Perlisten speakers

[Matthew J Poes]

Hard crossover? Is that a technical term? David Smith put a lot of effort into determining how to place the speakers geometrically, and what order crossovers to use in order to create the dispersion pattern he was after. The results is a consistent Maximum Response Axis across the band of interest, with attenuated response at angles away from the MRA. That's a beam. He achieves this largely by using an array of multiple transmitters (aka beamforming). Compare the dispersion he achieves with his expanding arrays to a pair of full range drivers stacked vertically, and you'll see how important all the driver interactions are.

I think you are nitpicking here in order to try to argue that somehow this is not novel and was done before. Again, the way the expanding array works is not the same. I am not sure what part of that you are not seeing. In the expanding array, there is a traditional crossover between the midrange and tweeter. Overlap is typical of the order used. They are symmetric. It is a true crossover. In the Perlisten, the tweeter in the center along with the midrange drivers all have a highpass filter at 1khz so they overlap substantially. There is a point where the midrange is phased out so that eventually the tweeter is the only driver operating. But there is large overlap between those points. Something like 4-5 octaves.

I have never heard beam forming defined like you are. All speakers have a forward radiation lobe that is directional. MTM and WMTMW designs have created symmetric lobes and also have existed for a long time. But I never considered those beam forming designs. If you want to, that is fine. I think we are just mincing words.

the Snell and Perlisten approach are not similar or related. That remains true. Call them what you want.

 

[Matthew J Poes]

Have calculated that very likely just like you. Okay, the SBAcoustics driver had 7.5'', the PerListen driver has only 7'' and a bit more linear excursion.

There is no problem to check my "calculation". A 7'' driver has a cone area of 130-150cm² - the thicker the surround, the smaller the cone area. So the 8mm excursion should rather require a thicker surround. Nevertheless, let's be generous and calculate with 145cm² cone area.

Vd = 4 x Sd x Xmax = 4 x 145cm² x 0.8cm = 464cm³

For comparison, let's consider the displacement volume Vd of 10'' subwoofer drivers.
One from the Dayton Reference series (mixture of sub and woofer) and one from the Ultimax series (pure subwoofer).

RSS265HO: Vd = 430cm³
UM10-22: Vd = 620cm³

Thus, the PerListen S7t has a displacement volume that is comparable to a 10'' subwoofer. With this comparison, I just want to prevent people from getting the wrong idea about what the speakers are capable of in terms of low frequency performance.

Update: Of course, you can doubt the manufacturer's specifications about Xmax. To be completely sure, you would have to measure all drivers yourself.

We could. But in this case we know who the manufacture is and how they test. I don’t doubt. Everything was done with Klippel and I know these guys. I believe the specs to be valid.

 

[Matthew J Poes]

I live only a few minutes from him…

It was surreal news back last year when the first Covid case in our town was reported and I saw his name/face in the article.

Ah we should all meet up at his house. Well I now live less than 2 hours away in Sarasota at this point. Maybe we can have you come over for the listening test between the Revel, Paradigm, Focals, and Perlisten .

 

[Matthew J Poes]

Maybe different requirements for mains/surrounds? Maybe the pass as "surrounds".

Ok I now know the answer and once explained it all made sense.

THX ratings are actually based on a system. Not an individual product. The system needs to meet the standard. As such there are different requirements for each speaker. That is why smaller subs may need multiples to meet the spec. Because they are ok with that.

the Monitor and bookshelf meet the standard as a surround speaker. Only the S7T and S7C meet the standard as an LCR. The others are still plenty capable. They just don’t meet the 92dB sensitivity requirement for the LCR standard.

finally, someone mentioned that nobody else applied for this standard. It’s been confirmed that what I knew is more true than I realized. A number of companies have applied and failed. So it is true to say that Perlisten is the first and only company to pass the standards. Others did apply and didn’t make it.

 

[Matthew J Poes]

Here is an in room response. It doesn’t show bass below 20hz like I thought it might, but shoes solid response flat to 25hz or so. I don’t think this suggests weak bass though. It looks pretty good to me.

I will update this with room volume when I get the dimensions. As well as how sealed the room is. I didn’t measure this. But I will have measurements from Genes room soon enough. Within a month or so I hope.

 

[Spocko]

... Only the S7T and S7C meet the standard as an LCR. The others are still plenty capable. They just don’t meet the 92dB sensitivity requirement for the LCR standard...

Matt, I'm still a little unclear about the the efficiency differences between the smallest S4b bookshelf and S7t towers when subwoofers are used. Given they employ identical tweeter/mid array, I assume that the efficiency given is for maximum bandwidth +/- 1.5dB down to the lowest frequency capability of each speaker. Obviously, performing at these lower frequencies is what pulls down efficiency numbers of the bookshelf, so if we use a subwoofer high pass at 120Hz for the S4b, wouldn't the S4b efficiency numbers be significantly higher?

 

[Spocko]

...
I wonder if they'd be willing to tell Erin's Audio Corner - a new upstart loudspeaker manufacturer - how to achieve their Dominus Rex status.

