Greets!
Yeah, nothing quite like a big organ in a (Ga.) marble, granite construction Cathedral.
One of my old 'stomping grounds'! 👍 😎 Definitely many decades ago ('50-'60s)!
Yeah, nothing quite like a big organ in a (Ga.) marble, granite construction Cathedral.
One of my old 'stomping grounds'! 👍 😎 Definitely many decades ago ('50-'60s)!
That’s the funny thing. He built several with small mids, including the one documented in this thread:
https://www.diyaudio.com/community/threads/synergy-horns-dayton-and-prv.244890/
But he has found nirvana with his latest build, featuring the dcx464 here.
You found nirvana by adding mids after building several with coaxial compression drivers. 😆
Yes, I was thinking it might be something like this. On the other hand, it’s tricky to attribute less good sound to distortion. And what really got me scratching my head is the dcx464 is a super powerful coaxial compression driver, capable of playing at concert levels for large groups of people outside. You‘re reporting it sounds best crossed all the way up to the 900hz region in a home environment. The pm90/pm60, conservatively crosses over 1 octave above the 300hz min—up in the 600hz area. You’ve built these as well, and many have reported that these things are super loud and clean.
This got me thinking that it might be the horn. From theory, we see that conical horns load bass poorly. So perhaps your high crossover corresponds to where the coax is starting to struggle at below 1kHz, as Keele’s chart shows.
So the perhaps the coaxial mid is struggling too much in a conical horn, leading to more audible distortion even in a home environment.
On the other, other hand…Scott’s latest design eschews the midwoofers for the compression coax. He crosses all the way down to 320hz, and reports it sounds better! He said it’s not suitable for outdoor pa levels, but that it will still blow your head off in a normal home sized room.
One thing I noticed about Scott’s design is that he was careful to keep the spacing within 1/4 WL. The 10” woofer ports are easily within 1/4 WL of the throat. But he’s also careful to keep the woofer ports 1/4 WL from each other. As you can see, he crams the 10” woofers as close to the dcx464 as possible, and the xo is 320hz. The diagonal center I’d of the ports hits 1/4 WL this way. I’ve seen popular 2 way designs that cross the woofers over much higher with this same spacing. So maybe he 1/4 rule for the diagonal distance is very important and audible.
You’ve noted in this thread as well that tightening up to 1/4 WL diagonal makes a difference.
So perhaps you both changed different variables and ended up with improved syn’s.
Mark, did your previous coaxial compression driver syn’s have the diagonal port spacing within 1/4 WL for each other?
It could be, you’re both right and want to both add small mids for better loading in the 100-900hz range to lower distortion, and keep the spacing super tight at 1/4 WL, even diagonally to get the perfect point source reproduction.
r
Thanks for that. I hadn't realized or forgot, that SpeakerScott is Scott Hinson.
I started out on the path of taking a CD staight to a large low-mid, after being able to discuss the idea with Tom D in person at a trade show in 2019.
I found both the BMS and B&C coax make it pretty easy to cross to 10"s or 12"s anywhere between 500-700Hz, port placements depending.
As you may have noticed in my series of syn 7 and 8 posts, I hugely advocated using a CD that could reach down so, for the ease in making a unity/synergy...(and not to mention getting the killer good sound) 🙂
Adding small-mids was really just an experiment to see if I was leaving anything on the table......
And yes to your question about 1/4WL spacing....on all builds all along, i've tried to keep all the various interactions within 1/4WL.
That's also been part of my steep xover strategy...to keep the entire critical frequency summation range within 1/4. Not so easy with shallow slopes.....
So who knows why Scott and i have different ending paths, that we currently each like best....
I certainly have no clue, and no reason to doubt anything he says.
Now that I've finally sorted out the 18"s in this syn11, I'm going to play with not using all the driver drivers sections.
I can skip either the 4" mids or the 12" lows super easy.
Heck, the CD can reach straight to the sealed 18"s, and skip both the mids and lows!
The other day when quickly making a new house curve tuning, I took the time to define what i think is the maximum usable bandwidth of each driver section.
By that, I mean bandwidths that took only moderate EQs, and the end points needing little to no boost, to maintain rolloff targets.
