CSS Criton 3TD-X Kit Speaker Review

[Vergiliusma]

That Philharmonic HT graph is from EAC, not ASR, please correctly credit where credit is due to avoid confusion.

Thanks--edited to provide attribution

 

[a4eaudio]

Without actually hearing the speakers and just looking at the graphs above, I'd be indifferent between the speakers. If I had to guess, I'd guess that I'd like the CSS a little better due to the slightly lower output in the 1khz to 5khz range. But, I typically like AMT tweeters better than dome tweeters, so that is an advantage for the HT Towers.

But at this point it probably comes down to budget, DIY abilities, and other preferences. The CSS is on sale for $1,800 and I can build a cabinet similar to the Philharmonic speaker for $300-$500. (I am not claiming the average DIY builder can, though. I have seen the Philharmonic speakers and they are quite nice.) So I could build the CSS tower in a nice cabinet of my exact choice to match my preferences/decor for about half the price of the HT towers. Saving $2,200 isn't going to get me a new car, but I can think of some things I would do with it. If someone cannot, or does not want to, do a DIY build then the CSS's (and every other DIY build on ASR) are simply not an option.

 

[cavedriver]

fwiw, Dennis of Philharmonic has said that he feels that roundovers make a clear difference, but I forget whether he indicated this was in measurements or he could hear it too. But I would not use the HT's as a comparison to the design of the CSS because the CSS is claiming they have a smooth directivity without the use of a waveguide on a tweeter, whereas the HT's use a planar tweeter which we know is likely to have a wide directivity making it easier to transition to the mids. It also uses the mids in a sort of D'Appolito array, so again a different topology. It would be better to find other simply three-ways that use dome tweeters to compare to the CSS's. Perhaps the Ascend's when they use that Seas Titan tweeter that seems to work so well in their design.

 

[Rick Sykora]

Only fair to compare finished speakers to other finished speakers. In which case, would much rather buy the HT towers with much better parts and better overall measurements than the CSS 3TD-X. Also would have another $1200 to spend!

As kits go, not a lot available that directly compares but the DXT-Wave looks more promising if you can make a cabinet. Also the SB Sasandu looks nicely designed but have not seen a solid set of measurements. Am hoping SB sends one soon.

 

[Vergiliusma]

But I would not use the HT's as a comparison to the design of the CSS because the CSS is claiming they have a smooth directivity without the use of a waveguide on a tweeter, whereas the HT's use a planar tweeter which we know is likely to have a wide directivity making it easier to transition to the mids. It also uses the mids in a sort of D'Appolito array, so again a different topology. It would be better to find other simply three-ways that use dome tweeters to compare to the CSS's. Perhaps the Ascend's when they use that Seas Titan tweeter that seems to work so well in their design.

Imo, it's still a fair comparison because if the drivers are decent, which in these cases it seems like they are, it mostly comes down to implementation. The difference, as far as directivity goes, will be more in the vertical plane for the dome in the CSS over the planar, but I don't know that the jury has come back with a verdict on how much that matters compared to the horizontal plane.

I've always looked at it like consistent directivity is more important than the amount. Of course, there's the variables of the room effects and personal preference.

 

[a4eaudio]

...Also would have another $1200 to spend!

Am I missing something?
CSS 3TD are on sale for $1,800 pair
Philharmonic are $4,500 pair (exactly 2.5x the price)

 

[Ktacos]

DXT-Wave

Those always looked great to me but how on earth did he do those chamfers?

 

[Rick Sykora]

Am I missing something?
CSS 3TD are on sale for $1,800 pair
Philharmonic are $4,500 pair (exactly 2.5x the price)

The finished CSS 3TD speakers are $5500…

SHOP | css-audio

www.css-audio.com

 

[375HP2482]

Personally I would choose the previously-reviewed Polk ES60 over this. Higher sensitivity, lower distortion, better looking, fewer $$$.

 

[Rick Sykora]

Personally I would choose the previously-reviewed Polk ES60 over this. Higher sensitivity, lower distortion, better looking, fewer $$$.

They are a decent value but for less than half the price of any of the other tower speakers mentioned, the Polk R700 competes well here too. Is actually less than the CSS kit.

From EAC...

