[stmblaster]

Hello!

This year I decided to try my first custom audio solution in a VW Touareg (2009) and I am having trouble with it. I had the hardware install done by a shop (everything works perfectly, also great looking) but it is me who has to do the software tuning.

The DSP is Mosconi/Gladen 6to8 Aerospace and the speakers are Gladen, subwoofer also Gladen with custom built enclosure.

The tweeters and midranges are installed in the A pillars (almost with no angle - facing each other) at ear level and the midwoofers low in the doors in their stock location. The subwoofer is in the trunk, simple box centered on the backseats' back.

All amplifiers were (physically) set to a maximum safe gain level (measured by the shop guy with oscilloscope for clipping and turned down to a safe place).

Things I have done so far:

• checked the polarity of the speakers with the guy that worked on my car -> all are correct;
• set the crossovers according to AudioFrog guide with LR24 filters (two BW12 for each LP/HP):

- subwoofer HP 20Hz, LP 80Hz (my subwoofer has a very steep slope so it sounds/measures better crossing it at 80Hz instead of 60Hz);
- midbass (6") HP 80Hz, LP 350Hz;
- midrange (3") HP 350Hz, LP 2500Hz;
- tweeter (1.5") HP 2500Hz;
- set the delays - measured with tape from the tip of the microphone to each speaker grille's center, entered the distances in the calculator AudioFrog provides and then the milliseconds in the DSP;

After these the image is focused in the center, everything sounds natural but the frequency response is crap, voices move around, some parts are harsh and others you can't hear etc., but the sound is somewhat listenable to. It sounds as if there's amazing quality, but **** tuning.

What I did next was try to flatten the curve towards AudioFrog curve which I managed to match within 1dB (except for big dips) by only cutting frequencies , but it sounded horrible. I calculated each Q factor with the recommended formula and tried to not do any surgical Qs (like 10) as much as I could.

The tweeters sounded as if they struggled to produce music and the midrange was shouty and harsh and felt like everything was screaming at me for dear life. The midwoofers and subwoofer blended in nicely, tight bass coming from the front of the car.

Now I have no idea what to do to make it sound better and I am asking for your input. I have measured the system with no EQ or digital Gain modifications so you can get a better view.

Left Channel:

Right channel:

Center (both channels playing):

Delays:

 

[vectrasoundz]

try setting your scale to 5db increments instead of 2db and apply 1/6 smoothing and post the graphs again.

 

[bbfoto]

^This.

Where are you located?

The tweeters sounded as if they struggled to produce music and the midrange was shouty and harsh and felt like everything was screaming at me for dear life.

The statement above and the steep roll-off of the high frequencies in your graphs above kind of makes me think that the channels from the DSP for the midrange drivers are being sent to the tweeters, and the signal for the tweeters is being sent to the midrange drivers. Have you confirmed that all drivers are receiving and playing the proper signals from the DSP and on to the appropriate amplifiers?

I think that some RCA cables may have accidentally been swapped somewhere between the DSP and the amplifiers...Or the speaker wiring for the tweeters and midrange drivers was swapped.


And which Gladen component speakers do you have? The Aerospace 28 tweeter, 80 3" midrange, & 165 6.5" midwoofers?

Is your main playback source the VW OEM stereo head unit, and if so, how is it connected to the DSP? Which head unit model is it? Have you used VAGCOM to set your head to output a clean, flat, unprocessed full-range signal?

If not, are all of the OEM head unit's audio settings set to "0" or neutral, and any effects such as "Loudness" etc set to Off?

What is the quality of music files you are using? And are you playing your music files via USB memory storage or Bluetooth from your smartphone through the OEM head unit? CD discs?

What specific music tracks are you playing that make the voices or instruments move or shift position in the soundstage, and what tracks sound especially harsh or weird?

Are you using the Gladen Aerospace 6to8 DSP Remote to control Master Volume, or the head unit's volume?

What Gladen/Mosconi amps are powering the system?

Can you take a few cellphone photos of the speaker locations in your doors and A-pillars and post them here?

Other than measuring, have you played band-limited pink noise or sine wave tracks of individual frequencies from 30Hz, 40Hz, 50Hz, 60Hz, etc to 18kHz to hear where the the dips and peaks in level are, and which frequencies shift the center image to the left or right of center?

If you're using the OEM head unit, do you have another source such as your smartphone that you can connect via digital or headphone output directly into the DSP inputs to bypass the head unit?

What measurement microphone are you using? Did you load its Calibration file into the measurement software?

 

[cathul]

If you have a two channel input to your subwoofer, you should disable one input when measuring each individual side.
That the tweeter has such a steep roll off in the highs is normal if you play them so far off-axis. My Gladen Zero Pro 165.2 DC shows the same behavior in my Mustang with tweeters in the stock location.
Are subwoofer and right midbass really the same distance?

What i would do:

First turn off the bars in the RTA window and use 1/48 octave smoothing while measuring. You can smooth to 1/6th octave afterwards after saving the measurement.
With the bars and 1/3rd octave you cannot see the real problem. If you want to adjust the EQ while playing Pink Noise disable bars and use 1/6 smoothing.

