By the way, Fulcrum Acoustics (Dave Gunness) notes that using his form of "temporal" DSP correction results in loudspeakers that have the same timbre, even though they may have very different physical properties. This apparently enables Fulcrum to use their loudspeaker models in a mix-or-match configuration without worrying about timbre differences:
I found out about this effect first-hand...well after Mr. Gunness found it, about 8 years ago in my case, but that decay-trail-correction FIR filters are not really needed to achieve the same effect (timbre matching) if using loudspeakers having reasonable "early decay time" full-range performance.
Chris
I found out about this effect first-hand...well after Mr. Gunness found it, about 8 years ago in my case, but that decay-trail-correction FIR filters are not really needed to achieve the same effect (timbre matching) if using loudspeakers having reasonable "early decay time" full-range performance.
Chris
Thank you all for your contributions.
Something i want to tell about BMS 4592/4596; directly from BMS.
"4591 and 4592mid are similarly suitable here, both have the same coil installed.
4596 has a copper coil installed, which is designed for high output and has a little more distortion."
My speakers will be a full active system with digital crossover (Acourate Software with Room correction and FIR Filters) maybe it is a important to know that. Everyone wants to build a 2 way speaker but I think it will be not good enough. Too many compromises.. even with a Woofer like AE TD15M. 600Hz in a housing where the Wavelength fits in many positions... mostly built as Bass reflex.. I don't know
Something i want to tell about BMS 4592/4596; directly from BMS.
"4591 and 4592mid are similarly suitable here, both have the same coil installed.
4596 has a copper coil installed, which is designed for high output and has a little more distortion."
My speakers will be a full active system with digital crossover (Acourate Software with Room correction and FIR Filters) maybe it is a important to know that. Everyone wants to build a 2 way speaker but I think it will be not good enough. Too many compromises.. even with a Woofer like AE TD15M. 600Hz in a housing where the Wavelength fits in many positions... mostly built as Bass reflex.. I don't know
https://www.eighteensound.it/en/products/hf-driver/2-0/8/ND2080A ? 2"
https://www.eighteensound.it/en/products/hf-driver/1-4/8/ND1460A ? 1.4" - (better off axis ?)
https://en.toutlehautparleur.com/compression-driver-faitalpro-hf1440-8-ohm-1-4-inch.htmlhttps://en.toutlehautparleur.com/un...ion-driver-faitalpro-hf10ak-8-ohm-1-inch.html ? 1" (enough for home hifi due to the less needed spl ?)
https://en.toutlehautparleur.com/un...ion-driver-faitalpro-hf108r-8-ohm-1-inch.html ? (no need of DSP or expensive DRC = flat enough for passive filter XO ?)
I assume better if no XO in the treble area i-BMS driver- but if you have expensive low noise digital filter like Cask05 ?
https://www.eighteensound.it/en/products/hf-driver/1-4/8/ND1460A ? 1.4" - (better off axis ?)
https://en.toutlehautparleur.com/compression-driver-faitalpro-hf1440-8-ohm-1-4-inch.htmlhttps://en.toutlehautparleur.com/un...ion-driver-faitalpro-hf10ak-8-ohm-1-inch.html ? 1" (enough for home hifi due to the less needed spl ?)
https://en.toutlehautparleur.com/un...ion-driver-faitalpro-hf108r-8-ohm-1-inch.html ? (no need of DSP or expensive DRC = flat enough for passive filter XO ?)
I assume better if no XO in the treble area i-BMS driver- but if you have expensive low noise digital filter like Cask05 ?
I believe that you know that the 4592ND (dual diaphragm--which I recommend) and 4592mid (single diaphragm, which I don't if absence of lobing is important to you) aren't the same. I don't recommend separate horn apertures and drivers at those frequencies.
Clearly, I don't agree, but that's okay. Not everyone has the same insights and experiences in this pastime. The journey of learning is half the value, at least (although I do prefer to listen a lot more than tinkering, and to invest in different hi-fi recordings of my musical tastes.
Acourate certainly will provide you with a wide playing field with which to learn more on your own. As a learning approach, I would take it incrementally, first correcting as much as you can using IIR filters, then apply overlay FIR filters to flatten the phase response. Later, as your expertise in the DSP grows using these type of drivers/loudspeakers, then excursions using more FIR filtering techniques can be compared to your base case.
The good thing about DSP is that it supports continuous learning coupled with critical listening.
Happy trails.
Chris
The published data seems to show the 4596nd having less distortion or a more manageable distortion profile. At 1khz it's the same level, the 4592-mid has a rising distortion lever going up. For your application I still think 4596nd would be better than 4292nd or 4592-mid. A higher output driver tend to have less distortion at the same lower spl level, it's not always true but often see and probably the reason we like horns and compression drivers so much.
My system is using a DEQX in a 3 ways with fir filter, perfect impulse response correct etc. I work with off axis measurement to find the proper spot to put the crossover. Lot's of fun and work but the results are very good.
