Using an SPL Meter to adjust audio

John is right, but I still think that ANY measurement you take is better than nothing... and I think even the test disc and SPL meter will tell you a lot. You are bypassing a lot of problem areas like sound cards, etc. No the SPL meter and test disc is not perfect... but neither is the interior of your car nor your system. It will, make an improvement and show you the problem areas.

See what CL65 wants for his setup used. Then lowball him and see his response... lol....

I wish I was there and could do this for you.

I will ask my acoustics engineer about those models you mentioned. He knows them pretty well.
 
In my experience, the RS meter + calibration disc and one hour will make a signficant improvement to the sound of your system. Much better than doing it by ear. Buy a test CD. Don't mess with a laptop. You'd have to feed the output of the laptop to your head unit via its AUX input. Your laptop's analog circuits are probably questionable. And then messing with a good microphone....If your speaker has a good tweeter - it should be close to linear above the crossover and your response variations will be mostly due to reflections in the car - and very position sensitive.

You'll eliminate the gross peaks and valleys that probably exist now. It is probably better not to have too many bands of EQ because people will over-adjust.
 
hofffam said:
In my experience, the RS meter + calibration disc and one hour will make a signficant improvement to the sound of your system. Much better than doing it by ear. Buy a test CD. Don't mess with a laptop. You'd have to feed the output of the laptop to your head unit via its AUX input. Your laptop's analog circuits are probably questionable. And then messing with a good microphone....If your speaker has a good tweeter - it should be close to linear above the crossover and your response variations will be mostly due to reflections in the car - and very position sensitive.

You'll eliminate the gross peaks and valleys that probably exist now. It is probably better not to have too many bands of EQ because people will over-adjust.

I thought the laptop was just to run the RTA software, plug in the mic, and analyze sound playing from the stereo? Why would I need to connect to laptop to the stereo?

While I appreciate all the feedback (no pun intended), this is simply crazy guys. This whole thing started with someone recognizing that my Kenwood EQ looked like it had incorrect settings. I have for a long time, wanted to get better sound from my existing equipment. Not that my stuff is top of the line, Alpine, JL, and Kenwood is not that bad, but can be tweaked for improvement, I have no doubt. The accoustics of this car are mind-boggling to me. I am 80% satisfied with the sound, and would like to improve it. I can not grasp how using an SPL meter can make it worth my while. I thought putting my laptop to good use, with a microphone, might be a worthwhile venture with some sprectrum analyzing software. This was my choice to explore a viable solution to improve my car stereo. Nobody is making me do this. I would simply like to:

1. Aquire the proper CD for playing through my stereo.
2. Find the recommended software and microphone to use on my laptop.
3. Analyze what my stereo is doing well, and not so well.
4. Make adjustments, specifically to the amp and Kenwood EQ to improve the sound. For example, play a specific tone or tones, see that my 500hz is high, my 1000 hz is low, reduce high frequencies, lower bass output based on the findings of the software.

For all experts involved, if this is going to be impossible, because I need a $300 microphone, a better sound card, $200 software, then no way. I am sure you see my point. And as always, this community is the best, and I appreciate all the help.
 
Last edited:
Best bet... stick with the RS SPL and graph paper. Get a basline, tweek from there by ear. Don't waste $$$ or alot of time. Cars in general are lousy accoustic enviorments. Lots of glass, bad speaker placement, road noise, engine noise, etc, etc. Plus any tweeking you do with a RTA and EQ are out the door when you roll down the window, take the top off, etc.

You all ready have the RS, just try and do a basline so at least you know where your big problem areas are... bet you that there is a big spike around 5k hz. :cool:

BTW: I don't remember prices, but I figure I have around $300 in mike, software, cables, etc.
 
The analyzer works by putting pink noise (among other tests) into the system and analyzing what comes out. That's why you need to connect it to the stereo. You can also analyze if you throw a disc in there with pink noise but it will limit the kinds of tests you can do.

