DiyMobileAudio.com Car Stereo Forum banner

1 - 20 of 26 Posts

·
Registered
Joined
·
22 Posts
Discussion Starter #1
Hi guys, im in need of some help tuning my system. This is my first attempt at tuning a 3 way and im having a hard time getting it to sound good. Everything looks good when i measure, but it keeps sounding quite harsh and shouty. Hopefully someone will be able to give me some ideas to try to calm it down a touch.

I am using a 3 way set of AF GB's in mostly stock locations in my Toyota 86. The Gb60's are in the bottom front corner of the front doors completely deadened with MLV and all that jazz, Ive glassed them into the door panels with a solid baffle made from several layers of 1/4in pvc and 6in schedule 40 pvc pipe to vent into the door and not be behind the door panels. The GB25's are in the corners of the dash up firing into the windshield using 1/4 in plexi baffles sealed up to the stock grills with some CCF weatherstripping. The GB10's are up in the A Pillars mainly on axis aimed a few feel behind the front seats center of car at the moment (I am still playing with positioning and want to figure this out before glassing them in and having to redo it). And a gb12.4 glassed into the trunk with 1200RMS. The 60's and 25's are being run off of a JL 600/4 getting 150 RMS each and the tweets are being run off an audio control ACM 4.300 getting 50 RMS.

I used Jazzis tuning tool to get some house curves and think i did a fairly good job at getting there, although i beat up the mid with a bit more EQ than i probably should have. So right now im using all LR/24db xo slopes. Im shooting for the acoustic slopes of 20-80 Sub, 80-300 MB, 300-3000 Mid, 3000-20000 tweet. The dash mids start rolling off kind of early and had a large low pass looking boost starting arounk 1k on either side. So i used a shelf filter on both to bring down the low end and extended out the low pass filter out to 4245 on the left mid and 3861 on the right mid to get the 3000hz XO frequency I was looking for. The tweeters acted like expected so their low end roll off looks good right at a 3k Usually this gets me close enough that some minor attenuation here and there will get me to where I want to be, but this time it just ended up so bright that im not too sure where to attack first.

Im going to attach some pictures of my measurements and my helix screens, so hopefully this will be enough info to give you guys an idea of where I am at the moment. But just let me know if some more measurements or whatever will help. And im sorry if I posted this in the wrong area, feel free to move it if there is a better spot on the forum for it!

Thanks in advance for the help guys!
 

·
Registered
Joined
·
22 Posts
Discussion Starter #2
Left mid and tweet 1/6 smoothing and unsmoothed

Left Speakers.PNG
Left Speakers Unsmooth.PNG







Right Mid and Tweet 1/6 smoothing and unsmoothed

Right speakers.PNG
Right speakers unsmooth.PNG







Left Mid EQ and XO

L Mid Tune.PNG






Left Tweet EQ and XO

L Tweet Tune.PNG






Right Mid EQ and XO

R Mid Tune.PNG






Right Tweet EQ and XO

R Tweet Tune.PNG
 

·
Registered
Joined
·
1,687 Posts
Left mid and tweet 1/6 smoothing and unsmoothed

View attachment 274977 View attachment 274978






Right Mid and Tweet 1/6 smoothing and unsmoothed

View attachment 274979 View attachment 274980






Left Mid EQ and XO

View attachment 274981





Left Tweet EQ and XO

View attachment 274982





Right Mid EQ and XO

View attachment 274983





Right Tweet EQ and XO

View attachment 274984
What house curve are you tuning for? Those mids seem to be quite high in the middle, I would expect that area (300-3k) to be closer to flat.

Sent from my SM-G935U using Tapatalk
 

·
Registered
Joined
·
1,048 Posts
Your mids aren’t ‘rolling off‘ you are getting a lot of reinforcement in the lower mid due to where you’ve placed the drivers in the corner of glass, sides of the car and the dash, it’s a fact of life of placing mids on a dash, there is a reason that andy recommends not putting mids up there, that being one good reason why not

mid also turn the mids down by 5-8db and readjust the eq so you have levels more matched... currently you have ten db more mid range than 4-5k and 20db more than 16k plus... that is not tuning to any house curve I can think of, look up the audiofrog curve and adjust to that, I’m not surprised it’s a bit Woolley and harsh sounding as it’s way way out of balance

have a read of the audiofrog guide and follow it, do not use jazzis sheet as that’s not how andys guide does things with individual speaker tuning
 

·
Registered
Joined
·
6,671 Posts
Thats almost the exact same set up and install that i have in my car. The GB25 that i have needs to be dialed down 6-8 db from the rest of the system. I'm thinking your's may need to come down too.
 

