Maybe a step response or impulse response measurement would be enough. We could then convolve this with a square wave signal (or whatever you like) to see what the loudspeaker response to the square wave (or whatever other signal) would be. Joe, can you upload a step response or impulse response measurement?
I did some reading on Joe’s website for the Elsinore. The XO plot does indeed look close to -6dB/octave around the XO frequency. Those are 3dB per division vertical scale.
With the waveguide for some setback, this might have a chance at being transient perfect.
But a Step Response would clarifying things right away.
With the waveguide for some setback, this might have a chance at being transient perfect.
But a Step Response would clarifying things right away.
Attachments
Last edited:
What's that hump in the SPL curve at 150 Hz?I did some reading on Joe’s website for the Elsinore. The XO plot does indeed look close to -6dB/octave around the XO frequency. Those are 3dB per division vertical scale.
With the waveguide for some setback, this might have a chance at being transient perfect.
But a Step Response would clarifying things right away.
Out of stock...does anyone know if that's a permanent thing or temporary. Sometimes kits just go away. Looks like a fun project though.
The bump gets smoothed with a contouring filter applied later. This was just a mid/tweet XO discussion.
More details here if you are interested:
Elsinore Computer Modelling
The 150 Hz bump is most likely floor bounce.
Room acoustics and speaker placement are largely responsible for what happens below 200 Hz or so.
No, the raw response is flat, the addition of 6mH low pass choke causes the bump at 150Hz.
Reading the excellent detailed explanation of the XO development by Joe Rasmussen is very helpful. Very few designers go through this much detail to explain their choices and design methodology for a XO.
Elsinore Computer Modelling
Reading the excellent detailed explanation of the XO development by Joe Rasmussen is very helpful. Very few designers go through this much detail to explain their choices and design methodology for a XO.
Elsinore Computer Modelling
No X, the Elsinore Computer Modelling story applies to the first generation Elsinore (it was 9 instead 6 mH by the way).
The current (EL6) Elsinore has a different crossover with a 3.9 mH woofer choke and uses SB drivers instead Peerless.
Joe points out that any peak and dips below 1 kHz are primarily room related.
The current (EL6) Elsinore has a different crossover with a 3.9 mH woofer choke and uses SB drivers instead Peerless.
Joe points out that any peak and dips below 1 kHz are primarily room related.
I don't know...I just read a blip from '17 that said they closed the doors and would not reopen. Maybe things have changed.
The point is, the measurement was flat without the choke so it wasn't a room bounce. Anyone experienced at developing crossovers takes the measurement with proper time-gating and precautions to eliminate room floor-bounce for the critical FRD file used to synthesize the XO.
It's a well known fact that time gating is reliable for frequencies above 200 Hz.
Below 200 Hz floor bounce, room acoustics and speaker placement dominate as already stated.
I am pretty sure that Joe did not use an anechoic room for measurements.
This low frequency room/placement dependancy is not typical for Elsinore loudspeakers but equally valid for any loudspeaker.
Below 200 Hz floor bounce, room acoustics and speaker placement dominate as already stated.
I am pretty sure that Joe did not use an anechoic room for measurements.
This low frequency room/placement dependancy is not typical for Elsinore loudspeakers but equally valid for any loudspeaker.
Normally a nearfield or in-box measurement is used to measure the SPL curve at low frequencies. No need for gating or anechoic rooms. In fact, gating would not yield any data at all below 300 Hz or so, because the impulse response would be too short for that.
Developing an x-over based on echoic data from a specific room would hardly be useful, because the resulting x-over would reflect the acoustics of that specific room.
I therefore think (hope?) that the low-frequency part of the SPL curve shown in the above plots was determined from a nearfield or in-box measurement.
Developing an x-over based on echoic data from a specific room would hardly be useful, because the resulting x-over would reflect the acoustics of that specific room.
I therefore think (hope?) that the low-frequency part of the SPL curve shown in the above plots was determined from a nearfield or in-box measurement.
I guess so.
Joe used nearfield measurement for LF and farfield measurement for MF/HF.
He merged the measurements into a complete picture.
From website Classe seems to be in business.
Classe Audio - Home Page
News reports are that key staff got it refunded in 2018.
dave
Classe Audio - Home Page
News reports are that key staff got it refunded in 2018.
dave
I wondered why the had all this gear for sale. I kinda jumped in on the 2200i because it looked like a solid piece but certainly power overkill for something like the Elsinores. I think that's going back to the dealer.
Phase is related to the derivative of magnitude. Phase will therefore be rather flat.
