Posted in Humor, Music, Physics

Honda Needs a Tune-Up

This is the story of how Honda engi­neers screwed up a big expen­sive project with a sim­ple arith­metic mis­take, tried to fudge their result with sound edit­ing soft­ware, and con­grat­u­lated them­selves for being totally awesome.

When I was a kid, my fam­ily used to drive up to The Pin­ery in Ontario, a beau­ti­ful park by Lake Huron. Very scenic. My favorite part, though, was a stretch of road a half-hour out­side of the park. To dis­cour­age reck­less Cana­di­ans from bar­rel­ing past the houses and barns, the local gov­ern­ment carved five sets of grooves in the road before every stop sign. Drive over them, and the car would vibrate: “vbvb­vbvb… vbvb­vbvb… vbvb­vbvb… vbvb­vbvb… vbvb­vbvb.” The faster you drive, the higher the pitch.

My Dad is a musi­col­o­gist, with a par­tic­u­lar inter­est in tun­ing. So there was no way he was going to pass up the chance to exper­i­ment with this instru­ment. Every time we approached some grooves, he’d start fast over the first set, and try to slow down by the last set, to play a descend­ing scale: G-F-E-D-C. If there was no oncom­ing traf­fic after the stop sign, he’d swing over to the other side of the road and play an ascend­ing scale as we sped up.

Ratios of speeds cor­re­spond to ratios of vibra­tion fre­quen­cies, which cor­re­spond to inter­vals between notes. To play an ascend­ing scale C-D-E-F-G, you need to drive at these ratios to your start­ing speed: 1 — 9/8 — 5/4 — 4/3 — 3/2 (for exam­ple, 24 — 27 — 30 — 32 — 36 mph)[1].

Play­ing a scale with a ’95 Toy­ota Pre­via is not easy. The notes tend to come out a lit­tle wonky — we’d get the half-step between E and F too wide, and with not enough space between F and G. It usu­ally sounded kinda modal… but still awesome.


So imag­ine my delight when I heard about this musi­cal road [CNET] that Honda built in Lan­caster, CA.. A team of engi­neers carved some grooves into a high­way that were care­fully spaced to play the William Tell Over­ture as you drive over them at a con­stant speed. Awe­some, right? The prob­lem is, it’s spec­tac­u­larly out of tune.

Here’s the orig­i­nal melody:

Audio clip: Adobe Flash Player (ver­sion 9 or above) is required to play this audio clip. Down­load the lat­est ver­sion here. You also need to have JavaScript enabled in your browser.

And here’s the Honda road again:

Audio clip: Adobe Flash Player (ver­sion 9 or above) is required to play this audio clip. Down­load the lat­est ver­sion here. You also need to have JavaScript enabled in your browser.

The Honda ver­sion isn’t sim­ply “out of tune”… the notes are just wrong. The orig­i­nal starts with a ris­ing 4th, F-B♭[2], and even­tu­ally reaches an octave above the start­ing note before descend­ing to the tonic F-E♭-D-B♭.[3] But Honda’s ver­sion starts with a ris­ing major 3rd, and its top note is a major 6th above the start­ing note. Some might have noticed that the last few notes in Honda’s com­mer­cial sound OK. That’s because they edited over them! I can prove it.

Basic melody in the William Tell Over­ture (schematic)

The CNET arti­cle above spec­u­lates that Honda designed the road specif­i­cally for the Honda civic dri­ving at the speed limit, and other cars might need to drive at a dif­fer­ent speed to make it sound bet­ter. But if you’re going at a con­stant speed, all that mat­ters is the spac­ing between grooves. Speed­ing up or slow­ing down just trans­poses every­thing. It would be the­o­ret­i­cally pos­si­ble to “cor­rect” the melody by dri­ving at dif­fer­ent speeds (like on the road to the Pin­ery). But the notes on the musi­cal road are too closely spaced for all but con­sum­mate musi­cian Mario Andretti.

It also doesn’t mat­ter what car you drive[4]. The vibra­tion fre­quency is f = v/d, where v is the car’s speed, and d is the dis­tance over which the road pat­tern repeats. There’s no place in the equa­tion for wheel spac­ing, tire size, side-impact airbags, etc. All of these things affect the qual­ity of the sound, but not the pitch.

