Honda Ridgeline Owners Club Forums banner

Measuring road noise

6K views 25 replies 8 participants last post by  RZilla 
#1 · (Edited)
There's a few reasons for this initial test.

First is stepping further up the learning curve with REW software - which - in the longer term is part of acoustic experiments related to an ever growing mobile audio system.

Second is 06 is due for new shoes. IME: noticing ride and handling improvements after replacing tires is readily apparent, but road noise is another topic altogether. The psychology of "feel" in road behavior deeply affects the subjective human sense of sound.

I wanted to have something more than my brains influence over my ears to gauge the upcoming change.

Bridgestone "Duelers" as of this morning:

Tire Synthetic rubber Automotive tire Tread Auto part


The measurement system consists of:
- REW software
- Behringer ECM 8000 measurement microphone
- Blue Icicle XLR to USB A/D
- Standard 25mm microphone shock mount
- Vintage Atlas mic stand
- 3" open cell foam pad insolating mic stand base from vehicle floor
- Series of bungee cords limiting mic stand shaft movement while maneuvering vehicle.

In addition to windows software and mic system, 2 additional devices were used during the drive as a confirmation/sanity check:
- Calibrated measurement mic and Android RTA application
- Galaxy handheld SPL

To the degree microphonic induction is minimized in this configuration, vehicle vibrations *should not* be present in the body of the A/D and measurement mic.

Vehicle Motor vehicle Car Automotive exterior Auto part


Sorry for the dust. This Alas stand sits in the home studio most of its life holding up a sweet 60's era Shure SM-55. It didn't occur to me it needed cleaning until the flash made it stand out like a Las Vegas casino sign. SHEESH!

Automotive wheel system Wheel Black-and-white Tire Automotive tire


In the driveway, windows up, engine off. Pure ambient spectrum and pressure as displayed by REW SW and Behringer Mic.

Text Line Font Technology Screenshot


Android RTA and handheld SPL meter (C weighted) sitting in the driveway, windows up, engine idling.

Two-way radio Electronic device Technology Electronics Measuring instrument


Same conditions above as displayed by REW/Behringer mic.

Text Line Font Screenshot Technology


Continued....
 
See less See more
6
#2 ·
Spectrum and level of road noise @ freeway speed. See notes below.

Text Line Pattern Design Font


In the image above, average is configured "Forever", meaning the curve displayed is the sum of averages across the audible spectrum, gathered over approximately 10 minutes of driving. The software began capturing samples as I left the driveway, drove down the hill from the neighborhood, negotiated about 1 mile of surface streets, then entered the freeway. The point being: averaging in this regard tends to flatten out peaks occurring time. Sort of like taking a snap shot of fuel consumption where MPG appears high but when averaged over the life of a tank, drops significantly.

Anyhow... Next run I'll turn averaging off and take several spectral samples @ freeway speed which should be a more accurate representation of road noise.

When 06 gets new boots, we'll see how much - if any - they make in road noise leaking into the cabin.

The Android app holds 1/3 octave peaks. Here's what that looks like:

Text Colorfulness Technology Font Electronic device
 
#3 ·
According to hand held SPL meter in MAX mode, peak pressure reached 95db C weighted. 74db A weighted. If I recall previous casual glances, they were lower. That would have been approximately 30K miles ago, suggesting the Bridgestones are getting significantly louder as they approach end of service life.
 
#4 · (Edited)
Having owned my RL since June '05, I remember the Stock Michelin Tire LTX, and I thought it was quite good. At the time of needing replacement I was planning a Road Trip to Salt Lake City. I opted to change to the Michelin Latitude, IIRC, at the time the tire was considered an "eco" tire from Michelin, and it would be considered a LRR tire if you search for it Today. Granted my tires had 6 years of desert environment and were starting to show their age. But the change to the Latitudes was as if the road noise was cut in 1/2, Very noticable.

I believe the LRR tires in our environment provide an excellent tire choice. If you'd like to get together and swap them for a test drive PM me and lets discuss.

