Next level Active DSP Crossover

...no other type of loudspeaker technology can.
Dunno that its quite so absolute. Years ago I regularly put together systems using JBL and Altec horns, including large bass bins. They speakers all sounded different, different versions of diaphragm surround design sounded different, how the systems were configured made them sound different, if the bass bins were fitted with wings they sounded different, etc. More recently I had some JBL M2 speakers here on long term loan, so I know what those sound like. The other speakers here are relatively new large panel Sound Lab ESL, with stereo REL subs. A fellow forum member visited here a couple times and described the system here verses other systems he is familiar with: https://www.diyaudio.com/community/threads/dac-recommendation.376015/post-7560022

Here is a brief quote:
...The first thing I noticed was that the electrostats could play loud with little to no distortion. Much more so than my horn speakers, which can go plenty loud. I had heard electrostats before and was always disappointed that I couldn't turn it up without hitting the limit. On these, the limit is high enough that you can enjoy realistic sound levels and stunningly low distortion...
This low distortion from the speakers resulted in a lot of transparency and ability to hear down into the preceding components.
 
I think last part of the quote I just finished up editing makes clear that it does include hearing down into the electronics at pretty much all signal levels.

Anyway, in case that wasn't clear yes, its possible to hear very low level detail although of course it takes more power than horns would for the same SPL.
 
Yeah, i dunno either....about fully horn-loaded speakers needing different/better electronics, ones that can run very cleanly at low signal levels.
Made the point earlier, but it seems to be worth repeating......ime/imo, proper gain structure rules.

Keep line level signal reasonably hot all the way to the amp. Either have a good way of attenuating the post DAC analog signal into the amp, if such attenuation capability isn't built into the amp, or choose an amp with lower gain.
Theoretically, best gain structure is when everything in the chain clips at the same time.

Also think gain structure is an issue for not just horn-loaded speakers, but any speaker using a compression driver or other high-efficiency device.
So I'd say electronics' concerns are about the same for many direct radiator speakers, just like for fully horn -loaded.
 
If I'm reading the posts correctly, you are simply using the PEQ and/or shelving filters to limit out of band energy much like a "crossover" would, but just by attenuating the frequencies of interest in the stop band? I understand the natural bandwidth and roll-off comments.

Also, with these settings in mind, listening at >110dB may not be a good idea (driver wise for FM/AM).
I may be remembering wrong, but wouldn’t an appropriately large capacitor protect against out of band bass frequencies to the mid or tweeter without introducing a phase shift in the pass-band ?

After reading through a bunch of this stuff I’m definitely going to give it a shot on my LS soon.
 
I may be remembering wrong, but wouldn’t an appropriately large capacitor protect against out of band bass frequencies to the mid or tweeter without introducing a phase shift in the pass-band ?

I know that the reactance on the amplifier output terminals will increase significantly if you do that. I'm not a big fan of that approach, to be honest. If you've got the kind of amplifier that can often fail with DC on the output terminals (i.e., tube-type amplifiers), then the capacitor-blocking output is okay.

One thing that I think is difficult to convey in a way that will stick--is that the compression drivers that I use are many times more robust than the typical K-77 tweeter of the PWK era (i.e., the only tweeter that met PWK's audible performance requirements of that day/age and that had high enough sensitivity). I think that experience is very much present in a lot of Klipsch owner's minds that have blown K-77 diaphragms in the past.

I can relate a story that might help to alleviate some of that simmering fear: the first guy that I helped to dial-in his loudspeaker remotely (a K-402-MEH in the same setup as his 1st-Gen Jubilees) also chose to tri-amp one or more of his Belles, and he kept popping K-77 diaphragms using a Xilica XP series. I proposed to set a fast limiter on the Xilica to a good setting for him to prevent that from occurring. He said that all the problems with blown K-77 diaphragms completely ceased after that, and he could not detect any difference in the audible output of the Belle.

A success story, if I might say so myself. Consider this approach if you have any fears of blowing HF driver diaphragms. There is no effect on the output audio fidelity if the limiter is set properly. If you need to set a limiter for your system, let me know in a PM (i.e., a "conversation" on this forum).

As far as the out-of-band bass frequencies affecting HF compression driver output at crazy-high drive levels (i.e., above 110 dB continuous), first I'd say that you need to factor in the response of your compression driver at bass frequencies before worrying. Second, setting another attenuating PEQ or shelf filter in the compression driver's stop band to drive the relative signal level below -40 dB is pretty easy to do. If you've run out of PEQs in the output channels of your compression driver, simply copy one of the compression driver output channel PEQs that isn't close to the crossover point(s) into the input channel PEQs to free up an extra PEQ in the output channel, then add a big attenuating PEQ or shelf filter deep into the stop band to assuage your fears.

Chris
 
Last edited:
  • Like
Reactions: 1 user
Unfortunately I'm currently in the same boat as Gigantic as I'm using the same Dayton DSP-408 to run things at the moment.
It was a good deal at less than $100 for an open box but it obviously has a number of limitations.
The PEQ and high/low shelf settings only have a range of plus or minus 12dB.
I might try setting the shelf filters where they need to be and then for the mid and tweeter I could try the regular high-pass crossover filter set maybe an octave lower and see how it measures and what the phase response looks like.