So why is the musi­cal road so unmusical?

The Error

Honda posted a series of 5 ridicu­lous videos: [Part 1][Part 2][Part 3][Part 4][Part 5], in which they talk about all the hard work they did and con­grat­u­late them­selves for being so awe­some. There are lots of com­pli­cated sound­ing num­bers, there’s a “Mathematician/Musician,” and plenty of experts. I’m sure some peo­ple behind the project under­stood what was going on. But I think they failed to antic­i­pate a basic mis­un­der­stand­ing on the part of the groove-designers.

In the fourth “mak­ing of” video, they men­tion that the ini­tial note, a low F, has a spac­ing of 4 inches (4in) between grooves (1:47):

From the video, it looks like the grooves them­selves are about 1in wide. Now, sup­pose you want to make the B♭ a 4th above F. A per­fect 4th is a fequency ratio of 4/3, so you should mul­ti­ply the width by a fac­tor of 3/4… But the width of what?

Based on the Civic’s 106.3 inch wheel­base, we can see from this pic­ture that s+g is about 5 inches. Honda says the low­est note has a 4 inch spac­ing, so that’s con­sis­tent with 1 inch grooves.

The width that really mat­ters is the total width of the spac­ing plus groove (s+g). That’s the dis­tance over which the road pat­tern repeats, so that’s the dis­tance over which the car com­pletes one vibra­tion.[5] Sup­pose you didn’t know this, and only changed the spac­ing, from s = 4in to s’ = 3/4 × 4in = 3in. Then the fre­quency ratio is (s+g)/(s’+g) = (4+1)/(3+1) = 5/4, a major 3rd, not a per­fect 4th. What about the octave above the start­ing note? An octave is a fre­quency ratio of 2/1, but if you only changed the spac­ing to s’ = 1/2 × 4in = 2in, you’d get an actual ratio of (s+g)/(s’+g) = (4+1)/(2+1) = 5/3, a major 6th, not an octave.


mak­ing an octave, incorrectly

There are two ways you could cor­rect this problem:

  1. Adjust the groove width g as well as the spac­ing s. For instance, to make an octave, use a spac­ing s’ = 2in and a groove g’ = .5in, giv­ing a fequency ratio (s+g)/(s’+g’) = 5/2.5 = 2/1. This is prob­a­bly hard with typ­i­cal cut­ting tools. Also, the engi­neers may have found that they need to make the grooves big­ger than some min­i­mum width to get a good sound. So on to method 2…
  2. Over-adjust the groove spac­ing so that the total g+s is cor­rect. For instance, to make an octave, adjust the groove spac­ing to s’ = 1.5in, so you get a fre­quency ratio of (s+g)/(s’+g) = 5/2.5 = 2/1.

mak­ing an octave, correctly

The Coverup

Armed with this the­ory for why the musi­cal road sounds so bad, I crunched some num­bers in Math­e­mat­ica, and was able to repro­duce Honda’s result, sort of…

Here’s Math­e­mat­ica play­ing the cor­rect William Tell Overture:

Audio clip: Adobe Flash Player (ver­sion 9 or above) is required to play this audio clip. Down­load the lat­est ver­sion here. You also need to have JavaScript enabled in your browser.

And here’s Math­e­mat­ica pro­grammed to make the mis­take I think Honda’s engi­neers made:

Audio clip: Adobe Flash Player (ver­sion 9 or above) is required to play this audio clip. Down­load the lat­est ver­sion here. You also need to have JavaScript enabled in your browser.

And here’s honda’s com­mer­cial ver­sion again:

Audio clip: Adobe Flash Player (ver­sion 9 or above) is required to play this audio clip. Down­load the lat­est ver­sion here. You also need to have JavaScript enabled in your browser.

Notice that a few notes in the com­mer­cial sound dif­fer­ent from Mathematica’s ver­sion. Par­tic­u­larly at the end. Honda’s last few notes are sort of… in tune! Turns out that’s a bit of Hol­ly­wood magic. Here’s a record­ing I stole from a dif­fer­ent video of some­one dri­ving down the Musi­cal Road[6]:

Audio clip: Adobe Flash Player (ver­sion 9 or above) is required to play this audio clip. Down­load the lat­est ver­sion here. You also need to have JavaScript enabled in your browser.

