Three years to get around to my own install.

[geolemon]

Listening settings - Phone DAC EQ set to "normal" (flat, no DSP).
I was running tweeters for this first set (Pro Tech time lense)


Brothers in arms (Dire Straits) - a few notes located at the left array only.

Sultans of swing (Dire Straits) - some notes and stereo effects were notably at the speakers.

I Will Remember (Toto) - mostly great. Some low voice introduced mild stereo gap

Avratz (Infected Mushroom) - good detail reproduction. Stereo effects heavy song - disabled the midbass to evaluate. Low midrange slight separation, improves as you go up.

I just died in your arms (Cutting Crew) - some vocals and some guitar locateable, bad recording (can hear in details)

Get Lucky (Daft Punk) - largely really good, good detail. Imaging good, other than "We" in hook is slightly locateable.

Layla (Live) (Eric Clapton) - emotionally good. Hi-hat locateable on left but that's because the tweeter was behind the array. Sometimes Clapton seemed left or right, but not from moving.

Something just like this (Chainsmokers/Coldplay) - also emotionally good and great imaging - possibly from the heavy effects in the music - but it worked.

Adventure of a lifetime (Coldplay) Opening high guitar and entire singer range are dead nuts on. Very complex music like this is easier to image.

Madness (Muse) - guitar @2:40 is a bit to the left but no gap. Same with vocals after solo. Higher "I need to loooooove" emotionally good and centers up - don't like that transition but is showing frequency based image shift (BMR? too low I suspect)

Keith Don't Go (Live) (Nils Lofgren) - guitar great. Higher vocals so really good. Very little to complain about - sounds "Live"!
Ether Sunday - lower piano slight gap, as with others upper mid is great. Vocals slight left but no gap.

Polaris (deadmau5) - the electronic stereo effects danced back and forth - no separation! As each 16-beat element came in, it was seamless. Energy lull/keyboards at 4:45 were all well centered, possibly helped by heavy effects (higher frequency components). Full beat resumption at 6:48 - I couldn't switch tracks. That good.

Sit next to me (Foster The People) - lots of upper vocals, predictably good imaging - but I could hear the lower mid gap if I listened for it - but the bass guitar is all midbass (and these anarchys do that with detail)

The Sound Of Silence (Disturbed) - signature low vocals - definitely locateable but these speakers still conveyed enough detail to sound very "live" and realistic (I'm going to have to sign up for Spotify Hi Def)

Still of the night (Whitesnake) - there's a strong presence to EQ out - makes hair metal feel thin, though admittedly I'm running my midbass intentionally low with this flat EQ. Using EQ the "power" is there.

I followed fires (Matthew and the Atlas) - basically all my same observations from Sit Next To Me, but better simply because there's nothing in that lower midrange.

At this point I ran them full range-
** Aimed in, no tweeters (500hz and up, removed JL comp set Xover from equation). **

Isn't She Lovely (Livingston Taylor) - the whistle was spot on great. Vocals and tonality impressive. Really missing no "sparkle" in this song, but still good.

Bust A Move (Infected Mushroom) - here I thought a little 'sparkle' was missing, but only barely. And impressive as I turn the volume up - until I detect enclosure resonance. Eight of them can get decently loud.

Polaris (deadmau5) - note I'm stressing these with electronica to try to push these things... And push I did I ran this whole song with about everything my Yamaha receiver had and no stress at all. I found an air leak in my cabinets I have the Anarchies in, yet had no sense of stress from these widebands. Wow.

I actually hadn't pushed the Anarchies much prior to this track - but they brought impact, even with an efficiency and power deficit that would have me bringing the midbass up and widebands down, if I we're EQing rather than testing/trialing.


Sent from my LM-G820 using Tapatalk

[SiW80]

I found the Tectonics good but missed some body and sounded thin on the lower end.

They were firing up off the windshield and passed 1.25kHz up.

Swapped out directly for the Alpine SPX clones from eBay and found the missing body and sparkle.

If you have the funds suggest you try the SPX drivers


Sent from my iPhone using Tapatalk

[geolemon]

I do appreciate that- I've got some other traditional dynamic driver here, I might still check those out. I can't say I've ever so much as heard that driver, but I'm sure worth looking at.

Bear in mind there's some nutty professor concepts here to really focus on imaging...
The array aspect here was to increase the stage width, to simulate a single driver of further distance - and also to create multiple sound path in the car that creates four direct-to-ear pathlengths (again why the compass was used to locate them, although I am sure I altered that slightly by angling the whole panel), while each of their reflected paths will be different. That does somewhat re-enforce the primary signal, and helps reduce and diffuse the secondary signals - plus seems to increase the stage width, I believe for purely psychoacoustic reasons.
And of course, the whole "BMG" factor, for that omnidirectional effect.

