DiyMobileAudio.com Car Stereo Forum banner

121 - 140 of 169 Posts

·
Registered
Joined
·
1,301 Posts
so what you're saying is someone familiar with electronic circuits was able to bread-board a solution for his particular car and install, that kicked butt without using a digital delay of any kind?

I can believe that.

that's the premise behind the Linkwitz Pluto and it's response shaping, what you essentially get is time alignment without using a digital conversion, analog all the way.

I would think that if you could avoid the D/A, A/D thing altogether, and the jitter and clock crap, and whatever else buggers the signal you could have a more true, approach.

whether that's audible or not, or just an idea that keeps people up at night, I couldn't say, but that kind of mastery of time and space is worth seeking out.

maybe approach audio with a "nuclear bomb resistant" angle, where all circuits that can be fried on a PCB are swapped out for glass tubes, and everything is in the technological chasm of the 1950's.

How neat would that be?
STEAMPUNK DIYMA....... that would atleast look cool....
 

·
Registered
Joined
·
719 Posts
PLD left to right will cause a time alignment issue left to right.
PLD mid to tweeter will cause a phase issue which will cause a time alignment issue between the two drivers and the frequencies they are playing.
Reversing the wiring will change the phase and may or may not help with time alignment.
Locating drivers the proper distance away from the listener will prevent some need for time alignment.

Will a DSP like MS-8 delay the entire frequency range of each channel the same or will it delay differently across the frequency range depending on how it reaches the listener?
If it delays differently across the range, multiple drivers could be on each channel as long as they covered different frequencies, like a passive component system.
If not, more than one driver on each channel will have to have each located correctly to keep phase and PLD correct to the listener. This would not work well with passive component systems.
 

·
Banned
Joined
·
4,394 Posts
STEAMPUNK DIYMA....... that would atleast look cool....
had to look that up, haha..

but it wouldn't be steampunk, it would be the actual, the original, the analog without the digitized skin..

of course, you'd have to have analog masters, playback source to go along with it.

maybe that's out of the question, but in an off-axis thread, it's all out of the question, haha..
 

·
Registered
Joined
·
4,941 Posts
Here's the separation data for on-axis aiming compared to aiming parallel to the drivers door glass. Keep in mind this is for the left side which is pretty sensitive:




Here's the data for on-axis compared to aiming at the rear view.



These tweeters show no drop in output at 45 degrees off-axis until you get out past about 5k hz, so that may explain a little of what we see in the first graph.

I think this process should be a little cleaner on the left side tweeter and give more distinct results.
Interesting... About 10dB separation at the optimal spot. We should perceive that as a half the volume or so. I'll run some tests in my cars when I got the systems running.
 

·
Banned
Joined
·
4,394 Posts
If they were using all pass filters then they were using a proaudio processor.
Linkwitz has links on his site to the schematics to build your own all pass filter, and it's not a pro audio processor that you are working with?

He builds one into the Pluto's electronic crossover to delay the Whisper's output to the acoustic center of the omni driver, while putting the physical position of the Whisper at the edge of the omni's surround, reducing diffractive distortion.

If I was up to the task I'd make the calculations in a car, then build an all pass filter using the same techniques to delay the tweeter to meet with a door woofer's longer path length.

but alas, I'm not up to that task, and anything requiring large scale changes may be out of the all pass's ability in an analog circuit, to make that adjustment.

Or, in the crossing range of the Pluto, using an all pass to delay everything above 1Khz is practical but trying to do the same thing with mid bass transition, or large path length differences of 10" or more, may not be realistic or practical.
 

·
Moderator
Joined
·
11,394 Posts
I really doubt anyone out there in the competition scene is building analog filters. But the proaudio processors like the Rane and Behringer do have all pass networks. I know who Kelvin is talking about, and he was using a Rane.
 

·
Premium Member
Joined
·
6,214 Posts
Would really like to know what Andy thinks could be our options if we don't plan to have T/A available in the system building...

Could be an interesting read.

I'm all for T/A since I managed to create a wonderful soundstage with rear OEM speakers without having to lower the level + create a golf ball center in my girlfriend's car using horns.

Kelvin
 

·
Registered
Joined
·
719 Posts
If they were using all pass filters then they were using a proaudio processor.
This gets discussed some over on diyaudio.com. An All-Pass can just be a HP or LP out of the frequency range you are listening to. Like a HP @10Hz or a LP @ 40KHz. It won't cut anything you actually hear but the design will still have the phase effect. 1st order, 6db slope, 90 degree phase shift. This is usually done on Home Audio to get the mid/tweeter aligned, not L to R.
 