Since they don't tell you "how to achieve" THX certification, but rather the guidance is "do your best and send it to us - we'll tell you if its THX Select, Ultra, Dominus or None-of-the-Above" then you pay the exorbitant testing qualification fee only to receive the "REJECTION" stamp - that's gonna hurt unless they tell you where you failed so you can improve that one measure and maybe this is where the NDA comes in.

 

[hardisj]

then you pay the exorbitant testing qualification fee only to receive the "REJECTION" stamp

I'd be pissed if it were a guessing game and they didn't give me a target.

maybe this is where the NDA comes in

Yep, most likely.

 

[blueone]

I wonder if they'd be willing to tell Erin's Audio Corner - a new upstart loudspeaker manufacturer - how to achieve their Dominus Rex status.

My guess, based on what I've read here, is that you have to approach them as a designer and OEM of speakers, and sign a non-disclosure agreement to see the specs. The definitions of their specs almost certainly aren't patentable, so a non-disclosure agreement is their only recourse, if they really believe they need to be secret.

Just my opinion, I suppose worth to THX what they're paying for it, but as a member of their target market, I think this designation is completely worthless, and the hype around it is a turn-off. The first thing that comes to mind when I read about THX is MQA.

 

[amper42]

Here is an in room response. It doesn’t show bass below 20hz like I thought it might, but shoes solid response flat to 25hz or so. I don’t think this suggests weak bass though. It looks pretty good to me.

I will update this with room volume when I get the dimensions. As well as how sealed the room is. I didn’t measure this. But I will have measurements from Genes room soon enough. Within a month or so I hope.
View attachment 132083

Matthew - was EQ used with this measurement? It looks way too good to be without some EQ on the low end. What were the electronics used?

Tweeter drops off rather quickly after 11K. Any idea what type of tweeter that is? Thank you!

 

[MZKM]

Matthew - was EQ used with this measurement? It looks way too good to be without some EQ on the low end. What were the electronics used?

Tweeter drops off rather quickly after 11K. Any idea what type of tweeter that is? Thank you!

As you can see from the graph, the highlighted response is an average of the others. Upper treble always slopes down for in-room response.

I haven’t, but if we look at the Early Reflections part of the Spin, it should match the in-room he measured.

Now, the graph does span 110dB, so it does smooth out the look.

 

[Matthew J Poes]

Matt, I'm still a little unclear about the the efficiency differences between the smallest S4b bookshelf and S7t towers when subwoofers are used. Given they employ identical tweeter/mid array, I assume that the efficiency given is for maximum bandwidth +/- 1.5dB down to the lowest frequency capability of each speaker. Obviously, performing at these lower frequencies is what pulls down efficiency numbers of the bookshelf, so if we use a subwoofer high pass at 120Hz for the S4b, wouldn't the S4b efficiency numbers be significantly higher?

Your mixing up efficiency and sensitivity. Have to be careful on the terms we use here. Efficiency is something different.

Reference sensitivity is a measure of a speaker when a 2.83v signal is sent to a speaker and measured at 1 meter. The average SPL over a range of frequencies is then used to calculate the sensitivity. It is supposed to be given as a free space or anechoic sensitivity spec, but some companies (cough, Klipsch, cough) give a fudged number based on in-room and includes a boost caused by boundary reinforcement and reflections.

So let's take this speaker. The S7T has 4 midbass drivers and an array. The S4B has only one midbass driver. The tweeter is around 112dB 2.83v/1m with no padding. So its way hot. We can't use it like that. Let's pretend the sensitivity of the midbass driver is 90dB. 2 of them would be 93dB. 4 of them would be 96dB. You might think you end up with a 96dB sensitivity rating in that speaker. However, we have a problem, baffle step. As sound wraps around the baffle, we get a boost in the response in the midrange, and then it drops off toward the low frequencies, creating a shelf. It's around 6dB, but in practice we will get variation depending on how the speaker is designed. In any case, our 96dB sensitive speaker is now back to 90dB once we fully account for it.

With the bookshelf, we have a problem, we don't get any boost from the multiple drivers, so now we are stuck with a sensitivity of just 84dB after designing the crossover such that the baffle step has been addressed. So you can see, even if you add subs, it won't restore the sensitivity. The speaker is still only putting out 84dB at, say 500hz with 2.83 volts.

Of course these numbers don't directly map onto Perlistens, I just made them up. It was just to illustrate the issue you run into with designing speakers and sensitivity. There are a number of nuances that impact how precise the "theoretical" numbers match reality. For example, an array of 4 midbass drivers often combine more coherently at midbass and bass frequencies than they do at midrange frequencies, so there is a slight natural compensation for the baffle step losses. Some designers prefer a leaner design under the assumption that subs will be used. Personally I think that leaves things too lean between the 100hz and 500hz range, but I've seen it a number of times. I think it is because they are chasing SPL.

 

[Matthew J Poes]

As you can see from the graph, the highlighted response is an average of the others. Upper treble always slopes down for in-room response.

I haven’t, but if we look at the Early Reflections part of the Spin, it should match the in-room he measured.

Now, the graph does span 110dB, so it does smooth out the look.