Below are the traces and what I think are the usable f-6 ranges.
18" subs purple. 35-425Hz
12" lows blue. 65-850Hz
4" mids green. 200-1150Hz
dcx464 lower section red 390-5400Hz
dcx464 higher section orange > 3200Hz
Lots of room to try things...lots of room to get good and confused Lol
Thanks for that. I hadn't realized or forgot, that SpeakerScott is Scott Hinson.
yes, and he clearly states the dcx464 build is his best yet.
It is funny, the reverse paths we took, isn't it?
I started out on the path of taking a CD staight to a large low-mid, after being able to discuss the idea with Tom D in person at a trade show in 2019.
I found both the BMS and B&C coax make it pretty easy to cross to 10"s or 12"s anywhere between 500-700Hz, port placements depending.
As you may have noticed in my series of syn 7 and 8 posts, I hugely advocated using a CD that could reach down so, for the ease in making a unity/synergy...(and not to mention getting the killer good sound) 🙂
Adding small-mids was really just an experiment to see if I was leaving anything on the table......
And yes to your question about 1/4WL spacing....on all builds all along, i've tried to keep all the various interactions within 1/4WL.
That's also been part of my steep xover strategy...to keep the entire critical frequency summation range within 1/4. Not so easy with shallow slopes.....
So who knows why Scott and i have different ending paths, that we currently each like best....
I certainly have no clue, and no reason to doubt anything he says.
Now that I've finally sorted out the 18"s in this syn11, I'm going to play with not using all the driver drivers sections.
I can skip either the 4" mids or the 12" lows super easy.
Heck, the CD can reach straight to the sealed 18"s, and skip both the mids and lows!
The other day when quickly making a new house curve tuning, I took the time to define what i think is the maximum usable bandwidth of each driver section.
By that, I mean bandwidths that took only moderate EQs, and the end points needing little to no boost, to maintain rolloff targets.
Below are the traces and what I think are the usable f-6 ranges.
18" subs purple. 35-425Hz
12" lows blue. 65-850Hz
4" mids green. 200-1150Hz
dcx464 lower section red 390-5400Hz
dcx464 higher section orange > 3200Hz
Lots of room to try things...lots of room to get good and confused Lol
Im here for all the fun. Owning a pair of SH96s I'm sure there are ways to improved even slightly on the design.
The world may never know. One more thing I didn’t quite understand from Scott’s paper. In page 26 He shows the series parallel L-pad he uses on each the mid and hi portion on the coax compression driver. On pp. 31-33, he discusses that one reason this is needed in his active 3 way design is to fix the problem of the measured output being the same as the output in use, since the 2 diaphragms are acoustically aware of each other.
Is this something you considered for your coaxial compression designs, and do you think it would make enough of a difference to close the gap between your designs with mid woofers?
The midwoofer version does have some good benefits. The woofer mids are more efficient than the mids from a coaxial compression driver through the magic the port being further toward the mouth. Also, I might make a corner horn pair, and using midwoofers allows for getting closer to a corner with a physically smaller tweeter & shallow mids.
Super useful. I might try to go from a 1” CD to 4” mids to as large a woofer as possible, given the spacing limits and 1/4WL rule.
Your experiments have been super educational and I really appreciate you writing it all down!
That's what i expected...it takes a lot of cones to outrun a good CD in a syn/unity, huh? 😀
What CD is in yours? I've heard the difference between the 96 and 96HO is mainly a slightly stronger BMS coax CD. To your knowledge, is that correct?
Also wondering if you've heard if the dcx464 made it into any of their boxes. I've kinda quite following along so much....
Eminence 314t.
After speaking with a few people there is really no difference other than a little bit more control over the 15s. I havent heard one using coax drivers yet however. To me the weak point is from the 4s to the 15s. Almost feels like a gap is there because the 4's are asked to work over too many octaves similar to what you mentioned in your analysis..
I've owned a few PM60's so they are comparable however i now bring less boxes to cover the same area..
After speaking with a few people there is really no difference other than a little bit more control over the 15s. I havent heard one using coax drivers yet however. To me the weak point is from the 4s to the 15s. Almost feels like a gap is there because the 4's are asked to work over too many octaves similar to what you mentioned in your analysis..