I have some BG Neo3S panels that I want to try next - I have two mounted firing up for the next set of listening, but I'm already leaning towards trying four of them firing in (probably recycling this same "enclosure" panel).
I did make a 3D printed set of domed reflectors, to emulate those Mirage speaker's approach to "omni-directionals", which was what started this. No one to my knowledge has ever tried them with magnetic planars... so why not? Maybe there's reasons.

And also, I have some Dynaudio 3" domes from a Volvo that I'll be trying out - that's actually what I modeled the reflectors off of, to truly try the Mirage-style approach.

Of course - those reflectors can be used with regular dynamic drivers also - that's what Mirage did for the mids, after all. Maybe worth a shot - I have some Aura whispers I was going to try it with, but now you have me curious in general about SPX clones...
So, not legit Alpine? They aren't Tang Band by chance, are they?

[SiW80]

Hope this link works to the other place for more info: Hot eBay deal - Alpine SPX PRO tweeters https://www.diymobileaudio.com/threa...eeters.286809/

They look genuine or if not a very good copy. Anyway for $30 or so they sound great


Sent from my iPhone using Tapatalk

[geolemon]

On to the next...
Bohlender Graebner Neo3S magnetic planar drivers, mounted in cabinets, firing up flush from the dash top.
These were just some "I already own them" and have some unique things people do to them with yarn and things to change their dispersion... and I was already 3D printing some Mirage-like onmidirectional reflectors - so why not try the BG's and omni reflectors? Never seen that done, so let's try...

I'll post pics of them when I post my listening for these.
Spoiler alert - Ultimately I did less listening to these than the 2" BMRs because I just wasn't meeting my imaging goals with this location, reflectors or not.
Tonally though, I was pretty impressed. I really didn't feel like I needed to augment them with tweeters. I'm still committed to running tweeters - but if I were pairing them with these, I'd probably really only use them as supertweeters - 10khz or so and up, at the lowest. And for that, I'm not sure if I'd use my TN52's or consider that a waste. These time lense tweets are super accurate.

So I'll start with the measurements:

Again, I did three because I was doing my listening both with and without the midbasses, back and forth, gauging their performance independently and "as usual", and also trying to make judgements on how I'd rank the tonality and dynamics of them.
Green - the BG Neo8S's alone.
Blue - the Anarchy 6.5's alone.
Red - Both playing.

I had been thinking they didn't seem to play as low as the array of 2" drivers, but looking at the measurements, I think that's just that the BMR's had a smoother response, where the BG Neo3S's have a hump between 500hz and 1khz, then a dip between 1khz and 2khz that would need to be EQ'd out. They seem to have plenty of overlap to let me pick an Xover point between 500hz and 1khz, although I'm not sure if I really want my doors playing that high.

I didn't really ever do the "crank it to full blast" test - I'll do that next, since I want to test these BG Neo3Ss in a simulated pillar setup, and possibly also with a 15 degree angle (borrowing from the Mirage's engineering) from the corner of the windshield toward the listening space. More to come on that - listening results and pics below are of them installed in the dash top:

[geolemon]

I started with the drivers firing up (there's cabinets below). I liked the drivers, but felt there was a gap at the center, and the image height was low. Width was great at least.


I listened to a couple songs, but quickly decided it was not up to par from an imaging standpoint, so I added some domed reflectors that I designed and 3D printed to see what would happen.
...bear in mind I didn't design it for this speaker, I designed it for a 3" Dynaudio dome mid pair I own, and/or a 3" Aura Whisper set of inverted domes. But I wanted to see if it would do the same for these as the Mirage's.

With those over the top (centered, but just slightly back) I saw a dramatic image improvement - mostly centered, although a few items still opened a gap at the center, and some higher pitch elements were locateable. Stage width was also still great.


I dramatically raised the image height - but unfortunately with the locateable aspects here and there, it caused a bit of that dreaded image-smearing, reality-snuffing rainbow effect. Possibly I could iron that out with some tweeters taking over the highs at some point, but this dash-top install would be my most difficult to pull off, so I'm not particularly eager to cling to it.

As mentioned, I ended up listening to fewer songs, although I listened more than this may seem, since I was backing these songs up repeatedly to try different things... Midbass on, off, no EQ, EQ to make it closer to flat, with the reflectors, tweaking the reflectors, sliding them side, back, fourth...
Lots of listening was done:

[geolemon]

BG neo3
<B>- facing up, no reflectors</B>
Brothers in arms - midrange imaging great, some gap in the middle, but the guitar solo is great with no gap. Cymbals locateable.

Isn't she lovely - some separation of the opening whistle and vocals. Turning off the midbass, it was fairly locateable.

<B>- facing up - added reflectors. </B>
Used same track to position them, noting slight shifts made significant changes to image. Centered but just slightly back of center works best - I suspect this is very similar to Mirage's 15 degree angle - directing slightly more sound energy forward. Will have to try 15 degree angle.

