Based on what I have read so far in this thread, many people severely underestimate the effect of the room (on recording AND playback) even when presenting evidence of how much it WILL change the sound talking about echo and youtube videos. If you don't find creative ways to deal with the room, everything else is throwing money at a problem that gives you an insignificant return on investment.
Maybe one day, enough lemmings will fill the trench, and people can cross over to the promised Nirvana.-some wiseguy.
Based on what I have read so far in this thread, many people severely underestimate the effect of the room (on recording AND playback) even when presenting evidence of how much it WILL change the sound talking about echo and youtube videos. If you don't find creative ways to deal with the room, everything else is throwing money at a problem that gives you an insignificant return on investment.
Dang about time you came back here.
While I'm *talking* about building audio projects you're actually *doing* it!
Love the BMS coax project you have in your car.
In my 25 years of working in this industry, I've listened to thousands of cars and I can count the ones that sounded great on two hands. Most of them have serious problems and some of the worst ones are IASCA winners... 50% of these guys have plans to change all the equipment in their cars because they don't sound good. In every case so far, none of the equipment has been the cause of poor performance. In every case, it's the installation, the adjustments or the system design. AW
Based on what I have read so far in this thread, many people severely underestimate the effect of the room (on recording AND playback) even when presenting evidence of how much it WILL change the sound talking about echo and youtube videos. If you don't find creative ways to deal with the room, everything else is throwing money at a problem that gives you an insignificant return on investment.
Yeah Bro That is very true! I always remember the principle "Acoutic First Then Audio". Especially in car audio environment with lots of obstacles (i.e diffraction, reflection, absorbtion, diffusion etc) in order to sound right and accurate
This is one of the things that's really vexed me about horns; the only practical way to get them to 'sync up' with a midrange is to use on that's the same size. I seriously do not know of any other solution besides that. EQ can't fix it, DSP can't fix it. They just have to match, physically.
Even on the Thiel, you can see a shallow waveguide on the tweeter, designed to 'mate' the woofer and the tweeter.
So I think frequency response is important, but power response is a pretty big deal too. A lot of these 'audio classics' have good power response in the midrange. This could be due to any number of things. Thiel used a gentle waveguide on the tweeter to match the power response to the midrange. Magnepan just uses a ginormous midrange, so large that the crossover is waaaaay lower than normal. Same story with Quad and Soundlab.
Here's the thing that I want to explore in this thread. Does phase matter? That's the crux of this thread. (Leave it to me to post two pages of nonsense before getting to the point.)
KEF REFERENCE 207/2's - YouTube
Now compare the Thiel with a Kef 207/2. To me, the Kef sounds a lot like the JBL, and the Linkwitz Orion.
Now what do the JBL, KEF, and Linkwitz have in common? You guessed it - high order crossovers.
Here's a quick synopsis of this thread so far:
#1 - I've built a lot of horns, both for the car and the home. I have a bit of a bias against 'regular' loudspeakers
#2 - At this years CES, I heard a speaker that did a lot of the things that I like about horns. Basically it had a midrange that didn't sound sibilant or nasal, like a lot of regular speakers do.
#3 - After talking to the designer (Herron Audio), it got me interested in doing something similar for the car.
The 'trick' seems to be phase-coherency. So this thread will explore that.
I don't have any real info on the Herron speaker, but the Thiel CS3.6 is a well docuemented loudspeaker that's known to be phase coherent. So let's reverse engineer it.
Here's a picture of the midrange and tweeter. They're both from Vifa. Less than $100 worth of parts right there.
A little bit of a waveguide on the tweeter. If you play with A.R.P.E., you'll see that this waveguide is actually functional. It helps to reduce the interaction between the tweeter and the woofer in the octave from 3000hz to 6000hz.
Here's the simulated response of the two drivers. You may notice a couple of things. First, the tweeter is beaming quite a bit above 10khz. (the red line is 45 degrees off axis, peach is 15 degrees.)
But the overall response is pretty darn good I'd say. I think this is attributable to a couple things. First, the phase behavior is quite well behaved with these low order filters. The yellow line in this graph is the phase response, and you can see it's very well behaved. It fits in a sixty degree window. At the crossover point, that translates into less than 0.75" of delay. So basically, this is pretty close to a point source, if it's aimed carefully.
The devil is in the details though, and aiming becomes a pretty big deal with these types of crossovers
This is stereophile's measurement of 'the real deal' (Thiel CS3.6, from sixteen years ago.)
