Orion-like OBs with AE Dipoles

[NYCOne]

With the availability of low cost DSPs (minidsp and others), and ARTA and home testing relatively straight forward, have folks played with Orion-like OB builds?

I was thinking about Orions, but was not convinced the woofers were ideal. I've been drawn to the AE TH15 or Dipoles. One build that seems to be heading in the direction I was considering was done by Sdodds. He did his Bob with an Orion top and AE 12" IB woofers. Has anyone tried the AE Dipoles for a woofer section?

One could imagine a wonderful outcome with them.

[454Casull]

When in doubt, go big.

[preiter]

Originally Posted by NYCOne

With the availability of low cost DSPs (minidsp and others), and ARTA and home testing relatively straight forward, have folks played with Orion-like OB builds?

I was thinking about Orions, but was not convinced the woofers were ideal. I've been drawn to the AE TH15 or Dipoles. One build that seems to be heading in the direction I was considering was done by Sdodds. He did his Bob with an Orion top and AE 12" IB woofers. Has anyone tried the AE Dipoles for a woofer section?

One could imagine a wonderful outcome with them.

I'm not convinced that a high-QTS driver is the way to go with an Orion-like speaker. Higher QTS generally means a smaller magnet and lower sensitivity, with a resonant bump at Fs. The alternative is a lower QTS speaker with higher sensitivity without the bump.

The benefit of the high QTS driver is that is provides some compensation for the OB roll-off without requiring any EQ. If you are going to be using active EQ anyway, you can get better results with a lower QTS.

The advantage of building the Orions as they are is that you get years of careful engineering and fine-tuning by a very experienced designer. If you start swapping drivers, you throw much of that away.

[454Casull]

Originally Posted by preiter

I'm not convinced that a high-QTS driver is the way to go with an Orion-like speaker. Higher QTS generally means a smaller magnet and lower sensitivity, with a resonant bump at Fs. The alternative is a lower QTS speaker with higher sensitivity without the bump.

The benefit of the high QTS driver is that is provides some compensation for the OB roll-off without requiring any EQ. If you are going to be using active EQ anyway, you can get better results with a lower QTS.

The advantage of building the Orions as they are is that you get years of careful engineering and fine-tuning by a very experienced designer. If you start swapping drivers, you throw much of that away.

Linkwitz Orions beaten by Behringer.... what!!?

Don't think too much of "staying true to the design", FWIW.

[twest820]

Short answer: it's been discussed extensively (just search for Orion) and there are two basic camps; those that like to stay close to Linkwitz's design and those who feel speaker design's moved on enough that if one's going to the trouble one might as well pursue additional optimizations. Personally, I started out with intent to build Orion clones and ended up here instead as I had the design leeway to select a geometry with more consistent directivity, simpler build, and less vibration. I was also able to select drivers more for transient response and directivity than for excursion as Linkwitz did with the Orion.

Somewhere in the middle of that there was quite a bit of back and forth with the proprietor of AE Speakers---you should be able to turn up the thread with the search function. To summarize, AE makes sense if you're locked into a design which requires large excursions. If you can avoid that, there are lower cost options which seem to measure as well or better.

Originally Posted by preiter

If you start swapping drivers, you throw much of that away.

Given reasonable replacement choices most of what's likely to be lost is the inconsistent directivity of back to back tweeters, Seas' odd tendency to design woofers which run hot on axis, and the undersized subwoofers. There's also the magnet shadow on the rear wave from the 8", though the cross to the tweeters is low enough the directivity mismatch between front and back waves from the shadowing is mitigated. In most cases this should be no great loss, though one obviously shouldn't go tinkering around with the design unless you know what you're doing.

[LineSource]

"When in doubt, go big".

I found this good advise for dipole woofers. I am currently using two Lambda Dipole15s wired in parallel in a 16" deep H-baffle for dipole bass, and I find this "adequate". My amplifier easily drives the low resistance of two parallel drivers, and the higher efficiency is important to get details from male vocals. An H-baffle with two 15" is very tall, and this limits M-T options for a 36"-40" seated listener height. A single 18" dipole woofer like the new Lamba now in a group buy would give enough volume with a lower height H-frame.

For the mid-tweet I find a conflict between two options:
1) The JohnK NOTE that optimizes constant directivity and power. JohnK's 5-way NOTE uses a BG3 planar tweeter mounted as close as possible to a 4" ScanSpeak upper midrange to get good directivity, plus two 8" low/lower midranges, plus woofers. From my experience, crossovers within the upper-mid to lower-mid frequency range are usually audible.

2) Optimizing the "vocal_range" like Linkwitz's 3-way using an 8" midbass, that even when crossed at 1,400Hz, exibits flaws in directivity. I use a Tang Band W8-1808 crossed at 1,700 to a 1.2" wide dipole ribbon tweeter and live with the directivity flaws in order to get the coherence of all voices on one midrange. There are good reasons why there is an entire "full range" forum.