I can write a really long reply here but I don't want to get too technical. If you don't want to make a big investment or make it too complicated, do what I suggested the first time around... your meter and test disc, pen and paper.

Look... even if you buy a $200 surround receiver these days, the unit asks for you to do time alignment and level calibration. Anyone that has ran through "test tones" on their receiver has used pink noise in the past... no one in the home surround sound world says "just setup the levels so that it sounds best to your ear".... why should it be so different for your car? You don't get an RTA with your receiver, but it does provide you with pink noise and almost all do require an SPL meter. Now the newer ones come with a mike and have the setup done automatically. I believe even certain car stereos do this now as DSP's are cheap these days.

This is not a terribly complicated process... you somehow think that getting software, a mike, and sound cards is less complex than spending 20 minutes in the car with a disc and a meter. Why? :confused:
 
TURBO2GO said:
This is not a terribly complicated process... you somehow think that getting software, a mike, and sound cards is less complex than spending 20 minutes in the car with a disc and a meter. Why? :confused:

Here are a few examples why...

Malibu Rapper said:
You really need an RTA (real time analyzer) to do this right. Even after tuning the system flat it can sound like crap. A good sound shop can tune it for you and get you pretty flat which I am not sure if that Radio Shack SPL meter is suited very well for doing so.

nsxtcy85 said:
AMEN. You dont need one of those very unaccurate spl meters, if your that concered go to a high end audio shop around pay 100 bucks and they will tune it with a very expensive spl/rta-sq meter.

CL65 Captain said:
if you graph the whole thing out... then you have basically the same thing as a RTA but a much slower process.
 
As turbo said - I assumed you would also use the PC to inject test tones into the stereo. Sure you could use it as a better sound level meter with charting built in.

But you're making this way too hard. You wrote the process yourself:

1. Buy a test CD with 1/3 octave warble tones
2. Make a table with Excel or similar so you can record your response
3. Play test tones, measure the levels with the RS meter (takes 10 minutes or so).
4. Adjust with headunit, EQ, whatever
5. Re-measure
6. Tweak one more time
7. Write down your settings so you can go back to them if someone messes with your EQ.

Remember to set headunit tone controls to zero and put EQ sliders at zero at the start of the tests.

My suggestion is to set the subwoofer level first using test tones. Then run through the sequence above.

In one hour you should have a system with far flatter response than before.

I did the above in my Honda Ridgeline. If you want to see what I did (all active crossovers) look here.
 
djdrock said:
Here are a few examples why...

Well I don't want to get into a discussion over the accuracy of the RS meter. Like I said I do this for a living and I know your situation and my suggestion will work just fine for you. You are not designing loudspeakers, and "inaccurate" is a relative term. You certainly can go to a shop and ask them to do this for you if they are experienced and have the proper tools but I would just do it myself and then tweak it after listening to it flat for at least a couple of weeks.
 
TURBO2GO said:
Well I don't want to get into a discussion over the accuracy of the RS meter. Like I said I do this for a living and I know your situation and my suggestion will work just fine for you. You are not designing loudspeakers, and "inaccurate" is a relative term. You certainly can go to a shop and ask them to do this for you if they are experienced and have the proper tools but I would just do it myself and then tweak it after listening to it flat for at least a couple of weeks.

Here is one site that compared the RS meter to a B&K reference device.

http://www.digital-recordings.com/audiocd/radio.html

Note the final statement:

"The results from the Radio Shack Sound Level Meter are very impressive taking into consideration the fact that the meter costs only about $49.99 Can. It appears that this meter can be used in some situations that demand less accuracy with a high level of confidence. Certainly, it can be used to check the free-field levels created by STS signals from the AUDIO-CD."

It is WAY better than doing it by ear.
 
Hofffam that was a beautiful job on the ridgeline.
 
TURBO2GO said:
Unfortunately Hugh's viewpoint is rather common, because most people have no idea what correct sound or video even is. They have no frame of reference at all... and so who cares if there is a 10 db peak at 80 Hz.........................