Aaaaaand here's my sim of the VERTICAL response of this loudspeaker. See that giant suckout in the midrange? That's due to these funky 1st order crossovers. Because both drivers are covering a great deal of the same octaves, aiming becomes very critical. For instance, the crossover point is at 3khz, so both drivers are interacting significantly in the two octaves from 1500hz to 6000hz.
IMHO, the Stereophile measurement shows a giant suckout because the microphone was a bit above or below the axis where the speaker would measure best. Basically, if you move your head up or down a few inches, the sound will change significantly.
That kinda sucks for home stereo, but is a bit more tolerable in a car, where we know exactly where everyone will be sitting.
Here's the frequency response of the speaker, but substituting fourth-order Butterworth for 1st order.
See how the frequency response is MUCH better off axis?
That's (one of) the great things about high order crossovers. There's less driver interaction.
But check out the phase - the two drivers are waaaaaaaay out of sync.
You can see this in a waterfall plot, if you compare something like a Dynaudio or a Thiel to a Focal. The Focal will have a delay between the midrange and the tweeter, due to the high order crossover.
There's about a bazillion things I skated over here - this article would be a good one to read if anyone's curious about investigating it further. It's by John Kreskovsky, who's also one of the authors of the ARPE program that I used to do these sims.
In a post coming soon, I'll start to explore actual parts that we can buy to implement a front stage that's phase coherent.
In my 25 years of working in this industry, I've listened to thousands of cars and I can count the ones that sounded great on two hands. Most of them have serious problems and some of the worst ones are IASCA winners... 50% of these guys have plans to change all the equipment in their cars because they don't sound good. In every case so far, none of the equipment has been the cause of poor performance. In every case, it's the installation, the adjustments or the system design. AW
Here's Stereophile's measurements of the horizontal polar response and the vertical polar response of the Thiel CS3.6, respectively.
A few things to note:
My sims aren't an exact match, but they're in the ballpark. The real speaker is complicated by the driver's natural rolloff, the rounded shape of the baffle, and the small waveguide on the tweeter.
The vertical polar response is horrendous. A small MTM instead of a TM would yield some improvements. Thiel went with a coax, which is a far superior, albeit expensive solution.
If you were clever, you might use the awful vertical polars to your advantage, by 'aiming' the lobe in way that off-axis sound is reduced. Aiming these would definitely require more care than a 4th order design.
I'm still learning about phase here. The ARPE program helps a lot. One thing I noticed is that even a 1st order crossover has some phase rotation. Based on what I read, this is due to the driver's resonance, and the natural rolloff of the driver. In other words, if you REALLY wanted to have perfect phase, you'd have to crossover at a point that's well above the resonance, or use DSP to shape the response.
This might explain why the JBL Control Now array, that I posted a video of, rings so bad. Besides using a high order xover on the tweeter, the horn will increase the rate of rolloff. So you might end up with something on the order of 5th or 6th order rolloff, which would entail a great deal of phase rotation at the xover point - which is right in the midrange frequencies, where it's most obnoxious.
In my 25 years of working in this industry, I've listened to thousands of cars and I can count the ones that sounded great on two hands. Most of them have serious problems and some of the worst ones are IASCA winners... 50% of these guys have plans to change all the equipment in their cars because they don't sound good. In every case so far, none of the equipment has been the cause of poor performance. In every case, it's the installation, the adjustments or the system design. AW
-Speakers are spaced away from the back wall. Pass
-Left speaker near sidewall, right speaker near open space. Fail
-Stereo recorded with a monophonic mic. Monophonic mics usually fail miserably in dealing with reflections. This however does not mean monophonic cannot capture the required information, it just means you need to be careful.
-Speakers against back wall. Fails stereo requirement
-Right speaker near sidewall, left near open space. Right fails stereo requirement.
-Monophonic recording of a highly reverberant space.
Quote:
KEF REFERENCE 207/2's - YouTube
Now compare the Thiel with a Kef 207/2. To me, the Kef sounds a lot like the JBL, and the Linkwitz Orion.
It's interesting that both Kef and Thiel are on the 'no diffraction bandwagon.' So this sibilance I'm hearing likely isn't an enclosure thing. It's also consistent with what I've found in my own experiments. Reducing diffraction improves imaging, but sibilance and articulation is a time-domain problem, not a frequency response problem.
Seems pointless to worry about diffraction if you are also not worrying about nearby surfaces.
