DiyMobileAudio.com Car Stereo Forum banner

21 - 40 of 54 Posts

·
Registered
Joined
·
56 Posts
Discussion Starter · #21 ·
1. Unless you have a DSP with a few bands of parametric EQ, you will have difficulties. But you may be able to make broad brush improvements.
Yeah, I know that PEQ with 10 bands would be great to have, struggled a bit with a peak at around 3 kHz, since I could adjust 2.5 kHz 4 kHz but nothing in-between. The Kenwood product page mentions something about being able to adjust the Q but I haven't found that. Might be an error on their website.

2. Log-sweep measurements are much better than pink noise - they reduce the influence of background noise and get you phase info.
Hmm I'll have to look into the differences between the pink noise and sweep methods. Thanks for the tip!

3. That dip at 550hz or thereabouts - I bet is a phase issue. Could be crossover induced, speaker position induced, or acoustics induced. Have you looked at the phase plot?
No, haven't looked into it yet. I too suspect it to be a phase issue. My tweeters don't go that low so it is probably the woofers. Maybe it will flatten out with some time alignment (head unit has some basic TA that I need to look into). I'm guessing the dips and peaks between 1 kHz and 2 kHz might me because of phase as well, as my tweeters are crossed over at 2 kHz according to the spec sheets.

4. You should probably not try to fix phase issues with an EQ bump.
That is actually one of the first things I learned about EQ, so all of my adjustments have been cuts only.

5. DSP will not just be better at EQ, it will allow time alignment Left-Right, which will affect EQ (due to phase) massively
6. Going active will open a new world of possibilities.

Just don't get too hung up on a graph if you only have a graphic EQ. Use it to know where to make some level cuts, fine tune it by ear, and live with what you have, until you get a DSP. That would be my advice.
Yeah, I am convinced, a DSP and going active has been added to my build plan. As my wallet is a little bit on the light side, however, it just might take some time.
 

·
Registered
Joined
·
56 Posts
Discussion Starter · #22 ·
It looks about as good as you're going to get it using your current equipment. If you can, bring 300Hz down a few dB. But, besides that looks OK.

Ge0
Thanks, then I guess I have done something right at least. That 300 Hz is one of those hard-to-reach frequencies with this GEQ, as I can only do 250 Hz or 400 Hz. Might try dropping 250 Hz one step and see what happens.
 

·
Registered
Joined
·
2,911 Posts
Most of my experience is in home audio, so I'm going to keep it a bit vague - I don't want to sound too knowledgeable about car audio specifics.

1. Unless you have a DSP with a few bands of parametric EQ, you will have difficulties. But you may be able to make broad brush improvements.
2. Log-sweep measurements are much better than pink noise - they reduce the influence of background noise and get you phase info.
3. That dip at 550hz or thereabouts - I bet is a phase issue. Could be crossover induced, speaker position induced, or acoustics induced. Have you looked at the phase plot?
4. You should probably not try to fix phase issues with an EQ bump.
5. DSP will not just be better at EQ, it will allow time alignment Left-Right, which will affect EQ (due to phase) massively
6. Going active will open a new world of possibilities.

Just don't get too hung up on a graph if you only have a graphic EQ. Use it to know where to make some level cuts, fine tune it by ear, and live with what you have, until you get a DSP. That would be my advice.
Log sweep measurements are problematic in cars (all the reflective surfaces?), but are usually used for troubleshooting phase issues. Moving mic averages taken with the RTA are generally better for eq'ing in cars.
 

·
Registered
Joined
·
56 Posts
Discussion Starter · #24 ·
I ended up buying the Helix M Four DSP and have played around with it. Thousand times better than running passive with the limited DSP of the HU. However, I've got some terrible phase issues now. I measured the distance between the center of each speaker and the center between my ears and plotted it into the DSP software. After this, I did some EQ on each speaker to get them as close to each other as possible, aiming at a house curve. This involved some pretty steep EQ cuts on some frequencies for each speaker. After doing so, I have a null around 180hz, before the sound goes from being centered to pulling right at 200, back to center, hard right around 400Hz, back to center and so on. Does the EQ shift the phase? If so, how do I combat this? Less EQ? All pass filter?
 

