Compression Drivers for Music at Home?

[dallasjustice]

That's an older waveguide. There's a new crew designing waveguide speakers at Harman.

So horns give you the directivity control, and they are efficient. But, as the Stereophile review above says: "the JBL 1400 Array's cumulative spectral-decay plot is not as clean as I would have liked to see, presumably due to reflections of the driver outputs from the edges of the horns" - and the mid and treble anechoic plot is a long way off the flatness and smoothness that direct drivers give you. In other words, colouration. Some people don't seem to notice it, and some types of music don't show it up as clearly, but I think it is always there.

 

[dallasjustice]

That's a stoopid price. I'd bet the 708p would smoke me for less money.

Nice set of Synthesis 1400's for sale on ebay for $4998,88
http://www.ebay.com/itm/JBL-Synthes...rs-/132080557295?_trksid=p2141725.m3641.l6368

 

[Sal1950]

That's a stoopid price. I'd bet the 708p would smoke me for less money.

You with your 4367's, or the 1400s ??

 

[Blumlein 88]

So horns give you the directivity control, and they are efficient. But, as the Stereophile review above says: "the JBL 1400 Array's cumulative spectral-decay plot is not as clean as I would have liked to see, presumably due to reflections of the driver outputs from the edges of the horns" - and the mid and treble anechoic plot is a long way off the flatness and smoothness that direct drivers give you. In other words, colouration. Some people don't seem to notice it, and some types of music don't show it up as clearly, but I think it is always there.

I would be interested if anyone here could do some of the REW measurements on some horns and show us the unsmoothed response or at least 1/48th smoothing. I have done some experimental room correction/eq with some older horns a few years back. The horns do tend to have some spiky output. Maybe newer better horns don't. I was doing this on K-horns, Advents, and Electrovoice speakers. By spiky I mean narrow large spikes and dips in the response worse than you see with other types. Knocking down those spikes with an early Tact unit (which I seem to recall worked on 2 hz increments or something like basically a 32k fft view) made a positive change. Response was smoother and even, yet it still had good attributes of being dynamic in a muscular way and sounding finally like low distortion when you cranked on the volume.

 

[RayDunzl]

Send me some horns and I'll do it.

Who else posts?

DallasJustice has the ability and the JBL 4367, but is usually busy arguing elsewhere (and getting paid for it).

 

[amirm]

I might get the 708p when they come out so perhaps we can revisit then.

 

[oivavoi]

I would be interested if anyone here could do some of the REW measurements on some horns and show us the unsmoothed response or at least 1/48th smoothing. I have done some experimental room correction/eq with some older horns a few years back. The horns do tend to have some spiky output. Maybe newer better horns don't. I was doing this on K-horns, Advents, and Electrovoice speakers. By spiky I mean narrow large spikes and dips in the response worse than you see with other types. Knocking down those spikes with an early Tact unit (which I seem to recall worked on 2 hz increments or something like basically a 32k fft view) made a positive change. Response was smoother and even, yet it still had good attributes of being dynamic in a muscular way and sounding finally like low distortion when you cranked on the volume.

Mirrors my own anecdotal experience. The nice sounding horn I heard was heavily eq'ed. The bad sounding horn had no eq applied at all.

 

[watchnerd]

When you guys refer to "horns" are you including waveguides or only referring to "true horns" like Tractix?

 

[Dynamix]

When you guys refer to "horns" are you including waveguides or only referring to "true horns" like Tractix?

I'd include waveguides.

 

[RayDunzl]

Looking at pictures of waveguides, I see anything from full horn to a little decorative dimple...

Besides, the topic of this thread was Compression Drivers.

 

[mitchco]

Here is a near field, unsmoothed frequency response of a BMS 4540ND compression driver on a QSC 152i waveguide:

The BMS 4540ND driver has real response past 20 kHz. Sweep is from 1 kHz to 30 kHz. As it is a constant or controlled directivity waveguide, it requires compensation. The QSC is a really nice sounding waveguide with excellent, controlled off axis response.

I have moved on to a different compression driver and waveguide setup. Happy with the waveguide, but trying a different compression driver as it is not as sweet sounding as a BMS and I require a 1.5" exit as opposed to the 4540ND 1" exit.

I really liked this combo, it sounded very detailed, yet dynamic, with precision pattern control. Kinda missing it now

 

[Frank Dernie]

If you understand French!

http://www.tuneaudio.com/wp-content...-in-French-Stereo-PRESTIGE-IMAGE-Magazine.pdf

I have a pair of these, they don't sound very coloured at all, not quite as good on speech as BBC LS3/5a listening to Radio 3 or Radio 4 (BBC is my normal check for mid-band colouration since our ears know what to expect with voices) but massively better on stuff with dynamic range, as would be expected.
I had heard people say how coloured horns are and had never heard any but after listening to some at the Scalford show decided to listen to a few and ended up choosing and buying these, though I also keep using normal speakers for some stuff.
For me the big shortcoming I have seen from measurements of horns is the somewhat relaxed waterfall plots but this doesn't create any noticeably detrimental (by me) effect.

BTW the Stereophile and HiFi news standard microphone location for speaker review measurements is entirely unsuited to measuring this sort of horn of course, given horn directionality and distance between units. For good integration between drivers a minimum listening distance of 3m is recommended iirc, I listen at around 5.5m from mine. The mid range horn is about 2' diameter.

Mine are burr Myrtle btw not paint!

