Rough tune on your EQ - Page 4

XxAndyxX · 11-23-2006, 08:38 PM

any suggestions for this ugly response curve?

I can send some a higher resolution picture if you give me your email.

JasonWW · 11-23-2006, 08:41 PM

Adjust the x-over freqs.

You've got 2 big gaps around the common x-over freqs between the sub and mid and between the mid and tweet.

Give me some info on your entire setup and how you took those measurements.

e-mail is in profile.

XxAndyxX · 11-23-2006, 10:36 PM

email sent...

JasonWW · 11-24-2006, 03:44 AM

To try and figure this out I would need to know a lot more info. From what I can gather you have 2 12's in the back hatch and a mid/tweet component set with a passive x-over between them.

Did you turn off all the bass and treble boost stuff on the amps and on the PC? Is the EQ set flat or bypassed?
How did you measure this response? Where was the microphone?

Since you don't have much control over the mid and tweet lets start with them. Are you using the front door locations? Where is the tweeter? There should be a tweeter level control on the passiv x-over, what is it set at?

With all the tone controls set flat, your subs turned off and your amp gains adjusted properly, tell me how it sounds, good, bad, whatever. If you have some 6.5" mids cross those over at 90-110Hz or so. Do you hear all those drop offs in the response? The big drop above 16K is no big deal and fairly common, I wouldn't worry about that. I see the amps don't have a high pass filter. Do you have some kind of high pass filter elswhere?
What Audio Control EQ do you have?

The big drop between 1K and 4K is pretty strange though. A decent component set should be much smoother than that. It's as if the mids are starting to roll off too soon and then the tweets kick in. Those speakers are pretty pricey, I would expect better of them.

Basically you want your response curve to look less jaggy and keep the 1/3 octaves within 3 or 4 dB of the band next to it. The mid/tweets will form their own curve and then you adjust the sub x-over and gains to bring the bass up to a matching volume so everything sounds good. When you see a sharp peak or dip in the curve, say 5dB or more, then you can flatten it out a little (not completely). Do as much of the response smoothing you can with the crossovers and amp gains. Sometimes you will need to alter the speaker positions or mounting angles to get a smoother response. This can be hard with the mids, but moving the tweeter is usually possible.

The peak you have at 50Hz is just the car, no real problem there. You can cut a few dB at 40Hz and 50Hz and add a few at 70Hz to smoothen it out a little, but that's all I would do there. A bit of a hump there can sound pretty good and is more of a listener preference. You can adjust the amp gain to raise or decrease the subs level to make it blend in better with the mids.

It really is the mids that seen to be problematic. They have weak midbass and roll off too soon (I'm guessing on that one). Unless you measured them wrong, then they should sound pretty crappy. Are they all that bad sounding to you?

XxAndyxX · 11-24-2006, 10:46 AM

All the bass boost is turned off and the EQ is set flat (not bypassed). I measured the response with the meter from radio shack and it was located at the passenger side head rest. I sat in the back of the car during testing to make sure not to block the sound coming from the doors and get bad measurements. The speakers are located in the stock firebird locations. There is not tweeter level control that I know of.

My NINe.2X amp does have a HPF, I adjusted it to get rid of the 125-200 dip. And I don't know why I said I had the HD-62...It was getting late. I actually have the ES 07.1 which are about twice as expensive and much nicer and should have more midbass (7" mid-woofer). The equalizer is an Audio Control EQT

Here are some pictures of the setup:

Placement of the meter (ofcourse when testing, door and everything is closed)

This is my ES 07.1 in the door.

Audiocontrol EQT. I have them there so I can easily adjust the EQ, when I'm done I will move them to a more secure location. As you can see in the picture I have boosted some of the frequencies; they were set flat for all the response tests I did, I just adjusted them since then.

The xover I'm using for my tweets is the SatNet 560i
Link for the ES 07.1 series

Also, should I adjust each EQ (and speaker) independently? Like just unplug the left speaker. And adjust the right....Then try to match the right's response curve with my left speaker? Because I'm also getting imaging problems. The system overall sounds decent. But for what I paid, it shouldn't just be decent, it should be much better.

