Hey X, as always impressed. Thanks for digging... I feel the case for Karlsons at least below 500Hz is becoming much stronger thanks to all this analysis and research...And I'm still reeling from that K15 predictive plot you posted... It immediately reminded me of ER4p (IEMs). I'm afraid after 3 years of spying on Freddi and associates I'm slowly being inducted into the Kult.
Interesting that Metro's T15 notching is at 250 and 400 respectively... I'm probably gonna build our test box in a modular fashion to allow us to try a few configs including the metro dimensions and see if that graph is marketing hyperbole or indoors response like you suggested, or what
Did you try different Qts drivers or even a 12" to see if that has any effect? So mysterious.
Interesting that Metro's T15 notching is at 250 and 400 respectively... I'm probably gonna build our test box in a modular fashion to allow us to try a few configs including the metro dimensions and see if that graph is marketing hyperbole or indoors response like you suggested, or what
Did you try different Qts drivers or even a 12" to see if that has any effect? So mysterious.
Last edited:
Hey X, probably a lark, but have you tried moving the measuring position, vertically?... Floor bounce playing a role?
Looks like the notch's depth and center was influenced by where Freddi measured:
Looks like the notch's depth and center was influenced by where Freddi measured:
that graph was with the mic 1/4" from the cone
here's the effect of a 5" deep front shelf on the same cabinet with mic laying on the ground and ~2M - there may be some bounce path effects - one trace shows the 5" deep board moved to the rear chamber as one might do as a back panel brace. K15 was somewhat smoother than this test-coupler. With the front board inserted, this cabinet sounded much like K15.
I'd like to know if he Acoustic Control 115BK scaled to K15 bulk would be smooth
here's the effect of a 5" deep front shelf on the same cabinet with mic laying on the ground and ~2M - there may be some bounce path effects - one trace shows the 5" deep board moved to the rear chamber as one might do as a back panel brace. K15 was somewhat smoother than this test-coupler. With the front board inserted, this cabinet sounded much like K15.
An externally hosted image should be here but it was not working when we last tested it.
I'd like to know if he Acoustic Control 115BK scaled to K15 bulk would be smooth
Hey X, as always impressed. Thanks for digging... I feel the case for Karlsons at least below 500Hz is becoming much stronger thanks to all this analysis and research...And I'm still reeling from that K15 predictive plot you posted... It immediately reminded me of ER4p (IEMs). I'm afraid after 3 years of spying on Freddi and associates I'm slowly being inducted into the Kult.
Interesting that Metro's T15 notching is at 250 and 400 respectively... I'm probably gonna build our test box in a modular fashion to allow us to try a few configs including the metro dimensions and see if that graph is marketing hyperbole or indoors response like you suggested, or what
Did you try different Qts drivers or even a 12" to see if that has any effect? So mysterious.
You are comparing K15 with Etymotic Research ER4s? Interesting... I have a set of Aurvana 3 dual driver balanced armature IEM's that I use daily on a bus commute. Love them.
All headphones lack that nice spatial phantom image in front of you as the image is in between your ears!
I had to re-read this, for others interested in what causes the notch, here's what XRK has ruled out so far (in addition to floor bounce/mic position):
A kind of similar series of Notches also shows up on tapped horns, but the scaling is different (much tighter, more radical)... The Karlson being higher bandwidth and less tapped would be less affected by this notch... I'm wondering if you move the koupler wings lower so they block more of the driver's exit (creating more of a tapped chamber) if that would increase the depth of the notch's? I would expect it would... I'm starting to think perhaps the tapped aspect of the T15 is much less active than that of the K15 because of the less direct path from back to front and the boxed slot vs rectangular holes in baffle... We can see the Notch on the K15 is much deeper which indicates out of phase waves combining... and the bandwidth of the notch probably tells us something else about the enclosure or speaker...