·
Banned
Joined
·
4,394 Posts
one of the car audio companies has a left/right image shifter that uses an all-pass analog delay, a little box, knob, whatever... who was it, Focal?

and I remember something about a chip, an analog, perhaps Texas Instruments? That was supposed to make adjustments in analog and had great potential, sort of the Raspberry Pi but audio oriented?

anyways, just rambling about, I feel like the complicated method of creating filters for a specific application and putting that into a car's audio, is just like hammering a screw into the wood, when a digital screwdriver gets it done with much less effort and drama.
 

·
Registered
Joined
·
62 Posts
Good points!
Home Theater is more of a captive audience, seated closer to the source. Room reflections are more of the later reflections which don't actually occur in an automotive environment due to the small size. Later reflections are not as undesirable as early.
Exactly

Is there now a Home Theater style/type driver?
Not sure what you mean by this?

The Horizontal MTM or Line Array has really no other purpose except for a center channel under a screen in Home Theater. These still use non-specific home audio style drivers.
A horizontal MTM is a bad choice as a CC in a HT because it promotes vertical dispersion while limiting horizontal dispersion. They're just used for aesthetic reasons.


I would also think a highly efficient tweeter aimed directly at the listener would/should/could also not need as much power nor be as loud which minimizes the energy available to be reflected in the first place.
I would think so as well, but according to mojozoom's graphs, it seems like aiming the left tweeter parallel to the drivers door is the way to go. Which isn't that bad, but may cause a minor issue with narrow directivity horn tweeter we're talking about.

Kind of combining the two thoughts we have going on here, in an attempt to limit horizontal dispersion, how about a horizontal MTM on the dash?
You've got limited horizontal dispersion (good), pretty much the same vertical dispersion as a horizontal MT (ok), and better efficiency (good).
The only downside is packaging - most people won't be able to fit 5-1/4" drivers. However, an MTM with Aura NS3's and a 1" smaller faceplate tweeter should fit fine....
 

·
Registered
Joined
·
62 Posts
Here's the separation data for on-axis aiming compared to aiming parallel to the drivers door glass. Keep in mind this is for the left side which is pretty sensitive:




Here's the data for on-axis compared to aiming at the rear view.



These tweeters show no drop in output at 45 degrees off-axis until you get out past about 5k hz, so that may explain a little of what we see in the first graph.

I think this process should be a little cleaner on the left side tweeter and give more distinct results.
Sorry if I'm a bit behind the times on this one, but are you only using the left tweeter as your source, with L and R microphones?
 

·
Registered
Joined
·
148 Posts
There is a huge misunderstanding here about on and off axis. For speakers used within their piston range, on axis and off axis produce the same response. For example, below 300Hz for a 6" midbass, there is no real difference between the on axis and the off axis response between 0 and 90 degrees. The same is true for a 3" midrange below 3kHz. Those are pretty standard crossover points. We use the tweeter all the way up to 20k, so we are forced to use it in a range where its dispersion narrows.

What this means is that AIMING IS UNNECESSARY IF YOU CHOOSE CROSSOVER POINTS IN RANGES WHERE THE SPEAKER'S DISPERSION IS WIDE. Aiming of the tweeter may be helpful.

The second big misunderstanding here is the idea that there are early and late reflections in the car. While reflections from the dash and windshield are shorter than reflections from the back of the car, they are ALL too short to be perceived as stage and room. Think about it. IF you go to a concert at a venue like radio city music hall, the ENTIRE car could be parked on the stage. While the early and late reflections from the concert space might be included in the recording, they CANNOT BE REPRODUCED realistically by two speakers mounted in the car. If they are included in the recording, it's like the entire space has been reduced to the space inside the car. If you want to attempt to reproduce the larger space, you have to have a way to create the late reflections and the early reflections of the larger space. Since some of them come from the back, you need rear speakers with lots of delay that reproduce ONLY that ambient information. Adding more speakers with appropriate delay and steering of the ambient information contained in the recording can create a better reproduction of the original space, but only if the ambient information is included in the recording.

A room synthesis algorithm adds an arbitrary space to the sound of the recording--like changing the size of the car to the size of some other listening room in which the two stereo loudspeakers are placed.
So if we were to apply this to an example from an earlier post about the BMW with the woofer under the seat, if it's crossed below 300Hz, aside from the fact that there is a big physical obstruction (your ass) on top of the driver, would/could it sound and blend as good with the 3" mid high on the door and tweet in the sail compared to a woofer that was, say, on the bottom of the door panel?

