Perlisten speakers

[Dennis Murphy]

The frequency response drop towards low frequencies really corresponds to that of a closed speaker.
Here is a comparison of a HP with second order Butterworth @50Hz.
The match is not quite perfect, since the speaker cabinet provides too little volume for the four woofers, which is often deliberately chosen to make the low bass appear "more powerful" (SPL hump around 100Hz).
View attachment 131534
So the loudspeaker corresponds very roughly to a closed loudspeaker with an f3 of 50Hz. Where under 35Hz probably a little more sound pressure level is available because of the low BR tuning around 30Hz (if the ground plane measurement conditions were optimal and wind did not affect the measurements).


Looking closer at the ground plane measurement, an f10 at 25Hz is not exactly what I would call "significant output down to 22hz".
If the speaker is placed freely, the bass should sound rather "soft/slim/thin" - which does not have to be bad.
View attachment 131540
Source: Audioholics



If we assume that the woofers are SBAcoustics MW19TX-8, then each woofer has a displacement volume of about 103cm³.

With four woofers (Vd 400cm³), this is significantly less than a "reasonable" 10'' subwoofer can provide in displacement volume - Vd around 480-600cm³.
For such a slim and small speaker this is certainly okay, but for speakers with a pair price of $16000 the low bass performance is not special.

I would take the ground plane measurements with a grain of salt. I (and others) have discussed this with the reviewer after puzzling over results for other speakers that didn't seem to comport with published specs or nearfield measurements. My best guess is that for whatever reason he's not picking up the full output of the port(s).

 

[thewas]

Ground plane measurements of (usually large) loudspeakers standing directly on the ground partially also capture the boundary reinforcement of the floor which other anechoic methods don't, but on the other side this is not wrong as they are anyway almost always placed like that in practice.

 

[ctrl]

I would take the ground plane measurements with a grain of salt. I (and others) have discussed this with the reviewer after puzzling over results for other speakers that didn't seem to comport with published specs or nearfield measurements.

A ground plane measurement under optimal conditions and performed correctly is quite reliable - as @hardisj has shown us with his ground plane measurements. The correct execution is crucial.

A measurement in an anechoic chamber has a certain error tolerance (e.g. +-1dB), which increases below the actual measurement limit (occurrence of room modes, usually below 50-100Hz) due to the correction used for the room modes.

With the summation of near field measurements also some things can go wrong.
When measuring the near field of the woofer and the BR port, no dual-channel measurement is made and only the minimum phase is used for the summation, 1-2 dB errors are easily possible (and at very low frequencies more). The time difference between the BR port and the woofer is not taken into account - for example, a large floor-standing speaker with the BR port at the back and bottom and the woofer at the top front, this can easily cause an error of 1 dB or more. The sound level correction for the BR port is not calculated correctly,...

And as we have now seen several times, it also depends on a "correct" execution of the Klippel NFS, in order to achieve good results in the low frequency range.

I don't know the back story to your post, but I am skeptical of all (low frequency) measurement methods and measurements - except my own, of course

 

[blueone]

One thing I am confused about is how all their speakers and subwoofers appear to have the THX Dominus certification, including the R5t, S5m, and S4s. I am truly curious about this. Does this mean they can create more clean/loud output than all other speakers before them in equivalent categories (because no other speakers earned this certification)? I say equivalent categories because the surrounds got a THX dominus surround certification. If so, that is quite remarkable.

In the review of the S7T the reviewer mentioned the S7T was the only speaker with this THX certification, only adding to my confusion.

Links for proof (scroll to the bottom for dominus certificate):
S5m Monitor — Perlisten Audio
R5t Tower — Perlisten Audio
S4s — Perlisten Audio

Also confusing, both tower products are specified as either "bass reflex" or "acoustic suspension", which I assume means plugging the downward firing ports and creating a sealed box configuration for the woofers. I've always read that there is a significant difference between an acoustic suspension design and a sealed design, especially in the characteristics of the bass drivers. And, to top off the design confusion, Audioholics even has an article on their site making it clear that sealed designs and acoustic suspension designs are not the same:

https://www.audioholics.com/loudspeaker-design/acoustic-suspension

Very odd, considering that the author of the review (James Larson) is not known for factual oversight or technical errors. Larson calls the port plugging turning the S7t into a sealed design, not a ported design, in the review. And, I'm sorry to be such a nit-picker for discussion integrity, but Larson even included a editorial note in the above-referenced article about acoustic suspensions:

