I've tried several different CD's with different horns now and a zillion
different crossover points and I cant get them to sound good to my ears.
Too harsh and shrill. My intuition tells me its from the low xmax and super high sensitivity.
Just doesnt sound natural and I also noticed that my tinnitus would flare up horribly.
Switched back to putting cones in horns and waveguides and my tinnitus no longer flares up.
What gives??
different crossover points and I cant get them to sound good to my ears.
Too harsh and shrill. My intuition tells me its from the low xmax and super high sensitivity.
Just doesnt sound natural and I also noticed that my tinnitus would flare up horribly.
Switched back to putting cones in horns and waveguides and my tinnitus no longer flares up.
What gives??
Last edited:
I recently bought a new pair of JBL speakers that has JBL 2446H compression drivers and Iwata waveguides and a dbx 234 active crossover. I had connected everything incorrectly and thought the combination didn't sound very good.
After doing a little research, I figured out that I had the connections all discombobulated. I've also got some tinnitus, but not affected by the speakers. Having a few beers while listening makes me turn up the volume.
Also realized that these are my favorite speakers, best I have ever personally heard and owned.
After doing a little research, I figured out that I had the connections all discombobulated. I've also got some tinnitus, but not affected by the speakers. Having a few beers while listening makes me turn up the volume.
Also realized that these are my favorite speakers, best I have ever personally heard and owned.
Most of the cd's I've tried have needed a fair bit of 'shaping' to get them sounding sweet / non harsh. (The BMS 4550 I'm using at the moment has been particularly hard work to get sounding 'nice')
The best 'out of the box' one I own is the Beyma CP380. It was the 1st compression driver I tried and I went to it after using scanspeak revelator tweeters. It sounded really good and I've never considered a 'normal' tweeter since.
Rob.
The best 'out of the box' one I own is the Beyma CP380. It was the 1st compression driver I tried and I went to it after using scanspeak revelator tweeters. It sounded really good and I've never considered a 'normal' tweeter since.
Rob.
for one inch format -depending upon driver, and xover point say 1K6 or higher, try a "K-tube". Typical would be 1 inch inside diameter, 5.3" long half ellipse based slot with a 1/8" starting gap. A quick paper tube could be used for POC. The tube would point up ~30-30 degrees. A baffle could be used at the driver exit to give some support to lower F.
Since there's little to no gain on the low end, no contouring is needed in the highpass network. Directivity collapses on the top octave but the impression is very wide and open sound which can be pleasant even up close.
This commercial product had a 2 degree downward slope to its sidewalls. A straight pipe version will give pretty much the same result. Schedule 40 pvc works well.
Since there's little to no gain on the low end, no contouring is needed in the highpass network. Directivity collapses on the top octave but the impression is very wide and open sound which can be pleasant even up close.
This commercial product had a 2 degree downward slope to its sidewalls. A straight pipe version will give pretty much the same result. Schedule 40 pvc works well.
I think what is happening is the compression driver horn combinations have a lot of high Q frequency spikes that an not obvious with normal listing, but will be several dB louder than the average level of the horn output. These narrow output spikes excite your cochineal hair cells to the pint of distress. As others have stated, virtually all horns or waveguides need trap filters or active EQ. Other really important area is the transition between the CD throat and the waveguide. It is critical the angle is the same and the transition is smooth. Bug screens are evil, most reflect way to much sound back down to the CD. A good screen should be almost invisible or it should be removed.
You are right, of course.
Your ears don´t lie.
Yes, they *can* be tamed, EQ´d, etc.
All of which confirms the existance of the problem, doesn´t it?
Yes. There is clearly something to what you state here. Its hard to put into words
but its very obvious after 30+ minutes of listening. I found that the CD's always
put me 'on edge'. They are unequaled in realism from what Ive experienced but the tradeoff
is a certain harshness that to me isnt worth it. I cant relax into the music...like you say there's
some type of 'spike' that is too loud, I even tried using compression but it still doesnt fix it.
Snares for example, can bang like firecrackers rather than snap. Cymbals crash rather than shimmer...
My dome tweets invite me to turn up the volume without punishing me. I put an integra 424 morel mid/tweet into a waveguide and Im very pleased with it.
Morel's (and Dynaudio) are polar opposites to CD's..
Have you tried this one?:
Your next compression driver?
This is critical with any design (be it soft dome, hard dome, CD, etc..)
Subjectively the listener listens as an "average" much closer to the top of the "spikes" in freq. response than often what the measurement software suggests is the average. (..this becomes even more of a problem if the raw/non-averaged result has enough spacing between "spikes" (as a opposed to a "dense" spectra of "hash"), and CD's often are worse in this respect than many dome tweeters.)
Because we are generally capable of detecting .5 db changes in spl above (about) 1.5 kHz, just a small deviation from what we actually perceive as flat vs. what we typically measure (as an averaged result) can be substantial.
It's always best to look at a non-averaged result before deciding on final "voicing" of the crossover.
To compound matters:
increasing directivity (as with a waveguide) subjectively RAISES the spl-level relative to a less directive (more "omni") midrange or midbass driver.
-so it's often a double-whammy of a result: to loud because of an average response that is shown by your measurement program to be "flat" (when the "spikes" are substantively higher in pressure than that average) and ALSO to loud because your waveguide tweeter is more directive than the midrange.
Last edited:
The inconvenient truth is...
Compression drivers are difficult, and often expensive, to get right, and most people have never heard a really good implementation.
There's so much one can get wrong, from diaphragms that break up and resonate, to abrupt discontinuities at the horn throat, to flimsy / diffractive /too small horns used too low /too high/etc... to the ALL-IMPORTANT crossover! The list just goes on and on.
But, for whatever it's worth, my personal informed opinion is that when everything falls into place correctly, they are THE BEST way to reproduce the critical midrange, bar none.
Marco
Compression drivers are difficult, and often expensive, to get right, and most people have never heard a really good implementation.
There's so much one can get wrong, from diaphragms that break up and resonate, to abrupt discontinuities at the horn throat, to flimsy / diffractive /too small horns used too low /too high/etc... to the ALL-IMPORTANT crossover! The list just goes on and on.
But, for whatever it's worth, my personal informed opinion is that when everything falls into place correctly, they are THE BEST way to reproduce the critical midrange, bar none.
Marco
Very high praise!
That or I went to too many metal gigs as a teenager
To be clear I didn't say it was better than the scanspeak, just to my ears 'as good as' and slightly drier/more direct presentation which I prefer. The ability to go as loud as ever needed without strain is also a bonus for me.
Cheers,
Rob.
With horns it's more work to get them "right" as has been stated already. With a simply system like let's say 8" widebander with tweeter crossed at 5k it takes me 3 minutes to get the attenuator for the tweeter to where it needs to be. With horns 2 ways rarely are an option so it's 3 ways for a start and getting the levels right so they are right for the whole range of listening levels is rather taking 3 months. And it can be frustrating along the way but pay off nicely. Like when I feel something is broken because I seemingle can't crank it up and then measure SPL and the meter shows 90dB but it's so clean that it doesn't screach or hurt. Quite the opposite from the first post.
Most horns have at least 1% 2nd harmonics at any level. Many cone drivers (which includes domes) at normal living room listening levels do considerably better than that.
IMO, only at high SPL, horns become better than cone drivers because either cone drivers start to distort even more, or can't reach the desired SPL to begin with.
IMO, only at high SPL, horns become better than cone drivers because either cone drivers start to distort even more, or can't reach the desired SPL to begin with.