Isn't she Lovely - impressive improvement. Mostly eliminated separation, though still mild separation at vocals, less so if I shut off the midbass. Image height is raised significantly - until the left-only and right-only guitar - essentially the rainbow effect. That goes away nearly entirely with the midbass playing along, but the plucky snaps on both left and right solos are locateable.

Seven Nation Army (Scott Bradlee's Postmodern Jukebox) - far right finger snaps locateable. Mild center separation of female vocal. Far right snare locateable.

Suzie Blue (Ben Harper) - opening was convincingly mono and with no gap. Decent stereo imaging, although the oboe(?) on far right made the speaker locateable. It is right-speaker only, to be fair.

Open up (leftfield) - good center, stereo effects panning subtly back and forth really come out well. John Lydon' s voice is centered - I'm noticing it's locateable to the speakers when he's singing his lowest (which isn't low). Once I hear the rainbow though, I can't unhear it.

Really want to try these drivers firing en from the sides. That will be next - probably recycle the BMR array cabinet for that.

Sent from my LM-G820 using Tapatalk

[geolemon]

Minor disassembly of the test rig, and new assembly to try the Neo3s in an a-pillar install

Decided to re-use that cabinet I used for the BMRs to try the Neo3Ss in a phase-aligned line like I had the BMRs in - so figured I'd try four of them for two reasons -
1) same as the BMRs, ie. phase aligned array, simulate a further out single, stronger direct sound path, weakened reflected paths.
2) to possibly consider a pair of Neo8s, which extend further down than these Neo3s can.

Stay tuned.

Sent from my LM-G820 using Tapatalk

[geolemon]

I ended up building in a little angle to them, using a little foam weatherstripping so I could easily swap in the baffles for the 3" BMR drivers later. Jenky but functional.

This was an improvement to either up-firing alignment - the height was greater, there was no "rainbow" even with pure left or right content, and the width was still decent. There really wasn't a "gap", but still there was a width, a spread rather than a distinct focus, of some of the lower mid - I think right around 500hz, so tough to troubleshoot since that's my passive Xover and sure could be phase shift. The two vertical lines I drew shows the breadth of area, for that.

Bear in mind I'm "using four" rather than "using two".
They are mounted on that same curve (more of a line here) that I had the 2" BMR array on. Close enough for direct path phase alignment.

They are wired in series, giving the receiver 8 ohms, but were still more efficient than the pair @ 4 ohms. I ended up doing most of my listening this time with my EQ on my phone, basically boosting the underpowered Anarchies and cutting this hyper-efficient quad-Neo3S setup, as shown.

Seems extreme, but bear in mind the passive Xover is 500hz, so this really did work - it underscores that there's a big difference in efficiency here... and also why I'm separately second-guessing my front stage amp.

What I mean is - I like matching amps, I might be able to use a whole 100w/ch on the Anarchies - but I probably won't need 100w/ch if I went with these drivers, maybe more like 25w/ch.

Also - even before I set these up, from my last listening session I was already thinking "two per side would be a good indicator of whether the big brother Neo8s would be any even better fit". But these Neo3s sound great.

Sent from my LM-G820 using Tapatalk

[geolemon]

Here's the measurements, for the four Neo3's - you can see the efficiency gained compared to the last graph, but otherwise no real surprises here, nothing really changed from a response perspective - even though they are certainly firing from a different angle.
Red - both drivers playing
Purple - Anarchy only
Green - Neo3's only



As you can see, the Neo3's really drop off quick, they just don't play below 500hz. The Anarchy are fine up to 500hz at least - and really the imaging isn't bad with them in my little sealed cabinets here, but if I could get the dash (or pillar) speakers to play lower, that should be better since I am getting good imaging at this location.

So I'm thinking of the Neo8s rather than the Neo3's, even though I already own the four Neo3's.
Advantages of the Neo3's - I could do that phase-aligned array with one Neo3 on the sail panel (where factory tweeter is) on the door, and one in the A-pillar. And it's pretty clear that I don't need tweeters with these. Heck, they basically market them AS tweeters.

With the Neo8's I'm not necessarily prevented from that either - I could do an A-pillar that stretches down over the sail panel, I don't think that would block the door opening, and wouldn't block any vision that's not already blocked by the A-pillar. Would be on me to make that look factory rather than weird, but I think that's doable.
What I don't yet know is if "I don't need tweeters" with the Neo8's. That's also not a problem since I am sitting on a bunch of tweeters, and if I could even run the Neo8's up to 10Khz and then run some tweets to function only as super-tweeters, that's fine. But of course that also impacts if I need 2 or 4 channels, so that needs to be sorted.

Here's the response from their white papers, of the Neo3 vs Neo8 response:


It almost seems like the Neo8 reaches even higher than the Neo3, but really it's an unknown since I don't own them.