11-29-2016, 06:58 PM | #23 |
Captain
345
Rep 801
Posts |
Thanks for the response. As far as the measurements go, I'm completely aware of needing multiple measurements and averaging. Being that the measurements would vary so greatly between spots, I would have no way to recreate each one without some form of jig to do so. When I do this on concert systems I use about 3-6 mics. I take many traces and do averages to come up with EQ corrections that I apply. Hell, I take it one step further and create target curves for different engineers I work for from different averages I take over the course of working together. This way we can reach their desired outcome quicker, giving them more time to spend on the creative side of things. By correcting one spot in the venue doesn't mean it will translate to the next. The exact same can be said for phase alignment and coherency. At which frequency do you want to align at and where in the room.
This was something extremely basic with the very basic tools available in the car to shape the sound of the system. It was more or less meant as a demonstration with one mic position and how the different settings affected the one position. I'm also very aware of what the Fletcher Munson curve is. I did not know BMW's did any sort of intended loudness curve, so thank you for clarifying that. Flat or with the EQ I've given it I still think it does a very poor job at this. Especially in the low end. It's quite overpowering at quieter volumes. It's odd how it doesn't adjust the high end to compensate when turned down. I haven't taken any of that as negative criticism and thank you for the comments! Again, I can't stress enough that my measurements were not to be taken as anything absolute as what's happening in the vehicle. On something so basic as this, at the end, I just trusted my ears. Last edited by HKD126; 11-29-2016 at 08:03 PM.. |
Appreciate
0
|
11-29-2016, 08:48 PM | #24 |
Second Lieutenant
82
Rep 252
Posts |
Thanks Kevin for your input on this, I was hoping you'd drop in! Looks like you got the CLIO rig, I got one here as well, though I use it mostly in my lab for calibrated measurements, I use Smaart a whole lot more in the field because of the nature of my profession at many different types of venues (recording/production studios all the way to stadiums, and everything in between). I'd love to see how you place your mics in the car, I'm sure there was a lot of thought and experimentation that went into determining what works best for you.
This fletcher-munson thing is very interesting, and particularly the way Harman implemented it in the car audio system at different loudness levels. I'd be interested to know what the SPL levels were at for each of those measured responses were, and how the inverse of the response curve compares to the appropriate phon curve. Peculiar that they modified the curve somewhat, I'd love to pick the engineer's brain on why they did what they did. I too suspect part of it has to do with the effects of background road noise masking. However I would wish I could turn this "feature" off, though that's just me being a bit anal about it. Yes, definitely do spatial averaging when you're measuring a car audio system, particularly inside a small space such as a car, the more the better. Because of the small space, one thing you need to watch out for is the modal behavior of the space, I would venture to guess that some of it reaches up to the 800-1,000 Hz region given the smallest dimension of the car is governed by roof height. This is definitely evident in the measurements you posted since that is the area of greatest variance with plenty of cancellation happening here. To those who are trying to understand this behavior, think of the typical car acoustically as a giant bowl of jello below 1000 Hz, you poke it (or stimulate at those lower frequencies) and the whole space resonates! Move the mic in different places within the car and you get a wide variance in the response of this behavior, because it is modal (and position dependent) in behavior. Add to the complication that most definitely BMW engineers took considerable efforts to mechanically dampen the lower frequency behavior of the car to reduce the effects of road noise in the cabin, it's a science of it's own, and quite remarkable what they are able to accomplish these days. Above this modal behavior in frequency you're getting into the effects of reflections, multiple arrivals, on/off axis response of the individual drivers, a mix-match of the direct field coming from the loudspeakers and a whole ton of specular reflections to complicate things. Bottom line, it's a mess to measure! It's not an easy job to measure a car audio system. You can use an RTA (real time analyzer) and just look at the frequency response at various positions throughout the car, and ignore what's happening in the time domain (ignorance is bliss), and yes it will get you reasonably close to a good result if you're careful. Or you can use a higher level dual channel FFT and zero in on what exactly is happening both in frequency and time, and make informed decisions on what to equalize with the system, and yes it can be very challenging to understand. That's the science part of what we do. Finally add a healthy dose of experience and a "good ear" and adjust the tonal balance to suit your taste. The grey matter in between our ears is nature's best analyzer, that is the subjective nature of how we hear as humans. It's a careful balancing act between art and science. Whether it's a $1K, $10K, or $500K sound system, the principles measuring, evaluating, and optimizing a system are pretty much the same, though in practice you need to adapt to the varying conditions.
__________________
2017 430ix Gran Coupe
Last edited by Arthurrs; 11-29-2016 at 08:58 PM.. |
Appreciate
0
|
11-30-2016, 11:14 AM | #25 |
Enthusiast
466
Rep 1,791
Posts
Drives: F87 M2 / F31 330d
Join Date: Nov 2005
Location: Bavaria
iTrader: (0)
Garage List 2004 E46 BMW M3 [0.00]
2017 F87 BMW M2 [10.00] 2013 F31 330d Touring [9.50] 2009 E89 Z4 Sdrive 30i [9.00] 1989 E30 320is 2-door [10.00] 1986 E28 M535i [9.00] |
Many thanks to the OP for doing the measuring and posting it here!