Third Critique:
-Left speaker near back wall, the right floating. Might be saved by the horn's ability to focus the direct sound resulting in a better direct to reflect ratio, but still a horrible room mismatch
-Looks like the right might also be further from the sidewall.
-monophonic recording with highly reflective room again.
Quote:
Originally Posted by jainbaby
You can just tell that this sounds amazing in person...
"Accuracy" is a mythical creature. Who defines this "accuracy"?
It's easier to try to achieve dynamics..something a good system should be capable of. It will sound more believable. I think most people would chose something that sounds "louder" and doesn't really sound stressed in the process. These puny systems have a hard time sounding believable. Been there done that. A nice big 3 way can easily achieve dynamics and have good power response. This stuff doesn't have to be hard, people make it hard.
Maybe one day, enough lemmings will fill the trench, and people can cross over to the promised Nirvana.-some wiseguy.
did Hendrix sweat the exact driver/s in his amp/s when he was mid show?
did tina turner check thd stats before going on stage?
did bob marley insist on rear ported cabs only??
no,
why,
because its not about 'the system' .......
music is emotion not science
I don't think any of those had the mental capacity to comprehend this stuff. Plus ...they were about 'entertaining' ...not sounding 'great'. Sounding 'good enough' is all they cared about.
>^..^<
If money is the root of all evil, why do churches beg for it?
Give me liberty ...or I'll get up and get it myself.
An armed man is a citizen ...an unarmed man is a subject.
I don't think any of those had the mental capacity to comprehend this stuff. Plus ...they were about 'entertaining' ...not sounding 'great'. Sounding 'good enough' is all they cared about.
>^..^<
better to keep your mouth shut and let people think you are stupid then open it and remove all doubt.
It's easier to try to achieve dynamics..something a good system should be capable of. It will sound more believable. I think most people would chose something that sounds "louder" and doesn't really sound stressed in the process. These puny systems have a hard time sounding believable. Been there done that. A nice big 3 way can easily achieve dynamics and have good power response. This stuff doesn't have to be hard, people make it hard.
"If all you have is a hammer, everything looks like a nail."
The reason I announced my bias in the very first post, is that I generally veer towards projects which have anything but mass appeal. After going ambio with my home system, the conventional stereo triangle seems a bit antiquated.
In fact, a few weeks ago I started a thread where I basically offered to design some HLCDs and give 'em back "free" to the forum, but the thread didn't get a single reply!
So I'm definitely a fan of dynamics and high efficiency and ambio, but I'm not sure if there's a big audience for those types of systems.
This project is definitely an attempt to explore something a little bit more conventional than my Unity projects. (Although the engineering is quite similar.)
In my 25 years of working in this industry, I've listened to thousands of cars and I can count the ones that sounded great on two hands. Most of them have serious problems and some of the worst ones are IASCA winners... 50% of these guys have plans to change all the equipment in their cars because they don't sound good. In every case so far, none of the equipment has been the cause of poor performance. In every case, it's the installation, the adjustments or the system design. AW
I've considered a bunch of options for this project; everything from conventional two ways to horns.
I'm leaning towards a three way, using a faux-axial midrange/tweeter, along with a midbass. Very similar to the Thiel PCS.
The main advantage of pairing a faux-axial with a midbass is that we can keep our crossover points low. For instance, let's say I wanted to do a 3000hz first order crossover from a tweeter to my midrange. If I wanted to keep the center-to-center spacing down to one-quarter wavelength, the tweeter and the midrange would have to be 1 1/8" apart (!!!)
So that's the beauty of the faux axial - it lets us achieve an ultra-tight spacing between tweeter, midrange, and midbass.
You can see this really easily in the measurements of the various Thiel speakers.
Here's the horizontal polar response of a Thiel 3.6, which uses three woofers. The dips and raggedness is due to the shallow xover slope and wide spacing. We deal with this in the car, due to the spacing between tweeter and midrange, but it's far more severe for us. This is probably one of the reasons people have had good results with the Founteks. (they're virtually immune to this.)
The Thiel 2.3 improves upon the old speaker's polar response, mostly due to the use of Thiel's faux-axial midrange/tweeter
By moving the midbass much closer to the faux-axial, the Thiel PCS achieves the best polar response of all three speakers.
If anyone's curious - here's why the PCS measures so much better: Even though the 2.3 and the PCS use the same faux-axial, the center to center spacing on the PCS is much tighter. The smaller woofer in the PCS starts to beam about 2khz, AND it's closer to the faux axial. So the 6.5" in the PCS blends a lot more seamlessly, and the measurements prove it. The big woofer in the 2.3 will play deeper, but the measurements demonstrate that it's not as well matched, particularly in the octave from 500hz to 1khz. Ideally, the polar plot would be a flat line that descended smoothly from zero to -24dB.
