Beyond the Ariel

[Lynn Olson]

Originally posted by agent.5
How many bands of overlaping parametric EQs do you think we need for each channel of the midbass/bass?

I have some vintage Rane parametric EQs with 3 bands for each channel. I think that should be enough. But maybe not and I need to start bidding on some Klark Teknik 5 band EQs.

There is such a thing as over-equalization - in this case, all I plan to do is remove peaks from the response. 2 or 3 parametric EQs should be plenty - the few times I twiddled with a parametric EQ, 2 sections usually got rid of the worst peaks. A room would have to be pretty bad to need 3 peaks removed in the narrow 80 to 240 Hz range - bad enough that room treatments might be a better approach.

My attitude towards holes in the response is "live with it", or at worst, do something about the room. Dumping energy into a room null is a bad idea all around - it wastes power, greatly increases driver excursion (really bad), and still sounds weird and "pressurized" when it measures flat.

Channel-by-channel peak removal works pretty well, though, clarifying the bass region wonderfully. The improvement in the stereo image is also quite noticeable and appreciated.

[Lynn Olson]

Can Line Arrays Form Cylindrical Waves?

Beam Steering with Time Delays

HF Driver Design

[KBK]

My attitude towards holes in the response is "live with it", or at worst, do something about the room. Dumping energy into a room null is a bad idea all around - it wastes power, greatly increases driver excursion (really bad), and still sounds weird and "pressurized" when it measures flat.

thank you. I took alot of heat on the AVS forum a while back for telling all those high end guys that their 'digital room correction' was a still birth at best.

[Zene Gillette]

Some yeas ago I remember reading of a mfg using very high quality and great sounding speakers built a 3-way system.
The crossover was passive with the very best parts, soft and hardware available at the time.
It was massively complex.
They got it flat to 0.5dB.
Bottom line it sounded like s***.
Zene

[terry j]

Originally posted by KBK
"My attitude towards holes in the response is "live with it", or at worst, do something about the room. Dumping energy into a room null is a bad idea all around - it wastes power, greatly increases driver excursion (really bad), and still sounds weird and "pressurized" when it measures flat."

thank you. I took alot of heat on the AVS forum a while back for telling all those high end guys that their 'digital room correction' was a still birth at best.

Sure, but Lynn doesn't say "don't cut", all he says is to "watch the boost".

DRC certainly has it's place and can produce excellent results if done properly.

[Lynn Olson]

Originally posted by KBK
"My attitude towards holes in the response is "live with it", or at worst, do something about the room. Dumping energy into a room null is a bad idea all around - it wastes power, greatly increases driver excursion (really bad), and still sounds weird and "pressurized" when it measures flat."

thank you. I took alot of heat on the AVS forum a while back for telling all those high end guys that their 'digital room correction' was a still birth at best.

It depends where you do the correction. From 200~300 Hz on up, the ear can distinguish between the direct-arrival spectra and the spectra of the room reflections. The microphone is too stupid to do this, and sums them all together.

There's still some controversy over this, but most speaker designers optimize the direct-arrival spectra, since the consensus opinion (with a few noisy dissenters) is that first-arrival spectra determines the sonic "color" of the speaker, with the room playing a secondary role.

With nearly all speakers, the room energy diverges to at least some degree from the first-arrival spectra, with the room energy tending to roll off at higher frequencies. You can see where this is going. If you mindlessly equalize the combined spectra from the direct-arrival + room spectra, the direct-arrival will be "hot". That's why the prosound folks so commonly build in a HF rolloff starting around 3 kHz.

I'm old-school enough to aim for a subjectively pleasing direct-arrival spectra (slight bass lift) and trying to get the room spectra fairly close to that. This is the traditional BBC method, and I'm sticking to it. It works for symphonic music, and sounds good on rock, too.

Now for frequencies below 200~300 Hz, the wavelengths are getting so large they are starting to merge with room modes, making them difficult to isolate. This is where EQ makes sense - the distinction between direct-arrival and total room energy disappears at these frequencies.

So I believe in limited EQ - below 300 Hz, and only to remove the biggest peaks on a per-channel basis. The same EQ compensation should NOT be applied to both channels! One of the biggest benefits of individual per-channel EQ is the ability to get rid of the worst LF channel imbalances - so speakers can be located for best image quality, and not be as constrained by avoiding LF artifacts.

The EQ addicts just slam everything through the auto-EQ, and of course that gums up the direct-arrival sound, wrecking the frequency response and making a mess of the time coherence. When I apply the auto-EQ from my Denon HT receiver, I have to undo more than half the EQ - basically everything above 300 Hz.

So I take a purist approach above 300 Hz, designing the passive crossover and the drivers to have an intrinsically flat first-arrival response, and with as good a time coherence as I can achieve. Below that, modest active EQ works fine, and transistor amps don't sound too bad either.