What hap­pened to the end­ing? It’s all funky again. Go back and lis­ten to the Math­e­mat­ica ver­sion that mim­ics Honda’s mis­take. Same funky end­ing[7]. Who­ever put together the Honda com­mer­cial must have edited over the end­ing, assum­ing that as long as the last few notes were cor­rect, no one would notice any­thing wrong.[8]

What I don’t under­stand is: if they were going to doc­tor the sound, why didn’t they just cor­rect the whole thing? It’s not that hard. My dad did this ver­sion in about 20 minutes:

Audio clip: Adobe Flash Player (ver­sion 9 or above) is required to play this audio clip. Down­load the lat­est ver­sion here. You also need to have JavaScript enabled in your browser.


I learned some­thing else kind of ridicu­lous from this analy­sis: if Honda didn’t doc­tor the over­all pitch of the melody in their com­mer­cial, then they were speed­ing. The open­ing fre­quency is about 238Hz, which cor­re­sponds to a speed of about 67mph if the road pat­tern repeats over 5in. But they men­tion in one of the videos that the speed limit is 55! Crap.

In fact, in this youtube video, where they explic­itly state they’re going 55mph, the melody starts a minor third below the Honda com­mer­cial. A minor third is a fre­quency ratio of 6/5, so this is con­sis­tent with Honda’s dri­ver doing 6/5 × 55mph = more than 10mph over the speed limit…

Another funny point is that some of the inter­vals you get from Honda’s mis­cal­cu­la­tion are pretty bizarre. The D, a major 6th above the start­ing F, should have a fre­quency ratio of 5/3 above the start­ing fre­quency. Instead, it has a ratio 5/(4 × 3/5+1) = 25/17. This isn’t really in the west­ern scale. It’s about 2/3rds of the way between an aug­mented 4th and a pure 5th. Micro­tonal com­posers like Easley Black­wood might have found a use for it, but I don’t think it’s what Honda was after.

If I were them, I’d seri­ously con­sider paving over the road. In fact, it seems like some local res­i­dents might do it for them. There is another option, though. If they bring in the bull­doz­ers, and shuf­fle around a few chunks of asphalt at the end of the road, they might get a decent ren­di­tion of “When The Saints Go March­ing In.”

Audio clip: Adobe Flash Player (ver­sion 9 or above) is required to play this audio clip. Down­load the lat­est ver­sion here. You also need to have JavaScript enabled in your browser.

Update [12/30/08]: Added pic­ture com­par­ing grooves to Civic wheelbase

Update [5/2/11]: I am both sorry and delighted to hear that they rebuilt the musi­cal road (see, e.g., here), and they fixed noth­ing. Here it is on April 28, 2011:

Just… wow.