The findings are very interesting and back up what I'm hearing in my F31 wagon with HK. I would really appreciate if someone would do the measuring on an F31 wagon and post optimized EQ settings, as I think it differs a lot from the sedan.
__________________
2017 F87 ///M2, Long Beach Blue, 6MT
2013 F31 330d Touring, ///M-Sport, Alpine White, Alcantara/Cloth, 19" Style M403, 8AT 2009 E89 Z4 SDrive 30i, Titanium Silver, Coral Red Leather, 19" Style 296, 6MT 1989 E30 320is, ///M-Technik II, S14 Engine, Sterling Silver, Black Leather, 16" BBS RS 003, 5MT 1986 E28 ///M535i, Agate Green, Pine Green Velours, 16" ALPINA, 5MT |
Appreciate
0
|
11-30-2016, 01:04 PM | #26 |
Captain
345
Rep 801
Posts |
Added a whole bunch more to the main post!
Side note - Arthur, tell Rational Acoustics to hurry up and come out with a multichannel Smaart iO already! 8 channels and no linking the current one together! Last edited by HKD126; 11-30-2016 at 01:12 PM.. |
Appreciate
0
|
11-30-2016, 02:43 PM | #27 | |
Second Lieutenant
82
Rep 252
Posts |
Quote:
Did you download Smaart 8.1 yet? Vastly improved trace capture capabilities!
__________________
2017 430ix Gran Coupe
|
|
Appreciate
0
|
11-30-2016, 03:01 PM | #28 | |
Captain
345
Rep 801
Posts |
Quote:
I have not downloaded it yet. I completely forgot it came out. My biggest complaint about 8 has been the recapture function and file hierarchy. Both drive me absolutely bananas! That and the inability to use the find time function with the remote client window.... I thought it would finally be my solution to walking around with a mic and doing transfer functions.... And SPL not being available in the remote client window.... I'm about to update right now though! Thanks for the reminder. As a rule, I update nothing while on tour. Maybe that's why I forgot. Anyways... Cars. Ha. |
|
Appreciate
0
|
12-02-2016, 11:26 AM | #30 |
Captain
353
Rep 687
Posts
Drives: 22 TR X3MC
Join Date: Aug 2015
Location: New Jersey
|
glennQNYC
__________________
2022 X3M Comp
2013 335i xDrive (Sold) |
Appreciate
1
Dog Face Pony Soldier9719.50 |
12-02-2016, 01:25 PM | #31 |
Private First Class
118
Rep 178
Posts |
Hey, we're Integral Audio (not Integra Audio)! Appreciate the shout out though!
Only had time to give a quick glance at the updated/added measurements, but a couple quick observations: Summing/Averaging: Looks like you summed the multiple measurements within Smaart. That's going to retain phase in the calculation. In order to reflect what you'd hear you need to strip out phase and sum/average only the magnitude response. Otherwise it's like simply summing microphone outputs. Smoothing: Without knowing exactly what's being done, I think you are over-smoothing. Smooth anything enough and it'll look good. Not sure if you're smoothing multiple times, or what, but those graphs look far too clean to be even 1/3rd octave. There are some structural limitations that should prevent you from achieving the response curves you're getting, they shouldn't be that smooth. Also, I can say for certain that the high frequency rolls off much more than you're getting above 10kHz. Not sure what/why/how you're getting there, but just mentioning it because if there's a problem there something is happening somewhere in the process that's causing it. Again, if the goal is using this data to meaningfully adjust settings, the data needs to be accurate or the corrections you apply will be wrong. The whole garbage-in, garbage-out thing. 1/3rd octave most closely resembles what our ear/brain hears, but there are limits. In many cases that's too much smoothing, or at least that's what we've found. A good caveat to keep in mind is that the audibility of any peak or null is proportional to the area underneath the peak/null. I.e. a low-amplitude, low-Q peak can easily be (and usually is) more audible than a high-amplitude, high-Q one, even though it doesn't look nearly as dramatic on paper. Phase: I know this was discussed with differing opinions, but my recommendation would be not to focus on phase so much. The phase data you are seeing isn't measured, it's computed. Phase is the slope of the magnitude response, and it's being computed here from the frequency response data via Hilbert transform. And in this context it's meaningless, because what you are actually seeing is a jumble of multiple speakers (some of which are duplicating frequency bands) and reflections, etc. You aren't getting real data, and even if you were you don't have the ability to apply any filter that would do anything about it. If you were measuring individual drivers and designing crossovers (like we do) that would be one thing, but in this context what are you going to do with any of it? You can try to knock down any peaks that sort of align with whatever eq control you have, and that's it. Loudness: You asked about why they didn't boost the high frequencies in their loudness contours. They actually do, it just isn't very apparent in the graphs I posted. Here's the actual loudness that's being applied: View post on imgur.com It looks like it's much lower at the high frequencies, but it really isn't. That's the electrical output. The acoustic response will be much greater due to shorter path length, fewer obstacles (i.e. seats) to attenuate, and probably most importantly, the higher sensitivity of the tweeters relative to the underseat woofers. Also, we haven't measured the HK amp output, but FWIW I can tell you that Harman seems to like to do roughly the same thing across the board for BMW/MINI, and this is what the HIFI amp output looks like: View post on imgur.com
__________________
Kevin Bennett
Integral Audio - We make the finest automotive audio systems on the planet. Have you seen our SoundstageDSP™ for BMW 2 series and SoundstageDSP™ Ultra for BMW 3 & 4 series? www.integralaudio.com |
Appreciate
2
Dog Face Pony Soldier9719.50 badbread332.50 |
12-15-2016, 04:44 PM | #34 |
Enthusiast
466
Rep 1,791
Posts
Drives: F87 M2 / F31 330d
Join Date: Nov 2005
Location: Bavaria
iTrader: (0)
Garage List 2004 E46 BMW M3 [0.00]
2017 F87 BMW M2 [10.00] 2013 F31 330d Touring [9.50] 2009 E89 Z4 Sdrive 30i [9.00] 1989 E30 320is 2-door [10.00] 1986 E28 M535i [9.00] |
This is the HK System...
