Tonality vs RTA

[Jscoyne2]

Haha. I sit spl guys in my car all the time. Give them sq demos and then turn up the subs and have some 142db at 25hz fun. Show them its entirely possible to have both worlds.

Sent from my XT1710-02 using Tapatalk

[blockrocker]

Yes, the best of both worlds is my ultimate goal. Its been a fun process. Just the other day I got JimmyDee in the truck after a much improved tune and gave him the full meal deal. I think he was pleasantly surprised. I've never measured this system before (though I should) and believe I'm just over the 140 range as well. My highest ever was 148, many years ago, with 2 15" cerwin vega strokers. An old friend sold me two brand new "old stock" in like 2002 or so. Even then those things smashed hard. Ran them on some Lanzar Optidrive amps. Fun setup.

If I had to put some goals on the current build, I want a 150 and an SQ win at a judged event. I'm not sure I'll ever get the chance to put the truck in a show, but I put some feelers out this winter - so hopefully I get an email or two in the spring. I miss the days of social stereo shows. I'm not a very competitive guy, but do enjoy some bragging rights every now and then.

[Babs]

May I ask how picky you get with smoothing your response to curve? IE: 1/3, 1/6, 1/24 smoothing etc. May I also ask for a few details on how you get everything in phase? Is it just measurements and tracerite calculator or is there a more advanced trick? Its great hearing from guys with so much experience.

In the Helix RTA I typically run 1/6 but occasionally I'll double check at 1/12 octave. I dunno that I have a whole lot of experience but I too am continuing to learn. I just throw in a Helix RTA custom curve from Jazzi's spreadsheet output approximating the crossovers applied to flat, then I go. That said, I am going to try a few different approaches for giggles.. Never set down too hard on just one process, or you might miss a great opportunity to learn. Try different things and learn from them if possible.

Ditto to what BlockRocker said!!

What do you mean "electrical target which includes the slopes." I understand the difference between 'electrical' & 'Acoustical' Freq Response. So are you referring to using just the 'electrical' settings to flatten out the Freq Response before you apply any EQ'ing ? I 'm having a hard time to fully understand your method.... Could you expand a bit and explain with a lil more detail ?
'

Electrical target means the target curve I pull over into the RTA, such as the Helix RTA or you can actually generate one within REW's EQ function. Acoustic means whatever you've measured coming from the system under test.. your speaker(s). What I mean is when I do an individual driver for example, my electrical target will be flat but with a slope applied to look like a 24db response should look like. So lets say I have a tweeter highpassed at 3500hz. The target will be a flat line down to 3500 with a downward slope from there. The full response bandwidth the driver should theoretically be playing. I'll EQ to that.

[blockrocker]

Wonderful replies Babs! I love trying new techniques and learning how other guys get to the results they have. I've also experimented with Jazzi's speadsheet and found it to be a very cool tool. In fact, Justin seems to be one of the most knowledgeable guys on the forum. Lots of good suggestions from lots of experienced people in this place. Thanks for your input!

[Babs]

Thanks. Well that said, I’m finding I simply cannot tune each midbass driver alone and expect the same results because when modal issues in the car take over the two drivers can be aligned well but still have phase issues in their individual response due simply to car “room” bass resonance (room mode cancellation etc). So I will have to take the approach of EQ’ing the two midbass drivers together for at least a good response, and phase will have to be what it is.. They mate up in phase nicely at least in the crucial crossover region between midbass and mids, so it will be a tougher bit of work for sub stage.


Sent from my iPhone using Tapatalk Pro

[blockrocker]

Believe it or not, its very reassuring to hear other guys have to deal with the same issues as I do. Sometimes I convince myself I'm doing something wrong to get the responses I get from the RTA, when in fact its something everyone has to deal with i a car environment, and its just a matter of experience, knowledge, and compromise to achieve the best result. The more times I see a cat skinned a different way, so to speak, the better I get too.