A kind of similar series of Notches also shows up on tapped horns, but the scaling is different (much tighter, more radical)... The Karlson being higher bandwidth and less tapped would be less affected by this notch... I'm wondering if you move the koupler wings lower so they block more of the driver's exit (creating more of a tapped chamber) if that would increase the depth of the notch's? I would expect it would... I'm starting to think perhaps the tapped aspect of the T15 is much less active than that of the K15 because of the less direct path from back to front and the boxed slot vs rectangular holes in baffle... We can see the Notch on the K15 is much deeper which indicates out of phase waves combining... and the bandwidth of the notch probably tells us something else about the enclosure or speaker...
Another thought, did you try making the rear chamber smaller?
My theory at this point is the notch is related to the distance from rear to front (which will only be noticeably adjustable if there is actual proper loading of the tapped aspect of the cabinet)... And my guess is the t15 is a lot less tapped than the k15 and that we'll be able to actually move the notch around (from box changes) when the front chamber is more tightly engaging the tapped mode of the cab. So my proposal to change the koupler wings to follow a tight log curve that enforces much stronger front to back coupling might have some validity... In this theory I guess first evidence would be found by removing wings (predicted to reduced the notch depth) and then adding a rectangular front baffle like a tapped horn (producing deeper notches).. Once we have evidence of the effects of the tap, then we can try changing path lengths and angles to modify it...
on a wing here,
My theory at this point is the notch is related to the distance from rear to front (which will only be noticeably adjustable if there is actual proper loading of the tapped aspect of the cabinet)... And my guess is the t15 is a lot less tapped than the k15 and that we'll be able to actually move the notch around (from box changes) when the front chamber is more tightly engaging the tapped mode of the cab. So my proposal to change the koupler wings to follow a tight log curve that enforces much stronger front to back coupling might have some validity... In this theory I guess first evidence would be found by removing wings (predicted to reduced the notch depth) and then adding a rectangular front baffle like a tapped horn (producing deeper notches).. Once we have evidence of the effects of the tap, then we can try changing path lengths and angles to modify it...
on a wing here,
Last edited:
I revisited IG81's informative post on the Karlson here:
http://www.diyaudio.com/forums/full-range/213594-karlson-10.html#post3529033
IG81, Magnetar, GM and others have expressed that hi-fidelity use of the K15 is best when it's crossed below the notch... The notch I've been trying to figure out how to move lower
FWIW, moving it lower is probably a bad thing if everything above it starts to jump and jive like it's going through a comb filter, yet knowing what controls the hi-fi BW (and it's markers) could help us make improvements to the design, so I think it's still worth seeking data that explains what controls this.
The super smooth plot of the Kilomax-K15 that was posted a few pages back made me completely forget how ragged the response above this notch looks in many (but not all) other measurements... Hey X, is that graph smoothed or is that K15 simulated with stuffing in the back chamber or something? Compared to your first simulations of the K15 on the other thread it's unbelievably smooth. When I first saw it I looked up in the treble region and saw some ragged edges and concluded it wasn't smoothed.
Both Freddi and IG81 have stated that cone control extends below Fb so we can use a little EQ to safely extend the bass unlike a purely reflex cab. This is worth exploring as it could save between 3 and 6 inches volume depending how much extra control the Karlson offers, the LF target and how many watts we need to push. Since I don't run sound for a living, I'm gonna have to ask someone, but I'm guessing we want minimum 89dB (as per Mancuso's loft party specification) and max 95dB RMS - both these at 5 meters... I think 95dB RMS ends up around 119dB peaks.
what'd u say?
so yeah,
David Mancuso's approach of exactly 89dB at center of dance floor is really wonderful, but doesn't work when the temp drops below 50f. The more slam and weight we have the better cold bodies will respond. 4 T15's should do the job.
http://www.diyaudio.com/forums/full-range/213594-karlson-10.html#post3529033
IG81, Magnetar, GM and others have expressed that hi-fidelity use of the K15 is best when it's crossed below the notch... The notch I've been trying to figure out how to move lower
FWIW, moving it lower is probably a bad thing if everything above it starts to jump and jive like it's going through a comb filter, yet knowing what controls the hi-fi BW (and it's markers) could help us make improvements to the design, so I think it's still worth seeking data that explains what controls this.