·
Registered
Joined
·
56 Posts
Discussion Starter · #25 ·
Hmm doing a sine sweep with only one speaker at the time reveals that what I'm facing probably is an issue with the cabin itself. When playing the left woofer solo, the sound gets pushed to the right between 60-70 Hz, cancels around 180Hz etc. I guess there's no easy way to combat this, right?
 

·
Registered
Joined
·
1,772 Posts
How can the sound be "pushed to the right" when only the left midbass speaker is playing? I'm not following...

Just take pink noise measurements of each speaker individually, then each speaker "pair" (L+R midbass, L+R tweeter, etc) - then take "Whole side" measurements (L tweeter + L mid + L midbass - same for right). Then all together. This will show you any phase-related issues.

Use the "moving mic" method with the MIC pointing straight up and down and go back and forth to each ear a few times while REW is averaging the measurements. This will get you a good measurement. Using sweeps is really not the best way to take measurements in a car, IMO. It doesn't tell the whole story like moving-mic pink noise measurements will.

If you do have some phase issues, you can possibly address them with allpass filters.
 

·
Registered
Joined
·
56 Posts
Discussion Starter · #27 ·
I can't tell you how it's happening, as I'm wondering about the same thing myself. Maybe multiple reflections from the right side which are in phase with each other? Idk but the sound moves around even when it's only one speaker playing.

I already did the type of measurement and adjustment you mention, and even though it looked good in REW it sounded weird in the midbass/midrange area, as if some of the frequencies were out of phase. That's why I did the sine sweep to begin with. Some of the cuts I did to get the left woofer flat was around -6 or -7 dB and I was wondering if that may have caused the phase to shift in that area. I read something about IIR shifting phase and FIR not shifting phase and that the Helix used IIR, but I did not quite understand it. Hmm I might try to cut the left woofer less in those areas and see if that helps with the difference in phase.

Edit: I'll post some measurements tomorrow. It's 1AM here now, can't wake up my neighbors
 

·
Registered
Joined
·
2,911 Posts
I can't tell you how it's happening, as I'm wondering about the same thing myself. Maybe multiple reflections from the right side which are in phase with each other? Idk but the sound moves around even when it's only one speaker playing.

I already did the type of measurement and adjustment you mention, and even though it looked good in REW it sounded weird in the midbass/midrange area, as if some of the frequencies were out of phase. That's why I did the sine sweep to begin with. Some of the cuts I did to get the left woofer flat was around -6 or -7 dB and I was wondering if that may have caused the phase to shift in that area. I read something about IIR shifting phase and FIR not shifting phase and that the Helix used IIR, but I did not quite understand it. Hmm I might try to cut the left woofer less in those areas and see if that helps with the difference in phase.

Edit: I'll post some measurements tomorrow. It's 1AM here now, can't wake up my neighbors
IIR filters do change phase along with amplitude, but any changes made should be correcting both. Sounds like, what you said, you have some enviromental craziness going on.
 

·
Registered
Joined
·
56 Posts
Discussion Starter · #29 ·
Ok, as promised, here are some measurements. First left woofer...

299896


...then right woofer...

299897


... and this is how they look compared to each other

299898


I can see that the 180 Hz issue on the left side seems to be an issue with cancellations due to where the woofer is mounted. Seems to be a similar dip around 200 Hz for the right woofer.

Here's a measurement of them playing together (forgot to mute the tweeters but that shouldn't matter)

299899


Unless I'm reading this wrong, there shouldn't be any extreme deviations, yet the image is pulled hard right around 380-450 Hz. I tried playing with all pass filters, and while I managed to get the image fairly centered around the 400 Hz area, it opened up another can of worms. There doesn't seem to be an easy way to fix this. Unless anyone here has some trick up their sleeve that I could easily implement, I guess I'll leave it be for now while I read up on all pass filters etc.
 