  1. If anyone’s won­der­ing what hap­pened to the 1/12th pow­ers of 2 in this whole tun­ing dis­cus­sion, I’m using what’s called Just Into­na­tion, which is an (often better-sounding) approx­i­ma­tion to the Equal Tem­pera­ment sys­tem most peo­ple know. Actu­ally, it’s really the other way around: the rea­son we use 12 equal semi­tones is that it lets us approx­i­mate nice inte­ger ratios like 3/2, 4/3, 5/4, etc.. This is a long story that I’m not going to get into here. []
  2. Actu­ally, the start­ing note in the record­ing is around a B♭. I’m going to pre­tend like every­thing is in the key of B♭ (so the start­ing note is F), since that’s the key they talk about in the making-of videos.a picture of Honda's scorea pic­ture of Honda’s scoreSorry to the perfect-pitch peo­ple. []
  3. The orig­i­nal melody actu­ally has a run down to the B♭: F-E♭-D-C-B♭. Honda appar­ently decided this was too com­pli­cated and used a sim­pli­fied ver­sion. That’s what I’ll stick to here. []
  4. With one excep­tion that can’t fix the tun­ing. See my com­ment, below. []
  5. More pre­cisely, once you know the force dri­ving the vibra­tions is peri­odic with period T=d/v, it fol­lows that the vibra­tions them­selves have that peri­od­ic­ity, so the Fourier trans­form of any resul­tant sound is only nonzero at inte­ger mul­ti­ples of f=1/T. For more expla­na­tion, see the sec­ond com­ment, below. []
  6. I’ve actu­ally trans­posed it up to be in approx­i­mately the same key as the other record­ings in this arti­cle. By the way, there are hun­dreds of such videos on Youtube. []
  7. Aside from a sin­gle pass­ing note. If I change the clos­ing notes from F-E♭-D-B♭-D-B♭ to F-E♮-D-B♭-D-B♭, and apply the Honda mis­cal­cu­la­tion, it sounds almost exactly like the undoc­tored record­ing of the musi­cal road. So it appears that there are two errors at work here: the groove spac­ing mis­cal­cu­la­tion, and replac­ing an E♭ with an E♮. []
  8. It seems like Honda fixed up some of the other notes, too, to get a more pleas­ant sound. Some might object that it’s easy to make the notes sound bad by speed­ing up or slow­ing down as you drive down the road. How­ever, I don’t hear any­thing like that in the ran­dom person’s record­ing. The melody returns to pre­vi­ous notes with rea­son­able accu­racy, which it wouldn’t do if the speed were vary­ing. []

41 Responses to “Honda Needs a Tune-Up” Comment Feed, Trackback.

  1. Jeremy Heilman says: Dec 24, 2008 @ 11:27 am

    Bril­liant work, David. I must have just come home from church choir prac­tice when I first heard this com­mer­cial, because I recall shout­ing at the TV, “You think YOU have tun­ing problems.…”

    In addi­tion to your point that it is the total period (space + gap) that deter­mines the pitch, the ratio of space/gap will alter the over­tone series and the tim­bre of the road. For instance, if the gap and space were always equal (and half the total), the over­tone series would be that of a square wave (given by a Fourier series http://www.numerit.com/samples/fours/doc.htm) with the loud­est com­po­nents grouped near the fun­da­men­tal. I would think this would be the best reflec­tion of most instru­ments you could get from a road (but likely still pretty poor). How­ever, using a big space and a lit­tle gap (or vice-versa) would result in many of the high­est fre­quency har­mon­ics hav­ing large ampli­tudes, for a “tinny” sound. If any­one has ever dri­ven a Civic, they know it sounds pretty tinny already, and doesn’t need any help from man-made potholes!

    The other gross over­sight (this would have actu­ally got­ten them some major awe­some points in my book) would be to cut the grooves on a diag­o­nal, such that no two tires are in the same point with respect to the pat­tern at the same time. Then, each tire would sound indi­vid­u­ally, and you would have a rub­ber quar­tet play­ing in uni­son. Or cut half the pat­tern on the left side of the road and half on the right, to make diads!

    Of course, I think if the music you make sounds worse than the begin­ning band, you’d be bet­ter of sell­ing insurance.

  2. davidsd says: Dec 24, 2008 @ 4:50 pm


    Some good ideas! I can’t imag­ine what coun­ter­point would have sounded like if they didn’t get a sin­gle melody straight…

    Some peo­ple have asked me about some­thing that you touch on in your com­ment, so I thought I’d answer it here. There is a way that the kind of car can affect the pitch: the peri­od­ic­ity of the road guar­an­tees that only inte­ger mul­ti­ples (called “har­mon­ics”) of the fun­da­men­tal fre­quency f=v/d can appear. The rel­a­tive strength of these har­mon­ics deter­mines the qual­ity or tim­bre of the sound, and it can depend on what shape the grooves are, the dis­tance between the front and back wheels, etc.

    An extreme case of this (as my friend Matt points out) is that the fun­da­men­tal fre­quency f can be sup­pressed or even dis­ap­pear. For instance, if the wheel spac­ing is a half-integer times the groove dis­tance d=s+g, the road would vibrate the car at twice the rate, and it’s pos­si­ble that the fre­quency 2f would be stronger than the fun­da­men­tal (though I think it’s unlikely the fun­da­men­tal would dis­ap­pear entirely).

