I've seen measurements of one Larsen model that was not pretty.
https://www.stereophile.com/content/larsen-hifi-8-loudspeaker-measurements
One of those few occasions where JA1 couldn't find something nice to say about the product: "the Larsen 8's measured performance reveals its audio engineering to be flawed." The big hump around 600Hz certainly looks worrying, but it would be interesting to see how these measure on a Klippel.
Yeah, the 8’s seems flawed.
John Atkinson apparently DID NOT measure the Larsen 8's in their DESIGNED-FOR location! They were DESIGNED to be used with their backs up against the front wall, but Atkinson says that he placed them up against the SIDE walls for the in-room measurements and for his listening!
The Larsens were designed for a SPECIFIC acoustic situation, and imo reliable conclusions cannot be drawn from in-room measurements made and listening done when they were placed improperly.
Figure 4, the quasi-anechoic frequency response, shows a large peak centered around 500 Hz and a deep dip one octave higher, at 1000 Hz. IF the speaker was placed with its back up against the FRONT wall (as intended), the wrap-around round-trip reflection path length for the midrange driver would be about thirteen inches. This would delay the reflection off the wall by about one-half wavelength at 500 Hz, relative to the direct sound. So there would be a cancellation dip centered on about 500 Hz at the listening position. Can you see where this is going?
Likewise the round-trip reflection path length would delay the reflection by one wavelength at 1000 Hz, resulting in a reinforcement peak in that region.
So the frequency response of the Larsens was designed to ZIG where the anticipated strong early wall reflections ZAG. If they are NOT placed up against the front wall as intended, their in-room frequency response will be AWFUL, just as Atkinson documented (Figure 7).
On the other hand, the boundary reinforcement loading in the bass region seems to have worked quite well up against the side walls, which would be expected because the first reflection path lengths are very small relative to the wavelengths. The quasi-anechoic frequency response is about -15 dB at 20 Hz, yet the in-room frequency response holds up well down to 20 Hz, and even deeper if we discount the nasty in-room peak around 600 Hz due to their incorrect placement.
I'm NOT saying that the native frequency response of the Larsen 8's is PERFECT (ime deliberately designing in a response peak is risky), but it is MUCH BETTER suited for its INTENDED placement than what Atkinson's (imo flawed) in-room measurements imply.
Anyway can you see the POTENTIAL inherent in Larsen's approach? Take another look at Figure 7 and mentally correct that in-room peak with proper placement (or EQ or a refinement in a subsequent version) and all of a sudden you've got something very interesting.