Like you I don't really get the attraction of a two way specially with large horns. I tried and got to listen to many iteration and I find the 5-10khz and up too compromised and the directivity in the highs kill me. I have a pair of faital hf10ak in azura horns 1100hz full roll back that are not bad up to 10khz and could do it with a 10" maybe a 12" with some compromises but I much prefer a 3 ways setup where I can cut the direct radiation as low as possible. 3 ways is already a compromise where you have to work hard to find proper drivers and horns so I think 3-4 ways is the sweet spot between SQ and complexity personally.
Interesting! Up to 2,6kHz. When I look at the pdf from BMS 4596 I could see that the distortion increases at about 3 kHz. Are these the chamber resonances?
A comparison with Axi2050 would be very interesting! Let us know about it 👍
I have 6x TD-4002 with hardly used DP-4001 diaphragms which I can let go for 2500€/pair. After cleaning, measurement, etc.
I have written a long text about my speaker ideas and hit the wrong button.... 😵😤 so everything is gone...
I will write later again... but for now another question:
Bjorn told me that the crossover point at 1600Hz is a bad choice...
So my question: where is the best place to crossover? I know it always a compromise but maybe a better one?
I will write later again... but for now another question:
Bjorn told me that the crossover point at 1600Hz is a bad choice...
So my question: where is the best place to crossover? I know it always a compromise but maybe a better one?
Attachments
In my direct experience, 230-300 Hz with an Axi2050 on a K-402 horn works well.
Using other 2" drivers, I recommend something around 500 ±100 Hz. I don't recommend crossing higher than 900 Hz-1 kHz--for some pretty important psychoacoustic reasons (starting on slide 17)--if clarity of presentation is important to you. Getting the throat apertures within 1/4 wavelength at the crossover frequency is pretty important for performance reasons, in my experience.
Perhaps you should consider stating your loudspeaker design requirements--rather than inserting apparent design solutions...which perhaps much less clearly imply the requirements assumptions that you're making.
Chris
Last edited:
I don't know if it's the case but I can't measure any resonance and would had guessed this to be a problem with the coaxial versions not the single diaphrams ones. Up to 2600hz only because after that the response is attenuated softly and would need eq correction to remain flat which I avoid when I can.
I have to find a pair of axi2050 first but I'm in no rush with the improvement the 4596nd brought me.
I don't really buy that a crossover between x and y is a bad choice. I can cut up to 300db/oct, if needed, in FIR with time correction and you would never be able to guess the crossover point, specially at the listening position but also off axis to reasonable angle if I do my job well. This concept is probably a general idea that was developed using passive or low order crossover points.
The crossover point should be where one of the driver start to get low on SQ or dispersion doesn't match properly in my opinion. If this happen to 1600hz and it's done properly I don't see any problem.
Some of my horns/drivers can be crossed over from 1500hz to 5000hz and there next to no SQ difference because both can do a good job in this range. Having some latitude for the crossover point can be useful but not necessary if the selection is done properly.
I made the comment because I've had 4590s for about 2 decades; according to BMS at the time, mine were the first pair to go to a customer south of the equator. The factory xovers were very off sounding and they never quite sounded right. Enter a very early DSP I got and tweaking it with the decent but not '22 grade measurement tools I had, it came into focus. At the time everyone was telling me the 4001 was better, but as I had the TADs and JBL2446 with Radian phrams, I didn't agree. My notebooks from the time are in storage so I can't give the settings.
I've been using Unitys since maybe 09 and I'm not sure. I'm about 1/3 of your K402MEH thread on Klipsch. I'm going to try a smaller one for my new home system and maybe a PA for my Church so I'll try your approach there. I have some form factor limitations that make the K402 size not work. I still have all the 2" drivers except the 4001s because someone someone offered me an obscene amount of money for them in maybe '05. Not popping $US650ea for the Radian Bes for the 2446s though, I can buy 4592s for that.
Hello,
Interesting thread.
After quite a lot of tests, I ended up 3 way with a waveguided 6.5" and a 1" compression driver on faital horn. Crossover at 1600hz. I use cheap BMS 4524 and it is already very good to me. I ended up with quite directional philosophy because the room gives a lot of reverb. I am also starting to work on the room.
Anyway I wonder which driver you would choose for this use. I bought the faital horns because faital recommended these to me paired with hf108 for home use.
Interesting thread.
After quite a lot of tests, I ended up 3 way with a waveguided 6.5" and a 1" compression driver on faital horn. Crossover at 1600hz. I use cheap BMS 4524 and it is already very good to me. I ended up with quite directional philosophy because the room gives a lot of reverb. I am also starting to work on the room.
Anyway I wonder which driver you would choose for this use. I bought the faital horns because faital recommended these to me paired with hf108 for home use.
I guess this is the next step after this
https://audioxpress.com/article/tes...rom-sb-audience-a-new-brand-from-sb-acoustics
https://audioxpress.com/article/tes...rom-sb-audience-a-new-brand-from-sb-acoustics