If you do not exactly clone the Linkwitz Orion, then I think these two design options are worth further research.



AES Speakers Lambda Dipole18

Dipole18 D16 - coils in parallel
Fs: 22.9hz
Qms: 8.85
Vas: 885L
Cms: .42mm/N
Mms: 115g
Xmax: 14mm (one way)
Xmech: 23mm (one way)
Sd: 1218sqcm
Vd: 3.4L p-p
Qes: .89
Re: 6.8ohm
Le: 0.35mH
Z: 8ohm
Bl: 11.24Tm
Pe: 150W
Qts: .80
1W SPl: 92.8dB
2.83V: 93.5dB



Vocal_range Speakers

WIKI: Vocal range is the measure of the breadth of pitches that a human voice can phonate. The most common application of the term "vocal range" is within the context of singing, where it is used as one of the major defining characteristics for classifying singing voices into groups known as voice types.

The following are the general vocal ranges associated with each voice type using scientific pitch notation where middle C=C4. Some singers within these voice types may be able to sing somewhat higher or lower:

Soprano: C4 – C6
Mezzo-soprano: A3 – A5
Contralto: F3 – F5
Tenor: C3 – C5
Baritone: F2 – F4
Bass: E2 – E4

In terms of frequency, human voices are roughly in the range of 80 Hz to 1100 Hz (that is, E2 to C6) for normal male and female voices together.

Fundamental Speech frequency
The voiced speech of a typical adult male will have a fundamental frequency from 85 to 180 Hz, and that of a typical adult female from 165 to 255 Hz.

[ScottG]

Originally Posted by NYCOne

Has anyone tried the AE Dipoles for a woofer section?

One could imagine a wonderful outcome with them.

Forum member Studiotech has.

He has since gone with a solution I've been advocating for a long time.. A much more efficient pro midbass driver in a much more limited pass-band (say 60 Hz up to around 300 Hz), with a sealed servo sub filling in below that.

If there is enough over-lap (between the sub and the dipole mid-bass), it tends to provide not only the clarity of the open baffle experience, but also tactile quality of the sealed sub. Depending on the design, the overall nonlinear distortion can be substantially improved as well (when compared to a driver or drivers operating as low as 20 Hz as a dipole).

He uses a B&C 15".. but it could just as easily be any number of pro drivers, including perhaps the AE 15M.

[gainphile]

- 8" drivers on narrow baffle like Orions, crossed at 1.5khz-ish are fine. There are plenty of directivity plots to show this. The issue is the transition to dome tweeters. They are not dipoles and will exhibit widening-narrowing directivity a.k.a. "Klingon battleship" polar response. Whether this is audible is up for discussions, but clearly not desirable.

- With woofers, as long as they can produce target frequency response of smooth 2nd order rollof (e.g. fs=20hz, Q=0.5), any types will do. WIth low-qt woofer just use 2 shelving lowpass, and for high-qt woofers just use Linkwitz biquad transform.


ps. example of my bird-of-prey plot, taken from those dipole on my avatar pic

[ScottG]

JohnK's NaoII directivity plot (which is I believe a bit better than the Orion +, and is substantially similar in design to the Orion +), can be seen here (if you put your cursor over the Note polar plot):

NaO Note Details

Here you can see what happens when you radically "shrink" the baffle to "NO" baffle on monopole drivers "back-to-back" run out of phase:

Investigations

[Drew Eckhardt]

Originally Posted by 454Casull

Linkwitz Orions beaten by Behringer.... what!!?

Don't think too much of "staying true to the design", FWIW.

Only in regards to "auditory scene" which is a superset of imaging, sound stage, and the reverberation characteristics of the source venue when used with subwoofers and after being equalized to match the Orions in their optimum position.

All other speakers were equalized to match the Orions in their optimum position. Sub-woofers were used. Also tested were speakers constructed of four Radio Shack in-wall speakers one per cabinet side, with the pair firing towards the front wall at full level, inward pair at -6dB, and outward pair at -12dB which were found to be an improvement with regards to "auditory scene" creation.

http://home.provide.net/~djcarlst/SLReport10.05.pdf

All configurations (reference Behringers, Orion in 3 locations, Radio Shack things, Behringers in the room corners) were assessed in six categories on a scale of 1 to 10

1. Speakers disappear
Sound is independent of speaker location

2. Local acoustics not heard
Acoustics of the room your are listening in are supressed

3. Images lateral localization
Frontal, distribution, stability and not split are most important

4. Images depth localization
Usually relates to audibility of venue acoustics

5. Ambience non-localized
Recorded venue reverb spatial quality

6. Freedom of movement
Reasonable "sweet area" AS stable with small head movements