Editorial Note about Sealed vs Acoustic Suspension Enclosure by James Larson:
As excursion for bass drivers have grown larger, high compliance suspension of the type needed for acoustic suspension design becomes more difficult to engineer for reliable performance. The thin suspension components needed for high compliance suspension are susceptible to self-destruction at high excursions, not to mention increased distortion from rocking modes and deformation. Furthermore, air doesn't have enough damping force against the effects of progressive suspension for high excursions. At high excursions, the suspension must be able to damp itself, and that means that it has to have lower compliance than an acoustic suspension system would call for. Acoustic suspension was a more viable solution in older eras of loudspeaker design when one-way linear excursion could be around 5mm, but modern designs aim for much larger excursions where high compliance suspensions become impractical.

I definitely think it's possible for the Perlisten towers to be very good speakers, given Larson's measurements, but it would bolster buyer confidence to see Perlisten's marketing materials and specifications be more technically sound and complete.

 

[Dennis Murphy]

A ground plane measurement under optimal conditions and performed correctly is quite reliable - as @hardisj has shown us with his ground plane measurements. The correct execution is crucial.

A measurement in an anechoic chamber has a certain error tolerance (e.g. +-1dB), which increases below the actual measurement limit (occurrence of room modes, usually below 50-100Hz) due to the correction used for the room modes.

With the summation of near field measurements also some things can go wrong.
When measuring the near field of the woofer and the BR port, no dual-channel measurement is made and only the minimum phase is used for the summation, 1-2 dB errors are easily possible (and at very low frequencies more). The time difference between the BR port and the woofer is not taken into account - for example, a large floor-standing speaker with the BR port at the back and bottom and the woofer at the top front, this can easily cause an error of 1 dB or more. The sound level correction for the BR port is not calculated correctly,...

And as we have now seen several times, it also depends on a "correct" execution of the Klippel NFS, in order to achieve good results in the low frequency range.

I don't know the back story to your post, but I am skeptical of all measurement methods and measurements - except my own, of course

That all sounds reasonable. I've only had one speaker reviewed by Audioholics, and in that case the modeling was done by very sophisticated software and the nearfield measurements replicated the modeling. However, there was a pretty substantial difference between those and the ground plane measurements. I don't know whether the software took into account timing differences between the port and woofer.

 

[Spocko]

I would need to see more detailed data. Maybe. But most speakers eventually become tweeter limited. The raw tweeters are often maybe 90-96dB 1w/1m. I need to double check this number, but I recall Dan telling me the tweeter array is 112dB sensitive and padded down a lot. So the Perlisten has a ton more headroom in the tweeter section than would the Revel. It’s midbass limited. But those midbass drivers are already more output than Revel has in that area.

all to say, I highly doubt the Revel can do 117dB even above 100hz.

DANG I didn’t realize - I’m impressed!

 

[Spocko]

I’m very interested in the S5M monitors as they appear to have that beam forming shape without the height/weight expense especially if you’re going to use subs below 80Hz

 

[richard12511]

I would need to see more detailed data. Maybe. But most speakers eventually become tweeter limited. The raw tweeters are often maybe 90-96dB 1w/1m. I need to double check this number, but I recall Dan telling me the tweeter array is 112dB sensitive and padded down a lot. So the Perlisten has a ton more headroom in the tweeter section than would the Revel. It’s midbass limited. But those midbass drivers are already more output than Revel has in that area.

all to say, I highly doubt the Revel can do 117dB even above 100hz.

Somehow missed this post, but I think it addresses many of my questions.

About most speakers becoming tweeter limited, I didn't know this, but I can see now how it changes things. I actually always thought the opposite. My thinking was that - due to the avg spectrum of music - that tweeters are usually playing ~20db down or so and should almost never be the limiting factor. FWIW, my horn loaded speakers do seem to do better(less painful) at really high volumes than my cone/dome speakers, and the extra harshness from the dome speaker does seem to be treble related. This is without the clip lights going red, which is weird. It's only at really really high volumes, though.

