Beyond the Ariel

[Lynn Olson]

Discussion of 2 -> 3 channel stuff has been moved over here. Good comments, and no offense to the Trinaural folks intended. I encourage everyone to try 3 phase-matched speakers with a static-coefficient decoder some time - you might be surprised. Have fun playing with the decoding coefficients and center-speaker level - you can make dramatic changes in the perspective with small adjustments.

[dobias]

Klaus,

Maybe we ought to label the new thread "Pandora's Box".
I didn't realize I was opening it.
I had planned to use a third Wharfedale Super 12/CS/AL I have for the center. As dominant as they are, I couldn't imagine using anything else.
For what it's worth, the Ambisonics group believes it best to avoid a separate amplifier for the center channel.

dobias

PS: I, too, have had patents (that made fortunes for my employer).
I learned the only good patent is one that can withstand lawsuits.
I defended my patents for 5 years in Federal Courts. In fact I have the dubious distinction of having the last Federal trial , all the way to the Appeals Court, before Patent disputes were resolved by an arbitration board.

[Lynn Olson]

My first foray into commercial audio was the invention of the Shadow Vector SQ decoder - this idea was my ticket out of an unpleasant retail-sales job in Los Angeles, moving to Portland, Oregon, and working for Audionics. What a wild ride that was! The 2~3 years of developing the Shadow Vector prototype, meeting the bigwigs at CBS and demonstrating what a little Oregon company could do, was definitely worth it.

I was especially amused that the CBS team had to hire a professional mathematician to describe the operation of their Paramatrix - while I, the math dunce, the one that never got a decent grade in math in college, reduced the operation of the Shadow Vector down to 4 equations. The Scheiber-sphere figures 14 through 22 on the patent were the essential concept - the rest was a reduction to practice and building the thing. Surprisingly, the first time I heard it, it did sound like the sound I'd had in my head the whole time - spacious and clear, with smooth, natural reverberation all over the room.

I'm not sure the CBS team ever understood the concept of equal-energy decoding, which prevented the decoder from "hotspotting" and concentrating too much reverberant energy in any one area of the room. A very real risk for any dynamic decoder is optimizing too heavily for max separation at the speakers, while ignoring the stability and quality of the phantom images between the speakers. Unlike many other decoders, the Shadow Vector had equal separation for all directions, phantom or not. Dolby Pro-Logic I had a problem with "detenting", for example, while Pro-Logic II in "music" mode is greatly improved - although the energy distribution still doesn't sound completely even to me.

Even though quadraphonic never went anywhere, it was a fun ride while it lasted - and the experience came in handy for designing loudspeakers. Since I'd spent 2~3 years on a system that was designed to have smooth distribution of reverberant energy - while maintaining crisp localization - the key concepts were just carried forward.

[Robh3606]

Hello Lynn

“I detest the sound of JBL CD horns in movie theaters, but I'm not going to blame the compression drivers for the defects of PA horns and appalling THX signal processing.”

I have heard some really terrible sounding installations as well. I have also been in the same venue for both acoustic and electric bands and had some real eye opening experiences there. Saw the acoustic show first with guitars and vocal with some drums. I was surprised just how good that PA sounded. Went back for another show with high hopes and it sounded just awful. I was amazed it could sound that bad after seeing the previous show.

Actually in Pro and Cinema applications I think it’s a combination of both the drivers and the horns. I think Earl Geddes is on to something with his waveguides designed to minimize the effects of HOM modes in the horns. From what I have read the audibility of HOM is dependent on SPL.

In that application you have worst case conditions with both HOM and the compression drivers. If you look what CD compensation is doing you are adding between 6-10Db of active compensation so you power input is between 4 to 10 time greater right in the region where these diaphragms are in these various break-up modes.

If you look at what happens in a typical home installation it’s apples and oranges. At home we can use a passive network for CD compensation to attenuate the midband response. This reduces both the diaphragm excursions in the compression driver and avoids dumping power into the drivers above the 5K and up range. I most cases depending on what the woofer sensitivity is you would be lucky to be using more than several watts of power. You will be well below the drivers power rating and excursion limits.

With the reduced SPL’S the HOM are not as audible and may be below your tolerance threshold for them. I have pushed some horns at home and sure enough you get past a certain point and it gets nasty for lack of a better word. At the typical lower SPL’s used for home listening I thought they sounded good. By typical my average is about 85db where I sit but I have the capability to hit a clean 115db on peaks so I have 30db of available headroom above the average. That headroom makes a big difference on drums, horns, and percussion IMHO.

I don’t think it’s a fair assessment of what performance you can get using Pro applications as a reference. I think that the potential is there for horns and compression drivers to offer better performance in a home installations than we typically hear when used in Pro sound applications.

Rob

[Lynn Olson]

Hmm - probably because of my previous work in quadraphonics and low-diffraction loudspeaker design, I may have sensitized myself to loudspeaker artifacts that damage the stereo perspective and perception of ambient impression. Much of the work on Shadow Vector was focussed on retaining the ambient impression without distortion (truncation, fluttering sensations, what we call "front-to-back depth" but over the whole 360 playback arc). If the venue was large, it had to sound that way, and without "detenting" effects or shadowing (portions of the room that lacked depth). Most other quadraphonic decoders had almost no sensation of depth at all, or a very unnatural perspective that was "speaker-centric". Since a lot of monitoring was done on pro monitors with very poor depth rendition (but good dynamics) the recordings themselves could be quite variable.