I've owned a few PM60's so they are comparable however i now bring less boxes to cover the same area..
Last edited:
Yes, the 2 CD diaphragms are definitely aware of each other. Just as the 4" mids are very aware of the CD's lower diaphragm.
I think anytime two driver sections have raw overlapping band widths, they can't help but interact..especially when coupled on a horn.
It always amazes me how much an unshorted CD can effect raw mid measurements.
Which i think is why I think anyone using a passive xover has got to have some circuitry in place before even semi valid raw measurements can be made.
I mean, the unmeasured section has to either be shorted (directly or by amp) or have some kind of shunt across it for dampening.
That said, I'm a fish out of water when it comes to passives.... I just know you damn well have to short driver sections that arement being measured.
It is kinda amazing how sensitive/efficient the four 4" mids of my syn10s and 11 are.
And yep, using the small mids does make it easier to locate separate low ports more optimally (that using one large low-mid)
I like what you suggest.... that may be the easiest, most practical unity build out there.
And thanks for the kind words !
I'm working on showing some measurements I made, trying to get a handle port interferences for each section...
I've owed Steffen too, some of this info for a while....it gets into the whole on-axis vs 10 deg off stuff. etc...
Thanks for that. So my guess is the bottom end of the 4's are the chokepoint.
If that's correct, it's cool you brought that up. One of the things I've been doing with syn11 in trying to decide the bandwidth spans for each section, is go back to the good ole piston excursion calculator, and see how low each section can honestly go with unclipped headroom.
It's kinda crazy on syn11, that I still realize a section can say uh-oh...I really don't want to go that low,haha
Yep, I imagine you had to splay a couple of PM60's per side, for relatively larger gigs. A single SH96 has to be easier..
I think Peter nailed it with the PM90, one box per side for the audience it fits.
@mark100
Awesome work! Great to hear the bass issue is solved.
The information you shared on this forum and Scott Hinson's MEH document is the reason I want to try DCX464.
While doing gating measurement on the HF part I found a high Q dip around 16kHz. In your experience, is it better to EQ it back in the frequency response?
Hope I have the technique and space to build MEH someday.
Scott Hinson describe in his article about the design challenge of mid tap and the coaxial compression solved this problem for the ability to cross the whole midrange with woofer.
I don't know how much weight/size would it add to add another way to the system but this box is also design to demonstrate on burning amp 2022 so a degree of portability is required I guess.
Awesome work! Great to hear the bass issue is solved.
The information you shared on this forum and Scott Hinson's MEH document is the reason I want to try DCX464.
While doing gating measurement on the HF part I found a high Q dip around 16kHz. In your experience, is it better to EQ it back in the frequency response?
Hope I have the technique and space to build MEH someday.
Scott Hinson describe in his article about the design challenge of mid tap and the coaxial compression solved this problem for the ability to cross the whole midrange with woofer.
I don't know how much weight/size would it add to add another way to the system but this box is also design to demonstrate on burning amp 2022 so a degree of portability is required I guess.
Thanks yys310,
I'd say do whatever you want with the dcx464, as far as EQ. Just comes down to what horn it's on, as to what it needs, ime.
My test for valid EQs is how well they hold up across polars. If they don't hold up well, usually best to ditch em.
Hey, this morning I just ordered a second dcx, so i can make another syn11 !
I never really thought there was much of a chance i would want a pair of the oversize syn11s...thinking it would be a big standalone outdoor mono speaker.
But it simply sounds so good indoors too, it's a leap up from anything I've built prior.
So stereo here i come !
I could just use my bms coaxes for a pair of syn11s and save some coin.
But I really don't yet know why the syn11using the dcx sounds so extraordinary.....so I'm making sure I can duplicate it in all details.
I'd say do whatever you want with the dcx464, as far as EQ. Just comes down to what horn it's on, as to what it needs, ime.
My test for valid EQs is how well they hold up across polars. If they don't hold up well, usually best to ditch em.
Hey, this morning I just ordered a second dcx, so i can make another syn11 !
I never really thought there was much of a chance i would want a pair of the oversize syn11s...thinking it would be a big standalone outdoor mono speaker.
