Quote from: Fast SHO on December 23, 2013, 03:27:55 PM
I was having an issue with the sound being a bit muted on the highs.  Couldn't figure out how to fix it until I changed the woofer low pass to 18KHZ.  That is high, but the woofers are rated for it.  The result is that it really filled out the sound.  Here is how my freqs are set.

I know most people say to low pass the woofers at 2500 with a low pass overlapping the sub, but for some reason it just sounds better with it at 18KHZ and tweeters high passed at 2500.

spot on. sounds like a terminology problem...... you got it now!!  experiment with the 18Hz between the subs and mids.... at high volumes this could be hard on the mids...... Most would shoot for a flat line without "valleys" across the top of the curve you posted, but if you like the sound that's the key!

I'll work on the valley's.  I try and get it where it is technically supposed to be then move it around based on what sounds good.  It is the EQ settings I'll tackle later after more reading, and then time alignment.

I have the mids low passed at 18,000HZ, doing that filled in my highs.  Previously they sounded muted when I had them low passed at 2500 (i.e. playing everything below 2500hz down to the high pass of 70hz).  For some reason relying on the tweeters starting at 2500hz to 20,000hz without the mids was causing that problem.

Thing is, any time you overlap you introduce potentially devastating issues.

Example of one, is your woofer and tweeter in their overlap are now operating at slightly different distances from your ear, so it creates interference patterns.  Another, the tweeter's timbre playing those frequencies, and the woofer's timbre playing those, are different.  More interference (distortion), and so on. 

Minor overlap is potentially a great thing, a way to "fill".  But everything is a series of compromises.

Quote from: Lanson on December 23, 2013, 03:50:44 PM
Thing is, any time you overlap you introduce potentially devastating issues.

Example of one, is your woofer and tweeter in their overlap are now operating at slightly different distances from your ear, so it creates interference patterns.  Another, the tweeter's timbre playing those frequencies, and the woofer's timbre playing those, are different.  More interference (distortion), and so on. 

Minor overlap is potentially a great thing, a way to "fill".  But everything is a series of compromises.

Spot on Lanson!   That's some of the reasons that ultimately the final tuning is based on what the listener's ear likes and not an RTA which hears all frequencies equally.

I know the overlap is significant but for some reason when I cut the woofer off at 2500 and had the tweeter pick it up from there I was getting an odd sound I didn't like.  "muted, muddled" etc... would be a good description.

Once I set the woofer to go up between 18-20KHZ it solved the problem.  No idea why, but it sounded better.  My brother was listening and said the same thing.

I can adjust time alignment also with this DSP but the changes don't seem to be retained, so I need to figure that issue out.

Looks like I can adjust the actual delay or just use distance.

I was set on installing a sub on the side of the trunk but I love your install so much I'm actually thinking of chucking the spare.

Something to keep in mind when you start eq'ing is to pull down peaks and not boost valleys.  You can run out of headroom VERY quickly boosting and it may even be a dead spot you will just have to live with. 

The best thing I can recommend is getting the software roomeqwizard from hometheatershack.com.  There are also pretty good cheap mics you can get that just plug into a usb port or you can spend a little more and get a calibrated mic and powersupply/usb sound card.  The software is fantastic!  You use a spl meter to calibrate the software first.  You turn the volume till the spl meter says 75db then you tell the software what its reading now is 75db.  After that you can run sweeps or any tones of your choice and it will measure response, delays, reverb, you name it; all with very nice graphs.

Close to flat response is achievable but sounds pretty shitty.  You'll be wanting to dial in a "house curve".  At lower volumes our ears are more sensitive to midrange as stated earlier but as the volume goes up it gets flatter.  The "loudness"  button was mad to compensate for loss of bass and treble perception at low volumes but everyone left it on at high volumes and it sounded "better".

Quote from: Fast SHO on December 23, 2013, 04:38:06 PM
I know the overlap is significant but for some reason when I cut the woofer off at 2500 and had the tweeter pick it up from there I was getting an odd sound I didn't like.  "muted, muddled" etc... would be a good description.