JasonWW · 11-24-2006, 10:22 PM

OK, I used to have a set of the EQT's, so I'm familiar with them.
On the x-over they say there is a tapped inductor and that the adjustor is labeled by driver size (5.25 and 6.5). I don't know what they mean, maybe a jumper you can set for different mid range sizes?
When looking on the ED site they only showed a LPF on the nine.2 amp. So it's does have a HP as well? What was it set at before and what now? Did that give you some more midbass kick?
There's no need to be in the back when testing the response. It may even be more realistic to sit up front. What audio tracks where you playing to get the sound you measured? Are you using "slow" response and "C" weighting?

So you have the analog signal coming out of the sound card and then to the EQT's? What sound card are you using? What audio player are you using?

I can tell you right off the bat that in these cars the stock tweeter location is not very good. An a-pillar install is the way to go, but this is something to look into later if your serious about sound quality. For right now I would turn off the left speaker and map the right and vice versa. If you want even more accuracy, disconnect one of the tweeter wires and map just one mid, then the other, then map just the tweets. That way you can see what the x-over is doing.

Still, I can't understand why you are getting such big drops in the response. It could be the source material. If you have a CD player laying around you could plug it in as a source and listen to a particular song and then listen to the same song through the PC sound card and see if they sound identical. If so, then you can eliminate the sound card as a problem area.

If after all these checks and the mid/tweets still have these huge swings then I would suggest finding a sound shop that has an RTA and let them take a look. They can play Pink noise which contains all the freqs at the same level and you can see the whole response curve at once. Typically you wouldn't use any of the EQ bands more than absolutely necessary. Most changes should be under 4 or 5 dB with maybe some at 7 or 8 dB, but your curve has huge 20 dB dips which no EQ can cure. This tells me either the measurements are wrong or something else in the equipment is not right. You can try placing the meter on the center console and play back the 800 and 2K notes and see if they still have a 23 dB difference.

Hey, the mids are not hitting the door panel when moving are they? They look prety close in the pics.

If you want to play around with the tweeter location to raise the sounstage and improve the imaging you can try something like I did which is to silicone some flexable metal straps with the tweets on the ends and move them around the a-pillar location.

I found that it sounded so good I flush mounted them in the a-pillar.

It will definetely give you better sound and the money you've spent so far shows you want real good sound. Right now we focus on the mids and then mess with the tweets later.

XxAndyxX · 11-25-2006, 11:55 AM

The xover did have a jumper, but it was set correctly. Before I had the HPF on the NINe.2X set pretty high (probably what gave me that gap), I adjusted that and it gave me some more midbass, so that's good. It doesn't label the frequencies very well so you just have to kind of guess. I have it set to were I marked in red here

Here is the sub's amp, you can see that I have it set to the red mark.

I had the meter set to slow and C weighting. For my audio card, I'm using a DAC (EB-801). I'm playing these tones through winamp (without any equalization). I downloaded the tones from a reliable source (http://realmofexcursion.com/). The mids a definitely not hitting the door, I made sure of that. Now I'm getting more of a 9db difference between 800 and 2000 which is bad but not anything like 23db. The problem I'm having is getting an accurate reading. The meter is all over the place. It's hard to even get an avergage. Check out this video I uploaded of me trying to to a reading at 6k. http://www.youtube.com/watch?v=Ytu9HfcBMPU Sometimes it swings even more than in that video. And it never comes to a stop.

JasonWW · 11-25-2006, 12:22 PM

Do you think maybe the meter is not working right? Does moving the meter to a different location stabilize it any?

Try playing that same 6K test tone on your home PC and speaker setup just to see if it jumps around or is consistant.

Now that you know what the different freqs sound like, can you hear the gap from 1K to 3K in music? I'm wondering if there is really that large a gap or if the meter is just not reading accurately.

BTW, did you hear your component speakers in any other car before you bought them? Did they sound really good then, but never sound that way once you put them in your car?

XxAndyxX · 11-25-2006, 01:58 PM

It jumps around in any location. I took it inside and tried it on my laptop and it still ranges quite a bit (85-92dB). Most frequencies the meter will lock dead on. In the lower frequencies, it does really well. It will start to vary once it gets into the higher one's though. What is the biggest range that you get on yours?