There is so much blah blah blah; the theoretical and real world are so different sometimes. It's hard to sort out, and prioritize factors for installs.

LOL; just beat your head against a wall. Less frustrating sometimes.
 

·
Premium Member
Joined
·
6,214 Posts
This idea that the correct polarity is the one that sounds good is a little silly. If reversing the polarity sounds better, then I think more investigation is required. If it's an out of phase issue between left and right due to a distance in difference that you're correcting with a polarity change, then delay is a better fix. Reversing the polarity just exchanges one problem for another. Swapping the polarity of BOTH right and left due to crossover phase shift (i.e. 12dB/octave butterworth filters or distance between mids and tweeters) is a completely different exercise.

When I hear things like "you have to experiment to find out what works best", I cringe. When I hear, "you gotta play with the settings", i throw up in my mouth a little bit.

:)
Would like to ask again:

For people that don't have access to T/A, what would you suggest then?

Have access to a L/R EQ? Install drivers so that PLD is as small as possible? Swap polarity of 1 mid anyway?

Kelvin
Andy? Care to help me? :)

Thanks,
Kelvin
 

·
Premium Member
Joined
·
4,235 Posts
Discussion Starter #136
Well, if you don't have access to TA, then choosing speaker locations that minimize the need for TA is the next best thing. A center channel and an upmixer are also helpful and minimize the need for TA, since the center image doesn't depend on a perfect acoustic sum from right and left.

IN reading this thread, I noticed some blowback about tweeter response. When you measure a tweeter, the baffle is a big deal. The height of the dome creates a dip in the response. If you're mounting the tweeter in a box or on a flat baffle, you'll see that dip. In an A-pillar or sail panel, the dip will go away or at least be very different. To get an idea of what the response of the tweeter is, measure it without a baffle. A simple ground plane measurement from 2M is similar enough to anechoic that you'll at least be able to see what's up before bashing speaker manufaturers. In the graph of the Hertz tweeter above, those two peaks are probably the result of a little ringing, and a highQ alignment and ALSO that dip.
 

·
Premium Member
Joined
·
6,214 Posts
Well, if you don't have access to TA, then choosing speaker locations that minimize the need for TA is the next best thing. A center channel and an upmixer are also helpful and minimize the need for TA, since the center image doesn't depend on a perfect acoustic sum from right and left.

IN reading this thread, I noticed some blowback about tweeter response. When you measure a tweeter, the baffle is a big deal. The height of the dome creates a dip in the response. If you're mounting the tweeter in a box or on a flat baffle, you'll see that dip. In an A-pillar or sail panel, the dip will go away or at least be very different. To get an idea of what the response of the tweeter is, measure it without a baffle. A simple ground plane measurement from 2M is similar enough to anechoic that you'll at least be able to see what's up before bashing speaker manufaturers. In the graph of the Hertz tweeter above, those two peaks are probably the result of a little ringing, and a highQ alignment and ALSO that dip.
Thanks Andy for the reply - not what I was expecting though :p

Kelvin
 

·
Premium Member
Joined
·
4,235 Posts
Discussion Starter #138
Hmmm...what were you expecting? A diatribe on all pass filters? Those are a bit more difficult to do than even a simple delay, which is more effective. You can do an all pass filter as an active circuit--just sum a 12dB/oct HPF and a 12dB/oct LPF.
 

·
Premium Member
Joined
·
6,214 Posts
Hmmm...what were you expecting? A diatribe on all pass filters? Those are a bit more difficult to do than even a simple delay, which is more effective. You can do an all pass filter as an active circuit--just sum a 12dB/oct HPF and a 12dB/oct LPF.
Yeah something fancy that only you have the secret for :D

So doing an all-pass filter adds enough ms to one driver that it can put every frequencies timing back in phase?
How difficult would it be to implement time delay in passive Xover with today's technology? Even in home audio, they are still doing it mechanically.

I know RF has a remote for their passive Xover that can switch phase 180° on MIDs and TWs. But reading your post, it's not nearly the same thing as adding a delay.

Kelvin
 

·
Registered
Joined
·
193 Posts
This is exactly what I have been preaching and someone, today, called my ears to be iron ears cause I am listening to Focal on axis. I think, in short, you're saying, it is almost impossible to have a perfect sound in a car environment. Loud, yes, you can go loud as you can but cannot achieve the perfect sound.
 
121 - 140 of 169 Posts
Top