You can't just do this by ear or "what sounds good to you" unless you have years of experience tuning. Even then, equipment helps much to set things right at the beginning.

My point to you Hugh is this... what sounds good to an untrained ear sounds good for a short period of time... with a particular music selection... it's only a matter of time until you realize something is not quite right..

Blah blah blah blah blah blah.....

The human ear is not a very accurate instrument for "measuring" or listening to sounds. First of all it has over 1% of THD secondly anatomical variations from one person to the next have a profound influence on how that individual hears what he hears. A 20dB peak as read by a meter might mean nothing to a person who's ear canal produces a 20dB loss. Likewise for the higher frequencies which are the ones most commonly affected by hearing loss.

Some examples:

Averages for.....

Normal range or no impairment = 0 to -20dB
Mild hearing loss = 20dB to 40dB
Moderate hearing loss = 40dB to 60dB
Severe loss = 60dB to 80dB
Profound loss = 80 dB or more.

I've gotten into people's cars and they sounded horrible to me but fine to the owner of the car. The bottom line is that a mic and a graph don't mean squat when tuning a sound system. Of course you want the crossover points and gain levels set correctly. To me that's a properly tuned system. As far as tonal changes, that's best left up to the owner to find what sounds best to them.
 
Hugh said:
Blah blah blah blah blah blah.....

The human ear is not a very accurate instrument for "measuring" or listening to sounds. First of all it has over 1% of THD secondly anatomical variations from one person to the next have a profound influence on how that individual hears what he hears. A 20dB peak as read by a meter might mean nothing to a person who's ear canal produces a 20dB loss. Likewise for the higher frequencies which are the ones most commonly affected by hearing loss.

Some examples:

Averages for.....

Normal range or no impairment = 0 to -20dB
Mild hearing loss = 20dB to 40dB
Moderate hearing loss = 40dB to 60dB
Severe loss = 60dB to 80dB
Profound loss = 80 dB or more.

I've gotten into people's cars and they sounded horrible to me but fine to the owner of the car. The bottom line is that a mic and a graph don't mean squat when tuning a sound system. Of course you want the crossover points and gain levels set correctly. To me that's a properly tuned system. As far as tonal changes, that's best left up to the owner to find what sounds best to them.

..and, according to my girlfriend I never hear her, how many db's is that? :biggrin:
 
LOL Hugh... thank God that djddrock knows enough on his own not to take your crazy advice.

20 db here, 30 there, what's the difference... lets all throw EQ's in our systems and play away... its all about what you like. And we know Hugh likes... the smiley curve.

freakin' crossdresser... STOP CONFUSING EVERYBODY!!!
 
This is an age old debate for anyone who's been in car audio. Tune for flat or tune for fun? Anyone remember R.A.C.? These debates never get anywhere. I hear very similar debates about steaks and whether or not you can eat a steak with a sauce. Diehards say no, rebel culinaries say yes. I'm of the belief you do whatever you like in every facet of life and no one but yourself should decide what tastes you should have...
 
Malibu, I know that debate but that is not really the debate here. I don't think ANYONE of us has said that ruler flat will sound great. You and I have enough experience to know that it won't. A meter is a tool... it will point to the potential problem areas. Once those are laid out, its up to the owner/tuner to do with it what they will.

This is ALL I am suggesting djdrock to do: Take some measurements and see what is going on. Its very difficult to know what is causing a problem without measurement.

I think all the talk on this thread about the accuracy of the meter, the sensitivity of the human ear, and the debate of flat or not flat does nothing for an average person reading this thread and trying to improve their system. Although all those points made have validity, none of them take away from the fact that a simple Radio Shack SPL meter and a test disc can truly help someone make a serious improvement in their system.

My suggestion to djddrock and anyone else reading this thread was a simple one that is valid and works for most people and situations. Please for the sake of a non-audio person reading this, lets not make it more confusing. A valid inexpensive suggestion is getting drowned out in the tech talk.
 