EDIT: the Latitudes only have about 31k miles in the 5 years since install
 
#5 ·
Very interesting test OhSix. I'd also be curious how the Android app Sound Meter compares to your reference system. You are an Android user, right?
 
#8 ·
In retrospect, a more accurate title for this topic might be "Measuring cabin noise". When driving in silence, tuned into the various noises in a moving vehicle, it becomes apparent there are a multitude of vibration induced noises much more diverse than rolling rubber. Can I get a DUH?

Settings such as spectral width, upper/lower scales, octave resolution and a boat load of others options in REW are wildly diverse - and very different from hardware dedicated spectrum analyzers - so the learning curve continues.

One lesson learned, navigating SW settings, capturing logs (vs. screen shots), adjusting band width - dynamically - on the fly is important to acoustic analysis in a moving vehicle. Said another way, driving and using SW simultaneously isn't smart and doesn't produce measurements as detailed as could be if a driver piloted the vehicle while the passenger piloted the SW. Mak has expressed interest in furthering this experiment - and that's great because there's a shortage of geeks 'round here that would have the patience & willingness to participate in this kind of thing. :act006:

This mornings efforts:

Driveway, sitti8ng quietly. Red/yellow arrow points to start of tests. Staircase graph indicates me moving around, keys rattling, insert key in ignition, starter motor turning and engine start/idle.

Text Line Map Diagram Parallel


Spectrum of the events creating the pressure levels seen above. Note: red line is "peak hold" creating a record of frequency and pressure peaks. The black line is spectrum and pressure of cabin sound at the time the graph was captured. The peak appearing @ approximately 1,700hz is the starter motors signature.

Text Line Blue Font Plot


Leaving setting in the same configuration, the vehicle departs the driveway, climbs a short/steep hill, navigates 2 streets existing my neighborhood, then proceeds downhill reaching a max_speed of 40MPH on an asphalt surface. See time stamps: 10 second intervals between vertical graph lines.

Text Line Parallel Font Design


That kind of data is useful to understand basic cabin acoustics. But in my world, 06 spends most of her time @ freeway speeds, this is where we might find a difference pre/post tire swap. Before reaching any conclusions in the test so far, I'd like to define, save and repeat SW settings to make the test apple/apple when comparing the noise contributions of old to new rubber on the road. For example, the spectrum window is currently set wider than the conversion capabilities of the Blue Icicle A to D. In these first tests, there isn't much ambient noise activity above 10kHz, so why waste graph space in that area? On the low end of the scale, there's no value in leaving the window open to 2Hz - and at the current conversion rate of the A to D, the low end is cut off @ 15Hz. More wasted graph space. So fine tuning settings and saving them for future use should be valuable.

Below is the spectrum of cabin noise @ 60 to 70 MPH on a smooth freeway. The red arrow points to the affects of driving over a textured concrete (a bridge) - interestingly @ approximately 100Hz. The green arrow points to vibration transfer when tires engage raised lane reflectors, centered @ approximately 160Hz.

Text Line Plot Pattern Design


This is the SPL histo-graph of the freeway ride this morning:

Text Line Font Diagram Design


The hand held SPL meter - using A weighted scale and MAX (meaning remember highest value measured) confirm, over all cabin noise during the trip.

Electronic device Measuring instrument Technology Two-way radio Tool
 
#9 ·
Notes on Android RTA app with Parts Express calibrated mic:

Screen shot here is quiet interior, engine off. Phone is mounted in its air duct holder - making it susceptible to microphonic influences - which may explain low frequency energy in the graph.

Text Technology Font Colorfulness Screenshot


This shot is the result of this morning ride:

Text Colorfulness Font Technology Electronic device


SPL and frequency content roughly align with REW and Berhinger/Blue. For the cost of a free App and $16 mic, this is a more than reasonable mechanism for understanding sound around you. Fun, portable and cheap. A hard combo to beat.
 
#12 ·
Does a typical android device mic have the hardware capability to send accurate info to an app? Or are they typically limited above and below certain freqs?
 
#13 ·
Basically, Android and base modem SW doesn't filter audio outside of "in-call" TIA band limits - so the only real limitation *should be* frequency characteristics of the mic itself.