·
Registered
Joined
·
22 Posts
Discussion Starter #6
I'll pull some out of the mids and see how that effects it. I did forget to mention last night that I still get a very similar harshness when playing the mids only with those eq settings. If I pull like 10db out of 2.7k hz with a crazy high q it seemed to keep it somewhat in check, but still be somewhat noticeable. I'm assuming it's a window reflection but it seemed like something was off seeing such a massive cut in the measurements that I figured that something else is wrong.
As far as the house curve I am shooting for, it is a combo of my tastes and a half whitledge. I tend to like a bit of a pulled back top end so it doesn't end up fatiguing at the volumes I like to listen, and a boosted low end.
I also use the jazzi tool mainly to make sure i am getting my acoustic crossovers where I want them to be, less for the actual tuning for staging and whatnot.
As far as above 16k, I wasn't terribly worried about that since my hearing disappears after 16.5k. But I was also working under the assumption that the mic was rolling off up in that octave and the levels were roughly where they should be since I can't really hear them lol. But I'm also using the audiofrog mic so I'm not sure if that makes a difference.
Thanks for the reply guys!
 

·
Registered
Joined
·
358 Posts
I would also pull all of the 1k down , a ton
The 1k range is where you always get harsh sounds from


Sent from my iPhone using Tapatalk
 

·
Registered
Joined
·
1,048 Posts
I'll pull some out of the mids and see how that effects it. I did forget to mention last night that I still get a very similar harshness when playing the mids only with those eq settings. If I pull like 10db out of 2.7k hz with a crazy high q it seemed to keep it somewhat in check, but still be somewhat noticeable. I'm assuming it's a window reflection but it seemed like something was off seeing such a massive cut in the measurements that I figured that something else is wrong.
As far as the house curve I am shooting for, it is a combo of my tastes and a half whitledge. I tend to like a bit of a pulled back top end so it doesn't end up fatiguing at the volumes I like to listen, and a boosted low end.
I also use the jazzi tool mainly to make sure i am getting my acoustic crossovers where I want them to be, less for the actual tuning for staging and whatnot.
As far as above 16k, I wasn't terribly worried about that since my hearing disappears after 16.5k. But I was also working under the assumption that the mic was rolling off up in that octave and the levels were roughly where they should be since I can't really hear them lol. But I'm also using the audiofrog mic so I'm not sure if that makes a difference.
Thanks for the reply guys!
You’re assuming andys mic is bad but your system sounds awful, try actually tuning strictly to the half whitledge curve before putting your own twist on it... I’m not being funny but if you don’t know what’s causing issues and harshness you really aren’t at a level to be improvising a house curve to how you think it should be... as you clearly don’t know what part of the audio spectrum is what

a half whitledge curve has almost the difference between your mid and tweeter at 20khz between 20hz and 20k, your nothing like it apart from a downward slant, sorry if that’s harsh
 

·
Registered
Joined
·
22 Posts
Discussion Starter #9
Who said I thought the mic is defective?
I appreciate you taking the time to answer, but no reason to get defensive. I wouldnt be on here asking if I had the answer. Every time I've used someone else's mic before I bought my own from andy, I was told that I should expect the high end to have a downward slope. Granted ive taken their word for it since it was their equipment. And I am familiar with what a car tuned to a half whitlage sound like. I've heard others vehicles and my own tuned to the curve and I've found i like more bass and less high end. Is it unrealistic to the source? Yes, but is it what I've found that I like comparing to those systems.
I am aware the the tweets are very low in comparison to the mids volume in those measurements, the reason being it becomes pircing above that so thats where it is sitting at the moment to be listenable without being unpleasant.
My goal here was to have someone more knowledgeable with 3 ways, and dash mounted corner loaded mids than myself hopfully shed some info on why I have the balance the way it is at the moment to sound bearable. Maybe its common to cut in some areas because of reflections due to the location that dont show in averaged mesurements? I dont know thats why I'm asking.
Like I said, I appreciate you taking the time but throwing daggers doesn't make for a helpful dialogue.
 

·
Registered
Joined
·
1,687 Posts
Hi guys, im in need of some help tuning my system. This is my first attempt at tuning a 3 way and im having a hard time getting it to sound good. Everything looks good when i measure, but it keeps sounding quite harsh and shouty. Hopefully someone will be able to give me some ideas to try to calm it down a touch.