    This would make the melody jump up by an octave when­ever it hit a cer­tain note. This doesn’t hap­pen in any of the record­ings I’ve heard (the full range of the melody is only a major 6th), so it’s fair to say that this effect wasn’t part of Honda’s design. The key point, though, is that the peri­od­ic­ity of the road guar­an­tees that this is the only way that the make of the car can affect the sound, and it doesn’t hap­pen in prac­tice (nor could it fix the tun­ing at all).

  3. Dad says: Dec 24, 2008 @ 9:01 pm

    Great post, and inter­est­ing com­ments about the pos­si­bil­ity of an upper octave being heard if the wheel­base is some odd mul­ti­ple of 2.5in.

    As it hap­pens, the 2008 Honda Civic does not fall into that cat­e­gory, hav­ing a wheel­base of 106.3in for the sedan, and 104.3 for the coupe. There are some cars where it could apply, how­ever: the new VW Tiguan has a “unique” wheel­base of 102.5in, and the 2007 Aveo has a wheel­base of 97.5in.

    That could affect the sound of the low­est (i.e., first) note, but other wheel­base lengths could affect other notes, could they not? Since the pro­por­tions used by Honda for this project are not sim­ple ratios for s+g, it would take some inves­ti­gat­ing to see what works on the road. Any­one out in Cal­i­for­nia with a Tiguan or an Aveo for starters?

    Come to think of it, this would only work for longer notes, I think — longer than the wheel­base, at least. If the note only cov­ers, say, 8ft of road, it would sim­ply extend the fun­da­men­tal pitch by hav­ing the back wheels vibrate after the front ones were done.

  4. Paul Rapoport says: Dec 25, 2008 @ 4:43 pm

    This is,in the words of a for­mer high school music teacher, a ruddy riot. (He said that when we squawked in a Honda tun­ing or an unrea­son­able fac­sim­ile thereof.)

    Rea­son­able that you con­sider a major scale to have a 9/8 major sec­ond. I’m sure you’re aware of the “sec­ond scale degree” prob­lem, its not being justly in tune with both the 3/2 and the 5/3, and the pro­lif­er­a­tion of issues aris­ing from that.

    I like that you arrived one place at a 25/17. At 667.67 cents, it’s about 2/3 the way between cer­tain inter­vals, as you said, but I imag­ine there you’re think­ing in 12-tone equal. It’s a dif­fer­ent dis­tance (about 3/5) if you take the just aug­mented 4th of 10/7. All of which affects your point not a bit.

    As for Easley Blackwood’s inter­est in a 25/17, you’ll find some­thing very close in his 18-tone ET etude. But he’s never taken a com­po­si­tional inter­est in just tun­ings. Ben John­ston, on the other hand, has writ­ten music with har­mon­ics up to 32. Whether there’s a 25/17 any­where, I have no idea.

    Thank you once again for your les­son in Civics. Beats the usual kind. Do you think Honda’s ver­sion is any bet­ter if it’s painted in blues?

  5. Mano Singham says: Dec 29, 2008 @ 5:44 pm


    Very inter­est­ing post! It reminded me of some­thing I was told a long time ago but which I never tried to verify.

    You know those rum­ble strips by the side of a high­way that warn you when you are drift­ing off the road, say if the dri­ver drifted off to sleep? I was told that engi­neers once designed them so that they would say some­thing like “Wake up!” but found that on test­ing, it gave dri­vers such a shock to hear a dis­em­bod­ied voice that they aban­doned the plan.

    I sus­pect that this story is too good to be true but was curi­ous if you had encoun­tered any­thing like it in your research on this.

  6. davidsd says: Dec 31, 2008 @ 5:31 am


    I can’t seem to find any­thing about rum­ble strip voices, but I did come across this video about a musi­cal road in South Korea that was appar­ently cre­ated to keep dri­vers from falling asleep (unlike the one in Lan­caster, CA which was mostly an adver­tis­ing stunt).

