Katan SPL - This speaker only provides phase alignment between woofer and tweeter on the woofer axis, therefor all measurements were taken along this axis.

Individual Driver Slopes - This is a very unusual crossover topology with an extremely wide overlapping region, where both the tweeter and woofer are both contributing equally to the sound.

Reverse Polarity Null - When the tweeters polarity is reversed it should produce a deep symmetrical null. This null is deep, but very wide for the steep slopes and not even remotely symmetrical.

Step Responce - Fairly good behavior in the time domain, notice the slight glitch at the hand-off from tweeter to woofer and the slight irregularity of the woofers decay.

Cumulative Spectral Decay - This graph also shows the slight error in the time domain, notice how the woofers response stands forward of the tweeter. Other than that the speakers decay looks good but not great with the ridges at 2.5k and 1k both being about 15dB down and almost 1ms long.

Horizontal Directivity - The horizontal response is excellent, with the diffraction caused dip at 3k and bump at 4.8k working themselves out very quickly off axis due to the square corners and sloping sides.

Horizontal Directivity - Normalized to woofer axis

Vertical Directivity - I am surprised to see such would would usually be considered "bad" vertical response in this speaker. Immediately off axis from the woofer the phase correlation between drivers becomes very bad and large suckouts occur above and below this axis. In its defense the steep filter slopes used in this design make the off-axis suck-outs very narrow in frequency and they are probably not as noticeable in room as the graph would suggest.

Tweeter Distortion

Woofer Distortion

Tweeter and Woofer Distortion Overlayed - These distortion measurements had to be done in the nearfield due to lack of gateing provisions in such measurements. Also, the absolute distortion levels are slightly higher than in reality because of the distortion profile of the microphone and the extreme volumes of nearfield measurements. This measurement was taken at "listening levels" of about 1/2W. It is really meant to show the distortion "pattern" more than its level. It can be clearly seen that the woofer breakup distortion is not very well controlled by the high crossover point, as shown by the peak in 3rd harmonics around 1.9k and 5th at 1.2k.