[Jdunk54nl]

I posted this in another thread but find it very beneficial to realize how being just a little out of phase with time alignment alone can make a huge difference. Remember 180 degrees out of phase (so top/bottom, either 180 or -180, for one speaker and middle, 0 degrees, for another)

Minute 2:20 is where it gets good
https://www.youtube.com/watch?v=R_G4wrX1nk4

If you watch the top IR graph, you can see how that is lining up with the center fill as he delays it, the phase lines up, and then you can even see the FR get impacted by the phase as some frequencies go in and out of phase with each other and the FR goes up and down. Very cool stuff in that video and super important. (Note* I could be wrong about the FR graph because I don't know if he has both speakers playing and it might just be an artifact of changing the delay on the one speaker)


I think the big thing between tonality and what the normal RTA graph shows is definitely phase issues for most. REW isn't the easiest program to use to get phase information of good use. I think this is the one area where REW lacks and wish you could see the measurements in real time like in SMAART. That way, you would at least know that you did the best you possibly could. There is still going to be some compromises in car. The Phase we really have control over is between two speakers, not really going to change in vehicle issues with speakers and reflections.


Using Justin Zazzi's spreadsheet definitely helps get the phase better in the crossover region and gives a good target curve so volume phase issues should be well taken care of.

[blockrocker]

That was a very cool video! Its shocking to see how fine the line can be at times between in and out of phase. It is also cool to see how much of the bandwidth stays in and out. I had no idea how different one m/s to little would look so much different than one m/s too late. If anyone needs me, I will be googling SMAART. ;-)

[Babs]

Yeah looks like he was adjusting in tenths of a millisecond. I think it's important to note the phase change was much larger as you go up the frequency band, such as that wrap that occurred at 4-5khz when he went .1 ms too far.

Believe it or not, its very reassuring to hear other guys have to deal with the same issues as I do. Sometimes I convince myself I'm doing something wrong to get the responses I get from the RTA, when in fact its something everyone has to deal with i a car environment, and its just a matter of experience, knowledge, and compromise to achieve the best result. The more times I see a cat skinned a different way, so to speak, the better I get too.

Certainly, and I had forgotten some fundamentals.. When I posed the question on SSQ FB page, Erin hit me up and reacquainted me with Schroader frequency and the concept that about everything below it is modal and everything above it is reflection, due to the size of wavelength and dimensions of the "room".. In most cars this happens around the 300-400hz region which you can tell if you do sweeps moving a mic, it diverges from that point up, but stays fairly static from that point and below. This indicates even spatial averaging at the seat/head position reaches a point it's unnecessary in the bass region because it doesn't waver much at all, and it's more driven by modal standing waves in the car then reflective energy from surfaces.


All that for a point being we really kinda sorta can approach a car tune as though we're tuning two environments. One reflective from that frequency range and higher, and one modal from that frequency and lower. For the latter, you can simply tune both drivers together for response to a curve because in the end it's their combined response with the car's modal affects that are un-shakable, that must at least have an acceptable frequency response, even if you cannot achieve exact phase coherence throughout the midbass and sub region. It's a "best you can with what you got" approach.

[JimmyDee]

Here's what I've found;

I have a tune that I did in my BMW 440i, a few years back.
It was done with a Helix DSP-Pro II, and was aligned with the Helix house curve, and then (heavily) modified to suit my personal taste.

Since then, I have done a few more tunes (a couple in the old 440i, and a handful in the new X6-M).
All of them were almost perfectly aligned with the Helix curve. But it never sounded right to me...
I always went back to loading that old tuning file from my 440i, and 'tinkering with it'.

The same thing just happened when I installed my new DSP-Ultra.
I had it tuned with the RTA, and damn near perfectly aligned with the Helix curve... and I just didn't like the sound of it.
Sounded a bit hollow, and definitely lacking highs (which is opposite to what you've experienced).

I ended-up loading that old 440i tune, and making a few adjustments (time alignment distances, and a few minor EQ adjustments).
I gotta say; that old tune still sounds the best to me.

In all honesty; I know that old tune is nowhere near aligned with the house curve. But I think we just used to a 'certain sound'.
And since I've basically used the same components in my past few vehicles, I'm used to that 'certain sound' from each speaker.

At the end of the day; I would say that my tune is about 70% aligned with the house curve.
The other 30% is manual, and some of it is WAY OFF the curve.

Tune your system so it sounds good to you... that's all that matters.