I am working on tuning a car audio system using RTA software and a 30-band 1/3rd octave EQ. My understanding is that a person wants to smooth out the response to get a flat line initially for give a good starting point, but leaving it this way of course gives a flat/dry sound. I know there is some degree of personal preference involved, but I’m wondering what the graph should look like in the end for a good overall sound across many genres – my intuition is to create a “smile” across the spectrum, maybe a little heavier on the bass end than the higher end...

Looking for suggestions to start me finding my own preference...

Well as a professional sound engineer whenever I see a smile EQ it only means one thing, someone didnt know what they were doing. A RTA can not be used to EQ a system. Although lots of people think they can, The RTA does not take into account the surroundings and what is causing a dip or peak, it can not tell if something is reversed in polarity and it can tell time. For the cheap you could probably use Roomeq wizard. SMAART would be another way.Good luck.

here is a great article on not using a RTA you might have to register to read it but its free.Page 16. I just checked the link and it didnt need a registration.

http://livesound.ca.newsmemory.com/ee/livesound/default_ezine.php?token=LJoKWAB0gUsIAS0HK9KQl0Y3PQ

Edmonton,

doing IASCA or other sound competition?

If so you want TWO EQ settins.

1. For the RTA portion. The flatter the better. SQ is NOT a concern (and beleive me, besides the rare anomoly of a car that actually sounds good with a flat freq resp, most will sound horrid thanks to all the super reflective glass)!
2. For the SQ portion. You can use the EQ to get a baseline but the ear is where the final tuning will take place

as it looks from an RTA.

What I usually saw the sponsored cars trying to replicate was 3 bands and drop 1 db for the next 3 bands so by the time you hit the 12.5k, 16k and 20k you are down -10 dB from the 25hz, 31.5 40hz bands.

Also seen them not as drastic as every ear is different :)

I will agree with the article that a 31-Band "Auto RTA" sitting in the rack with your other gear in a PA setting is not very useful.

A software RTA is not the same thing, however. ;)

EdmontonAudio,
I believe Figgie got things right. :) I lot of people seem to like a bump in the 80-110Hz area. A bump at 4K for forwardness and at 8K for a bright high end are also really liked. Try not to bump 6K to much though.

Good Luck

Hmmm...not to sure I understand or if maybe I didn't ask the question right (LOOKOUT!! Another NOOOBE!!!)

Here's the background...many years had a stereo installed by a car audio shop who used a 30-band AudioControl RTA to tweak the 30-band EQ that was a part of the car audio system. I was not around when they did the tuning, but when I sat in the car and heard it for the first time, I couldn't help but laugh out loud because it was the most beautiful, crazy thing I had heard! It was loud, rich, and crystal clear. Friends would frequently bring their favorite CD to just sit and listen - be it classical, rock, rap, country...

Now, many years later, I am installing a system in my vehicle and I bought another 30-band EQ to tune it. I used a hand-held RTA off the bat just to get a very rough tuning...but even then, the change in sound was amazing. For fun and best accuracy (and since my career is in science & instrumentation), I thought I would tune it myself using a flat-response mic/computer/TrueRTA software combo (as outlined in these forums).

Ok...I better go read the article 'speco2003' posted...

I thought I would tune it myself using a flat-response mic/computer/TrueRTA software comboThis will work great. I was assuming you were talking about setting up for car audio competitions, but it doesn't matter the advice is the same.

The article speco posted is talking about something similar, but not exactly the same. He was dead on about the "smile" being bad though. There is no one size fits all setting.

Just set things to be flat at first on the computer. The EQ will be all over the place when you are done. Then if it isn't quiet what you want take note of the EQ positions, and start adding dBs to the EQ at the points I indicated before to taste.

Good luck. ;)

Here (http://64.233.169.104/search?q=cache:gQZWZ2xd80gJ:www.digitalprosound.com/2002/03_mar/tutorials/mixing_excerpt1.htm+MIDBASS+LOW+BASS+OCTAVES&hl=en&ct=clnk&cd=20&gl=us&client=opera) is a link to a page that has some more EQ suggestions. This should help you out, by giving you an idea on what you need to tweak to get the sound you want.