Typically, bill-of-material bean counters don't care as much about performance as they do line item supplier cost, so performance takes a back seat to $. Mic's in mobile devices are incredibly small these days - driven into the micro-world by the insatiable desire to pack features into an ever decreasing footprint - which is just a way to say "cheap and small" is the driving concern when sourcing components for mass production.

Having said all that, small and cheap doesn't necessarily mean crummy. Electro-mechanical transducers have come a mighty long way in the last couple years.

You've placed an interesting bug in my ear. Lemme run a test or two to confirm what I think I know. I'll use a calibrated mic with the RTA app, take a snap shot. Unplug the mic, then use the native Moto mic in my X. It'll be cool to actually know the delta (if any). My guess would be the external mic would have superior polar characteristics - but we'll see.
 
#14 ·
Six,

I love when a "bug" is placed in you ear. It means the "delta" for how much extraneous information we'll all receive is going to grow exponentially.

Everyone, strap on your seat belts, the ROC's Mr Wizard is going to teach us all something.
:grin:
 
#15 ·
I would expect that since the bandwidth for a phone conversation is limited, the in-phone mic will be optimized (cost-wise!) for that range.

I seem to recall the Bell System was set up for a 4 kHz bandwidth over their copper lines. Not sure what cell technology uses, but it's almost certainly chopped and compressed for transmission to the tower.

Chip H.
 
#16 · (Edited)
CTIA adopted TIA standards at the inception of mobile telephony. Bandwidth being what it is, over the air conservation and voice sample rates at the mobile station modem level still limit "quality" to 1940`s era voice quality band limits of 300 to 3.3kHz. Stinko!

The real tell in mobile mic quality these days is VoLTE. Voice over IP mobile (or high def audio) can be shockingly good. An end to end VOIP call is as amazing as other high def experiences. Burps and farts sound as though they originate in the room you are occupying. Thankfully that's just an analogy. LOL.

So, mics in modern phones have much expanded frequency character when compared to mic's of old.

Tests to follow shortly. :)
 
#17 ·
Well, I basically meant to ask if the frequency response of the tiny mic in the phone can provide measurably similar results to your reference standard gear. Or that's what I was trying to ask!
 
#20 ·
Getting to know REW a little better. Recently learning to save files for post event analysis and other fun stuff. In the raw files, time stamps are available, which gave birth to merging data from other devices - so been playing around with a couple things. More to come - but for now, integrating time stamps from Torque App logs provides a way to merge vehicle speed with SPL and spectrum graphs.

So... below is a first pass at associating speed with road noise. This is SPL characteristics of a (portion) of a 31 minute freeway trip to the office. Peak level reached 76db SPL C @ 9:35 + 1/2 second. With Torque logs set to 1 second intervals, the time stamp provides a cool way to isolate what the vehicle was doing at the time of peak interior noise. A whole bunch more data was collected - but parsing it is a bit challenging. As that process improves, we should see some nifty stuff. Nifty being a geek term for geeky. :act002:

Text Blue Font Line Parallel


More work to do on this, but for now - it looks to me the 3.5L VTEC acoustic signature hangs right between 38 and 42Hz. Note the short (.22mi) uphill drive mentioned has twin peaks @ 40 & 110Hz. RPM probably hit 3,5K or so but speed never exceeded 30MPH. The remainder of the trip, 40Hz remains prevalent. For now, I suspect the evil effects of resonance are at work. One way to confirm if the 40-ish Hz hump is always related to the engine *might be* to slip her into neutral on a safe stretch, see if that humps flattens out.

Text Font Line Green Blue


Isolating the freeway graph, we can see most of the action is happening below 2.8kHz or so.

Text Blue Line Font Purple


More fun stuff is coming. Gotta get the base line defined so when new rubber gets attached, the delta can be measured.
 
#21 · (Edited)
Dunno if anyone is as interested in this stuff as I, but learning REW and gaining some insight to road induced cabin noise while the RL is on motion has been instructive.

