DiyMobileAudio.com Car Stereo Forum banner

1 - 20 of 170 Posts

·
Premium Member
Joined
·
4,235 Posts
Discussion Starter #1
Re: Is off-axis better for front stage? - Hybrid Audio

Someone really needs to add something to this discussion of on and off axis. OK...I'll do it.

There's no mystery here and the differences between on and off axis placement are simple.

1. Sound is reflected from a bunch of surfaces in the car. We hear those reflections MOSTLY as changes in frequency response to the direect sound. They also contribute significant crosstalk (left ear hears right information and right ear hears left information). That crosstalk reduces the system's ability to sound "larger than the car".

2. The difference between on and off-axis sound from your speakers is very simply defined by the diameter of the cone. Frequencies that have wavelengths that are long compared to the diameter of the cone are radiated everywhere--to the front, the side and the back. Frequencies that are short relative to the diameter are radiated frontward.

2b. the high frequency response of almost all speakers has a peak. This is caused by cone distortion-a small part of the cone moves differently than the rest of the cone. That peak is "played" by a smaller part of the diaphragm, so it has a different dispersion than the rest of the cone. If you look at the off axis response of such a speaker, you'll see that the peak is not attenuated at the same rate as the rest of the high frequency response. This can be a HUGE problem and it's one of the criteria for choosing a crossover point.

3. If you mount a speaker on axis, the high frequencies will arrive at your ears direcetly and the lower frequencies will too, but they will ALSO arrive at your ears after being reflected off ADJACENT boundaries. The high frequencies will also be reflected by boundaries near your ears (side glass) but NOT from adjacent boundaries (because they are radiated FORWARD).

4. Reflections are attenuated because of the increase in pathlength (6dB for every doubling of distance). They are also modified by the material. Carpet doesn't reflect super high frequencies but glass does, for example.

5. So...a tweeter mounted off axis will direct more high frequencies into the reflecting surfaces and less into the direct sound. The same thing happens with a midrange (or a wide-bander).

6. When you equalize, you can't equalize the reflection differently than the direct sound from the speaker. That doesn't matter if your head is completely stationary because for every point in space, we hear the sum of the direct and reflected sound. When we move our head, we hear a different combiination. It's helpful to have the direct and reflected sound as similar as possible to be effectively equalized. This isn't possible in a car, so no matter how we mount speakers we'll still be confronted with this problem.

6. The problem is the worst when the off-axis response has a big hole and the on-axis response has a peak. This is common with 6-1/2" conponent systems where the woofer doesn't quite reach the tweeter. Sometimes, crossover designers who are focused on the on-axis response and ignore the directivity (off axis response) build a high-Q crossover to boost the response at the crossover. This makes this problem worse and can make the car sound bad when you listen despite a good curve.

All of this is why I suggest mounting tweeters on axis, using a small mid between a 6 and a tweeter, and keeping speakers away from the junction of the dash, side window and windshield.

Widebanders should ALWAYS be mounted on-axis if no tweeter will be used.
 

·
Registered
Joined
·
439 Posts
How would you go about adding a midrange to an existing two way component set. My component set has midrange/bass in factory position; front lower half of door, and my tweeters are mounted high on the door panel above the mid about 8 to 10 inches above my mid/bass. My tweeter is a dome type. I have thought about moving my tweeters to my sail panels and using my existing tweeter location for my midrange.

Any advice on how I should be approaching this?

Thanks for the helpful sticky. Very well explained, Sir.
 

·
Registered
Joined
·
22 Posts
You say keep the tweeters away from the junction of dash,windshield, and side window. Where exactly/and how far from this junction do you feel are ideal locations? I'm also wondering how important it is for a small mid range to be mounted on axis, how many degrees off axis should we try not to surpass?
 

·
Premium Member
Joined
·
4,235 Posts
Discussion Starter #6
I guess there are always those few exceptions out there.

OF course one can make a good sounding car by breaking all the rules. I'm just trying to make suggestions that add a measure of predictability and simplify this for as many people as possible. Winning IASCA contests means you win contests, it doesn't necessarily mean it's the best way to build a car that sounds correct (or as close to correct as is possible).
 

·
Premium Member
Joined
·
4,235 Posts
Discussion Starter #8
I think there might be a problem with the wording in the second number 6.

I'm going to assume that the crossover in question is the 6.5" low pass, and the high Q means a slight bump in frequency response at the upper end of the 6.5" response.

If the problem is a peak in the on-axis (combing/break-up), and a drop-off in the off-axis, (narrowed response from beaming/cone diameter function), then building a high Q in the response is going to accentuate the peak, and raise the off-axis response.

If building a high Q in the crossover is actually meant as creating a high Q filter, and is being used to create a notch at the response peak, then it makes sense that the designer is paying attention to the on-axis response.