1. remeasure each side with one input of the sub disabled (left side with right side input of subwoofer muted and vice versa) and EQ enabled
--> if you have both sub inputs enabled when measuring each side alone you don't have enough subwoofer volume when you play everything!
2. Adjust levels of subwoofer afterwards if needed
3. Relisten if it sounds better.

What do the individual speaker measurements show (1/48th octave measurement without bars and smoothed to 1/12 or 1/6 octave)?

 

[SkizeR]

I'm guessing you measured and EQ'd with the mic in a single position while you are outside of the car?

 

[dumdum]

try setting your scale to 5db increments instead of 2db and apply 1/6 smoothing and post the graphs again.

Why? Finer is ok… coarser is not

I agree on the smoothing… but if measured in a single point in space the results will never be optimal and will result in peaks and dips

But… above you have an example of why too many cooks spoil the broth

If you are going to follow the audio frog guide follow it… don’t freestyle it and decide you need an 80hz crossover on the sub because you think it needs it

Start at the start of the guide and follow it, religiously, then at least you have a start point, then you can start trying other stuff, but so far you have done what you think and partly don’t what the guide says and now you wonder why results are half arsed…

There are lots of guides to tuning, audiofrog isn’t my favourite with its static mic, but before slating the process and how it works try actually following the process fully

 

[stmblaster]

@vectrasoundz - I will remeasure and post as soon as I can.

@bbfoto - I am from Bucharest , Romania.

I don't know how to check if they are wired incorrectly, the channels 1+2 are tweeters and when I mute them the tweeters stop. I haven't found anything in the Mosconi 6to8 manual that indicates you have to wire specific speakers to specific channels, but the shop wired them likes this: 1+2 = tweeters (2 channel amp), 3+4 = midrange and 5+6 = midwoofers (together on a 4-channel amp) , subwoofer on a separate amp (mono).

If I mute any specific channel, the expected speaker mutes. I can isolate each speaker.

The components are the following:

Speakers: Gladen Zero Pro 165.3 https://www.gladen.com/GladenCompoCoaxPDFs/SpecificationsZERO165_3.pdf
Subwoofer: Gladen SQX 15 Extreme https://www.gladen.com/GladenWooferPDFs/SQX15Extreme.pdf
Tweeter amp: Gladen RC90c2 https://www.gladen.com/GladenAmpPDFs/RC90c2.pdf
Midbass+midrange amp: Gladen RC105c4 https://www.gladen.com/GladenAmpPDFs/RC105c4.pdf
Subwoofer amp: Gladen RC600c1 https://www.gladen.com/GladenAmpPDFs/RC600c1.pdf
Bluetooth to optical interface: Mosconi DSP-AMAS-LD4C https://mosconi-system.it/product/dsp-amas-ld4c/
Bluetooth controller: Mosconi DSP RC-Mini https://mosconi-system.it/product/rc-mini/

Playback is either via iPhone bluetooth to DSP-AMAS-LD4C (Apple Music Lossless) or via Android (LDAC - Apple Music Lossless). No EQ in either case, but Sound Check keeps turning on, on iPhone. I kept both phones 2-3 clicks away from full volume (as per recommendation from the installer - I guess Sound Check does the same thing?).

The whole system is not connected to OEM head unit in any way (only the rear speakers are connected to OEM unit, but never used - stock speakers).

I will post a few photos later when I take the measurements too.

I have not tried to play individual bands to try and hear where the dips and peaks are because I find myself highly unreliable to figure it out by ear. The best I can tune by ear is mostly adjusting bass/treble in OEM systems and even Harman/Kardon EQ is too much - every time I tried to use it the sound ended up great in the evening, only to sound like crap in the morning.

The measurement microphone I am using is from AudioFrog (UMI-1). I have both calibrated the included USB audio card by running the output into the input and sweeping it and also used the microphone calibration file provided by them (followed their guides precisely).

@cathul - This is probably why I thought I need 80Hz LP on the subwoofer. I always played left/right channels with both the subwoofer channels on and ended up lacking bass with both channels playing.

I've measured with tape the distances and they kind of are the same distance. Speakers were easy to measure precisely but the sub I measured a horizontal line straight to the cone of the speaker (with rear seats folded) and then added a few more cm to account for the vertical height distance between top of the enclosure and the speaker cone.

I will measure with only one sub channel playing and post later - I have not done this before because I didn't know if it's safe to do or if it was useful in any way. I will also measure each speaker individually and post later.

@SkizeR - Yes, I have measured and EQ'd with the mic in the ideal position - center of headrest where my ears are, while I was outside the car (can I measure while being in the backseat without influencing much? This would totally save some hot chocolate/tea and the winter experience I get when measuring this time of year).

@dumdum - I followed the AudioFrog guide fully with no personal "innovations" at least 5-6 times and I'm sure it works but I'm quite sure something is off with what I'm doing and I can't figure out what. Now at least I know I should have measured with only one channel of the subwoofer when measuring only one side - this is probably why I crossed the subwoofer at 80Hz instead of 60Hz - too low volume at the crossover point on the subwoofer due to my error.

