Ascend CBM-170SE2 Speaker Review

[Chrispy]

Excellent performance for the price.

Anybody know what the woofer is? It looks like a SEAS custom design of some sort...

David will probably fill you in, but probably info in the forums at Ascend's site too.

 

[PristineSound]

Excellent performance for the price.

Anybody know what the woofer is? It looks like a SEAS custom design of some sort...

All transducers are from SEAS as per Dave on his forum. In fact, all of Dave's transducers are SEAS today AFAIK.

 

[AscendDF]

Excellent performance for the price.

Anybody know what the woofer is? It looks like a SEAS custom design of some sort...

Fully custom and designed by me, loosely based on the original CBM-170 woofer which was an aerogel model made for us by Audax in France. It is not manufactured by SEAS, the tweeter is manufactured for us by SEAS. We are now at iteration # 4 of our custom 170 woofer.

 

[PristineSound]

Fully custom and designed by me, loosely based on the original CBM-170 woofer which was an aerogel model made for us by Audax in France. It is not manufactured by SEAS, the tweeter is manufactured for us by SEAS. We are now at iteration # 4 of our custom 170 woofer.

I stand corrected.

 

[Karmacoma]

Impressive quality/performance/cost ratio, as usual with Ascend. Kudos.
I have the hope this company will develop waveguides at some point.

 

[kyuu]

Can't wait for this company to start developing waveguides.

Yes, the lack of waveguides for the dome tweeters is about the only qualm I have with Ascend. Well, that and the lack of in-wall designs, but I suspect that one is mostly a me thing.

 

[dfuller]

Fully custom and designed by me, loosely based on the original CBM-170 woofer which was an aerogel model made for us by Audax in France. It is not manufactured by SEAS, the tweeter is manufactured for us by SEAS. We are now at iteration # 4 of our custom 170 woofer.

Nice! I'm sure having drivers designed with the behavior you want eases a lot of crossover difficulties.

 

[AscendDF]

Nice! I'm sure having drivers designed with the behavior you want eases a lot of crossover difficulties.

100% correct, and a major hurdle in the redesign of the 170SE to 170SE2. It is why it has taken us so long. I discussed this in the 170SE2 thread on our forum. To be honest, I almost gave up but so many people were emailing us about a "Klippel optimized" version so I ended up starting, stopping, starting, stopping so many times.

I am most pleased with the final result of this project, the speakers sound terrific. I noticed many people mentioning using them with the Wiim Amp Ultra, and that was exactly what I did for several few weeks until I exchanged the Amp Ultra for the streamer (needed Apple Air-play). It is a fantastic combination of components.

 

[Arthur Edwards]

It is unclear how they achieve a good-looking spin curve without a waveguide on the tweeter when it is crossed over to a 6.5-inch woofer.

I would like to provide additional context for my question.

The cabinet width is unchanged between the CBM-170 SE and CBM-170 SE2 models, and the tweeter remains flush-mounted. However, the Olive index increased from 6.55 to 8.73 with a perfect subwoofer.

Considering the speaker’s price point, the crossover network likely uses a limited number of components. It is unexpected that the crossover optimization software, which now has anechoic data from the Klippel system, produced these results. I would appreciate insights from the ASR community on this outcome.

It is unclear how they achieve a good-looking spin curve without a waveguide on the tweeter when it is crossed over to a 6.5-inch woofer.

 

[AscendDF]

I would like to provide additional context for my question.

The cabinet width is unchanged between the CBM-170 SE and CBM-170 SE2 models, and the tweeter remains flush-mounted. However, the Olive index increased from 6.55 to 8.73 with a perfect subwoofer.

Considering the speaker’s price point, the crossover network likely uses a limited number of components. It is unexpected that the crossover optimization software, which now has anechoic data from the Klippel system, produced these results. I would appreciate insights from the ASR community on this outcome.

This crossover network is quite complex, 18 elements, it is massive. It comes down to closely evaluating the individual directivity behavior of the woofer and tweeter to achieve the ideal crossover point that provides that best directivity match. Assuming that this crossover point is ok for both drivers (requires quite a bit of additional testing), from there it is about carefully determining the ideal filter slopes to keep things in phase and provide the best blend, which often require the slopes to be asymmetrical. It is complicated and our technique for this often doesn't work, thus requiring the many redesigns of this woofer (for example, reducing some inductance to achieve a bit more lower treble extension)

Without a typical waveguide, it does make things much more difficult, but I welcome the challenge and prefer the wider horizontal dispersion characteristics.

