I think the idea is to make the glass be the first reflection. Also, I think when the speaker is below beaming you hear sound from the cone itself and above beaming the sound you're hearing is the reflection. I could be completely wrong. I have 2.5" midranges in my dash firing up at the glass only playing from 1000-4500. They're crossed nearly an octave below beaming with a 24db slope. Tweeters are playing on axis from 5k and up. This is giving the best results so far.

 

i cant picture there being any pros to this.. yeah you might be able to fix the response a bit with eq, but the reflections would smear the image i would assume. never tried this location in a tuned install though

 

ASSumption is the mother of all screw-ups. Not trying will doom you to failure. Old sayings are still relevantThe tightest imaging I've ever heard had the midranges in the dash firing up at the glass. And yes, I've been around the block a time or two in some of the best daily driver sq cars in the country

 

From personal experience the sound is slowed due to the extra distance the waves travel up the dash and reflecting off the windshield before hitting the listener.

what do you mean sound is slowed? Sound doesn't slow down due to reflection...

 

I think what he meant was "delayed" and not slowed..

 

I'll take a stab at this.

Late reflections aren't too big of a problem, generally they are significantly attenuated and if anything add a bit of ambiance. Very early reflections are too bad either, but I think this is frequency dependent. It's the early reflections that cause the most trouble since they arrive late enough after the direct sound to skew the response, and their amplitude is still high compared to the direct sound.

Windshield angle, and distance to the speaker will play very big roles in determining whether the reflection is early, or very early, acceptable, or unacceptable.

 

All in car reflections are early reflections. The best you can hope for is a low over and close distance from primary source so that the initial impulse combine with the early reflection. That way you get 1/2 space, 1/4 space, ect loading. If either criteria fails, you will get combing of some type.

Your next best option for higher frequencies is to use a wave guide that takes the windshield out of the equation.

Can't say it would never work, but in most cases...

 

^^ correct

 

The effect of early reflections don't cause any significant issue related to timing. The brain takes all location cues from the direct sound and ignores the early reflections. This is why reflections have little to no effect on clarity of imaging. The effects of early reflections are heard in the response domain. Reflections make the incident sound seem louder. So frequencies above ~500 that are prone to reflections are perceived louder, it's one reason we have a downward sloping curve in a car.

Play a 300 hz and 1khz pink noise track both at the same amplitude, the 1 khz will sound much louder. Early reflections make perceived loudness > measured loudness above 500. Something easily cured with an eq. We worry too much about early reflections and attribute a lot of tuning or install issues to reflections.

In a car there are no late reflections.

 

The effect of early reflections don't cause any significant issue related to timing. The brain takes all location cues from the direct sound and ignores the early reflections. This is why reflections have little to no effect on clarity of imaging. The effects of early reflections are heard in the response domain. Reflections make the incident sound seem louder. So frequencies above ~500 that are prone to reflections are perceived louder, it's one reason we have a downward sloping curve in a car.

Play a 300 hz and 1khz pink noise track both at the same amplitude, the 1 khz will sound much louder. Early reflections make perceived loudness > measured loudness above 500. Something easily cured with an eq. We worry too much about early reflections and attribute a lot of tuning or install issues to reflections.

In a car there are no late reflections.

Old post, but I’d disagree with this 100%, the brain can’t calculate the difference between direct and reflected as it’s so close together, late reflections in a room the brain is able to work them out to some extent, early reflections are a fact of life in a car and the brain will hear all of them except the rear window I’d think in a long car

Reflections have a massive influence on imaging in a car, it’s very clear in most cars when you wind the windows down the soundstage shifts... if there are no reflections the brain uses why is that? That throws your statements out of the window straight away

 

going back to Lycan's approach the virtual driver created by the reflection is aimable, the place it is utilized most handily is the vertical expansion of the stage using a horizontal (read: easy) mounting location, and the slight horn loading effect of having the sound radiate to some degree encumbered by dash's baffle/plane in a constrained space.


the windshield rake is important, sometimes you don't want that bang-a-da bang from a long dash and low windshield, since it creates several virtual drivers that compete, AKA dilute the image possible.

with the right install parameters the dash-fired highs can create sumptuous sound, and hardly a DSP is required since the two-pronged manalishi of higher amplitude only sounds is constrained by the midrange which does dive deep into the distance portion, or surround sweepstakes where time-dependency lives, creating a symbiotic relationship that a full-range OEM oval cone can also be adept at, given the windshield bounce is throwing polars that only slightly smear the time-dependency and give most music a fair shake at it, a plus is you are able to throw an image up against the primary viewing target, without it being constricted by various crossover slope/level match issues.

A nice full range driver in the dash with good build quality is mostly throwing 88 db sensitivity and across the range up into the cymbal splash, and with the omni-fire of dash location giving us a homogenized upper end, we don't have the same issues that plague the 2500 hz tweeter insertion, by the by.

 

No but you have a lot more other problems to worry about than that.

What wideband is 88dB and can fit up high on the dash like that?

Jeff's original post was about dash board domes and comb filtering.

 

many 4" cone drivers have 88 db sensitivity in the dash, or at least their manufacturer-provided specifications say that they do.

what problems? I know a dash mat doesn't solve all ills but it can contribute to keeping those reflections down to a mild roar.

I've enjoyed the occasional dash mounted midrange, perhaps a little too much for this decidedly pickier crowd, I don't see as many problems in the approach as others might, just like with kick panel mounted midranges and their penchant for needing more on-axis build molds.


it's all a compromise and I've seen (heard) bad dash sound and good dash sound, not sure what you're tickling?

I'm pretty sure the virtual driver Jeff alluded to was not limited by dome tweeters?

 

When you put a driver near the windshield you get a reflection.

That reflection will set an upper limit on how high the driver will play. Here's the formula:

speed of sound / distance / 4

For instance, if you have a 3" woofer in a midrange pod and it's located six inches forward of the windshield, the upper limit on your midrange pod will be 563hz. (13,500inches per second / 6" / 4)

Now, a lot of people are going to say "I've heard midrange pods that sounded just find. How can you say that the upper limit is 563hz?"


Here's the thing :
The further away from the windshield you are, the less the reflections are going to matter, because the reflections will be attenuated by distance and the reflections will be scattered.

But do a gated measurement of the midrange pod, and you'll see the effect - the response is a complete disaster.


IMHO, there are only two ways to fix this problem:
1) Put the driver so close to the windshield that you "push" the limiting frequency up
2) Use a waveguide so that the high frequencies are directed AWAY from the windshield


BTW, DSP can't fix this. DSP doesn't chage directivity, or alter the laws of physics.

 

How far up the range will having a midrange loading directly off the glass and pillar be pushed? This is how my 2.5" midranges are firing. In my stock dash locations the Doppler at around 8khz has always been a problem that no amount of eq dumping will fix. I'm even thinking about moving my corner loaded and on axis tweeters to the sails to see if that fixes the problem. Same thing with my door locations somewhere below 100hz. Luckily I can somewhat fix the midbass problem by pushing the crossover points up with shockingly good results.

To touch on your bolded part, it's a good thing my seat rails are really long from the factory. Scooted up so my short legs can reach the pedals everything goes to hell. I just deal with it and do all my real listening with the seat back and engine off.

 

Comb filtering is a ***** , it really sucks when you go to time align and all it does is make a Doppler sound when trying and never gets pinpoint . A dash mat is your friend , and adding a small flap like Gary's car to just help give the sound a nudge in the right direction is worth while , especially at power responce of any given driver

 

Love all the input! Learning a whole lot. I have some Auditable Physics that I'm unhappy with in the doors on axis. The thought of mounting in the far corners of the dash firing up into the windshield crossed my mind but wanted some insight first.

 

Thg has weighed in on this but I'm trying a pretty wideband speaker (neo8) sideways on my dash. Will be testing this within the next two weeks. I might also try to shield the glass with a small felt lined baffle. I will first try it without the SEMIT and run it 800hz-20k

 

Not to be bearer of bad news . And I sincerely hope I'm wrong for your install, but I had a set of those neo8s I did a dash mount in my van a while back with them, I hated them firing into glass, maybe won't be bad for you. But yeah it was gay actually, I pointed them right at me and they sounded amazing . They're kinda wierd because they are unusually good with phase coheriance , but that kinda makes them bad for reflection or horn loading or whatever .


Anyway sorry to sound negative , I would try them tho and take a listen before you cut you dash , . But I didn't read all of this post I just saw pic . But ... It will sound really good tho , that setup is kinda cool.

 

"it was gay actually".. im not even sure how to take that in terms of how something sounds

 

Lol

Sent from my SM-G900V using Tapatalk

 

I can't get the right pictures to post I'm going to bed

 

If I could figure out how to post the proper pics from my iPad, I can show an example of reflections DIRECTLY measured during the installation........allowing informed decision making regarding tweeter faceplate orientation (the faceplate had little "waveguides" on them) so depending on which way you turned the tweeters in the mounts you would alter the amount of direct energy reflecting off the side glass and windshield. (The tweeters were in dash pods top of dash, on axis, Jeep Wrangler)

There is an easy super fast way to measure this, so that when its time to tune you won't be trying to EQ out something that should have been fixed in the physical/acoustics domain in the first place.

I'll have to wait till I get to work....

 

OK, here we go.......

 

and here is the same exact EVERYTHING, just with the tweeter waveguides oriented side to side instead of up-down

 

listening test confirmed that the orientation that resulted in a better "coherence" measurement was indeed better. Mainly in the anchoring in of the center image. like, how do I say this.....better defined?

 

I love you !!!

 

oh yes, i should just make clear that this is a SINGLE tweeter being measured

It was a Focal KRX2 tweeter

 

Frequency response measurements alone would have told me NOTHING about how allowing more energy from the tweeter to reflect off the glass changed the sound. Of course, listening tests would, but in my world. Time is money. I can tell you which orientation is gonna be better very quickly.

 

Coherence is a very useful measurement. I also use it to help me decid on crossover freq. because sometimes, you'll have a scenario where a midbass (like in a 2way) might play up pretty high according to the FR measurement, but the coherence drops like a rock past 4K (just an example), so I'll place a cursor there, and that helps me narrow down pornetial crossover freq/slope selections

 

So wouldn't you agree that a flat planar driver that doesn't attenuate energy over distance as much as a cone is going to have much better coheriance because amplitudes will be more equal.

 

So I guess the point I was making sorry was if you have sets of speakers and they have behavioral qualities that I don't lose energy as much as other drivers you install those in a car and you need to delay the left side to get a good image you run your delays and noticed it's right biased because there's no energy loss and there's no need to volume down the left side so at the frequencies where level gives you spatial cues it's making it a different type of tune any conventional driver doesn't mean it can't be tuned it just makes it different

 

How are you measuring phase coherence with only one speaker playing? Not sure I'm clear on what you're measuring.

 

How are you measuring phase coherence with only one speaker jplaying? Not sure I'm clear on what you're measuring.

Its not "phase" coherence.

It's simply called "coherence"

Meyer Sound SIM

Rational Acoustics SMAART

AFMG EASERA

AFMG SysTune

All these measurement systems and more can measure coherence.

From the SysTune manual:

"........The coherency function is a well-defined entity in system theory, for the mathematical expression we refer you to standard text books. Essentially, it measures the correlation between signal and reference, output and input, respectively, of the system under test. This results in a frequency-dependent function with values between 100% and 0%, which represents the level of linearity and constancy of the measured system. There are some fundamental properties worth mentioning with respect to the software implementation........First, its evaluation relies on the set of available FFT blocks. Coherency is always 100% if there is only a single FFT block included with the measurement, which is 1 average. In SysTune for this situation, you may only see values different from 100% if the reference signal is not defined for a particular frequency. For these frequencies the coherency will be 0%.
 Second, coherency accumulates variation and occurrences of nonlinearity over time. That means, the more averages you include in the measurement the lower will be the coherency, as every additional second of measuring time will always pick up a little bit of noise.
 In practice, a coherency value of 100% or close to that will be difficult to achieve. For a small number of averages, normally a coherency of 50% is a good value, for a large number of averages about 25% will usually be sufficient.
 Because coherence is a good measure for the validity of the data acquired, it obviously makes sense to use it as a processing filter for capturing valid data only. This function was implemented as part of the patent-pending SSA filter explained in chapter 5.5 and it is named the coherence filter........."

There are LOTS of acoustic analysis techniques out there that 99% of the car audio world doesn't realize simply because they don't take the time to learn.

 

How many of you guys read Jeff's original post about dash board domes and comb filtering?