 

[oupee]

 

[ROOSKIE]

Regarding capacitors. I do know a a blind AB test where some participants could consistently identify the change in capacitors. It was not that one was "better" than the other, but that they were identifiably different. I am sure I personally would not be able to tell a difference and I don't use exotic capacitors. Some people DO prefer higher end capacitors as evidence by almost every vendor who sells DIY kits. I can't imagine a competent business who would refuse to sell an option that their customers desire and improves their profit margin.

This is nonsense. IMHO this is not good to have normalized this thinking.
I can think of thousands of things that I would never to sell to somebody just because they 'desire' them.
It is not 'improved profit margin' to take advantage of a buyer it is a swindle. I would say at the prices some of these components are selling for it is exploitation.
Educate folks, give them a really good product and trust that you made enough money. I realize all is not ideal, and if somebody really won't buy your product without $1000caps you sure are in a pickle but again I can think of many other things I would not package with the speaker or product to make a sale.

What blind AB test? Link it and document it. This has continued to be a contentious subject for years and I simply can't see why, never have the boutique cap producers shown one shred of evidence as to why they are objectively or subjectively superior. (I can assure they could certainly show you nice fluffy wallets after a sale.)

 

[JudgeandInjury]

Beginner here. I don't understand why this is getting positive votes considering the measurements are just ok and $2400 seems like you can get significantly better speakers for the same or less. KEF Q11 Meta, Paradigm Premier 800f, MartinLogan Motion Foundation F2, Bowers & Wilkins 603 S3, Monitor Audio Silver 200 7G, and even Sonus Faber Lumina III are all cheaper...

 

[Kerry Armes]

I want to point out that our cabinets are made in the US so costs for the finished versions are going to be significantly higher than what other companies who are using Asian vendors, like Philharmonic or Revel, are putting out. The benefit is we can hand select veneer for the customer and do things like this.

 

[ROOSKIE]

Beginner here. I don't understand why this is getting positive votes considering the measurements are just ok and $2400 seems like you can get significantly better speakers for the same or less. KEF Q11 Meta, Paradigm Premier 800f, MartinLogan Motion Foundation F2, Bowers & Wilkins 603 S3, Monitor Audio Silver 200 7G, and even Sonus Faber Lumina III are all cheaper...

With the notable exception of the KEF Q11meta, the speakers you list are unlikely to measure better than the CSS kit reviewed here. I be shocked if they they did, certainly not the B&W and the Sonus Faber ---> those are likely going to be poorly measuring speakers by the standards here.

I think this is getting a generally positive response as it is a mostly good performer from a small company which is to some degree a meaningful factor to some folks. I like supporting the little guy and especially supporting DIY. I'd agree that for the price they measure just okay, especially since as a DIY they will take quite some time to get up and running, and only skilled builders will get $5k looks.
Still well done over-all given the design constraint of needing cabinets that are easy(relatively) to DIY and finish (imagine a weekend warrior putting veneer on very curved cabs or even harder one with 3/4" or 1" round overs). Really that cabinet constraint is why the conversation turned to larger waveguided tweeters, as with the sharp edges the real constraint is that tweeter. This has been covered but this begs for a larger waveguide(not a small waveguide) for a lower crossover point and with a geometry that allows for much less edge diffraction (so not a small waveguide, at least 5"). In fact with the right geometry there sharp edge will be zero issue and directivity(off to on axis relationship) will kill with a good x-over. Also a waveguide could allow for a more constant directivity in the very high highs which I prefer (this falls off after 8k). This has been beaten out and asked to be put aside but the design is limited due to essentially making it easy to build. I think that is relevant.
One caveat of waveguides is that some folks just can't stand them and no matter how well designed these folks perceive some amount of a cupped hands effect. Though I'd wager that in a blind test things may go differently for them.

Those always looked great to me but how on earth did he do those chamfers?

Right? This is beyond the general skill sett of most (and of course why the CSS design is squared for normal folks). Someone with a CNC or computer driven cutting gear is good with that complex shape. They could be 3d printed and added into square cuts. I'm sure other ways exist but difficult for a newb or even many who are skilled.

... because the CSS is claiming they have a smooth directivity without the use of a waveguide on a tweeter, whereas the HT's use a planar tweeter which we know is likely to have a wide directivity making it easier to transition to the mids. It also uses the mids in a sort of D'Appolito array, so again a different topology. It would be better to find other simply three-ways that use dome tweeters to compare to the CSS's. Perhaps the Ascend's when they use that Seas Titan tweeter that seems to work so well in their design.