__________________
2017 F87 ///M2, Long Beach Blue, 6MT
2013 F31 330d Touring, ///M-Sport, Alpine White, Alcantara/Cloth, 19" Style M403, 8AT 2009 E89 Z4 SDrive 30i, Titanium Silver, Coral Red Leather, 19" Style 296, 6MT 1989 E30 320is, ///M-Technik II, S14 Engine, Sterling Silver, Black Leather, 16" BBS RS 003, 5MT 1986 E28 ///M535i, Agate Green, Pine Green Velours, 16" ALPINA, 5MT |
Appreciate
0
|
01-21-2017, 04:10 PM | #36 |
Lieutenant
200
Rep 536
Posts
Drives: 2013 335xi Alpine White
Join Date: May 2010
Location: GTA
|
|
Appreciate
0
|
01-22-2017, 12:09 AM | #37 |
Banned
354
Rep 1,986
Posts |
|
Appreciate
0
|
01-22-2017, 11:35 AM | #38 |
Lieutenant
200
Rep 536
Posts
Drives: 2013 335xi Alpine White
Join Date: May 2010
Location: GTA
|
Well try the settings posted in this thread and if you don't like them feel free to turn off L7. As for me, I'm a classical violinist, so I do know what an instrument sounds like in real life.
|
Appreciate
0
|
01-22-2017, 07:34 PM | #39 | |
Banned
354
Rep 1,986
Posts |
Quote:
One would think an orchestra member would be the first to realise this. |
|
Appreciate
0
|
01-30-2017, 01:35 PM | #40 |
Gas Man Extraordinaire
64
Rep 80
Posts |
I'm using the final settings with L7 on with one adjustment (Treble is 0 not +1) and it sounds great - fairly flat midrange and treble.
Contemplating adding a sub in the boot to fill in the lowest frequencies but overall very happy with these EQ settings as is .
__________________
F30 330d xDrive MSport+ BlackSapphire . HKaudio + 2xFocal IFBMW-SUB.V2 + RckFsg P300-12 . Gafferlicious.com
|
Appreciate
0
|
02-03-2017, 09:44 PM | #41 |
Major
566
Rep 1,137
Posts |
|
Appreciate
0
|
02-03-2017, 10:06 PM | #42 |
Brigadier General
11091
Rep 3,404
Posts |
First, thanks for sharing this work with all of us. Great level of detail. Very impressive.
I will try out your suggested settings with and without L7. Can't wait to hear how it sounds. I've always thought the sound game is so complicated. You've done nothing to make think differently |
Appreciate
0
|
02-09-2017, 03:21 PM | #43 |
Mr. Discombobulator
160
Rep 162
Posts |
Excellent data in this thread. I tested out both EQ settings for preference (with some subtle tweaks), and I'm liking L7 OFF for my sound needs. The clarity took a significant jump after the adjustment. Thanks!
__________________
'14 MSport 335i | Melbourne Red | MPE | MPBBK | MPS | MPLSD | xHP Stage 3 | MHD Stage 2
|
Appreciate
0
|
02-10-2017, 06:04 PM | #44 |
SEAL TEAM SIX
123
Rep 496
Posts |
These final settings work well for me.
Thanks OP.
__________________
DON'T MAKE ME ANGRY!
GO1 LCI PHEV 30e. M Sport plus Pack, Tech Plus Pack, Comfort Pack, Pano sun Roof, Sensatec dash with Blue stitching, HUD, Galvanic controls. |
Appreciate
0
|
Post Reply |
Bookmarks |
Thread Tools | Search this Thread |
|
|