But it simply sounds so good indoors too, it's a leap up from anything I've built prior.
So stereo here i come !
I could just use my bms coaxes for a pair of syn11s and save some coin.
But I really don't yet know why the syn11using the dcx sounds so extraordinary.....so I'm making sure I can duplicate it in all details.
Hi folks, I've been trying to put together a series of measurements to show some of the explorations I've been attempting, but somehow i keep getting a mental block....feels like I'm trying to write a full novel uhg.
Just took too damn many measurements with the multiple mic rig, tried too many things.
Anyway, for anyone interested, here's start at a few of the things I've tried to explore....
the 5-way offers so many experiments....
First, I've wanted to get a handle on how much do ports in the horn effect response vs how much does mouth termination effect response.
By 'effect response', I mean still produce anomalies across polars, after having applied DSP correction.
And I've wanted to separate port interference from mouth termination issues, section by section.
So first set of measurements was to block all ports with thin sheet metal and duct tape.
And then relatively close, within 1-2ft of the mouth, measure and correct each section's response.
The idea behind measuring close, is to try to be too close for mouth termination issues to get into the meas.
After making corrections, polars were gathered to see what I think the best hoped for response from each section could be (iow pure horn.)
The CD of course, was the only section that could get a pure horn (with all ports blocked.)
Then open and measure the mids, keeping low and sub ports blocked.
Then open and measure low, and last the sub section.
Here's 0-30 degrees for the CD, mids, and lows in order.
Again with all ports that are for lower frequency sections, blocked.
I think, and pls feel free to critique the idea, that this shows the best that can be hoped for for each section ignoring mouth termination.
As I unblocked port sections for the tests above, I also measure the effect of each unblock stange on the higher drivers sections.
For instance, after unblocking the mids, I tested the CD again with just the mids open. And then did the same after opening lows, then subs.
The CD's response did deteriorate as ports were opened, but really not as bad as I expected given there are 12 ports in the horn and the sub ports are big!
Here's the CD with all ports open.
Here's mid uncovered. followed by low uncovered
I use these as the start for determining how to use section spans / determine xover points.
For instance the above says to me the lows shouldn't be taken any higher than about 325Hz.
Next step was to try to gauge mouth termination issues independent of port interferences.
So I ran the whole port blocking exercise outdoors in the far field, thinking that's about the best way i can come up with such... (any other ideas folks)
Whole 'nuther set of data....enough in this post already...
hope this made sense...
Just took too damn many measurements with the multiple mic rig, tried too many things.
Anyway, for anyone interested, here's start at a few of the things I've tried to explore....
the 5-way offers so many experiments....
First, I've wanted to get a handle on how much do ports in the horn effect response vs how much does mouth termination effect response.
By 'effect response', I mean still produce anomalies across polars, after having applied DSP correction.
And I've wanted to separate port interference from mouth termination issues, section by section.
So first set of measurements was to block all ports with thin sheet metal and duct tape.
And then relatively close, within 1-2ft of the mouth, measure and correct each section's response.
The idea behind measuring close, is to try to be too close for mouth termination issues to get into the meas.
After making corrections, polars were gathered to see what I think the best hoped for response from each section could be (iow pure horn.)
The CD of course, was the only section that could get a pure horn (with all ports blocked.)
Then open and measure the mids, keeping low and sub ports blocked.
Then open and measure low, and last the sub section.
Here's 0-30 degrees for the CD, mids, and lows in order.
Again with all ports that are for lower frequency sections, blocked.
I think, and pls feel free to critique the idea, that this shows the best that can be hoped for for each section ignoring mouth termination.
As I unblocked port sections for the tests above, I also measure the effect of each unblock stange on the higher drivers sections.
For instance, after unblocking the mids, I tested the CD again with just the mids open. And then did the same after opening lows, then subs.
The CD's response did deteriorate as ports were opened, but really not as bad as I expected given there are 12 ports in the horn and the sub ports are big!
Here's the CD with all ports open.
Here's mid uncovered. followed by low uncovered
I use these as the start for determining how to use section spans / determine xover points.
For instance the above says to me the lows shouldn't be taken any higher than about 325Hz.
Next step was to try to gauge mouth termination issues independent of port interferences.