Maybe I don't understand the context of your statement. The wide directivity of the planer makes it HARDER to get smooth directivity with a typically beaming mid and harder to get a smooth transition on and off axis, that is why he chose the BMR which is still very, very wide at the crossover point. The wide tweeter was chosen(and the BMR) as the designer, Mr. Murphy likes very wide dispersion speakers.

As well, it is not possible to get smooth directivity between any naked 1" dome or concave tweeter(Focal) and a concave 5-7" woofer. Certainly not crossed @2700hrz. That is the whole reason waveguides came to be, it is not possible otherwise. Nobody can really do it, only close enough but no cigar. CSS should not be claiming this, they did well given the design, they did not achieve bragging rights smooth directivity without a waveguide. No naked tweeter speaker I have seen has truly done this UNLESS using a small alternative mid-driver ala the BMR or a nice dome mid or maybe a 3" conventional driver. If using the 3" conventional driver and not a BMR the crossover still needs to be well under 2k for smooth directivity probably about 1600hrz with a very shallow 3" cone midwoof. The math/geometry/physics of it all is 100% fixed. Using a BMR was genius.

I do have the Philharmonic BMR monitor and they do sound fantastic, truly superb. I worship them.

 

[Rick Sykora]

Those always looked great to me but how on earth did he do those chamfers?

I agree, and likely could use a simpler more angular, pinched baffle.

From his website…

“Relatively demanding built process, but still, with router, jigsaw and two handy hands, straightforward doable.
In case of doubt, the CNC milling service will help.”

Seems a good case for a pre-routed front baffle imo!

 

[Kerry Armes]

There may be some underreporting of bass extension in my measurements due to spread out port and woofers. I thought about optimizing it but takes a lot of manual effort so I left it be.

I have to assume that is some of what we are seeing here as well since the sealed F10 of these drivers in a leaky cabinet matches what is shown in the measurements. Here is what sealed vs ported looks like in this sized cabinet in the modeling.

 

[Kerry Armes]

As well, it is not possible to get smooth directivity between any naked 1" dome or concave tweeter(Focal) and a concave 5-7" woofer. Certainly not crossed @2700hrz. That is the whole reason waveguides came to be, it is not possible otherwise. Nobody can really do it, only close enough but no cigar. CSS should not be claiming this, they did well given the design, they did not achieve bragging rights smooth directivity without a waveguide.

I'm not sure why you are saying it's not possible. Here is the HT Tower overlayed over the 3TD-X graph with the same scaling but with the HT Tower adjusted down in SPL so that both match in the bass frequencies around 200 Hz and below.

 

[cavedriver]

Maybe I don't understand the context of your statement. The wide directivity of the planer makes it HARDER to get smooth directivity with a typically beaming mid and harder to get a smooth transition on and off axis, that is why he chose the BMR which is still very, very wide at the crossover point. The wide tweeter was chosen(and the BMR) as the designer, Mr. Murphy likes very wide dispersion speakers.

As well, it is not possible to get smooth directivity between any naked 1" dome or concave tweeter(Focal) and a concave 5-7" woofer. Certainly not crossed @2700hrz. That is the whole reason waveguides came to be, it is not possible otherwise. Nobody can really do it, only close enough but no cigar. CSS should not be claiming this, they did well given the design, they did not achieve bragging rights smooth directivity without a waveguide. No naked tweeter speaker I have seen has truly done this UNLESS using a small alternative mid-driver ala the BMR or a nice dome mid or maybe a 3" conventional driver. If using the 3" conventional driver and not a BMR the crossover still needs to be well under 2k for smooth directivity probably about 1600hrz with a very shallow 3" cone midwoof. The math/geometry/physics of it all is 100% fixed. Using a BMR was genius.

My point was that the example of the BMR doesn't answer my question of how CSS achieved a smooth directivity because the BMR uses a different topology, and is therefore not useful in assessing CSS's method on that specific question. I admit I was wrong about the significance of the planar tweeter in Dennis' approach- the BMR mid is the driver I should have addressed in my comment. I agree that the BMR's use of the BMR drivers is key to Dennis' design.

 

[Rick Sykora]

Actual levels of drivers/port is NOT accurate. I actually manually align each graph to stitch them together. In this case, I should have pulled the port way lower. In near-field measurements, slight distance difference to each radiating device can make a big difference and I have no way to dialing that out.

Understand the near-field challenge, but probably could use PIR to bound the bass rolloff and make this simpler. Not sure if you have the feature, but Direct Sound Separation would likely be even easier.