I guess what that means is the notch will cause the off-axis response to drop further and increase the effect of the 6.5" not meeting up with the tweeter.

But, if building a high Q filter into a crossover doesn't work for the on-axis response because it pushes the off-axis down, is it better to let the peak stay in the on-axis?

And, are we building crossovers with passive components or pushing software sliders?
Well, I suppose there is a problem: there are two number sixes.

My point is that if the woofer begins to roll off lower than the lowest frequency we can count on the tweeter to play, then there's a hole in the response. The hole will be deeper farther off axis. If we boost the output of the woofer with a high Q crossover so that the on-axis response is flat, then the off axis response will still have a hole. When you adjust the response with an EQ, you'll adjust the power response because the car is reflective and then the on-axis response will have a peak and the off axis response will still have a smaller hole.

Doesn't matter whether this is a passive or active network.
 

·
Premium Member
Joined
·
4,235 Posts
Discussion Starter #9
You say keep the tweeters away from the junction of dash,windshield, and side window. Where exactly/and how far from this junction do you feel are ideal locations? I'm also wondering how important it is for a small mid range to be mounted on axis, how many degrees off axis should we try not to surpass?
Tweeters are not as big a problem as midrange speakers in that location.
 

·
Premium Member
Joined
·
6,214 Posts
I don't understand what you're saying to do, to fix this issue.

if we can't add some high-side boost to the woofer to "meet" up with the tweeter, because the hole's still there, what else is there?

are you advocating the use of ellipsoid tweeter faceplates and/or bringing the crossover down to meet the woofer, thereby pushing more sound out of the tweeter but controlling the directivity/dispersion, and leaving the woofer's roll-off alone?

I saw Pioneer (PRS components) bring the crossover down and introduce a low-playing tweeter, but it seems like a lot of people didn't care for it crossing that low. They preferred the higher crossover with active control, letting the woofers play up higher and crossing steeper than the passives.


Now, the use of an ellipsoid horn load, seems like a bit of a tussle when mounting these suckers, where and how is the question?

optimum mounting locations, "tastefully done" implementation of the idea, and accessible vehicles, likely suspects for this kind of install?

because, horns add a bit of directivity and well, directing the output means it would be easier to mount the faceplates on-axis somehow, I haven't seen this done...

has it been done, and unobtrusively?

I did almost take a sledge and body hammer to the wheelwell/dead pedal area, once..
was going to put a cheap compression midrange bolted from the engine compartment, but it seemed a little crazy, so...

anyways, looking for solutions/stopgaps, for dealing with uneven power response in the front of the car.

My install consists of a 5.25" mid and a ribbon tweeter but I'm wanting to cross the ribbon a bit higher than the mid wants to "reach" so any input...
I thought that waveguides (ellipsoid tweeter faceplates) on a tweeter was also to widen dispersion down low in order to better "meet" the midrange around the Xover point...

Don't really know what your question is...
If the off-axis response has a dip, you add a high-Q boost on the midranges upper end to flatten the response up to the tweeter - then you also add some boost to the reflected energy which as Andy stated, isn't really a good thing.
Also, the high-Q boost usually works well with the driver's side (closest to you) - however that same high-Q boost accentuates the peak on the passenger's side (more on-axis than the driver's side)

Kelvin
 

·
Premium Member
Joined
·
4,235 Posts
Discussion Starter #11
I thought that waveguides (ellipsoid tweeter faceplates) on a tweeter was also to widen dispersion down low in order to better "meet" the midrange around the Xover point...

Don't really know what your question is...
If the off-axis response has a dip, you add a high-Q boost on the midranges upper end to flatten the response up to the tweeter - then you also add some boost to the reflected energy which as Andy stated, isn't really a good thing.
Also, the high-Q boost usually works well with the driver's side (closest to you) - however that same high-Q boost accentuates the peak on the passenger's side (more on-axis than the driver's side)

Kelvin
No, in fact, they do just the opposite. They narrow the tweeter's dispersion at low frequencies to hopefully match the narrowing dispersion of the midrange at the crossover point. The result should be a power response with gradually diminishing high frequency response without peaks or dips. The "roundover" at the edge of the waveguide (if it exists) helps to scatter the high frequencies from the tweeter to widen dispersion at those frequencies.

The rest of this is correct.
 

·
Premium Member
Joined
·
4,235 Posts
Discussion Starter #12
Off-axis will almost always have a dip with a component mid/woofer and tweeter combination.

I don't know if Andy is advocating for 3-way implementation or some other format that deals with the off-axis hole/narrow dispersion issue.

and how are you implementing this "high-Q boost" if you don't have parametric equalization and adjustable crossover with level controls?

just want to understand where the traditional fail modes in installs are being corrected, and by what tools/methodology, since it's important enough to create a thread to discuss the observation/occurrence?