I'm currently running some Selenium D220TI for my tweeters so they are much more robust than the old K-77's which I think I sold off over 15 years ago.

-Josh
 
  • Like
Reactions: 1 user
The PEQ and high/low shelf settings only have a range of plus or minus 12dB.
That's a significant limitation for my use. I ran into that with the ElectroVoice Dx38. I would factor that in if anyone else is thinking about using the same model number. It's about as limiting as only having 6 PEQs per output channel (like the Dx38). I regularly use big attenuating PEQs in my dial-ins via email when I'm helping others--notably using either Xilica XP/XD series or miniDSP (2x4 HD or Flex).

In those cases, a big cap in series with the compression driver may be your only approach if trying out zeroth order crossover filters (i.e., PEQs/shelves only to limit phase growth), and you expect to use your setup at extremely high output levels (i.e., above the levels that I would use my setup).

I'm currently running some Selenium D220TI for my tweeters so they are much more robust than the old K-77's which I think I sold off over 15 years ago.
Yes, I'd always encourage others to look at the natural response of the drivers themselves before panicking about out-of-band modulation distortion.

One thing that I can say is this: you can hear modulation distortion, and if you find you're experiencing this when you really crank your volume control, you know exactly where to go and what to do. It doesn't hurt the driver, per se, but only sounds bad (if excessive).

Another way to look at it: if you're hearing modulation distortion while using while using zeroth order crossover filters in your setup, then this is an indication that, in advancing years, you may want to turn it down a bit...today... 😂 Sort of an automatic "turn it down" indicator. 😁

I do know that at least one guy that I helped in the past started talking about how it "sounded bad" when he cranked it up after I'd taken some effort to get his phase response flattened. When I made some inquiries as to how loud he was playing his basement system, I found out that he was running out of amplifier power (dual McIntoshes)- using Jubilees. My guess is that he was listening at near 120 dB (1m) in his basement. I believe that he could probably file for disability now for hearing loss. The human hearing system simply can't take that level of SPL without permanent damage occurring almost instantaneously--permanently. I know my own father had permanent hearing loss, but in his case, it was jet engine noise (he was the lead stability and control engineer on the XF8U3 project in the late 1950s--"the best fighter that the Navy ever cancelled"...in lieu of the F4 Phantom II. I think that phrase was correct...the following from Wikipedia):

...NASA pilots flying at NAS Patuxent River routinely intercepted and defeated U.S. Navy Phantom IIs in mock dogfights, until complaints from the Navy put an end to the harassment.[11] All of the Crusader IIIs were later scrapped.

Chris ;)
 
Last edited:
  • Like
Reactions: 1 users
The correction file phase correction can probably be done to any microphone that has a correction file.

Thank you the Motu + measure microphone is the solution I have searched for. Using acoustic timing reference from a speaker is a hassle.
It can probably be done with several different USB DACs but I have not had the time to also solve this challenge.

I have been thinking of buying a Motu but I have some questions.
1) can you direct an analogue input stereo signal from Motu into a USB computer with a convolver software like Hang Loose Convolver. And then back 8 channels to Motu through the USB interface. This way I would not need to have any other line connection input to the computer. Motu would be the preamp with input and output connections. Streaming of course from the computer so that there would not be a back-and-forth loop.
2) How easy is it to make a decent interface with Motu? Switching between signal sources and using the volume control.
This would preferably be done with an iPad app.
3) If you are running the CamillaDSP on the computer can you also switch input channels on the Motu with the Camilla interface? Or are you running two separate interface apps.
 
I assume then that you reject what I said about the FusionAmp FA122 experience, Mark (Mark100)? (I hope I don't have to repeat that experience again.)

Chris

Nope, how can I possibly reject the experience you had. I appreciate you reporting it.

However, I do reject that your experience applies universally to others, even in similar gear/setup situations.

ALL of our experiences, mine certainly, and yours certainly too...are just that....what we we each individually experienced.
They all need a YMMV....(your mileage may vary disclaimer)....
 
spalmgre, your questions on the Motu UltraLite Mk5:
1. I had not heard of the Hang Loose Convolver before your post, maybe the vendor will be able to tell you if their software works with USB Audio Interfaces. With REW connected to the Motu M4 for measurement, the M4 shows up as both an input device for the microphone and an output device for the speaker.
2. Look through this post
https://audiosciencereview.com/forum/index.php?threads/rpi4-camilladsp-tutorial.29656/post-1038294
It is part of Michael's tutorial, I followed it and now use a FLIRC and an old Squeezebox remote to switch sources and adjust volume. I linked to a picture in an earlier post. I also use the volume display (but in a cheap black plastic 1 unit half width rack case)
3. Switching channels - when I am testing and only want one speaker (in a tri-amped speaker) then I use the CamillaDSP GUI, click on the Mixer tab and mute the channels that are not being measured.
 