I am using a 3 way set of AF GB's in mostly stock locations in my Toyota 86. The Gb60's are in the bottom front corner of the front doors completely deadened with MLV and all that jazz, Ive glassed them into the door panels with a solid baffle made from several layers of 1/4in pvc and 6in schedule 40 pvc pipe to vent into the door and not be behind the door panels. The GB25's are in the corners of the dash up firing into the windshield using 1/4 in plexi baffles sealed up to the stock grills with some CCF weatherstripping. The GB10's are up in the A Pillars mainly on axis aimed a few feel behind the front seats center of car at the moment (I am still playing with positioning and want to figure this out before glassing them in and having to redo it). And a gb12.4 glassed into the trunk with 1200RMS. The 60's and 25's are being run off of a JL 600/4 getting 150 RMS each and the tweets are being run off an audio control ACM 4.300 getting 50 RMS.

I used Jazzis tuning tool to get some house curves and think i did a fairly good job at getting there, although i beat up the mid with a bit more EQ than i probably should have. So right now im using all LR/24db xo slopes. Im shooting for the acoustic slopes of 20-80 Sub, 80-300 MB, 300-3000 Mid, 3000-20000 tweet. The dash mids start rolling off kind of early and had a large low pass looking boost starting arounk 1k on either side. So i used a shelf filter on both to bring down the low end and extended out the low pass filter out to 4245 on the left mid and 3861 on the right mid to get the 3000hz XO frequency I was looking for. The tweeters acted like expected so their low end roll off looks good right at a 3k Usually this gets me close enough that some minor attenuation here and there will get me to where I want to be, but this time it just ended up so bright that im not too sure where to attack first.

Im going to attach some pictures of my measurements and my helix screens, so hopefully this will be enough info to give you guys an idea of where I am at the moment. But just let me know if some more measurements or whatever will help. And im sorry if I posted this in the wrong area, feel free to move it if there is a better spot on the forum for it!

Thanks in advance for the help guys!
Just a quick test for you, if you can, lower all of the inputs by 6 db and see if the harshness persists. If it goes away, you're probably hearing digital clipping from boosting too much.

Sent from my SM-G935U using Tapatalk
 

·
Registered
Joined
·
1,048 Posts
Who said I thought the mic is defective?
I appreciate you taking the time to answer, but no reason to get defensive. I wouldnt be on here asking if I had the answer. Every time I've used someone else's mic before I bought my own from andy, I was told that I should expect the high end to have a downward slope. Granted ive taken their word for it since it was their equipment. And I am familiar with what a car tuned to a half whitlage sound like. I've heard others vehicles and my own tuned to the curve and I've found i like more bass and less high end. Is it unrealistic to the source? Yes, but is it what I've found that I like comparing to those systems.
I am aware the the tweets are very low in comparison to the mids volume in those measurements, the reason being it becomes pircing above that so thats where it is sitting at the moment to be listenable without being unpleasant.
My goal here was to have someone more knowledgeable with 3 ways, and dash mounted corner loaded mids than myself hopfully shed some info on why I have the balance the way it is at the moment to sound bearable. Maybe its common to cut in some areas because of reflections due to the location that dont show in averaged mesurements? I dont know thats why I'm asking.
Like I said, I appreciate you taking the time but throwing daggers doesn't make for a helpful dialogue.
It’s not daggers, just try tuning it to the actual curve 👍🏼 What other people claim is tuned to something may be very different to what your tune to that curve results in, moving mic vs static, upwards pointing mic vs forward pointing will all give different results, so one mans tune may not equal another mans due to how mics measure differently, my point is to get a good start point I would tune to a half whitledge curve and see what it actually sounds like instead of discounting it before you’ve even tuned to it 👍🏼
You had said you thought the top end wasn’t accurate... andy doesn’t seem like the kind of guy to supply stuff without an accurate mic correction curve. I don’t get why you’d discount it straight away when you’ve bought something new and not cheap that’s all

If the tweeters become piercing it’s likely your overdriving them, so use them as the limit for the half whitledge curve and match the mids back to the level the tweeters aren’t piercing, if it’s not loud enough then you have other issues, but that’s likely why it’s sounding a bit rough, just trying to be honest and say what I see
 