Yeah exactly. When you consider the perspective (dB range) and fact that it is an in-room average, this is pretty expected behavior. The response of an in-room measurement shouldn't be flat. It will/should slope down a bit in the treble and have a bit of a wedge/slope shape from 20hz to 20khz.

This isn't necessarily reflective of what you hear either. The treble timbre is largely set by the listening window, not the in-room average. For example, speakers that have tweeters that beam like crazy tend to show aggressive sloping but wouldn't sound it. It's just an artifact of an in-room steady-state measurement.

 

[Matthew J Poes]

My guess, based on what I've read here, is that you have to approach them as a designer and OEM of speakers, and sign a non-disclosure agreement to see the specs. The definitions of their specs almost certainly aren't patentable, so a non-disclosure agreement is their only recourse, if they really believe they need to be secret.

Just my opinion, I suppose worth to THX what they're paying for it, but as a member of their target market, I think this designation is completely worthless, and the hype around it is a turn-off. The first thing that comes to mind when I read about THX is MQA.

I know three companies that applied, one succeeded, as I understand, all are aware of exactly what specs they are supposed to hit. They all signed an NDA for that information. It's iron clad, they can't share it. I asked Dan, he shared the same thing, can't share, NDA.

 

[Spocko]

Your mixing up efficiency and sensitivity. Have to be careful on the terms we use here. Efficiency is something different.

Reference sensitivity is a measure of a speaker when a 2.83v signal is sent to a speaker and measured at 1 meter. The average SPL over a range of frequencies is then used to calculate the sensitivity. It is supposed to be given as a free space or anechoic sensitivity spec, but some companies (cough, Klipsch, cough) give a fudged number based on in-room and includes a boost caused by boundary reinforcement and reflections.

So let's take this speaker. The S7T has 4 midbass drivers and an array. The S4B has only one midbass driver. The tweeter is around 112dB 2.83v/1m with no padding. So its way hot. We can't use it like that. Let's pretend the sensitivity of the midbass driver is 90dB. 2 of them would be 93dB. 4 of them would be 96dB. You might think you end up with a 96dB sensitivity rating in that speaker. However, we have a problem, baffle step. As sound wraps around the baffle, we get a boost in the response in the midrange, and then it drops off toward the low frequencies, creating a shelf. It's around 6dB, but in practice we will get variation depending on how the speaker is designed. In any case, our 96dB sensitive speaker is now back to 90dB once we fully account for it.

With the bookshelf, we have a problem, we don't get any boost from the multiple drivers, so now we are stuck with a sensitivity of just 84dB after designing the crossover such that the baffle step has been addressed. So you can see, even if you add subs, it won't restore the sensitivity. The speaker is still only putting out 84dB at, say 500hz with 2.83 volts.

Of course these numbers don't directly map onto Perlistens, I just made them up. It was just to illustrate the issue you run into with designing speakers and sensitivity. There are a number of nuances that impact how precise the "theoretical" numbers match reality. For example, an array of 4 midbass drivers often combine more coherently at midbass and bass frequencies than they do at midrange frequencies, so there is a slight natural compensation for the baffle step losses. Some designers prefer a leaner design under the assumption that subs will be used. Personally I think that leaves things too lean between the 100hz and 500hz range, but I've seen it a number of times. I think it is because they are chasing SPL.

OK, this totally makes sense now - thanks!

 

[richard12511]

My guess, based on what I've read here, is that you have to approach them as a designer and OEM of speakers, and sign a non-disclosure agreement to see the specs. The definitions of their specs almost certainly aren't patentable, so a non-disclosure agreement is their only recourse, if they really believe they need to be secret.

Just my opinion, I suppose worth to THX what they're paying for it, but as a member of their target market, I think this designation is completely worthless, and the hype around it is a turn-off. The first thing that comes to mind when I read about THX is MQA.

THX certification has always been a negative in my eyes. However, that was due in large part because I figured it was just a volume thing. Knowing now that it requires certain standards for linearity, off axis linearity, distortion, etc., I'll probably see it much more favorably.

 

[Laserjock]

Could see spin-offs meeting the requirements without the fee but also without the “advertising” rights for knowing they do the same.
Just thinking out loud... on a keyboard.

hmmm

 

[blueone]

THX certification has always been a negative in my eyes. However, that was due in large part because I figured it was just a volume thing. Knowing now that it requires certain standards for linearity, off axis linearity, distortion, etc., I'll probably see it much more favorably.

If consumers don't know what the certification criteria are, IMO it's worthless. Just raises prices.

 

[MZKM]

If consumers don't know what the certification criteria are, IMO it's worthless. Just raises prices.

Most don’t care. The average person still thinks Bose is excellent because they see them all the time with their NC headphones, car stereos, etc.

Maybe not the younger generation, but people will simply see THX on the product and equate it to being good (I really do miss the robot with that cow noise maker).

 

[Everett T]

If consumers don't know what the certification criteria are, IMO it's worthless. Just raises prices.

Measurements should tell you what that is, just not layed out. I use to laugh off avr and amp ratings, but the devil was in the details. Benchmark provided that for me. Speakers and subs were easier to see as there were more measurements over the years.