  7. Robert K says: Mar 06, 2009 @ 12:16 am

    After see­ing the TV ad I was lit­er­ally furi­ous that Honda had made a self-congratulatory com­mer­cial about a project that had so obvi­ously failed. So I Googled it, but no one seemed to have noticed that the melody was totally wrong. I thought maybe I was going crazy. The worst part was that the math behind design­ing it should’ve been so sim­ple, I couldn’t under­stand how they could’ve screwed it up.

    I just wanted to thank you for post­ing this and for doing the analy­sis to con­firm what went wrong, although I’m still amazed that they could have made this mis­take. I don’t even really know what this blog is about, but I had to express my grat­i­tude. It’s good to know that some­one else noticed.

  8. Ben says: Mar 26, 2009 @ 7:57 pm

    Fan­tas­tic. I want to sec­ond exactly what Robert K said (boy, did I think I was going crazy!). So nice to know that there’s an expla­na­tion for what went wrong. As a foot­note, I used to sell blue cheese to Easley Black­wood on a sur­pris­ingly reg­u­lar basis.

  9. davidsd says: Mar 26, 2009 @ 8:19 pm

    @Ben, Thanks! That is awe­some about the cheese.

  10. Nick says: Apr 01, 2009 @ 6:40 pm

    I’m so glad that this both­ered some­one else as much as it did me.

  11. Nosmo says: Apr 10, 2009 @ 12:19 pm

    A coworker com­mented that there must be some sort of soft­ware that cor­rects pitch, and it could be that the cor­rec­tion they did was run the sound through that soft­ware and it just changed the fre­quen­cies to the near­est note? That may explain why the cor­rec­tion they did was not done prop­erly. Does this make any sense?

  12. coreopsis says: Apr 14, 2009 @ 7:30 pm

    Awe­some analy­sis! Some­one should have paid you to write this article.

  13. Adam says: Apr 15, 2009 @ 1:35 am

    I’m glad I found this. Only after watch­ing that advert about ten times did I even fig­ure out that it was William Tell. It sounded noth­ing like it until I used some imagination.

  14. Zach says: Apr 16, 2009 @ 12:36 am

    Excel­lent write up. I searched for an expla­na­tion to the screw up and found this site. The com­mer­cial always frus­trated me because it did not makes sense why Honda would choose a song that was not com­monly rec­og­niz­able. I fig­ured maybe the engi­neers were a bunch of indie rock lovers and picked some obscure cult song (p.s. i like indie rock). It wasn’t until about the 10th time hear­ing the com­mer­cial that I rec­og­nized the tempo of the song they were appar­ently try­ing for. This engi­neer­ing screw up clearly is not some­thing they should be mar­ket­ing at all let a lone dur­ing every com­mer­cial break on every channel.

  15. John says: Apr 18, 2009 @ 10:06 pm

    Amaz­ing analy­sis, David. Add my name to the list of peo­ple who had NO idea what the orig­i­nal Civic-produced ditty from the com­mer­cial was sup­posed to rep­re­sent. I didn’t even think William Tell Over­ture until I read it here. What a joke. In hind­sight, I par­tic­u­larly like the col­lege kids in the com­mer­cial cheer­ing in cel­e­bra­tion of sev­eral months spent pro­duc­ing a song which sounds absolutely noth­ing like the original.

    Remem­ber kids, real­ity is what the man on the TV told you. That is EXACTLY how William Tell is meant to sound, because the faux-college kids from Honda Adver­tis­ing TOLD you so. hahaha

  16. PJ says: Apr 19, 2009 @ 1:51 am

    Thank you! I needed to find some mean­ing or jus­tice behind the trash­ing of the William Tell melody. If they didn’t get this right, what does this say about their newest cars?