·
Registered
Joined
·
22 Posts
Discussion Starter #12
Finally off work so I will give some of these things a shot and see what i can come back with tonight.
I take no offense, no worries. Im happy people are actually taking the time to try to help since ive never been a forum poster, just a reader. But your point of the mic makes a lot of sense now that I actually put some thought into it. I never really thought about it since what ive used in the past probably werent calibrated, so I just went upon my merry way like I was used to. But that is probably a major point of the problem right there. My assumption that there should be a roll off on the mic most likely shot me in the foot.
So far what I have gotten to work is to pull a low q (1.5) swath of like 6db out of 3.5k. Which would make a lot of sense looking at the graph knowing i should be looking for a MUCH shallower slope up top since my mic is calibrated. I feel stupid after it clicked hahah. So im gonna give that a shot first and then swing a notch filter through a bit to see if i can locate the offending frequencies a bit better.
I can also bridge the amp im using and give them 125rms instead of 50 to see if im over driving the amp channel, cause i doubt im over driving the tweets themselves. Its really not all that loud before it gets rough to the ears. Compared to my old cars 2 way setups atleast.
 

·
Premium Member
Joined
·
4,710 Posts
...
I can also bridge the amp im using and give them 125rms instead of 50 to see if im over driving the amp channel, cause i doubt im over driving the tweets themselves. Its really not all that loud before it gets rough to the ears. Compared to my old cars 2 way setups atleast.
Personally I would be looking at the digital gain levels (way up front in the DSP) before looking at increasing power to tweeters.
 

·
Registered
Joined
·
22 Posts
Discussion Starter #14
As of right now, i have the gain in the dsp in the negative on all the speakers to try keep from any sort of clipping within the dsp and keep my volume at 75% in the headunit for the same purpose. that being said, the highest i have on one of the speakers is -1.25 db with no boosts on that channel. Is it still something that I should be worried about?
 

·
Registered
Joined
·
22 Posts
Discussion Starter #15
I figured 50rms would be more than enough for the tweets, but i figured I would make it known that I do have the option. I planned on using the other 2 channels to play with differential rear fill in the future after getting the front dialed in, since ive never heard a car with it nor played with it myself. So I would prefer to not use them for this case unless it is needed.
 

·
Registered
Joined
·
22 Posts
Discussion Starter #16
I had some time this evening after work to finally take a look and I have good and bad news. Good news, I figured out the issue. Bad news I am an absolute idiot. It turns out that it actually was the mic kind of, but not becasue the mic is bad but rather because my novice is showing very heavily. Im actually quite embarrassed and dont even want to comment this, but you guys who offered some help should know where I am and what the reason was lol.

I was playing around all weekend trying to figure out why it sounded so bad and got no where. So i got frustrated and gave up and figured I would go back to it later. So tonight after work I figured I would poke into it again. So I thought i would try a different tactic and decided to load up the tune I did by ear before I got the mic in and measure that to see what it looked. Hoping to see if I could spot the differences that may be the culprit...and it looked AWFUL. So I was really confused for a bit as to why the measurements that looked somewhat good sounded terrible, yet the tune that sounded good was all over the place. After poking around a bit trying to figure it out, it hit me. The REW measurement tool defaulted back to spectrum instead of RTA. Why I didnt notice, I have no clue, they look completely different lol.
So i switched it back to RTA mode and remeasured the by ear tune, and BAM so very very much closer to the curve. Then I went back to the tune I started when I got the mic and re-measured it, and is no wonder it sounded so extremely bad. It also makes so much more sense why the tweeter looked so low in the last graphs, and why I had to pull 10db out of 3.5k to even listen to it without wincing. I spent 20 mins resetting the levels and a bit of eq work and we are back to sounding listenable. Far from good, but at least I know what the problem was and I can get it to sound good from here.

Thank you guys so much who spent the time to comment, I really appreciate it. If anyone is curious to see how awful the measurements looked when I went back to RTA mode, I am more than happy to post some pictures! Its honestly pretty comical! Or if anyone wants to know how things progress from here, I will be happy to keep the thread updated.

Thanks again guys!
 

·
Registered
Joined
·
1,687 Posts
I had some time this evening after work to finally take a look and I have good and bad news. Good news, I figured out the issue. Bad news I am an absolute idiot. It turns out that it actually was the mic kind of, but not becasue the mic is bad but rather because my novice is showing very heavily. Im actually quite embarrassed and dont even want to comment this, but you guys who offered some help should know where I am and what the reason was lol.