I didn't mean my post to come off as being unhelpful, I just wanted to elaborate on those specific technical points.

I do agree with Dave, that this is an area where knowledge is power. It never hurts to play, and a knowledgeable tuner can do more with less, even with less than ideal equipment. Even just hooking up a $50 used o-scope with a custom in-line adapter and properly setting levels on input/output stages can yield amazing results on the most simple of systems, and good luck finding a shop that would even go so far as to do that. Diagnosing and solving timing issues, problems at cross-over points, unwanted resonances, distortion, noise floors, imaging, etc... are all very do-able with time and desire. Like everything else with cars, no one cares about your ride more than you do.

If anyone has any interest... I still have all my course books from which are filled with practical info you just won't find anywhere else. If I could find the time and get a hold of a scanner, I'd be happy to have a birdie make some really detailed 600 page PDF's appear on installation, tuning, advanced fabrication, enclosure theory, and the like...
 
John@Microsoft said:
I'd be happy to have a birdie make some really detailed 600 page PDF's appear on installation, tuning, advanced fabrication, enclosure theory, and the like...

600 pages.. just enough to freighten everyone away for good John... :biggrin:

Notice we haven't heard from djdrock in a while... maybe he is out tuning... lol...
 
TURBO2GO said:
Malibu, I know that debate but that is not really the debate here. I don't think ANYONE of us has said that ruler flat will sound great. You and I have enough experience to know that it won't. A meter is a tool... it will point to the potential problem areas. Once those are laid out, its up to the owner/tuner to do with it what they will.

This is ALL I am suggesting djdrock to do: Take some measurements and see what is going on. Its very difficult to know what is causing a problem without measurement.

I think all the talk on this thread about the accuracy of the meter, the sensitivity of the human ear, and the debate of flat or not flat does nothing for an average person reading this thread and trying to improve their system. Although all those points made have validity, none of them take away from the fact that a simple Radio Shack SPL meter and a test disc can truly help someone make a serious improvement in their system.

My suggestion to djddrock and anyone else reading this thread was a simple one that is valid and works for most people and situations. Please for the sake of a non-audio person reading this, lets not make it more confusing. A valid inexpensive suggestion is getting drowned out in the tech talk.


I'll be Joe Schmoe here, since I don't know squat about car audio. Despite Hugh's frightening going-out attire, his "buy an equalizer, and play with it till you find what you like best" advice makes sense to me. The SPL meter results (have no idea what that means) would only get you to a good starting point quicker, right? Meaning, it's just a time saving device, no? :confused:
 
The advantage to using the SPL or RTA is this... would you know by listening to the stereo that there is HUGE spike at 5k hz. and how bad it was. Just by playing with sliders or in my case dials on my Audio Control EQL how much to cut and where to start making it sound better is close to impossible. The EQL has 31 bands... there are too many variables to start cutting this one or that one to find something that sounds right. The source material differs from song to song (as the songs themselves are EQed during the recording/producing process). Tuning it by ear would have been close to impossible. The RTA allowed me to find the problem areas like a 18db peak (which is HUGE) and fix them so I could then tweak by ear. You have to remember that changing one effects the others. As well, huge problems like my 5k peak would have masked any other changes I made by ear. THAT IS WHY USING A SPL (WITH GRAPH) OR RTA WILL ALLOW YOU TO FIX THE BIG PROBLEMS SO THAT YOU CAN TUNE TO YOUR LIKING BY EAR.

BTW: Once you get the stereo the way you like it, I highly recommend Chesky CDs to show off your new stereo. For those that don't know about Chesky... when it comes to natural sounding audio engineering and recording, he's the man!

Most CDs are recorded with multiple microphones in a dead-sounding studio, with reverb and overdubs added in the mix. Chesky's approach is purer: It records in the natural reverberance of St. Peter's Church in Manhattan, using custom-made vacuum-tube recording gear, a single microphone (in most cases), and no audio processing or overdubbing. The result is a spacious, natural sound you simply cannot hear in ordinary commercial recordings. Other audiophile labels do great recordings, too, but few capture the full-bodied tonality that Chesky achieves. And fewer still have signed such an eclectic array of artists.
 