I am not sure if this level of optimization would be possible without an NFS. I likely ran over 100 full scans with the NFS for this project.

I must get back to my day job now, but I wanted to answer as many questions as possible here. Please feel free to contact us directly and I will also try to be back later.

 

[DanTheMan]

I actually hope they avoid "waveguides". Broad dispersion sounds so good. Even 2 channel can feel enveloping. I am flabbergasted by the price to performance. For the guy who likes to DIY to save money, the jig is up. Just buy a great speaker that you'll want until you depart from this planet. Thank you @AscendDF for all the design work and manufacturing and thank you @amirm for the confirmation! I paid a similar amount for speakers back when I was in college30 years ago and I promise you they weren't anything like this. Kids these days have everything.

 

[RigorDude]

My reference tracks, one by one, impressed with no excuses needed to be made. Closing my eyes, I perceived a very large spatial image that helped in that regard. I am talking about a bubble that was 6 feet or more around and above the speaker! With a pair, the CBM-170SE2 will fill a large stage with delightful tonality and accuracy.

I’m assuming these tracks are two-channel if not multichannel. When you listen to them through a single speaker, how are the channels summed? (Also assuming you don’t listen to L only or R only.)

 

[PristineSound]

Over 100 scans. Jaw dropping. I think the effort you save has already paid itself off with just one project.

 

[amirm]

I’m assuming these tracks are two-channel if not multichannel. When you listen to them through a single speaker, how are the channels summed? (Also assuming you don’t listen to L only or R only.)

Stereo tracks and no summing. Much of modern music i listen to has similar content in both channels. So listening to one is sufficient without all the drawbacks of trying to convert to mono.

 

[amirm]

Over 100 scans. Jaw dropping. I think the effort you save has already paid itself off with just one project.

Indeed. A single scan performed by an outside company cost about $1,000. So we are talking about $100K opportunity cost!

I saw in the stats that I have tested 257 speakers. So that clocks at quarter of a million dollars in cost if we had to farm them out! And that is just for the CEA-2034.

 

[thewas]

I actually hope they avoid "waveguides". Broad dispersion sounds so good.

Using WGs doesn't automatically mean narrow dispersion but ideally smooth/continuous one as a waveguide can be even used to widen the radiation at some region, a ultra wide one isn't beneficial when is limited just to some regions.

 

[thewas]

In times of full spinorama simulation tools like VituixCAD2 really many NFS measurements are not necessary for crossover tuning as when you have "spinned" the drivers without crossover on the chosen baffle you can simulate with any crossover with high precision. Repeated full spinorama measurements were more necessary when designing waveguides although even there nowadays great simulation tools exists which aren't simple to use though. Of course its always good to validate the simulations in the end with the real measurement and possibly some minor tweaking.

 

[Waxx]

For it's size and price class that is very good, even not taking price in account it's quiet good. A waveguide could have made it better altough. But getting this result without is an archivement. But it's not availeble in Europe it seems, so for us Europeans not an option (importing it ourself will cost a lot).

 

[thewas]

It is interesting to compare the directivities to its predecessor on the same scale which Amir had named poor back then:

While the improvement on the early directivity index is rather small, the total directivity index (and thus sound power) has smoothed significantly which shows that despite of the lack of WG quite some improvement can be achieved by clever tuning of the crossover.

 

[Waxx]

It is interesting to compare the directivities to its predecessor on the same scale which Amir had named poor back then:

View attachment 549569
While the improvement on the early directivity index is rather small, the total directivity index (and thus sound power) has smoothed significantly which shows that despite of the lack of WG quite some improvement can be achieved by clever tuning of the crossover.

Off course you can, that is an old trick, matching the directivity of the tweeter and woofer by choosing the right crossover point and slopes (so the transition is smooth). It's a bit a lost art in times of waveguides i have the impression, but it used to make the difference between good and bad speakers. On DIYaudio this used to be (and still is in a lot of builds) an important factor in good speaker and crossover design.