As it stands, most higher-end metal and sandwich cones and kevlar based woofers are going to have some amount of snap at their roll-off.


so, going high-tech elevates the region where paper-based drivers smooth out the peak, but we want to enjoy that nice flat region as far out as possible, the issue is that peak is identifiable and noticeable when using active crossovers that do not have extra control, or DSP built-in.

so, a passive can easily have a notch filter that tames the out-of-band response, and that wonderful smooth response from the filter makes it hard to throw out that network when going active, but that's a drift...

back to the issue of beaming and narrowing dispersion, is this really that big of a deal?

Is the multi-element approach consistently able to outperform most 2 way installs?

the hard question for me is integration, the premium sound package available from JBL had the ellipsoid faceplates and I don't remember ever seeing them used, where they were actually same plane, same axis as a recording monitor might be, as a time-aligned active speaker might be.
Same plane. same axis isn't so critical. The GTi waveguides were designed for use at 35 degrees off axis, to facilitate installation in the doors.
 

·
Premium Member
Joined
·
6,214 Posts
how did this happen?

I understand the waveguides aren't off-center, they are symmetrically formed and have no adjustment of any kind, (at least, that's what they look like on an internet screen)

is the passive network adjusted for off-axis, or is the shape of the guide itself, the angle/pattern coverage tested for 35 degrees of uniform distribution, and that's the cut-off point at which one might be able to perform a door installation?

And, using the waveguides in the classic format seems like they would be best installed with their flange flush with the existing door panel, this is hard to achieve high in the door without really chopping up the panels, or maybe there's something I'm missing?

I always liked the clever idea/implementation for using a waveguide but finding them in use has been a bit of a unicorn thing...


maybe hang them from the bottom corner of the dash, like the ID horns but flush to the dash?

it's interesting and it's fitting the universal theme of recording monitors of today with a front baffle that has a smooth integrated guide for tweeter control, but in the car the placement options are so limited, and putting them in the dash, what about then? Is the reflection of a controlled dispersion tweeter off the glass going to coherently produce an image with the mid in the door locale?

is that why same plane/same axis is not as important?

I've never used the faceplates so I am unaware of their success in various install locations..
If you download the 660Gti manual, there's a good explanation about how to use the waveguide - not sure if it answers your questions but it's worth the read :)

Kelvin
 

·
Premium Member
Joined
·
6,214 Posts
Not sure in which thread I should write this so I'm writting it here:
A room that has a significant amount of reflected energy will often be perceived as more "spacious", and have more of a "they are here" characteristic. A monopole with controlled directivity involves the room less and tends to be more of a "You are there" characteristic.
Taken from: On Loudspeaker Directivity Part 2 Article By Jeff Poth

It's exactly what I've experienced with my 2 cars - Conventional cone and dome drivers VS horns

Kelvin
 

·
Premium Member
Joined
·
4,235 Posts
Discussion Starter #17
What kind of a reply are you looking for? I'd agree with the suggestion in theory. the use of a spherical waveguide on a tweeter isn't going to substantially change the perception of ambience, and there's very little opportunity to do any real pattern control in a car over a wide band of frequencies. Acoustic crosstalk is the big problem in cars if "you are there" is the objective. I find that "we're all are somewhere larger than this car" is a pretty good illusion for car audio.
 

·
Premium Member
Joined
·
6,214 Posts
What kind of a reply are you looking for? I'd agree with the suggestion in theory. the use of a spherical waveguide on a tweeter isn't going to substantially change the perception of ambience, and there's very little opportunity to do any real pattern control in a car over a wide band of frequencies. Acoustic crosstalk is the big problem in cars if "you are there" is the objective. I find that "we're all are somewhere larger than this car" is a pretty good illusion for car audio.
It just that you made such a great first post but I've yet to see people jump into this thread to give their point of view, agree with you or contradict some points in your post...

Guess I'm just bored :p

Kelvin
 

·
Registered
Joined
·
2,117 Posts
So are we saying a good place for midrange and tweet is the kick pannel?

Some of us liking both seats sounding good would have to go 35degree for both sides for a compromise.

In regards to center channels, it's pretty much always going to be close to one of those junctions.
 

·
Premium Member
Joined
·
4,235 Posts
Discussion Starter #20
There's a HUGE difference between mounting a small mid in the top of the door and mounting it in the dash or the A-Pillar. Kick panels have a couple of advantages and a couple of disadvantages. One of the advantages is that in some cars, the pathlength differences are reduced enough to provide good imaging in the midrange. The second is that they are FAR away from the junction of the windshield, dash and side window.

The disadvantage is that the shape of the footwell often produces a big cancellation in the mottom part of the midrange. The second disadvantage is that there are plenty of opportunities for obstruction from legs, etc. The third disadvantage is that it's difficult to make an enclosure that's big enough to allow the midbass to have flat response and good bass extension.
 
1 - 20 of 170 Posts
Top