Using the Torque App to log RPM, Speed, GPS LAT/LON and a couple other items I'm interested in - its possible to see where the vehicle was, how fast it was going & what surface it was rolling on at any point in time.

This is an example of the Torque log. The red arrow points to time resolution in the file, which is plenty granular for this effort.

Text Font Line Number Parallel


Importing lat/lon to google earth, searching time stamps in REW SPL logs then mapping SPL to spectrum, we end up with a geek party where road surfaces can be seen and text gathered from REW and Torque logs is added.

Map Ecoregion Land lot World Urban design


Turning to the spectrum graph, there isn't a whole bunch of energy in the 2K to 7K range - bracketed here in red. Level is pretty flat in that portion of the band with only slight +/- deviation around 40db, then dropping steeply 7K to 8K, flat lining from there out to ultrasonic. The red arrows point to 94db peak @ 40Hz. The RL cabin seems to have a resonant frequency down low. Check out that hockey stick @ the lower end, rising steeply @ approximately 500Hz.

Text Green Line Blue Pattern


Narrowing the upper limit down to 2kHz, a few peaks and valleys begin to appear. In the first graph, the red arrow points to a dip around 300Hz, so we'll use that to drill down farther in the 2nd graph below.

Text Green Blue Line Font

Text Line Font Screenshot Parallel


So, here's the most current thought. Most of the noise seen in these graphs is occurring below 1kHz. In the very low range, like down around 100Hz, there isn't much an owner can so to reduce over all noise intrusion that low. Certainly deflex materials will make a difference but the resonant frequency of the cab would take some serious attention to make a significant dent in low frequencies. The roof in the primary culprit for low frequency surface flexing, followed but door skins. Beyond that, there are harmonic contributions of all this metal flexing - the audible waves they generate combining in phase - increasing amplitude - hence, the hockey stick seen in the full range graph. There's a few minor things I have in mind for low frequency abatement in 06, but that stuff will hafta wait for now.

One of the purposes of this round of testing was establishing a base line so that when 06 gets new shoes, the difference between her old rubber and new could be measured. I'm pretty sure the data collected so far establishes what these tires contribute to cabin noise. My ears tell me most of the noise coming from them is well above (frequency wise - not amplitude) the low end of the spectrum seen in these graphs. I'm thinking the energy above 300Hz is going to drop with new tires. *I THINK* I hear a whole bunch of stuff around 1kHz when traveling @ freeway speed. But I can't discern tire noise from the other random symphony of moving vehicle noises.

It seems to me 06 is getting louder by the day - but that may be the result of driving in (relative) silence during these test - rather than tire surfaces wearing away.

At any rate - *I THINK* new tires will change the shape of the curve between the red lines in this graph (500 to 2kHz).
Text Blue Line Font Plot


We'll see soon enough. New rubber is right around the corner.

BTW: at the foundation of measured frequency characteristics is an understanding that the human ear isn't linear. We carbon units are most sensitive to the middle portion of the spectrum. For example, when thinking about frequency vs. pressure (SPL), know this: people can detect 3kHz @ 0db, but to hear 100Hz we need ~40db of pressure. So... "hockey sticks" in spectrum graphs aren't representative what is actually heard by **** sapiens. In the context of measuring cabin noise, all that low frequency energy seen in the graphs doesn't translate to actual sound. Looking into filters - if available, they would go a long way to in correlating graphs to audible sound
 
#22 ·
I've enjoyed your research into the characteristics and limits of the OEM subwoofer and I'm following this thread as well.

Hondas have a reputation for transmitting a lot of tire noise into the cabin, something they seem to be unable or uninterested in correcting. Your current tires have a lot of miles and years on them (less rubber between you and the road and the rubber has hardened over time), so I agree with your hypothesis that you will see a noticeable reduction in noise with new tires.

I've forgotten if you have applied any noise reduction treatment to your RL, or are your results thus far based on an unmodified RL. It would be really great to see baseline audio testing on a stock RL, then follow up test after applying dampening material (dynamat) to the outer door skins (~80% coverage), then retesting after the application of additional dampening material to the inner door skins (I think the results will be noticeable, but not significant). It would be good to see if you are correct about the roof being responsible for a lot of the low frequency noise as well.