So I ran the whole port blocking exercise outdoors in the far field, thinking that's about the best way i can come up with such... (any other ideas folks)
Whole 'nuther set of data....enough in this post already...
hope this made sense...
That is simply awesome! The last picture explains the dip I measured on my test horn mids when I added the 18" bass section with a single central port - the dip is lower when off axis (i.e. away from the port).
Hi Mark
Thank you for taking the time to write all this. Really a lot to comprehend! I need to read it a couple of times.
So it seems, that the midports affect the lows response too!? And I guess the same is true for the sub.
I'm not sure I fully understand your procedure? But did you keep ports open as you proseeded down in frequency, i.e. mids lows subs? Or did you close the ports again when proceeding to the next lower set of drivers?
Well, have to reread
Happy experimenting.
Steffen
Thank you for taking the time to write all this. Really a lot to comprehend! I need to read it a couple of times.
So it seems, that the midports affect the lows response too!? And I guess the same is true for the sub.
I'm not sure I fully understand your procedure? But did you keep ports open as you proseeded down in frequency, i.e. mids lows subs? Or did you close the ports again when proceeding to the next lower set of drivers?
Well, have to reread
Happy experimenting.
Steffen
Hi again
The really fantastic thing is, that it still is possible to create an extraordinary loudspeaker, in spite of all the port interferences!
It´s really going to be interesting to see what trade-off is the best? Few ways, few ports and deployment of maximum usable bandwidth or many ways, many ports but more narrow bandwidths!? Hmm.
I recently visited Uli Brueggemann the creator of Acourate, a software to measure and create FIR filters for speakers that you can use in his AcurateConvolver software. Pretty amazing two hours! Totally time-coherent. lots of headroom! I liked it very much and my wife too! He has a four way horn-setup, actually three separate horns and some woofers in big reflex-boxes.
Anyway, he started the conversation saying, that ideally he thinks a horn should not do more than two octaves! Some of his costumers have five and six way-systems.
Now, you deploy one horn, but it seems that each driver should not do more than two octaves.
Steffen
PS. Uli Brueggemann has developed a new way to create trinaural sound, i.e. left/center/right from a stereo-signal. He and his friends started out with Gerson/Pekonen, but took it further out, for instance taking into account the distance between the ears, to correct the delay on the center-speaker and much more. I can´t really explain it here, don´t get it all, I read about their adventure in a German forum. Maybe Mitch Barnett knows something about it?
The really fantastic thing is, that it still is possible to create an extraordinary loudspeaker, in spite of all the port interferences!
It´s really going to be interesting to see what trade-off is the best? Few ways, few ports and deployment of maximum usable bandwidth or many ways, many ports but more narrow bandwidths!? Hmm.
I recently visited Uli Brueggemann the creator of Acourate, a software to measure and create FIR filters for speakers that you can use in his AcurateConvolver software. Pretty amazing two hours! Totally time-coherent. lots of headroom! I liked it very much and my wife too! He has a four way horn-setup, actually three separate horns and some woofers in big reflex-boxes.
Anyway, he started the conversation saying, that ideally he thinks a horn should not do more than two octaves! Some of his costumers have five and six way-systems.
Now, you deploy one horn, but it seems that each driver should not do more than two octaves.
Steffen
PS. Uli Brueggemann has developed a new way to create trinaural sound, i.e. left/center/right from a stereo-signal. He and his friends started out with Gerson/Pekonen, but took it further out, for instance taking into account the distance between the ears, to correct the delay on the center-speaker and much more. I can´t really explain it here, don´t get it all, I read about their adventure in a German forum. Maybe Mitch Barnett knows something about it?
Ah good , glad the stuff helps some, and thanks guys.
Yes, I did keep the ports open for higher frequency sections, as I proceeded down in frequency.
Ideally, it would be nice to have only one port section at a time open for each section under measurement, to dig deeper into interactions.
But I've been swamping myself with data already.
Although my strong bet is using all the ways within bandwidths that each have the least polar variations after their correction curves.
That's been the motive behind trying to dissect what's causes the polar variations, port interferences vs mouth termination vs 1/4 WL throat bounceback etc.