Last edited:
Ski_Banker said:
I'll be Joe Schmoe here, since I don't know squat about car audio. Despite Hugh's frightening going-out attire, his "buy an equalizer, and play with it till you find what you like best" advice makes sense to me. The SPL meter results (have no idea what that means) would only get you to a good starting point quicker, right? Meaning, it's just a time saving device, no? :confused:

CL65captain said it but I will break it down and make it even simpler. A perfect human ear hears from 20Hz to 20,000 Hz. 1 Hz is one "cycle" of sound per second. One full wave going up and down. 20 Hz is the lowest bass frequency you can hear, and 20,000 Hz (or 20KHz) is the highest high. Most people's hearing however is much more limited than that.

An ideal stereo, would be capable of producing all these frequencies, all these sound waves, equaly. It would not exagerrate any particular part too much, or take any particular part AWAY too much. If you have no subwoofer for example, you probably won't get a lot of response from the system below 100 Hz. If you have no tweeter, you won't get much in the 10Khz area.

When you install a stereo system, you have to make sure that the tweeters, mids, woofers, subwoofers, all do their part right, and that their interaction with the "room", in this case the NSX cabin, produces a resonable response. When we talk about a "flat" response, all we are saying is that the frequencies are all produced equaly, at the same level (same loudness).

An SPL meter (Sound Pressure Level), just measures how loud a sound is. Thats all. The test disc, gives you tones... centered around frequencies such as 20 Hz for example, the next track a bit higher at 30Hz (a bit higher bass), and so on... all the way up. You simply sit in the cabin, and measure so that you know where a sound is louder, and where it is quieter. An increase in 3db (decibels) is roughly equivalent to a doubling of volume. So if you meausure and at 100Hz you have 90 dB, and everywhere else you are at 85 dB, you clearly have a "peak" at 100 Hz. It is too "loud" at around 100 Hz. And so you use the equalizer settings to lower the bass at 100Hz enough to bring it in-line with other frequencies.

Tuning a system is almost impossible by ear where there are a lot of problem areas, if you are not starting with a fairly clean sheet, a system with a fairly flat response. It is, like CL65captain said, very difficult to guess that you have a huge peak or dip at 5KHz, without measuring. All you know is something is not right. So instead of cutting at 5Khz, you may inadvertantly boost what is around it... at 7.5Khz, etc.

My advice for system tuning, is to just measure the response in your cabin. With a cheap SPL meter from Radio Shack (that I will state again is fairly accurate, especially for our use here), and a test disc like the stereophile (stereophile is a magazine) one to see what is going on. Once you have that info, there are plenty of experienced people on this forum that can look at the list of equipment, their placement and settings, and make suggestions as to what to do. If there is any area that a novice may screw it up, it is in this tuning area.

This is similar to for example sticking an wideband 02 sensor on your car, getting A/F ratios, plotting them, and then telling someone like Dynomike or the FactorX guys what you are getting. They can then help get the tuning right, and that is much better than sitting in your car with a laptop and "guessing" about how to set your A/F parameters.

This is a very general post and I know there are technical aspects that can be argued, I am refraining from doing so as I hope the more technical people here do as well.

I hope this was simple enough to understand.
 
All of the above means squat. What matters is the final sound to the end user. Once again, I repeat that the human ear is not nearly as sensitive as electronic measuring devices. If you really want to eq the system for the car's owner, first you should plot a frequency response curve for his particular hearing capabilities.

Botttom line, play with the tone controls until it sounds good to you. Anything else is a complete and utter waste of time.
 
TURBO2GO said:
Notice we haven't heard from djdrock in a while... maybe he is out tuning... lol...

He wisely unsubscribed from this silly thread, set his sliders to his liking and is out enjoying his car and his sound system.
 
Back
Top