I can hear small dips, but nothing like what it's showing. Infact, sometimes I think it dips and the meter says it get louder.

I never heard the speakers before I bought them, but read tons of good reviews and talked to the company for a while to see if it's what I wanted. I doubt they are the problem.

JasonWW · 11-25-2006, 02:31 PM

I went out and checked my response. My meter locks onto a spl and might flunctuate up or down one dB depending on the freq. It isn't jumping around like yours. I don't know what the deal is.

Maybe your source tone is not steady. Where did you get the 6K tone from?
Did you try the ones I linked to at the beginning of this thread? They are not great, but once above the bass range they seem OK.

XxAndyxX · 11-25-2006, 03:29 PM

I got my tones from http://realmofexcursion.com/downloads.htm, here's the link to the tones: http://www.realmofexcursion.com/jmac...test_tones.rar

I tried the tones you linked to on the first page, and at first it was okay. It would range between 95-96 for about 30 seconds and then 94-95 for another 30 seconds (better than what it was doing with my other tones). But I tested it again and it said about 91-92...So if I can't repeat the results then there's not point in doing it. Wierd thing is that both the ones you linked to and the ones I used are supposed to be 0db, but I'm getting a big difference between the two files (about 8-10db difference). I'm thinking it's the meter...I'm heading to rat shack to get another one.

On a side note, I was calling some shops and I was surprised to find out how few of them even knew what a RTA was? I don't know how you can work at an audio shop and have no clue what an RTA is. When I mentioned things like pink noise and response curve they made it seem like I was talking a different language...lol

JasonWW · 11-25-2006, 03:38 PM

It's not always going to be perfectly steady. As long as it is jumping between 2 or 3 numbers and no more, then just average them. You don't have to be exact as you are looking for overall tone and any pesky peaks and dips.

If it still looks like you may have a 9 dB dip between 1K and 4K then try raising those low freqs up 6 dB. That way they are only a few dB down and listen to the difference. Was it too much, not enough? Make it sound better or worse?

Your not trying to get the response flat, just a somewhat smooth transition from band to band without any sudden drops or peaks.

As far as finding a shop, you may have to look in the USACi and IASCA Forums for any well known shops in your area. If they compete, they should have an RTA.

XxAndyxX · 11-27-2006, 11:58 AM

Hey Jason thanks for helping me out. I'm back at college so it's harder to find a time and place to tune the system. Fortunately, there's a shop that has an RTA that charges $35/h. I talked to a friend though and he said they really don't know what they're doing, so I'm hesitant to take it to them. Basically if they don't let me do it myself then I'll just tell them exactly what I want. Play left speaker, smooth the curve out, and then match the right speaker with it. Then, ask them to show me both the left and right speaker's response curve. After that I can play with groups of bands to get the curve I want.

Before I spend the money of getting someone else to do it, I think I might try to do it myself again. I mean I did spend 50 bucks on this meter. After my last post, I tried it again using the integrated average function over a 30 second interval. I still got a variance of 3-4 dB's on some frequencies though. I may try a different approach and just use the max values of everything band and see how precise it is. It may just be precisely inaccurate (if that makes since). Which is fine because as long as the difference in dB is correct then I don't care if it's higher or lower than it's supposed to be.

JasonWW · 11-28-2006, 09:59 PM

You may want to be carefull measuring each side seperately. Sometimes the RTA or microphone will show a lot of differences between the sides that really aren't there. Let your ears be the final judge on that one. You can run a sinewave sweep that goes through all the freqs and then listen to see if the image moves left and right. Even though the left and right side are not identical with the steering column and such it may be better to do measure both sides at once. Listen to the sweep and if you hear the image move left or right at a certain freq, then find that freq and decide whether to raise or lower that one side to help compensate. Even then, you shouldn't have to adjust it much, maybe a couple dB.

FyreDaug · 12-05-2006, 04:34 PM

Radio shack here was bought by curcuit city and dont carry em, where can I find one of these units

User Name		Remember Me?
Password