The super smooth plot of the Kilomax-K15 that was posted a few pages back made me completely forget how ragged the response above this notch looks in many (but not all) other measurements... Hey X, is that graph smoothed or is that K15 simulated with stuffing in the back chamber or something? Compared to your first simulations of the K15 on the other thread it's unbelievably smooth. When I first saw it I looked up in the treble region and saw some ragged edges and concluded it wasn't smoothed.
Both Freddi and IG81 have stated that cone control extends below Fb so we can use a little EQ to safely extend the bass unlike a purely reflex cab. This is worth exploring as it could save between 3 and 6 inches volume depending how much extra control the Karlson offers, the LF target and how many watts we need to push. Since I don't run sound for a living, I'm gonna have to ask someone, but I'm guessing we want minimum 89dB (as per Mancuso's loft party specification) and max 95dB RMS - both these at 5 meters... I think 95dB RMS ends up around 119dB peaks.
what'd u say?
so yeah,
David Mancuso's approach of exactly 89dB at center of dance floor is really wonderful, but doesn't work when the temp drops below 50f. The more slam and weight we have the better cold bodies will respond. 4 T15's should do the job.
Last edited:
The smoothness of the response comes from the fact that I set the walls really far away (like a stadium) but if you have walls in the 6 ft to 20 ft range, you will start to see dips and wiggles due to room interaction. I will put a wall back in for you to see. 90 dB at 5 meters should be easy for the K15 - as that was my premise for this thread... I thought normally wanted to cross over below the notch - but thought you had a reason to do it at notch?
I do have damping in the back chamber now and that helps a little with the fine fuzz but larger dips is really from room effects. The K15 and probably the T15 do not like to be near a wall as you get really big dips due to room effects.
I do have damping in the back chamber now and that helps a little with the fine fuzz but larger dips is really from room effects. The K15 and probably the T15 do not like to be near a wall as you get really big dips due to room effects.
Last edited:
You have to shift the alignment's Fl, Fh lower like any other BP alignment, i.e. if the alignment is from ~50 - 400 Hz [~3 octaves] and you want it to have a ~250 Hz Fh, then Fl = ~250/2^3 = ~31 Hz.
GM
Indeed, but seems like doing so will ultimately result in loss of sensitivity and the low tunings won't allow the use of a K-slot for the front chamber anymore. We have to remember that the classic Karlson has no pretense of low-frequency performance (ignore the old sales pitch
) and is more of a midbass enclosure IMO, better leave it ~as is. A real f3=40Hz might already involve some sacrifices.In a low-tuned K15, we could get a 416-8C down to f3~30Hz, but system sensitivity is down to ~90dB, kind of a shame, even if it is still plenty usable.
IG
We have a couple PA use scenarios, one involves a horn crossing at 350-400, like the jbl2360, another involves crossing at 500hz to something like the altec mr945, and a later possibility involves crossing a DIY 90deg synergy horn...
The desire to shift the notch lower was based on a string of assumptions starting with the T15 will have less noticeable "reverb/cabinet reflections" than K15 making the useable upper bandwidth greater which then led to, "oohh, it would be nice if that notch was doing something useful for our mix rather than mucking with critical elements at 350". Since it's so common when mixing to notch out mud in the 190-250 range, it seemed useful if it could be positioned there somehow, but i was still thinking the bandwidth above the notch was good enough... Having not heard a K outside of youtube, I don't really know what to expect.
Last edited:
Indeed, but seems like doing so will ultimately result in loss of sensitivity........
......... and the low tunings won't allow the use of a K-slot for the front chamber anymore.
We have to remember that the classic Karlson has no pretense of low-frequency performance (ignore the old sales pitch
For a given set of driver specs, of course it will, same as any other alignment, so different specs are required to ~ 'have your cake and eat it too'. Not many 'free lunches' in audio.
Why not? It will need to be a different size and may even need to be deeper [thicker] to form a longer ‘flared’ vent, but can’t visualize ATM why it can’t be used regardless of alignment.
Yeah, in retrospect, I believe its downfall was primarily due to its ‘snake oil’ marketing hype.
Agreed, the original is already a good mid-bass/lower mids compromise for matching to a sub system and the many decades long well proven 350-500 Hz mids-up XO point, but apparently that’s not what Anthony wants/needs for his app.