 

[Tks]

Man, these look awesome (not the hardwood veneer ones). I know there's lots of folks here not liking the looks, but the checkered woofers also seem pretty cool.

 

[richard12511]

Man, these look awesome (not the hardwood veneer ones). I know there's lots of folks here not liking the looks, but the checkered woofers also seem pretty cool.

I actually really like the grills off looks(both white and wood veneer). Not a huge fan of those grills, though.

 

[AlexanderM]

I like the look, wondering about the price of the bookshelf speakers.

 

[Tks]

Wow made in China? These guys remind me of what they're doing in the speaker realm, as to what Matrix Audio does in the DAC/AMP realm. Performance looks good, and the product itself aesthetically looks good. Never would have pegged these as Chinese of origin (in the same way I never would have thought back in the day that Sabre DAC from MA was from a Chinese company).

Re-affirms my belief that if people actually were willing to pay for stuff, China most certainly can match (if not straight up outdo) the rest of the world's manufactured products. In fact, they already do since they can cater to both low-end products (the majority of what outsiders contract Chinese work for since product of that quality cannot be had anywhere else on the planet for such low cost), and obviously high end if given simply the right incentive (since they already have the tech and machines required to manufacture any mass-producible product in the first place, so it's not like you're going to find this anywhere else outside of hilariously skewed value propositions from all in-house producers).

 

[Spocko]

Wow made in China? These guys remind me of what they're doing in the speaker realm, as to what Matrix Audio does in the DAC/AMP realm. Performance looks good, and the product itself aesthetically looks good. Never would have pegged these as Chinese of origin (in the same way I never would have thought back in the day that Sabre DAC from MA was from a Chinese company).

Re-affirms my belief that if people actually were willing to pay for stuff, China most certainly can match (if not straight up outdo) the rest of the world's manufactured products. In fact, they already do since they can cater to both low-end products (the majority of what outsiders contract Chinese work for since product of that quality cannot be had anywhere else on the planet for such low cost), and obviously high end if given simply the right incentive (since they already have the tech and machines required to manufacture any mass-producible product in the first place, so it's not like you're going to find this anywhere else outside of hilariously skewed value propositions from all in-house producers).

Not many people realize that Chinese OEM houses can do high end as well as the best in the world because Apple, Samsung and Sony trained them very very well. As a matter of fact, I don't know of any country that has the capacity to produce a "high end" product at the scale and volume of OEMs in China other than maybe Korea or Taiwan for much more but not necessarily higher quality assembly/design. Just look at their boutique IEMs for $800 and you see where this is going.

 

[Spocko]

I like the look, wondering about the price of the bookshelf speakers.

I believe the S4B bookshelves MSRP for around $7,000 a pair which is a bit more than the $4,000 Revel M126Be but the S4B is 3-way with that impressive beamforming technology in its MTM tweeter array while the M126Be is a 2-way, both with beryllium tweeters. Neither the Perlisten nor Revel bookshelves are very efficient at 85 dB and 86 dB, respectively.

 

[Matthew J Poes]

The bolded part simply isn't true. Frequency-dependent pruning and thinning of arrays is standard beamforming practice.

man expanding array has it such that there is a hard crossover with no significant overlap between the tweeter and mid ranges. Is that beamforming by any definition?

what your describing isn’t a crossover. There still must be overlap in output of the beamforming drivers or they can’t Interact with each other.

 

[Matthew J Poes]

Not many people realize that Chinese OEM houses can do high end as well as the best in the world because Apple, Samsung and Sony trained them very very well. As a matter of fact, I don't know of any country that has the capacity to produce a "high end" product at the scale and volume of OEMs in China other than maybe Korea or Taiwan for much more but not necessarily higher quality assembly/design. Just look at their boutique IEMs for $800 and you see where this is going.

And in this case, the manufacturing facility is actually a partnership between the guys behind Perlisten and the facility. A joint venture. Perlisten is made within its own factory within the factory with dedicated trained staff.

 

[Matthew J Poes]

I'd guess there are many other HT speakers out there with even more dynamic headroom, but they just haven't applied for THX Dominus certification(which cost money). Some of the JBL LRX83X speakers(for example) should be able to play louder than these with subwoofer support(136dB according to spec).

there is a lot more to the spec than just output. That is only one part of it. Their response shape, bandwidth, and directivity must also meet the spec. While this was always true to some extent, the Dominis spec is more stringent and different. So while lots of speakers will play louder, they may not pass the tests anyway.

I know of two other companies currently trying to get Dominus status. We shall see. So far they haven’t passed. I only know that they sent in some data and suddenly went back to the drawing board. So I am guessing the spec is pretty hard to get. THX makes you not disclose the specifics so I only know little tidbits here and there that was shared. Nothing definitive.

 

[Matthew J Poes]

One thing I am confused about is how all their speakers and subwoofers appear to have the THX Dominus certification, including the R5t, S5m, and S4s. I am truly curious about this. Does this mean they can create more clean/loud output than all other speakers before them in equivalent categories (because no other speakers earned this certification)? I say equivalent categories because the surrounds got a THX dominus surround certification. If so, that is quite remarkable.

In the review of the S7T the reviewer mentioned the S7T was the only speaker with this THX certification, only adding to my confusion.

Links for proof (scroll to the bottom for dominus certificate):
S5m Monitor — Perlisten Audio
R5t Tower — Perlisten Audio
S4s — Perlisten Audio

so I actually have the same question about the S4s myself.

first, as noted a moment ago, there is more to the spec than just output and distortion. Directivity, response, and bandwidth all matter too. It’s different from the old spec as well.

I also know that all the subs other than the single 12 can meet the Dominus bass spec as a single. But the single 12 requires two. This is not unusual for THX to allow subs to pass in multiples.

I’ll find out why the smaller speakers appear to fail the requirements made public but get a certificate. There might be different requirements for certain types?

 

[Matthew J Poes]

The frequency response drop towards low frequencies really corresponds to that of a closed speaker.
Here is a comparison of a HP with second order Butterworth @50Hz.
The match is not quite perfect, since the speaker cabinet provides too little volume for the four woofers, which is often deliberately chosen to make the low bass appear "more powerful" (SPL hump around 100Hz).
View attachment 131534
So the loudspeaker corresponds very roughly to a closed loudspeaker with an f3 of 50Hz. Where under 35Hz probably a little more sound pressure level is available because of the low BR tuning around 30Hz (if the ground plane measurement conditions were optimal and wind did not affect the measurements).


Looking closer at the ground plane measurement, an f10 at 25Hz is not exactly what I would call "significant output down to 22hz".
If the speaker is placed freely, the bass should sound rather "soft/slim/thin" - which does not have to be bad.
View attachment 131540
Source: Audioholics



If we assume that the woofers are SBAcoustics MW19TX-8, then each woofer has a displacement volume of about 103cm³.

With four woofers (Vd 400cm³), this is significantly less than a "reasonable" 10'' subwoofer can provide in displacement volume - Vd around 480-600cm³.
For such a slim and small speaker this is certainly okay, but for speakers with a pair price of $16000 the low bass performance is not special.

as was already noted, you have the wrong driver. They custom make their drivers. Each has about 8mm of one way excursion. That’s 16mm P-P linear excursion. Not sure how you figure that a single 10” subwoofer could exceed four 7” drivers with 8mm of excursion. That single 10” driver would need 18mm of linear xmax to equal these four drivers at 25hz.

the measurments from James may be off in the bass. A more accurate microphone measuring it in the same way came to somewhat different results. It isn’t a huge error but is enough that if someone nit picks the data they could come to a different conclusion.

I already said the speaker rolls off at a 2nd order rate like a sealed box. This gives it very low group delay in the bass, good excursion control, and good in room extension. I will have in-room measurements of the speaker by Friday so I can show you what that looks like in the bass. We can see if they have good extension in room.

these have every bit as much bass as the Revel F328Be and Amir indicated that speaker had plenty of output. Everyone who has heard them without subs has been surprised by how good the bass is. I think it makes little sense to criticize the bass bases on a single measurement that seems to be misinterpreted.

 

[Matthew J Poes]

Thanks for the link. This should not change much in the result of the calculation as the drivers used by PerListen should have a similar cone area. Perhaps we will hear more about this from Matthew in the future.

the excursion difference is worth about 4dB at 25hz between those two and the Perlisten drivers handle about 4-5 times as much power. They have no relationship to the SB Acoustics drivers. All drivers were designed in house and custom made at their factory.