Here's some 'shop talk' about the faux-axial in these speakers:
Some of the reviewers complained that the top octave was too hot. This is actually unavoidable; I'll explain why in a future post. In a lot of ways, the extra energy in the top octave is a BONUS if you listen off-axis. (As this project will be designed to do; it's going to be cross fired.)
In my 25 years of working in this industry, I've listened to thousands of cars and I can count the ones that sounded great on two hands. Most of them have serious problems and some of the worst ones are IASCA winners... 50% of these guys have plans to change all the equipment in their cars because they don't sound good. In every case so far, none of the equipment has been the cause of poor performance. In every case, it's the installation, the adjustments or the system design. AW
Last edited by Patrick Bateman; 03-14-2012 at 01:56 AM..
"If all you have is a hammer, everything looks like a nail."
The reason I announced my bias in the very first post, is that I generally veer towards projects which have anything but mass appeal. After going ambio with my home system, the conventional stereo triangle seems a bit antiquated.
In fact, a few weeks ago I started a thread where I basically offered to design some HLCDs and give 'em back "free" to the forum, but the thread didn't get a single reply!
So I'm definitely a fan of dynamics and high efficiency and ambio, but I'm not sure if there's a big audience for those types of systems.
This project is definitely an attempt to explore something a little bit more conventional than my Unity projects. (Although the engineering is quite similar.)
There is an audience but many people have just never heard such a system. The other problem is that a system that uses these types of drivers has to be relatively sophisticated, which many people can't or won't spend their money on.
"If all you have is a hammer, everything looks like a nail."
The reason I announced my bias in the very first post, is that I generally veer towards projects which have anything but mass appeal. After going ambio with my home system, the conventional stereo triangle seems a bit antiquated.
In fact, a few weeks ago I started a thread where I basically offered to design some HLCDs and give 'em back "free" to the forum, but the thread didn't get a single reply!
So I'm definitely a fan of dynamics and high efficiency and ambio, but I'm not sure if there's a big audience for those types of systems.
This project is definitely an attempt to explore something a little bit more conventional than my Unity projects. (Although the engineering is quite similar.)
I dont know HOW I missed THIS one. Please point me back to it.
Basically my idea was to design a set of HLCDs, come up with a crossover, but detail the entire setup so that anyone else could duplicate it. I mean, how cool would it be if we had a series of 'open source' stereo recipes? For instance, what if Harry Kimura published *everything* he knew about his setup? The crossover points, measurements, frequency response, driver selection, amplifier, processing, etc? Basically laid it all out, so anyone could duplicate it.
My inspiration was the Econowave.
In my 25 years of working in this industry, I've listened to thousands of cars and I can count the ones that sounded great on two hands. Most of them have serious problems and some of the worst ones are IASCA winners... 50% of these guys have plans to change all the equipment in their cars because they don't sound good. In every case so far, none of the equipment has been the cause of poor performance. In every case, it's the installation, the adjustments or the system design. AW
Basically my idea was to design a set of HLCDs, come up with a crossover, but detail the entire setup so that anyone else could duplicate it. I mean, how cool would it be if we had a series of 'open source' stereo recipes? For instance, what if Harry Kimura published *everything* he knew about his setup? The crossover points, measurements, frequency response, driver selection, amplifier, processing, etc? Basically laid it all out, so anyone could duplicate it.
My inspiration was the Econowave.
my car is sitting wide open and I sent a PM asking you if you had any ideas you'd like to try out as Id be more than willing to give it a go, you didnt mention the HLCD's at that time, you said keep my eye out for this thread, just sayin
Basically my idea was to design a set of HLCDs, come up with a crossover, but detail the entire setup so that anyone else could duplicate it. I mean, how cool would it be if we had a series of 'open source' stereo recipes? For instance, what if Harry Kimura published *everything* he knew about his setup? The crossover points, measurements, frequency response, driver selection, amplifier, processing, etc? Basically laid it all out, so anyone could duplicate it.
My inspiration was the Econowave.
Being in the software world, open source really gets my attention Like a lot of free information, I think a lot of people would be reluctant to give it out, even though there are forums like this one which thrive on sharing of knowledge.