We recently bought a 2016 Accord Touring which has active noise canceling and is the quietest Honda I've ever owned by a wide margin. It's only a matter of time until aftermarket sound systems incorporate active noise cancelation technology, which may be our best hope for a truly quiet Ridgeline.
 
#24 ·
Thanks Robert. I know not everyone gets this geeky - it's nice to know there's a few of us hanging around here! :act006:
Yes, 06 front doors have been treated with Cascade Audio VB2. ~80% inside of outer door skins:
Automotive exterior Metal Rim Bumper Auto part


~30% on the inner door structural metal:
Auto part Automotive exterior Engine Vehicle Vehicle door


~20% inside door panels. Plus reuse of OEM padding.
Floor Leather

Vehicle door Car Automotive exterior Auto part Vehicle


http://www.ridgelineownersclub.com/forums/5-interior/128681-voltage-gauge-not-ashtray-lighting.html

http://www.ridgelineownersclub.com/forums/9-mobile-electronics/104465-couple-audio-upgrades-5.html

Subjectively speaking, there's a significant difference in sound the front doors make when closing vs. untreated rear doors. Rear doors sound "tinny", thin, light weight, front doors make a reassuring "thud-ish" sound when completing full door closure. It's a shame I didn't have tools at the time to capture the acoustic signature of front doors pre-treatment. But I do plan on doing that for rear doors when the time comes to treat them.

VERY interesting info on the active noise cancellation tech in your new Accord! I'd often wondered when the auto industry would adopt phase cancellation systems. I suspect its cheaper to deploy a cabin mic/DSP engine in a head unit than it is to treat key surfaces in a car with physical products. Once an OEM invests in an NRE project like that, it's application can be spread across hundreds of thousand's of units bring the cost down to low single digits per unit.

I agree, it shouldn't be long for the aftermarket to catch up. I would think a defeat key would be cool. Engage when needed but have the ability to by-pass signal processing when engaged in "critical listening", giving "purists" best of both worlds - although I don't see sonic validity in that, cuz EVERYTHING in the audio chain is SW controlled - very little pure analog left in mobile audio these days. Is the system in the Accord "always on"?

Six, offer still stands. I'm about as stock as can be found...
Awesome Mak! Let's git er' done.... A potential test plan:
- Capture acoustic signature of untreated front doors closing.
*Both inside/outside your 06
- Place this test rig in your 06.
- Install OBD/Bluetooth adapter your 06
- Drive local roads @ various speeds, making sure to capture black top & concrete road sonic signatures.
- Swap tires your 06 to/from my 06.
-Repeat test drives.

Another thought: with this base line in place, after 06 gets new shoes. Perform test above. That way, your "mid life" tires could be compared to new both in your 06 and mine. Haven't thought that all the way through. Open to any other suggestions! :nerd:
 
#23 ·
Six, offer still stands. I'm about as stock as can be found...
 
#25 ·
VERY interesting info on the active noise cancellation tech in your new Accord! I'd often wondered when the auto industry would adopt phase cancellation systems. I suspect its cheaper to deploy a cabin mic/DSP engine in a head unit than it is to treat key surfaces in a car with physical products. Once an OEM invests in an NRE project like that, it's application can be spread across hundreds of thousand's of units bring the cost down to low single digits per unit.
I agree, that in volume production Active Noise Cancellation (ANC) is really cheap to deploy compared to dampening material (DM). DM adds weight, takes up space and requires extra time to install. Assuming you are already installing an audio system with a subwoofer, ANC only requires the addition of a microphone or two, a processing unit and a little firmware.

I believe ANC is always on, to be honest I haven't spent too much time messing with the audio settings on the Accord. There are post on Drive Accord forum saying the system can be disabled and others saying the only way to disable the system is disconnect the ANC microphone(s).

Every time I visit Drive Accord, I am reminded of how awesome the ROC forum is. Lots of shouting and not nearly as much quality information on Drive Accord. :)

Can't wait to see how the testing progresses!
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top