The same applied to my later work at Audionics. I was firmly in the Brit-school of design, with an emphasis on the work of D.E.L. Shorter of the BBC and Laurie Fincham at KEF. Meeting them in 1975, and hearing BBC experimental quadraphonic mastertapes, left a strong impression which has influenced what I do to this day. The construction of a very low diffraction experimental speaker at Audionics (it looked like a big vitamin pill) in 1979 also left a strong impression - reduce diffraction enough, and the speaker effectively disappears. Increase the diffraction, there is more sensation of a distinct speaker-shaped sound source in the room, increase the diffraction some more, and soundstage "detenting" becomes apparent, center and part-center phantom images become diffuse, depth impression gradually diminishes to zero, and then, in the most severe cases, gross mid and upper-mid colorations become obvious - particularly with singers and massed chorus.

Some of this sensitivity must be individual - I can listen to systems that have gross amounts of diffraction and diffraction-related colorations, and other listeners in the same room are thrilled and rave about the "depth" and "realism" - and I hear no depth, no realism, and lots of tin-can and megaphone-like coloration. I've yet to hear a conical horn, for example, that doesn't have unacceptable levels of diffraction and mid-coloration - maybe they're out there, just haven't heard one for myself. By contrast, the Dr. Geddes and LeCleac'h waveguides that I've heard to date sound completely different - open, natural, and very little horn coloration.

What dismays me about CD horns is the intentional diffraction inside the horn - what I call a "kink" in the flare rate. Since horns already suffer from mouth diffraction and strong reflections off the hard surface of the phase-plug, adding an additional diffracting surface inside the middle of the horn just seems to be asking for trouble. This kind of trouble is very obvious in impulse-response measurements, but for reasons that must be historical, impulse measurements are very rarely used in the professional world.

I'm especially pleased that both Dr. Geddes and LeCleach'h are looking into time-domain performance for waveguide and horns. One of the more frustrating "could-have-beens" was that Dick Heyser (of TDS fame) published his first work in the early Seventies, and lived in Los Angeles. LA is a JBL/Altec company town, and these two industry-leading companies did their best to ignore him - "Not Invented Here".

If there had been less NIH attitude from JBL and Altec, we could have had horns with good time-domain performance for the last 35 years - and CD horns may have taken a very different direction. I think it reveals something that impulse response data on horns is still very difficult to find from the world's largest vendor of professional systems - 15 years after MLSSA became the industry standard for measurement - and impulse response is the very first measurement you see on the screen when MLSSA data is captured.

[Robh3606]

Rob, which of the JBL 1.5" exit drivers have the modern technology - and what's your take on the sound when installed on Tractrix or LeCleac'h profile horn? Very curious about this.

Hello Lynn

I split my reply to keep the post from going too long. The only JBL compression drivers I have heard on a tractix were reconditioned 2441’s listening to Bruce Edgars Titans. I liked them quite a bit.

For a 1.5” throat you would have 2430/2431/2435/2451 and 2452

The only modern drivers I have used from the current crop are the 2431 and 2435’s which have consumer equivalents in the 435AL and 435Be. The difference between the Vertec and the Consumer versions are the backcap and aquaplas used on the consumer versions. Other than that they are identical.

I was using a 2344A horn with a 2425 1” compression driver. Standard 80’s issue in the 4430/4435 monitors. Classic CD design with a diffraction slot with sharp transitions. That’s the horn used in Earls blind comparison testing for his Suma speakers. The only difference from stock was I had an aquaplas damped Ti diaphragm installed. I was happy with them for a couple of years. At the levels I normally listen at they sounded fine to me.

When 2435’s started showing up on E-bay I took the leap and purchased a pair. All you normally need to do with these is clean the gaps replace the Ferro run impedance plots and a quick measurement and you are good to go. One of that drivers main advantage as you know is freedom from the break-up modes present in the older Titanium drivers. I mated them up with the PTH 1010 Waveguide which are one of JBL’s newest waveguides.

The horn/waveguides are identical as far as coverage pattern, size, and frequency range in my application. I could just swap them into my existing set-up. The compensations had to be changed so I had to build networks but other than that only a change in level was required to drop them in.

I made the switch and was pleasantly surprised. The newer driver and waveguide was simply a better sounding combination. They were better in every way. I would encourage anyone looking at compression drivers to have a listen to the newer drivers. Not just JBL, any that are using newer phase plug designs or to try the newer waveguides on older drivers. If I was running 1" drivers I would be looking real hard at Earls waveguides.

Rob

[Magnetar]

Isn't aquaplas just latex paint?

Do you have a graph showing the titanium treated drivers having "no" breakup?

[Robh3606]

Hello Magnetar

"Isn't aquaplas just latex paint?"

No not at all. It's now called AntiVibe but still made by the same company. It's been around for quite some time. The white coating on the 4310/4311/L100 woofers and the LE-14 is Aquaplas/Antivibe. Do a search and you will get plenty of industrial supply house links. It used to damp ductwork and many other sound related applications.

I have a sample and it's a heavy paste that flows under preasure. In most applications it gets sprayed on. You can see the silica fillers in it. When it dries it's a matte finish with a "sparkles".

Tried doing some home grown experiments with it before and after. I will take a look to see if I have the CSD plots. It was a real head scrather. Reminds me of the Enable thread. One of those I know it's doing something I can hear it , now go and find what it is.

Rob

[Magnetar]

My Uncle used to be authorized repair for Lansing. He says it was called Lansauplas (something like that) and it was latex paint with shredded paper mixed in. He made his own when they quite suppling it to him - maybe they changed the formula

Looking at the JBL/Harmon stock it looks like they are about to shut there doors - it was a trading at 120 a fw months ago and now is trading around 43.

Hopefully they'll pull through- although Radian does make some nice aftermarket diaphragms

[Robh3606]

Hear are a couple of CSD plots for a 2344 driven by a 2425 and a PTH1010 driven by a 2435. I plotted from 1.5K and up because that the range I was/am using them over.

Rob