I have been playing with my measurement setup and programs and I decided to do a full set of measurements on my refrence speakers; the Linn Katan bookshelf speaker. I added some commentary, but I tried to let the measurements mostly speak for themselves. I got some very strange results, it really makes me wonder what these speakers are doing that I like so much. If anybody can make sense of the puzzling crossover topology, or has any other insights to explain this speakers subjective quality (at least in my eyes) please try and explain it.
MEASUREMENTS
Thanks,
-Chad
MEASUREMENTS
Thanks,
-Chad
I haven't heard the Katan, but have listened extensively to the Kan. I think that's the name of the speaker. It was promoted with the Linn Classik.
Anyway, my impression was excellent imaging and wide dispersion, and rich sound.
I always suspected that the rich sound was due to moderate amounts of harmonic distortion, (my partner also commented that it sounded rich in overtones) and the wide dispersion was due to excellent power response.
Anyway, my impression was excellent imaging and wide dispersion, and rich sound.
I always suspected that the rich sound was due to moderate amounts of harmonic distortion, (my partner also commented that it sounded rich in overtones) and the wide dispersion was due to excellent power response.
ch83575 said:
You probably like the due to the wide overlap augmented mid power response and tilted up highs, such speakers give spectacular sense of presence and detail. I remember the Kan sounding as such too. They are doctored to sound good against a wall. Wall reflects much mid, so a normal speaker would smear badly and would have sounded heavy if not bass light as it shows in your charts.. Boosting the presence in a vague lobe superimposes the reflection, but its still there, mixing in but smothered. Special speakers, ultimately colored though.
P.S. Why did you measure it on woofer axis? Just because it shows best phase integration there? Do you actually listen to it on its woofer axis?
I listen slightly above the woofer axis (about 1-2" depending on sitting/slouching) and that is the only place the crossover sums flat, so I thought the directivity plots would make more sense if they were normalized to the flattest axis. I dont like when a normalized graph looks awfull when actually the problems are all on axis and are remedied off axis (like the horisontal plot in these measurements).
I dont think I understand you fully on the topic of the vertical polar plot; are you saying that the off-axis suckouts are actualy to compensate for the wall reflection? I have known that these speakers used very little BSC (their woofer inductor is only .7mH) but I had no idea that close wall placement came into play in the crossover region as well. There are cirtainly things I do not like about these speakers (violins/strings for instance) but I would very much like to be able to understand exactly what they are doing that I DO like so I can incorporate it into my own designs. I do very much like the idea of close wall placement, even if it does have its compormises.
Thanks for the help,
-Chad
Better measure 1-2'' above woofer then so we can have a better idea of what you are listening.
I meant that the overlap (which ill effects the lobe) creates a non specific lobewise broadband, stronger mid presence region power response output that swamps the wall midrange reflection (non specific also) when the speaker is positioned very close to it. I have done such in such cases along minimum BSC. The end result is pleasing and near wall positioning compatible. Colored though.
In other words, use overlap crossing and 1dB BSC for up to wall placements.
I meant that the overlap (which ill effects the lobe) creates a non specific lobewise broadband, stronger mid presence region power response output that swamps the wall midrange reflection (non specific also) when the speaker is positioned very close to it. I have done such in such cases along minimum BSC. The end result is pleasing and near wall positioning compatible. Colored though.
In other words, use overlap crossing and 1dB BSC for up to wall placements.
- Status
- Not open for further replies.