That's a significant limitation for my use. I ran into that with the ElectroVoice Dx38. I would factor that in if anyone else is thinking about using the same model number. It's about as limiting as only having 6 PEQs per output channel (like the Dx38). I regularly use big attenuating PEQs in my dial-ins via email when I'm helping others--notably using either Xilica XP/XD series or miniDSP (2x4 HD or Flex).

In those cases, a big cap in series with the compression driver may be your only approach if trying out zeroth order crossover filters (i.e., PEQs/shelves only to limit phase growth), and you expect to use your setup at extremely high output levels (i.e., above the levels that I would use my setup).


Yes, I'd always encourage others to look at the natural response of the drivers themselves before panicking about out-of-band modulation distortion.

One thing that I can say is this: you can hear modulation distortion, and if you find you're experiencing this when you really crank your volume control, you know exactly where to go and what to do. It doesn't hurt the driver, per se, but only sounds bad (if excessive).

Another way to look at it: if you're hearing modulation distortion while using while using zeroth order crossover filters in your setup, then this is an indication that, in advancing years, you may want to turn it down a bit...today... 😂 Sort of an automatic "turn it down" indicator. 😁

I do know that at least one guy that I helped in the past started talking about how it "sounded bad" when he cranked it up after I'd taken some effort to get his phase response flattened. When I made some inquiries as to how loud he was playing his basement system, I found out that he was running out of amplifier power (dual McIntoshes)- using Jubilees. My guess is that he was listening at near 120 dB (1m) in his basement. I believe that he could probably file for disability now for hearing loss. The human hearing system simply can't take that level of SPL without permanent damage occurring almost instantaneously--permanently. I know my own father had permanent hearing loss, but in his case, it was jet engine noise (he was the lead stability and control engineer on the XF8U3 project in the late 1950s--"the best fighter that the Navy ever cancelled"...in lieu of the F4 Phantom II. I think that phrase was correct...the following from Wikipedia):



Chris ;)
I made some changes yesterday.
My ears tell me I’m on the right track.
I need to move some some stuff out of my living room to get some better measurements to see what’s actually going on.
 
  • Like
Reactions: 1 users
See https://www.diyaudio.com/community/...lee-acoustic-performance.396074/#post-7570882 for more discussion on the now-available miniDSP Flex performance (albeit with balanced connections and Dirac Live installed).

We found out that the one thing about the Flex that you can't really use with extremely high efficiency loudspeakers (i.e., fully horn loaded, like yours shown above) is the analog inputs--they sound horrible on a Heritage Jubilee. But note that the S/PDIF digital inputs (including TOSLINK optical) that also come with this unit haven't been tested yet. The person that I helped recently to replace the fixed Heritage Jubilee crossover (which has its own issues) didn't have a digital out on his preamp/processor, so he had to use balanced analog inputs. I really couldn't listen to the Flex's performance--it was that bad.

I think you're asking for unobtaium with your constraints. The better DSP crossovers define themselves by their low noise and hi-fi outputs, and they simply cost more. Think of this like acquiring tube amplifiers: you might get away with low-cost tube SETs, etc., but their limited output transformer size will come back to bite you.

I'd recommend thinking about this using a different paradigm. If you have extremely high efficiency/high performance horn-loaded loudspeakers, I've found that getting the on-axis SPL response flat to within ±2 dB (using psychoacoustic smoothing) across the board, and also getting the time alignment of the drivers to achieve flat phase to ±90 degrees (again using psychoacoustic smoothing) is a key enabler to a threshold of hi-fi performance that others have only heard using high-dollar electrostatics and certain very high dollar full-range drivers, in addition to all the other positives that fully horn loading brings to the table. Instead of thinking about inexpensive DSP crossovers, think first about getting a DSP crossover that has the required performance, then think about amplifiers and preamps. In other words, stop trying to use cheap DSP crossovers. If you can find one that has hi-fi sound and extremely low noise that also is lower cost, that's great, but I haven't found one yet (and I've tried a few over the years).

In reality, DSP crossover performance for use with fully horn-loaded loudspeakers (including MEHs) is a subject that the low-efficiency direct radiator loudspeaker crowd simply do not understand. The same thing goes for amplifiers having extremely good "first watt" performance--something that I found that many, many amplifiers used by the direct radiator crowd don't have... (Just ask me about my Hypex FusionAmp experience, for example.)

The problem now is the Xilica XP/XD series is apparently no longer being offered from Xilica. You have to increase your spending to Xilica Solaro levels (i.e., around $2K--$2.5K), which is on par with higher quality used amplifiers from Nelson Pass' First Watt, etc. You might get good enough performance using a lesser DSP crossover, but I think you're going to spend a lot of money along the way (money lost even if you resell the equipment) if you keep trying to find that cheap magic bullet of a DSP crossover. They apparently don't really exist yet for fully horn-loaded loudspeaker use. If I find one, I'll let you know, but I haven't found one recently.

Chris
I recently picked up a used Yamaha 2060 for just under $500...same model you suggested.
 
The Yamaha SP2060 runs at 96 kHz. It also has a higher noise floor level (using balanced analog inputs) than the Xilica XP series. You need to take care in setting up your gain structure. I haven't tested the digital input for noise, however. Other than the noise floor, it's a very hi-fi crossover.

Chris