Once I set the woofer to go up between 18-20KHZ it solved the problem.  No idea why, but it sounded better.  My brother was listening and said the same thing.

I can adjust time alignment also with this DSP but the changes don't seem to be retained, so I need to figure that issue out.

Looks like I can adjust the actual delay or just use distance.

If you're going to get this far into I REALLY recommend the rta software I mentioned.  It will show you what driver is capable of what then you can make decisions off the data.

the distance / ms in the delay is the same thing.  It just does the math from the distance you enter for you to get the ms.

Here, this helps




When you cross your woofer high, you'll introduce breakup distortion.  That's why usually we cut them at around 3k or so. 

Quote from: BiGMaC on December 23, 2013, 12:20:07 AM
I helped build 2 vehicles with Alma Gates (Google her.. seriously) and her son While I lived in Payson... one was a Bronco which was national SPL champ for several years in the early 90's (180DB+) and also the national sound quality champ, an F250 SC.

Man first the BMW and now Alma.  You're reminding me of my old TeamROCS days!  She passed away not too long ago if you didn't know.

Correction, here is the interactive one to help real-time

http://www.independentrecording.net/irn/resources/freqchart/main_display.htm

From here, you'll be able to move your mouse over each range, and get more info. 

Here's a guy I follow all the time, though he doesn't post too much any more.  The groundwork is already done.

http://zaphaudio.com/

Anyway, here's what cone ring and breakup look like on a waterfall RTA


And it manifests differently on a harmonic distortion plot


See, this woofer works OK (not great) up to around 2k or so, and then there's this NASTY breakup/ring that occurs around 6k.  And again, way up high out of the 20k bandwidth of most ears.  We would actually "feel" that speaker being grating.  This is a cheesy MCM-55 woofer btw.
freq response for reference


Here's a better one, a GR Research M165x

Waterfall

HD

and freq response


This woofer would probably cross much higher, with a lower "le" and a cone that is not susceptible to breakup, I'd wager 4k would be the sweet spot, on a steep slope.


By this same token, tweeters must be crossed properly too.  Not just excursion issues, but also distortion expectations at lower frequencies.



Another fun variable of multiple points of output for a given frequency is the nature of reflections/refractions.  A woofer in the door deals with those in a much different way that a tweeter on the A-pillar.  All adds to the fun.



LOL Now you guys see why I "push" the processor I do.  It gets maddening sometimes, thinking about all the variables.  Fun for the tweaker though.

Just not sure why playing the mid high like I did made it sound better?  When I low passed it at 2500 along with HP for the tweeter I was missing something.

Quote from: Fast SHO on December 23, 2013, 06:10:12 PM
Just not sure why playing the mid high like I did made it sound better?  When I low passed it at 2500 along with HP for the tweeter I was missing something.

My guess is that the woofer creates pleasing distortion, 2nd and 4th order distortion (even-order) and it is adding to the sound that you hear (and like.)

Which is fine.


I also think you may be experiencing a heavy drop in certain frequencies due to time alignment/phase relationships and by effect, polarity (180 deg out of phase.)



Consider, just play one speaker at a time.  Dead-flat.  Play accurate music like testing discs, start with LF woofer at full-pass, and then adjust crossovers slowly.  Take a notepad and write down what you hear, as you listen.

You'll get about an hour max before your ears are worthless, and then you take a break and go at it again.  I usually go 20-30min on, 30 min off doing something else in relative silence.

Play your music on a set of quality headphones (I use Sennheisers as my base test) and listen for the way they convey the sounds you are replicating. 

There's always more.

Quote from: dalum on December 23, 2013, 05:06:54 PM

Quote from: BiGMaC on December 23, 2013, 12:20:07 AM
I helped build 2 vehicles with Alma Gates (Google her.. seriously) and her son While I lived in Payson... one was a Bronco which was national SPL champ for several years in the early 90's (180DB+) and also the national sound quality champ, an F250 SC.

Man first the BMW and now Alma.  You're reminding me of my old TeamROCS days!  She passed away not too long ago if you didn't know.

Yea... there was a big write-up in the Arizona Republic... passing of a legend.