GM
It's a bunch of resonant cavities coupled together and the lower it's tuned the wider its resonance BW same as different tuning forks. What matters though is how we perceive them and as the BW shifts further down into our poor acute hearing BW the more our internal processor sums them, otherwise we wouldn't be able to tolerate being in a room with < ~250-300 Hz output bouncing around, over, through, us.
Anyway, personally, ~500 Hz/2nd order is the max for me in a stock K15 'party' speaker where the HF needs protecting more than having high fidelity and ~250 Hz/2nd order otherwise, but I'm apparently more sensitive to attack/decay than most; so for prosound loaded with a prosound mid-bass horn driver, I'm guessing you could get by with up to a ~1200 Hz/4th order XO point.
GM
even with a coaxial, imo , K15 is better than a party speaker -- by that criteria some other speakers would be relegated to background muzak speakers. AN10 in a reflex w/o compensation is pretty horrible - it sounds better overall in K15,
K15 has good intelligibility with coaxial and good fullrange. Like most other things, that factor can be improved somewhat by splitting up the spectrum. I would not recommend xover lower than 700Hz for K15. A big radial horn sounded nice (much more real than what I've heard so far from FR in boxes and BLH). Edgar's rectangular 250Hz tractrix would not blend.
they were made for coaxial/FR - a K-tube works well and blends well in the vertical dimension.
in contrast to the Karlson I dislike the Klipschorn for the way it smears the bass. I'm also not smitten with reflex.
K15 has good vocals - if appropriate speakers are used.
Magnetar to my knowledge never built K15 but went ahead to build two versions of his "Rosie" Klam which by form places the woofer further away from the aperture. Bulk for bulk a series vented klam would have no advantages over K15 save lower height.
one concern in making Ultra-Fidelity style K-couplers is the likelyhood of too much cavity peaking - I think that may be controlled by altering the height vs width but in general one does not want to lose too much punch.
according to Wayne Green, Karlson ruined the business by 1956. After John's death in 1973, his widow sold multiples of licenses. One licensee said JBL threatened litigation if their little company didn't cease production
T15 as a fullrange is an unknown.
K15 tuned low and loaded per John Tucker of Exemplar sounds great - measurement wise it is not great as the low end is dominated by the rear chamber volume and overall system tuning. A 1/4 wave design would have more sensitivity on the bottom.
Karlson are a musician/music lover's speaker.
should the bass reflex below be used only below its notch? - how do we know a lot of speakers touted as diy projects don't have anomalies in their FR?
K15 has good intelligibility with coaxial and good fullrange. Like most other things, that factor can be improved somewhat by splitting up the spectrum. I would not recommend xover lower than 700Hz for K15. A big radial horn sounded nice (much more real than what I've heard so far from FR in boxes and BLH). Edgar's rectangular 250Hz tractrix would not blend.
they were made for coaxial/FR - a K-tube works well and blends well in the vertical dimension.
in contrast to the Karlson I dislike the Klipschorn for the way it smears the bass. I'm also not smitten with reflex.
K15 has good vocals - if appropriate speakers are used.
Magnetar to my knowledge never built K15 but went ahead to build two versions of his "Rosie" Klam which by form places the woofer further away from the aperture. Bulk for bulk a series vented klam would have no advantages over K15 save lower height.
one concern in making Ultra-Fidelity style K-couplers is the likelyhood of too much cavity peaking - I think that may be controlled by altering the height vs width but in general one does not want to lose too much punch.
according to Wayne Green, Karlson ruined the business by 1956. After John's death in 1973, his widow sold multiples of licenses. One licensee said JBL threatened litigation if their little company didn't cease production
T15 as a fullrange is an unknown.
K15 tuned low and loaded per John Tucker of Exemplar sounds great - measurement wise it is not great as the low end is dominated by the rear chamber volume and overall system tuning. A 1/4 wave design would have more sensitivity on the bottom.
Karlson are a musician/music lover's speaker.
should the bass reflex below be used only below its notch? - how do we know a lot of speakers touted as diy projects don't have anomalies in their FR?
An externally hosted image should be here but it was not working when we last tested it.
Last edited: