Does anyone know the common peaks and dips in our cars. I'll be using a test disk and frequency analyzer but would like to know what frequencies to zero in on. Thanks

 

This question has been asked/answered on www.fbodyaudio.com I think.

 

Yep. Great info over there.

 

Sould just be able to use a sin sweep with all high bits to find this out. Are you trying to compete SQ or you just want it for your own listening pleasure? If its the ladder, I say just tune it however your ears like it the most

 

Just for my own pleasure. I'll probably only test the transition between the highs and mids to set the db level for the tweets on the crossover. After that I'll probably just go by ear. I'm thinking the a small boost at 1k is all I'll need since there is somewhat of a hole there for our cars.

 

How do you guys normally do this testing? I found some free spectrum analyzer software and used a good quality microphone in the car (not calibrated though) to play a known good white or pink noise file through the stereo. I found a number of peaks and valleys. There was a huge roll off after 12-14 kHz which is either the mic or the speakers. I tried calibrating the mic by looking at the spectral content of me coughing abruptly (don't laugh) which provided quite a flat response all the way up to 14-15 kHz. A true impulse would have a flat response, I understand people use starter pistols for this purpose.

I have also tried playing discrete sine waves at different frequencies and using a Radio Shack sound level meter to measure the levels.

What do you guys do?

I could post some images if that would help.

 

I'll be using a test disk and sound meter. Basically to make sure the tweeters are set right on the crossover, and to possibly fix the dip at 1k. Other than that I don't have a good enough crossover to do much else. I'll match the sub by ear.

 

I had used the sound level meter previously but its such a slow process! The spectrum analyzer is instantaneous, every time to adjust something you see it all at once. I just put on hearing protection and crank up the white noise.

I put in a Sony head unit that I had from another car and retained the Monsoon amp.

The upper graph is with the Monsoon speakers. The lower graph is with Infinity 6002i speakers. As you can see, due to the strange impedance expectations of the stock amp the Infinity's sounded like crap with no bass at all. Even adjusting the head unit equalizer couldn't fix the problem. Keep in mind the volume levels might have been different so the absolute scale doesn't matter.

The stock speakers sound really good with this head unit as long as I do not crank it up very far. If I do the bass sounds like the speaker cone is folding in half. So then I used the head units high pass filters to keep the bass from the stock speakers and added a sub (not part of the graph). Now I can go louder than I ever care to do normally and it all sounds really good.

I also tried the Infinitys straight off the head unit and they still sounded crappy so they went back to the store. I don't understand why they sounded so good in the store but so crappy in my car.

 

That's funny as hell. The Monsoon speakers seem to have a flatter and more even response than the Infinities, eh? I need to get a nice spectrum analyzer. When I do my addition and home theater I'll be getting one.


















Had to throw in the "eh" My moms side of the family is from up there. Dunn's.

 

When you say get a spectrum analyzer, do you mean a piece of test equipment. At work we have some that are worth $100,000. The piece of software I used was free, I can find a link if you are interested.

By the way, my relatives live in Virginia so I say "y'all" a lot.

 

Like yourself I'll probably get the software and use a laptop. If you can find the link that would be great.

I say y'all a lot myself. I was raised in Texas and dad was from West Virginia. lol

 

Go to this link and download WaveTools. Also download the dll that is also there and place it in the correct directory to make the tools work.

 

Thats some cool programs

I like the sonogram.


Remember that your test results will only be as good as the mic and recording device will allow.

 

woah, ok, you guys are scaring me.

Quick question to you SQ guys, are the stock front speaker locations (door) ok? Or should I get kickpanels? It's just for personal use.

 

The recording device is not a problem since any PC will able able to sample at 44 kHz which allows recording frequencies up to 20 kHz.

The mic does need calibration.

As I mentioned above, a theoretical impulse has infinite amplitude with infinitesimal duration. A spectral analysis of an impulse is perfectly flat across all frequencies. A common way of producing the impulse is a starter pistol. I did find a sharp cough is pretty close. Once you record an impulse, check the shape and you then know the microphones limitations. On mine, part of the roll off at the top end is due to the microphone, particularly above 14 kHz. That means the speakers' responses are flatter than the graph makes it look at the very top end.

 

Awesome, thanks man!

 

Im sure that you understand that EVERYTHING has a resonate frequency. This is especially true in electronics such as sound cards in your computer.

I understand the scope of this test but perhaps you should take a average of three different mics taken with 3 different sound cards to give more accurate results. Then you also have to consider the source such as the factory deck (has a resonate frequency), monsoon 'amp' (has a resonant frequency), and yep…you guessed it even the speakers in the car has its own resonant frequency.

Actually, because im not into competitions ill share with you a way that I won so many trophies.

Your car has its own natural resonant frequency. A frequency that it reproduces louder then any other naturally. Find this frequency then select a sub that is close to that frequency, tune the port 180 degrees out of faze on that same frequency, provide the subs with PLENTY of power. And youll take home many...many trophies.

 

Yes everything resonates, in my day job I design instruments for space craft, I model structural vibrations and interactions, I fight the inherent limitations of materials every day. The dips and valleys in the response curve are as much a part of the car body and interior flavouring the results as it is the stereo itself.

Using 3 different sound cards will just give you 3 uncorrelated measurements. A suggestion might be to measure the audio performance several times with the microphone in different places. Since high frequency sound is greatly affected by localized interactions with the objects in the interior, I tend to hold the mic near my head. I would always be affecting the interior by being in the car anyway, and I am tuning the stereo for my ears. If you calibrate the mic (coughing, pistol, etc.) you subtract the mic/recording device characteristics from the measurements you make. If the calibration shows a huge hump or dip in the middle you know its not the stereo. Now everything is accounted for. You do not need expensive equipment to do this, you just have to take careful measurements. One suggestion is to try and find response characteristics of the microphone you want to use. If that is not possible, I like to use a condenser style mic that uses a battery because the moving diaphragm is extremely small and light and therefore is able to respond to higher frequencies. Unpowered mics use the sound energy to power them and tend to give up at higher frequencies.

You mentioned resonances in the electronics, they definitely occur. Check out your spec sheets, an amplifier will have a value for damping factor from 30-100 for decent equipment. So while you have many resonances in the audible range, they are highly damped and basically inaudible. The digital portion of the head unit will have extremely good response characeristics.

Since all we care about is the final sound there is no need to account for separate components. Besides, the biggest single flaw in the system is the speakers themselves, perhaps only 0.5-1.0% of the electrical power entering the speaker becomes sound. 1W of acoustic power is nearly deafening. The speaker has a tough job to cleanly, across many octaves, absorb electricity and emit sound. That is why speaker selection is done by ear rather than by specification. Speakers have quite large and multiple audible resonances that are not highly damped. The speaker construction builds that into each speaker style. That is what makes some speakers sound "warm" and others "bright", etc. You basicaly are choosing one you like more than choosing a good or bad speaker.

Good suggestion, I do not compete in any way either, I just want my stereo to sound decent.

 

I agree with most of it. However placing the same mic in 3 different locations in the car does not solve the resonant frequency problem. You said that you would get three different results and thats exactly my point.

I think mainly you should tune it to your likings. Unless your just doing this for fun.

 

I guess I did not understand your point. Moving the mic around is just to eliminate a spurious high or low reading.

I'm just doing this for fun.