The perception being that they have spent a ton of money and/or time getting to a certain point (especially competition vehicles) and don't want to give away all their secrets. The flip side being that most people wouldn't spend the money or time anyway but would still like to know _how_ it was done, especially the tuning aspect.
Personally I'm all for it the idea. Having heard Harry's car, Richard Clark's GN and a few other HLCD equipped vehicles in the past I've always wanted to do horns and other semi-crazy installs. I even have to resist the temptation to find a vehicle that I could stuff a unity horn into. yea, thanks
Patrick/Jason- Have you heard the DE5? What is the smallest horn you can use with it that will go down to 3000-3500Hz? How does it sound compared to the AMT's?
Just literally got it 3 hours a go...right as I was walking out the door to go to work. US speaker and Pro Sound Service haven't sold many of them. Ideally I wanted the DE7 but there are none here in the US.
The horns I intend to put them on are small small unity-styled horns...like 4x6 rectangle to go where my planars are at in the upper parts of my front door.
I just got off the phone with USSpeakers asking about horns for the DE5---he has got nothing.
I then asked about a tweeter that was comparable to the Beyma CP-21 (slot tweeterO which I use at home and am very familiar with. He suggested that the Ciare 1.26Nd (neo mag-103db) conventional style tweeter was comparable to the CP-22 (bullet style horn) and the many people had replaced the 22 because they sound better! He also said it was comparable (but not quite as good) as the Beyma AMT which is badass and pushing $400.
3 1/2" dia. and 1.5" deep...Ever heard it or heard of it?
The Beyma is about as badassed as they come...the home guys say it's second to only the RAAL ribbon, but can go much lower. They were my first choice, but coming off of the hip for 3 of them with the wife being pregnant again, well, is impossible to justify right now. I am going to have about $250 (in speakers) wrapped up in the 3 of my micro-unities when I get them done. If I had more room to work with (which I might, but playing it safe for now), I could get the cost down A LOT more using the Goldwood tweeters over the Misco tweeters. I am trying to document this over on Patrick's site as I learn and go through it all.
id bet $50, in a darkened room, if we set up a curtain and some sound gear with gains/volumes/etc matched closely, and played a few songs,
very few (if any) would be able to determine which speaker design was playing repeatedly and accurately.
richard clark pwnt the whole world with this theory and amps,
id bet exact speaker design is about as hard to identify accurately, mayb e a little easier then an amp, but good speakers sound good, how you gonna tell them apart?
what does that mean?
it means spending all this time designing and building the 'perfect' speaker will actually leave you with a speaker that sounds almost the same as many other speakers that have existed for a long time.
Patrick/Jason- Have you heard the DE5? What is the smallest horn you can use with it that will go down to 3000-3500Hz? How does it sound compared to the AMT's?
If you want small, and you're willing to do some experimentation, I think I stumbled on a way to horn load the BG Neo 8.
But first a funny story -
I'm one of those guys who will totally nitpick your PA system. When I go into a club, if the PA sucks, I will totally make a mental note of it. This is actually one of the ways I first discovered the Unities - I was at Club RA in Las Vegas, and noticed that the club's PA system was clear as a bell, I checked out the speakers, and they were SPL TD-1s.
So I was at the Hoover Dam a couple years back, and noticed that their PA is killer too, and guess what it uses? BG NEO 3.
It's amazing the speakers can take the abuse - this PA is in the biggest room in the whole place, you could land an airplane in that room.
Anyways, over at diyaudio, in the Unity thread that Puggie started, I'm exploring the idea of using one driver at the apex of the horn instead of five. It's basically the same idea as what Thiel and Genesis did with their "faux-axial" drivers that a few people here bought. You basically isolate the dustcap of the woofer from the rest of the cone, and by doing that, it acts like a tweeter at high frequency, not a woofer.
But I think you could do the same thing with a NEO 8. You would use felt or close cell foam to partition the ribbon into quadrants, and then horn load the partitions seperately.
Basically turn the Neo 8 into a ring radiator, using felt or foam to subdivide it into rings
In my 25 years of working in this industry, I've listened to thousands of cars and I can count the ones that sounded great on two hands. Most of them have serious problems and some of the worst ones are IASCA winners... 50% of these guys have plans to change all the equipment in their cars because they don't sound good. In every case so far, none of the equipment has been the cause of poor performance. In every case, it's the installation, the adjustments or the system design. AW
Re: A Bateman Project That Isn't Patently Insane
Re: A Bateman Project That Isn't Patently Insane
Re: A Bateman Project That Isn't Patently Insane
Linear Mode