 

[hvbias]

http://www.zaphaudio.com/

Has plenty of measurements of standard drivers and some ribbons. There are clean and nasty looking CSDs (usually more of the latter). The same applies to horns and compression drivers; the worst looking ones will be narrow bandwidth horns with titanium diaphragm compression drivers. Once EQ'd constant/controlled directivity waveguides with good compression drivers with beryllium diaphragms have very clean CSDs. There are good, bad and ugly measurements for all types of drivers.

From everything Zaph has written it seems like you can get high fidelity midrange/tweeters from a variety of price points. With compression drivers powerful neo/alnico magnets are expensive as are beryllium diaphragms so the good ones (that I know of) are unfortunately pretty expensive. Be replacement diaphragms start at $300-500 and go as much as $1500, I can see why most commercial speakers go cheap on the compression drivers.

IME "that horn sound" is way more common than it is uncommon.

The consensus among the DIY speaker crowd seems to be that, as a general rule of thumb (and of course there are exceptions to such a rule), dome tweeters sound better at low-mid volumes, but start to suffer dynamic compression at high volumes, while compression drivers don't sound as good as domes at low-mid volumes, but stomp them at high volumes.

Would people generally agree with that?

Good speakers sound great at all volumes. Even playing at a volume where you can maintain a quiet conversation with others.

 

[amirm]

From everything Zaph has written it seems like you can get high fidelity midrange/tweeters from a variety of price points. With compression drivers powerful neo/alnico magnets are expensive as are beryllium diaphragms so the good ones (that I know of) are unfortunately pretty expensive.

That's exactly what I hear from Harman regarding high cost of larger beryllium compression drivers.

 

[mitchco]

notnyt over at AVSForum has measured and compared 4 different JBL 1.5" exit drivers and ranked them in order of preference. Includes Truextent Be diaphragm and the CD that is on the M2:
1) 2452H-SL + BeX4008
2) 2452H-SL / 2453H-SL
3) D2430K
4) 2432H
http://www.avsforum.com/forum/89-sp...pair-jbl-4722n-speakers-136.html#post42167329

Some measurements: http://www.avsforum.com/forum/89-sp...pair-jbl-4722n-speakers-121.html#post41520545
It is a long thread with many measurements, if one keeps reading.

In the end, it turns out the Aquaplased 2453H-SL is almost indistinguishable from the Be, but at a significantly lower cost...

 

[hvbias]

In the end, it turns out the Aquaplased 2453H-SL is almost indistinguishable from the Be, but at a significantly lower cost...

I believe he said that with home theater, but there was more difference with music. IMHO this will be more obvious with music with a lot of overtones in the higher registers like cymbals, horns, etc. How sensitive you are to diaphragm breakup will also play a role.

2452H-SL

With Be

The breakup seen above 8 Khz on the stock driver is a familiar sight on some non Be diaphragms. That said I would have expected more difference in the measurements.

While I think that adding Truextent diaphragms is a pretty good value, IMHO the drivers with native Be diaphragms (higher end Radian, TAD) measure cleaner. I'm not sure if this is down to superior driver design, being designed with a Be diaphragm in mind or something else.

 

[RayDunzl]

Most (85-90%) of the world's beryllium comes from one mine in Utah.

http://clui.org/ludb/site/brush-wellman-beryllium-plant

https://www.google.com/maps/place/S...03a73!8m2!3d39.7482779!4d-113.1821962!5m1!1e4

 

[watchnerd]

Today I was in the control room of my local symphony hall, and then in the nearby (small) mixing studio where practice mixes are made, as well as some of the live streaming.

Plenty of compression drivers in the hall itself. Zero compression drivers amongst the monitors in the control room or mixing studio.

I found that interesting.

 

[hvbias]

Today I was in the control room of my local symphony hall, and then in the nearby (small) mixing studio where practice mixes are made, as well as some of the live streaming.

Plenty of compression drivers in the hall itself. Zero compression drivers amongst the monitors in the control room or mixing studio.

I found that interesting.

Were these tower speakers or bookshelves? Size plays a significant role in loudspeaker design constrains. Compression drivers are almost never used without horns/waveguides, and those are pretty large even for fairly high crossover points, ie even the smallest Geddes speaker is huge for a "bookshelf". Pro audio is mostly concerned about power response and care little about off axis that Toole, Olive, Geddes, etc have written so much about. So marketing from well established brands stuck on old fashioned design can and does reign supreme. The various BBC monitor designs are still considered reference level speakers by many of the world's best recording studios.

 

[watchnerd]

Were these tower speakers or bookshelves? Size plays a significant role in loudspeaker design constrains. Compression drivers are almost never used without horns/waveguides, and those are pretty large even for fairly high crossover points, ie even the smallest Geddes speaker is huge for a "bookshelf". Pro audio is mostly concerned about power response and care little about off axis that Toole, Olive, Geddes, etc have written so much about. So marketing from well established brands stuck on old fashioned design can and does reign supreme.

I wouldn't agree about only being concerned about power response, neither in sound reinforcement (where stearable line-arrays are common), nor in mastering settings. In fact the JBL EON PA series uses the more recent waveguides and they make a point of showing the dispersion.

As for what I saw, these are studio monitors, so form factors are different.

The control room used ATC SCM 45A Pro on the bridge and soffit-mounted ATC SCM 110 ASL Pro for mid-field / main monitors, both fairly large:

The mixing studio used Dynaudio Air 20 as nearfield monitors: