticize," but I left both in. I knew you probably weren't dismissing the design. The ribbon Song Tower comes pretty close to your goals--5" woofers crossed at 2k. The dome version is crossed at 2.5k, so that might rule it out as a serious speaker for the constant directivity crowd. Personally, I always make it a point to wear earmuffs when I listen to it.

 

[monkish54]

Personally, I always make it a point to wear earmuffs when I listen to it.

If you'd cryogenically freeze your XOs you wouldn't have to!

On a serious note, does the SS8, 10, 12, and Phil 2 image more precisely than the SongTower? (In a large room...)

 

[GranteedEV]

If the early reflections are adequately delayed (>10 to 15ms) , you won't really hear them much in the "imaging". Spaciousness and overall balance, maybe, depending on the room properties.

... and ironically, it's speakers with "ideal" horizontal off-axis response - a la Salon2 or Philharmonic, that have the most unwanted energy in smaller rooms where the reflections are not adequately delayed.