That is so cool, you got to visit Uli B ! I've come very close to buying Acourate so many times, and have followed the Acourate forum posts for quite a while. I really like it's envelope function for choosing what measurement to correct, and have big interest in the LCR talked about in the forum.
And a very big yes to the idea of each driver not covering much more than two octaves.
Apart from pulling as much of the entire spectrum as possible into one "point-source" horn, ala a unity/synergy........
I think the increased number of ways is the only thing left to explain why syn11 sounds so dang good to me.
It's funny....syn11 polar measurements aren't any better than syn10, or 8 or 7 for that matter.
In fact syn7, where I did all the work to terminate the mouth with curved secondary flares had the best polars.
Syn 8 was the same thing as 7, but with straight flares for ease of construction and allow comparison to curved......it sounded the same as 7.
Syn 9 was a series of one-off prototypes that added the small mids, and simply sounded better enough to inspire syn10.
With syn10, I skipped secondary flares entirely, given how the syn9 protos which didn't have secondaries, sounded better than any previous builds which all have secondaries.
And now here's syn11, with yet another section, sounding even better still....and without polar improving secondaries.
So I dunno....one one hand I'm wondering how much do super polars really matter ...but on the other hand, all the measurement and processing work i"m doing is to try to continue to make the polars as good as possible.
Oh, one last thing I've learned that I enjoy with syn11, is the use of a separate sub despite the fact syn11 has two 18"s.
I low pass the separate closely located sub steeply at 80Hz, which turns the whole project into a 6-way Lol
It really helps dialing in bass on a track by track basis, to have separate sliders for the two sub sections.
The only thing I don't like about these 6-ways is even for me, it's gone over the top in terms of size and weight and room domination....kinda sucks.
I'm trying to figure out how to turn separate subs into a cart for syn11, but worried about the total weight.
But hey, what price glorious sound??? LOL
Yes, I did keep the ports open for higher frequency sections, as I proceeded down in frequency.
Ideally, it would be nice to have only one port section at a time open for each section under measurement, to dig deeper into interactions.
But I've been swamping myself with data already.
Yes, it will be interesting to see what trade-off is best.
Although my strong bet is using all the ways within bandwidths that each have the least polar variations after their correction curves.
That's been the motive behind trying to dissect what's causes the polar variations, port interferences vs mouth termination vs 1/4 WL throat bounceback etc.
That is so cool, you got to visit Uli B ! I've come very close to buying Acourate so many times, and have followed the Acourate forum posts for quite a while. I really like it's envelope function for choosing what measurement to correct, and have big interest in the LCR talked about in the forum.
And a very big yes to the idea of each driver not covering much more than two octaves.
Apart from pulling as much of the entire spectrum as possible into one "point-source" horn, ala a unity/synergy........
I think the increased number of ways is the only thing left to explain why syn11 sounds so dang good to me.
It's funny....syn11 polar measurements aren't any better than syn10, or 8 or 7 for that matter.
In fact syn7, where I did all the work to terminate the mouth with curved secondary flares had the best polars.
Syn 8 was the same thing as 7, but with straight flares for ease of construction and allow comparison to curved......it sounded the same as 7.
Syn 9 was a series of one-off prototypes that added the small mids, and simply sounded better enough to inspire syn10.
With syn10, I skipped secondary flares entirely, given how the syn9 protos which didn't have secondaries, sounded better than any previous builds which all have secondaries.
And now here's syn11, with yet another section, sounding even better still....and without polar improving secondaries.
So I dunno....one one hand I'm wondering how much do super polars really matter ...but on the other hand, all the measurement and processing work i"m doing is to try to continue to make the polars as good as possible.
Oh, one last thing I've learned that I enjoy with syn11, is the use of a separate sub despite the fact syn11 has two 18"s.
I low pass the separate closely located sub steeply at 80Hz, which turns the whole project into a 6-way Lol
It really helps dialing in bass on a track by track basis, to have separate sliders for the two sub sections.
The only thing I don't like about these 6-ways is even for me, it's gone over the top in terms of size and weight and room domination....kinda sucks.
I'm trying to figure out how to turn separate subs into a cart for syn11, but worried about the total weight.
But hey, what price glorious sound??? LOL