Remember though, less is more.

LOVE the link! Just what I was looking for!

So one more question, and again, I know the set points will have to be tweaked for best sound but...

Having six channels of audio - front component speakers, rear fill (mid range freq.), and subs, where abouts should I set the crossovers frequencies to start.

here is where it gets tricky as adding the rears will more than likely cause cancelations.

The trick is to do the eq's in steps.

First the sub
2nd the fronts.
3rd the rears alone.
lastly all three zones.

for the rear fills you want them to be BARELY audible when the fronts are playing. From memory they where usually about -6 dB below the fronts ref in white noise. Of course the design of the seats, rear window etc will have an effect on the dB value to attenuate the rear :)

Now what is going to happen is after you tweak everything alone and bring them only together. What you tweaked will probably not be there any more as the inevitable cancelations will rear thier ugly heads (cursed super reflective GLASS!!).

as for X-over settings for rear. I would say betwen the 110hz and 6khz tops but with a 1khz usually preffered.

Thanks for the eq'ing in steps...something else I needed to know!

...but what frequency range do I send to the front speakers, the rear fill, and subs? For instance, should I not filter out, say, 125 Hz and lower from the front component speakers, have the rear speakers at a bandpass from ____ Hz to ____ kHz, and the sub handle the ____ to ____ Hz range?

Thanks for the eq'ing in steps...something else I needed to know!

...but what frequency range do I send to the front speakers, the rear fill, and subs? For instance, should I not filter out, say, 125 Hz and lower from the front component speakers, have the rear speakers at a bandpass from ____ Hz to ____ kHz, and the sub handle the ____ to ____ Hz range?Again, there is no fixed answer without knowing what it sounds/measures like now. How can you prescribe a cure without a diagnosis?

It's going to take some trail and error on your part to get things ironed out, unfortunately.

There are two main schools of thought on how to set things up though.
1- Use the fronts from the sub xover (<~200Hz) up, and the rears as true filler drivers. Say from only ~4K down to the sub xover point.

2- Set things up like a 3-way speaker. Fronts down to say ~300Hz Rears from ~300Hz down to the sub xover point.

Both setups have their pros and cons, and will work better in different situations. The second option can work really well in setups that allow drivers that are bigger or higher xmax in the rear position. Try both ways to see what works best for you. Another common way is to mix the two and go say 200Hz up on the fronts, and 1K to the sub xover point on the rears.


Good luck.

Thanks! I just needed the general starting point (which you gave) so I don't put say, too much bass to my front speakers and damage them. Even though there is experimentation and taste, there is of course hard limits too...which I am looking to avoid when setting up. I mean, if I said I was trying to crank out 20Hz at my front component speakers :eek: ...not too wise a decision!

A flat RTA output is not actually flat sounding, because of the difference between first arrival, and early reflections. If there are significant modes and nulls in the listening position (and there are), then an RTA based EQ will over-compensate. A good start might be half-way between flat, and what the RTA tells you you "should" do.

Also, some people believe that it's more important to have an accurate and linear phase response, than a straight EQ curve. I give those people some credence. Try this, if you dare to do something different, that might actually make you win:
Get a number of digital delay lines. 24/96 precision would be ideal, as they'll give you good adjustability in delay time. One per speaker element in the car. Measure the distance from each element to the listening position. Tune all the delay lines so that the sound from each of the speaker elements will reach the ears at the exact same time.
After that is done, tune the EQ only slightly, to compensate for the worst modes (you can't really compensate for nulls). Give it a listen, and it just might be better than any strong EQ-based solution you can come up with!

Good luck!

Even though there is experimentation and taste, there is of course hard limits too...which I am looking to avoid when setting up.Most all woofers (read non-tweeters) can handle anything down to ~100Hz in car applications no problem up to ear damaging levels. If they are over 4" this drops to ~80Hz. 5-6" it's around 70Hz with the good ones dipping into the 60s. This is with at least a second order xover (-12dB per octave).

Sub-woofers tend to top out starting at around 80Hz for some to 150Hz on the better ones. Most tend to be in the 90-100Hz range. Also second order.

Hope this helps.