One thing that confuses me is that after setting delays/crossovers correctly (not much to speculate or try to change here), I have to set speaker levels before EQ. How should I set speaker levels (the logical thing that comes to mind is just leave all DSP gains at 0 and cut with EQ - clearly didn't work - speakers sound like they're struggling to survive).

If you cut wide with EQ (say the whole tweeter range) is it the same as cutting gain? Because even if the frequency response of both channels with the gains set to 0 and everything EQed is almost right on the target curve, everything sounded off - no focus, right side sounded louder (center vocals coming almost from the right speaker), stage extending all the way to the B pillar on the right side and on the left side between the center of the dash and the left A pillar everything was silent (except for completely left panned sounds which came straight from the left speaker).

 

[cathul]

EQ each speaker individually and remeasure without altering the volume knob or the amplifier gains during the whole process.
If you then see that your f.e. left mid is 2db hotter than your right mid and both are set to 0db in the Mosconi interface, either raise the right tweeter by 2db or even better, lower the left tweeter by 2db.
If you do this for each channel you should have a perfectly matched volume on all speakers.
If you set the amp gains with -5 or -10 db testtones never go above 0db in the Mosconi interface for any driver. In this case have your "hottest" individual driver set to 0db in the Mosconi software. The other drivers should be below 0 db in this case to keep the relative level between drivers.

 

[stmblaster]

EQ each speaker individually and remeasure without altering the volume knob or the amplifier gains during the whole process.
If you then see that your f.e. left mid is 2db hotter than your right mid and both are set to 0db in the Mosconi interface, either raise the right tweeter by 2db or even better, lower the left tweeter by 2db.
If you do this for each channel you should have a perfectly matched volume on all speakers.
If you set the amp gains with -5 or -10 db testtones never go above 0db in the Mosconi interface for any driver. In this case have your "hottest" individual driver set to 0db in the Mosconi software. The other drivers should be below 0 db in this case to keep the relative level between drivers.

Alright , this is something I have not tried. Mosconi software doesn't allow me to set any gain over 0dB, only negative. Amp physical gains are already set and I never touched them since. Even with -10dB on everything there is still more than enough volume so that won't be a problem.

I'll try it and report back as soon as I find the time.

Thank you and also everyone else for the input!

 

[stmblaster]

A few pics of the setup:

 

[stmblaster]

And a pic with the internals of the box (while being installed):

Also the front doors while being insulated:

Forgot to mention I never heard a proper SQ setup in a car before, only extreme SPL (not my thing - I literally covered my ears in fear on the first drum kick, out of defensive reflex - to me it sounded like a grenade went off). The only ones I had enough time with are Burmester in Porsche and Harman/Kardon in BMW and high end headphones.

My expectations are to at least get the Burmester sound which I hope is realistic (except for the rear fill they do which sounds great but maybe at a later time).

 

[SkizeR]

@SkizeR - Yes, I have measured and EQ'd with the mic in the ideal position - center of headrest where my ears are, while I was outside the car (can I measure while being in the backseat without influencing much? This would totally save some hot chocolate/tea and the winter experience I get when measuring this time of year).

Try getting spatial averages. This one mic method doesn't yield great results to an experienced ear

 

[stmblaster]

Finally managed to measure. I positioned the mic horizontally pointing straight towards the windscreen and measured in 9 points (center,left,right,up,down,up-left,up-right,down-left,down-right in the same vertical plane).

Should I also measure 9 points again twice, once with the headrest a few inches forward and once with it a few inches backwards? (I don't change my driving position tho). It would end up being an average of 27 points , kind of like a cube instead of a rectangle of points, but is it worth it? I would like to have a solid base before I start using EQ.

This contains the measurements and is .mdat format from REW:

9-Point Average.mdat

1 file sent via WeTransfer, the simplest way to send your files around the world

we.tl

I have done these measurements with the sub crossed at 60Hz and midbass at 80Hz. I have found that lowering all speakers by -4.5dB aligns them with subwoofer output, but there is still a big gap @ 60-80Hz and if I cross the subwoofer at 80Hz this gap gets better (add a boost of 5-7db to 70Hz to both sub and midbasses and it is almost gone).

Which way is the best to EQ? I would do it manually like in the AudioFrog guide (calculate Q) but if there are better ways I'd like to try them.

Another question: Should I have gotten DIRAC instead of all this? Seems like it does everything I am doing now but quicker and better and automatically.

 

[vectrasoundz]

Looks like your time alignment between sub and midbass is off....

 

[lisuku]

And a pic with the internals of the box (while being installed):
View attachment 317650

Also the front doors while being insulated:
View attachment 317652

View attachment 317653

View attachment 317654

Forgot to mention I never heard a proper SQ setup in a car before, only extreme SPL (not my thing - I literally covered my ears in fear on the first drum kick, out of defensive reflex - to me it sounded like a grenade went off). The only ones I had enough time with are Burmester in Porsche and Harman/Kardon in BMW and high end headphones.

My expectations are to at least get the Burmester sound which I hope is realistic (except for the rear fill they do which sounds great but maybe at a later time).

I woul like to know what treatment is that black foam over the butil insulator, how is it called