I was playing around all weekend trying to figure out why it sounded so bad and got no where. So i got frustrated and gave up and figured I would go back to it later. So tonight after work I figured I would poke into it again. So I thought i would try a different tactic and decided to load up the tune I did by ear before I got the mic in and measure that to see what it looked. Hoping to see if I could spot the differences that may be the culprit...and it looked AWFUL. So I was really confused for a bit as to why the measurements that looked somewhat good sounded terrible, yet the tune that sounded good was all over the place. After poking around a bit trying to figure it out, it hit me. The REW measurement tool defaulted back to spectrum instead of RTA. Why I didnt notice, I have no clue, they look completely different lol.
So i switched it back to RTA mode and remeasured the by ear tune, and BAM so very very much closer to the curve. Then I went back to the tune I started when I got the mic and re-measured it, and is no wonder it sounded so extremely bad. It also makes so much more sense why the tweeter looked so low in the last graphs, and why I had to pull 10db out of 3.5k to even listen to it without wincing. I spent 20 mins resetting the levels and a bit of eq work and we are back to sounding listenable. Far from good, but at least I know what the problem was and I can get it to sound good from here.

Thank you guys so much who spent the time to comment, I really appreciate it. If anyone is curious to see how awful the measurements looked when I went back to RTA mode, I am more than happy to post some pictures! Its honestly pretty comical! Or if anyone wants to know how things progress from here, I will be happy to keep the thread updated.

Thanks again guys!
I made that mistake once, I guess I looked in the right place for the problem sooner than you did. Seems like RTA should be the default instead of Spectrum.
 

·
Registered
Joined
·
7 Posts
Correct me if I am wrong but, it looks like you are running an RTA measurement on each speaker then setting the EQ on each speaker separately, is this the case?

Also, what final system curve are you shooting for, post EQ?
 

·
Registered
Joined
·
384 Posts
I had some time this evening after work to finally take a look and I have good and bad news. Good news, I figured out the issue. Bad news I am an absolute idiot. It turns out that it actually was the mic kind of, but not becasue the mic is bad but rather because my novice is showing very heavily. Im actually quite embarrassed and dont even want to comment this, but you guys who offered some help should know where I am and what the reason was lol.

I was playing around all weekend trying to figure out why it sounded so bad and got no where. So i got frustrated and gave up and figured I would go back to it later. So tonight after work I figured I would poke into it again. So I thought i would try a different tactic and decided to load up the tune I did by ear before I got the mic in and measure that to see what it looked. Hoping to see if I could spot the differences that may be the culprit...and it looked AWFUL. So I was really confused for a bit as to why the measurements that looked somewhat good sounded terrible, yet the tune that sounded good was all over the place. After poking around a bit trying to figure it out, it hit me. The REW measurement tool defaulted back to spectrum instead of RTA. Why I didnt notice, I have no clue, they look completely different lol.
So i switched it back to RTA mode and remeasured the by ear tune, and BAM so very very much closer to the curve. Then I went back to the tune I started when I got the mic and re-measured it, and is no wonder it sounded so extremely bad. It also makes so much more sense why the tweeter looked so low in the last graphs, and why I had to pull 10db out of 3.5k to even listen to it without wincing. I spent 20 mins resetting the levels and a bit of eq work and we are back to sounding listenable. Far from good, but at least I know what the problem was and I can get it to sound good from here.

Thank you guys so much who spent the time to comment, I really appreciate it. If anyone is curious to see how awful the measurements looked when I went back to RTA mode, I am more than happy to post some pictures! Its honestly pretty comical! Or if anyone wants to know how things progress from here, I will be happy to keep the thread updated.

Thanks again guys!

I did the exact same thing, took me weeks to figure out. I did exactly what you did to figure it out too lol.
 

·
Registered
Joined
·
22 Posts
Discussion Starter #20
I missed your guys replies, sorry. I was measuring each speaker individually and EQ'ing each alone to find my curve and best XO settings to match the acoustic XO im looking for before matching levels. Luckily now that I have gotten the RTA issue figured out things have been coming together quite well! I was also shooting for half whitledge as a base. I usually take a cut out of 2-4k and raise the low end (20-250) a touch over that as my final.
Im running into a new issue now though, I have a large dip at 185hz with both gb60s playing together. Everything is TA and is summing well everywhere but 184-187 hz. When i flip polarity of one speaker the dip fills out as it should, but the rest of the response is canceling, so im assuming its a phase issue due to the car dimensions. Ive never used an All pass filter before since my previous mini-dsps didnt have the option, but from what I understand this would be a prime candidate for an AP filter. Ive tried some fiddling around with one but Im having an issue filling it in without it pulling heavily to the drivers side and also messing up the surrounding response.
Does anyone have any words of wisdom here? Or am I barking up the wrong tree here with the AP filter idea? Thanks guys!
 
1 - 20 of 26 Posts
Top