Better Klipsch RP-150m off axis measurements

[yepimonfire]

Thanks to @shadyJ for measurement technique tips and for modeling the polar plots.

The graphs labeled “axis” were taken at .5m and a 4ms window was applied, the graphs labeled “klipsch axis 2” were taken at 1m. One is windowed and one is not. Shadyj applied 1/12 octave smoothing.

The off axis dispersion remains relatively flat all the way to 20khz out to ~20 degrees, and flat to 14khz at +-45 degrees. A steep roll off appears at 10khz out at 60 degrees, with a deep notch in response appearing from 13khz-19khz. It’s interesting to note that after the notch, beyond 45 degrees, dispersion at 19-20khz remains fairly even, not sure why this is.

Overall, the coverage of the newly designed hybrid tractrix horn is much better than previous models, providing very wide and even off axis coverage void of beaming.

I will add FR graphs tomorrow morning.




Sent from my iPhone using Tapatalk

 

[ATLAudio]

@yepimonfire

"Overall, the coverage of the newly designed hybrid tractrix horn is much better than previous models, providing very wide and even off axis coverage void of beaming."

You gained this perspective by singularly measuring the new design? Wouldn't you want to compare it with the same measurements to the older one? If you do please show me the measurement of the Reference II "beaming."

 

[shadyJ]

Hey Yep, the notch you are referring to is almost certainly a diffraction effect from the horn. I would also note that the irregularities at the 90 degree angle in the graphs labeled 'Axis' are very likely some kind of reflection that is getting into the measurement, so it is an measurement issue and not behavior by the speaker itself. I probably should have trimmed the 90 degree angle off that one, but it makes for such a pretty graph!

 

[Swerd]

All those graphs do look entertaining and impressive. Thanks.

I'm not too concerned with the loss of off-axis SPL in the higher treble range, approaching 20 kHz. In my experience, that wouldn't be too noticeable, and there is nothing one could do to increase the off-axis response of a tweeter in that range.

But I am concerned with the noticeable loss of signal, roughly 5-10 dB especially off-axis, in the range of 1 to 2 kHz. FWIW, the crossover frequency of that 2-way speaker is 1.5 kHz. Loss of output in that range, especially off-axis, would be very noticeable if it were directly compared to another small 2-way speaker that didn't display that off-axis SPL trough. I remember one AH poster who, in the past, described this off-axis discontinuity as a "mushroom cloud of sound".

I wonder why a smaller mid-woofer, 5½" diameter (?), such as in that speaker begins to beam at so low a frequency. Many other mid-woofers of a similar size are known to perform well off-axis above 2 kHz.

Because this seems to be centered around the crossover frequency, I wonder if some choice the designer made while designing the crossover network, led to that discontinuity?

 

[shadyJ]

All those graphs do look entertaining and impressive. Thanks.

I'm not too concerned with the loss of off-axis SPL in the higher treble range, approaching 20 kHz. In my experience, that wouldn't be too noticeable, and there is nothing one could do to increase the off-axis response of a tweeter in that range.

But I am concerned with the noticeable loss of signal, roughly 5-10 dB especially off-axis, in the range of 1 to 2 kHz. FWIW, the crossover frequency of that 2-way speaker is 1.5 kHz. Loss of output in that range, especially off-axis, would be very noticeable if it were directly compared to another small 2-way speaker that didn't display that off-axis SPL trough. I remember one AH poster who, in the past, described this off-axis discontinuity as a "mushroom cloud of sound".

I wonder why a smaller mid-woofer, 5½" diameter (?), such as in that speaker begins to beam at so low a frequency. Many other mid-woofers of a similar size are known to perform well off-axis above 2 kHz.

Because this seems to be centered around the crossover frequency, I wonder if some choice the designer made while designing the crossover network, led to that discontinuity?

The measurement where that 1.5 kHz dip is most pronounced was made at a half meter, so it might be too close to the speaker for the sound of the drivers to integrate. I wouldn't necessarily blame that on the speaker.

Also I don't think this speaker is beaming at low frequencies, I think that is a consequence of the measurements and the way the graph scales might look. For the graphs labeled 'Axis' think that acoustic reflections are affecting the off-axis measurements, and making it look like it is beaming when it is just cancellation dips.

 

[Swerd]

In the OP, I see three different graph labels: Axis, Klipsch Axis 2, and Klipsch 2 Windowed.

Two were defined:
Axis – .5m (microphone distance) – 4ms time window.
Klipsch Axis 2 – 1m – no time window limits.
What does Klipsch 2 Windowed refer to?

The measurement where that 1.5 kHz dip is most pronounced was made at a half meter, so it might be too close to the speaker for the sound of the drivers to integrate. I wouldn't necessarily blame that on the speaker.

That could also indicate the woofer and tweeter may have been partially canceling each other out between 1 and 2 kHz, as if one were out of phase with the other. Separate measurements of the woofer alone and the tweeter alone while connected to the crossover might help understand if that is the case.

Also I don't think this speaker is beaming at low frequencies, I think that is a consequence of the measurements and the way the graph scales might look. For the graphs labeled 'Axis' think that acoustic reflections are affecting the off-axis measurements, and making it look like it is beaming when it is just cancellation dips.

I agree at 1 meter without a time window (Klipsch Axis 2) there is less of an SPL loss in the 1.5 kHz range than in the Axis plots. But a loss of as much as 5 dB (depending on the off-axis angle) is still there. That could be explained by the absence of a time window during measurement, where late arriving boundary reflections better fill in the trough. What would graphs look like if the microphone was 1 m away and a 4 msec time window was used? And how would they compare to the measurement at 0.5 m with a 4 msec window? That might answer my question.

This loss is in a frequency range where our ears are quite sensitive to peaks or valleys smaller than 3 dB.

I also see a similar loss in the graphs labeled Klipsch 2 Windowed but I don't know what its measuring conditions were.

 

[yepimonfire]

The measurement where that 1.5 kHz dip is most pronounced was made at a half meter, so it might be too close to the speaker for the sound of the drivers to integrate. I wouldn't necessarily blame that on the speaker.

Also I don't think this speaker is beaming at low frequencies, I think that is a consequence of the measurements and the way the graph scales might look. For the graphs labeled 'Axis' think that acoustic reflections are affecting the off-axis measurements, and making it look like it is beaming when it is just cancellation dips.

The mic was placed in line with the tweeter, since I was only interested in how the horns dispersion behaves, I’ve taken several other measurements that do not display that dip, so it likely is a consequence of the way it was measured. I may need to take some newer measurements of the whole speaker itself and see what happens. I’m sure I’d notice if it were an issue, as I have a pretty well trained ear, and can pick out problem frequencies with pretty good accuracy.

Klipsch generally crosses their speakers over at a point to maintain good directivity between the woofer and tweeter, which is why their smaller speakers that may use identical drivers for the tweeter are crossed over at 2200hz, and the large speakers, such as the RB 81 are crossed over at 1400hz, this keeps the wavelength well below the diameter of the driver and maintains similar dispersion between both the woofer and the tweeter.

Subjectively speaking, the speaker sounds identical to the on axis sound when sitting 45 degrees off axis on the sofa along the left wall, moving around the room while music is playing, or standing up vs sitting at speaker level, I hear almost no change in timbre except at bass frequencies (which is obviously unavoidable). There are no audible suck outs anywhere within the specified dispersion pattern.


Sent from my iPhone using Tapatalk

 

[yepimonfire]

In the OP, I see three different graph labels: Axis, Klipsch Axis 2, and Klipsch 2 Windowed.

Two were defined:
Axis – .5m (microphone distance) – 4ms time window.
Klipsch Axis 2 – 1m – no time window limits.
What does Klipsch 2 Windowed refer to?
That could also indicate the woofer and tweeter may have been partially canceling each other out between 1 and 2 kHz, as if one were out of phase with the other. Separate measurements of the woofer alone and the tweeter alone while connected to the crossover might help understand if that is the case.


I agree at 1 meter without a time window (Klipsch Axis 2) there is less of an SPL loss in the 1.5 kHz range than in the Axis plots. But a loss of as much as 5 dB (depending on the off-axis angle) is still there. That could be explained by the absence of a time window during measurement, where late arriving boundary reflections better fill in the trough. What would graphs look like if the microphone was 1 m away and a 4 msec time window was used? And how would they compare to the measurement at 0.5 m with a 4 msec window? That might answer my question.

This loss is in a frequency range where our ears are quite sensitive to peaks or valleys smaller than 3 dB.

I also see a similar loss in the graphs labeled Klipsch 2 Windowed but I don't know what its measuring conditions were.

Klipsch 2 windowed was at 1m with a 2.3ms time window. All measurements were taken above the woofer directly centered on the tweeter.

Somewhere I have a 1m measurement with the mic placed at the center level of the speaker (vs the tweeter) with a 3ms time window which does not display the dip, displaying a +3dB,-2dB response above 500hz.


Sent from my iPhone using Tapatalk