·
Registered
Joined
·
56 Posts
Discussion Starter · #30 ·
I used one all pass on the left at 396Hz with a Q of 13 and one on the right at 406 Hz with a Q of 11 and somehow the phasing is no longer as bad as it was. Idk if that is the correct way to fix it but at least now music doesn't feel "unbalanced" in the midrange.
 

·
Registered
Joined
·
364 Posts
I used one all pass on the left at 396Hz with a Q of 13 and one on the right at 406 Hz with a Q of 11 and somehow the phasing is no longer as bad as it was. Idk if that is the correct way to fix it but at least now music doesn't feel "unbalanced" in the midrange.
I dont know a ton about APF, but you are in theory they should basically cancel each other out.
 

·
Registered
Joined
·
56 Posts
Discussion Starter · #32 ·
The Q is really high and there is a 10Hz difference, so the phase changes enough in relation to each other to get closer to 0 degrees. The bandwidth is around 30Hz for the 396Hz APF and around 37Hz for the 406Hz APF. Using an all pass on only one side causes cancellation and terrible phase shift, but doing it this way reduces the phase difference in the troublesome range. If there was a way to rotate the phase a certain amount of degrees in a specific frequency range, I would have tried that instead of using APF this way.
 

·
Registered
Joined
·
364 Posts
The Q is really high and there is a 10Hz difference, so the phase changes enough in relation to each other to get closer to 0 degrees. The bandwidth is around 30Hz for the 396Hz APF and around 37Hz for the 406Hz APF. Using an all pass on only one side causes cancellation and terrible phase shift, but doing it this way reduces the phase difference in the troublesome range. If there was a way to rotate the phase a certain amount of degrees in a specific frequency range, I would have tried that instead of using APF.
What order APF are you using?
If it's getting 360° range of rotation, you should be able to tighten up the q even more and use only one in the region you need
 

·
Registered
Joined
·
56 Posts
Discussion Starter · #34 ·
Second order filter. I'll give it a try when I have time!
 

·
Registered
Joined
·
45 Posts
I used one all pass on the left at 396Hz with a Q of 13 and one on the right at 406 Hz with a Q of 11 and somehow the phasing is no longer as bad as it was. Idk if that is the correct way to fix it but at least now music doesn't feel "unbalanced" in the midrange.
Do you have a new graph of the response?
 

·
Registered
Joined
·
56 Posts
Discussion Starter · #36 ·
I'll go do some measurements and post them soon!
 

·
Registered
Joined
·
56 Posts
Discussion Starter · #37 ·
LW:
LW.png

RW:
RW.png

LW and RW on the same chart:
LWRW.png

Midwoofers summed:
MW summed.png

All measurements on the same chart:
All.png

Edit: I think I gotta cut 200-400Hz by 4-5 dB later but I guess that's not what I posted these measurements for anyway
 

·
Registered
Joined
·
56 Posts
Discussion Starter · #38 ·
The dips around the 500Hz area has been bugging me, and having the tweeters and woofers crossed at 2500 isn't optimal. I'm considering buying a pair of Hertz ML700.3 but I'm not quite sure where to place them. Would surface mounting them in the door panel be a bad idea? Would I have to fiberglass/strengthen the area after cutting a hole the mids? And would it make sense to move the tweeters from the sail panels into the door as well?

mids.png

I don't have OEM mounting options for a 3-way, and I don't want to mess with the A pillars, as there are airbags there. The sail panels are kind of small so I don't know if I would be able to fit the mids there without some serious work beyond my skill level.
 

·
Registered
Joined
·
1,772 Posts
I may have missed this, but what speakers do you currently have installed and where are they installed? Mainly, I'm curious as to the size of your current speakers and where they are installed - and if they are OEM locations or not.
 

·
Registered
Joined
·
56 Posts
Discussion Starter · #40 ·
I currently have Pioneer TS-Z65CH (6.5" woofer and 1 1/8" tweeter) installed in OEM locations. Woofers in lower door and tweeters in sail panels. Active crossovers, woofers LP at 2.5kHz 24dB slope, tweeters HP 3kHz 24 dB slope. Idk if doing an underlap is a sin or not but it was a compromise.
 
21 - 40 of 54 Posts
Top