Wally Malewicz of WallyTractor fame has designed his own speakers, which he sells on a custom basis. He takes the same approach, using a Fane parametric EQ for the bi-amped bass modules, and using a passive crossover for the mids and highs. I spent a very enjoyable couple of hours at the last RMAF chatting with him, and I can tell you he's a very sharp guy, and knows what he's talking about. His speakers are first-rate and very musically balanced, in the Siegfried Linkwitz class.

[gedlee]

Just to close out my previous posts I want to say a couple of things.

First, I am not claiming to be right and others wrong about anything (not in these last posts at least), I am only asking that those who are making the claims support them with scientific data.

Clearly this is not a position that one can hold in this forum as was so clearly pointed out ("swimming upstream")

As a scientist, I have no business in discussions where the scientifc method is not held as the only viable road to truth. To me a subjective test that is done without scientific controls and statistical analysis of the results is no test at all. The fact that the reported results of such tests are often self serving to the discussion makes them even more suspect in my eyes.

I simply cannot have a discussion where there is no coaberating evidence and/or the data is so loosely obtained. Others feel perfectly comfortable with that situation, I am not. So I will leave these types of discussion to those who enjoy them.

[Lynn Olson]

From the ALMA webpage:

"Earl Geddes has been in the audio industry for nearly 40 years. He holds a BS and MS in Physics from Eastern Michigan University and a PhD in Acoustics from Penn State. He has worked in audio his entire life, mostly with Ford Motor Company but later at Knowles Electronics, always in the area of audio transducers and systems. Dr. Geddes has received numerous scholarly awards, more than 25 patents, and he has authored two books. Earl is currently President of GedLee LLC a consulting firm in Acoustics that he runs with his wife, and this year he joined up with Kenny Jackel to start Audio Intelligence (Ai), a loudspeaker manufacturing company located in Bangkok, Thailand. He has been a very active AES member since 1978, and an ALMA member since 1999."

Meanwhile, back at the Hogwarts School of Magic and diyAcoustics, Magnetar made an interesting post:

Originally posted by Magnetar

I have built quite a few 'hybrids' and all of them seem to fall flat on their face compared to mating the compression driver with a good bass horn. It doesn't really need to be loud to be significant. I have used JBL 2220, 2226, TAD 1601, 1603, EV 15L, 15B, Altec 912B, 416 and 515 as direct radiators crossed electronically to compression drivers in 300 Hz round tractrix horns almost anywhere between 600 and 1500 cycles. The direct radiator bass systems all seem to kill the dynamics compared to a bass horn.

15" + CD/horn systems trace their lineage back to the 1937 Iconic - it has become the sound we recognize as the "studio monitor", since so many have followed that format. Drivers come and go, horns and compression drivers change, but the basic format is now in it's 70th birthday.

Magnetar, what's your take on the cohesion issue? Do you think the combination of CD's and midbass drivers are doomed to fail, or maybe the transition to direct radiators needs to be deferred to much lower frequencies where horns become unreasonably big? Just curious about your take on the whole Iconic/studio-monitor format for loudspeakers.

Having been properly put in my place by a senior scientist at Ford Motor Company and Knowles Electronics, I'll put on my wizard's cap and ask the other magicians here what they think of the subjective-integration issue. Dissimilar driver technologies, working at the ends of their ranges? Trouble with changes in radiation pattern at the crossover frequency? Wrong choice of crossover frequency? Lack of scientifically-controlled double-blind testing with an objective test panel of listeners using a statistically-valid methodology?

Not that I'll be taking anyone's advice, mind you. I'm slowly writing my own book of spells, and as the good Doctor has so thoughtfully reminded me, I have a great deal to learn yet.

[Russell Dawkins]

I suppose a certain bias is explicit even in the implications that the chosen brand name "Audio Intelligence" carries with it.

[KBK]

Originally posted by gedlee
Just to close out my previous posts I want to say a couple of things.

First, I am not claiming to be right and others wrong about anything (not in these last posts at least), I am only asking that those who are making the claims support them with scientific data.

Clearly this is not a position that one can hold in this forum as was so clearly pointed out ("swimming upstream")

As a scientist, I have no business in discussions where the scientifc method is not held as the only viable road to truth. To me a subjective test that is done without scientific controls and statistical analysis of the results is no test at all. The fact that the reported results of such tests are often self serving to the discussion makes them even more suspect in my eyes.

I simply cannot have a discussion where there is no coaberating evidence and/or the data is so loosely obtained. Others feel perfectly comfortable with that situation, I am not. So I will leave these types of discussion to those who enjoy them.

Just be aware that the origins of the scientific method lie with Robert Hooke, Boyle, etc..and have nothing to do with the current norm as created by Newton, whom was someone who crushed Hooke and rebuilt the Royal society to his own desires. One of which was was to re-define 'scientific' methodology' to HIS particular 'way'. This actually made technological change and advance less occurant and less viable. Interesting stuff happened way back then. hhhmmmmm Like Newton's ego, for example. Ie, the idea of his math making reality, instead of the utterly realistic and sensible..other way around. Which the true origins of 'the scientific method' - is, and was, predicated upon.