  17. Elliott says: Apr 19, 2009 @ 4:03 pm

    David — In the early 90’s Tim Sprunger who worked at Walt Dis­ney World at the time, built a musi­cal road as a demo project that Dis­ney never picked up on. He’s an artist/musician and went through all the same cal­cu­la­tions you list above. His road played Zip-py-do-da of course and has a video of the test­ing. His design used raised ridges made of a spe­cial aggre­gate and epoxy for­mula. He found, unless the grooves were made in con­crete the sum­mer Florida heat would soften the asphalt and almost erase the grooves in one sum­mer. Tim lives Kissim­mee FL

  18. davidsd says: Apr 19, 2009 @ 5:28 pm

    @Nosmo, That’s a good idea — they def­i­nitely could have used some­thing like that dur­ing the edit­ing process. They cer­tainly made sure each pitch in the com­mer­cial is a note on the piano (as opposed to the way it really sounds, which has funny micro­tones you can’t play on a key­board). But I’m not sure they always mod­i­fied the pitch to the near­est note. For exam­ple, at the end, the honda com­mer­cial goes B♭-D-B♭, but I think the near­est notes would be A-C-A. Sounds like they were being a lit­tle more lib­eral with the pitch doctoring.

    @Elliott, very inter­est­ing! I do remem­ber read­ing some­thing about asphalt being a dif­fi­cult “medium” in hot cli­mates. That’s cool that Tim fig­ured out a solu­tion. I won­der if the city of Lan­caster (which appar­ently has rebuilt the William Tell road as a tourist attrac­tion) is aware of this.

    Every­one, thanks for the kind words! I totally agree that it’s ridicu­lous that Honda is high-fiving them­selves on every chan­nel for such an obvi­ous fail. It really doesn’t sound any­thing like William Tell. I like Zach’s orig­i­nal thought that Honda’s engi­neers were just a bunch of indie rock lovers :). PJ, I hope this writeup pro­vides some form of jus­tice. Boy, does Honda deserve it.

  19. silverx10 says: Apr 19, 2009 @ 10:31 pm

    …It’s a car com­mer­cial. They use them to sell cars. All your effort is for naught.

  20. davidsd says: Apr 19, 2009 @ 10:40 pm

    Silverx10, thank you for that. I had a good time writ­ing this arti­cle, and obvi­ously a few peo­ple enjoyed read­ing it… so I dis­agree. And I don’t really care at all whether Honda sells more or fewer cars as a result of their com­mer­cial. I actu­ally like the civic a lot — my par­ents have had one for over a decade, and it’s a great car.

  21. Gooberlicious says: Apr 21, 2009 @ 10:29 am

    As a Lan­caster res­i­dent, I had to try it. It was in the West­ern part of the Ante­lope Val­ley (1 hr north of LA in the high desert) on Avenue K between 60th and 70th Streets West. So I drive over the grooves and found that I couldn’t make head or tail of what it was! I’m sort of a musi­cian (been piano-ing for 31 years) and thought, “OK, cool, now what was that?” Due to some cranky res­i­dents in the area and nut-job dri­vers, they have since removed the grooves and moved them to Avenue G between 30th and 50th Streets West, which is con­ven­tiently on my way home from work. I decided to drive the speed limit (55) and try it again and again. The new grooves sound even WORSE! I thought maybe it was my speed under­stand­ing fre­quency changes with speed in this case. Being the good engin­erd that I am, I decided to exper­i­ment with dif­fer­ing speeds, at 1 mph incre­ments. To no avail. A col­league of mine sent me this link and now I have my answer. It’s all mucked up!!!!

  22. William says: May 14, 2009 @ 5:45 pm

    Bril­liant analy­sis. Win for David, his think-outside-the-box dad, and the world at large. I really enjoyed the musi­cal com­par­isons. I don’t know any­thing about notes or stuff like that, but I can cer­tainly appre­ci­ate a good and sci­en­tific presentation.

    Thank you.

  23. Kenton says: May 29, 2009 @ 2:34 am

    I hope the con­grat­u­la­tions aren’t get­ting old — I just wanted to echo the same sen­ti­ments expressed above. I live in Canada and just started see­ing this ad online thanks to a proxy server that allows me to watch Amer­i­can pro­grams. It took two lis­tens for me to get fed up and track this down on Google.

    I still can’t believe the final ad in the series where the engi­neers high-five after lis­ten­ing to a clearly failed attempt — even worse when you con­sider they were hear­ing the “pre-re-tuned” ver­sion in the car as they con­grat­u­lated each other. But I’m sure they took a gam­ble that the vast major­ity of peo­ple wouldn’t notice, and judg­ing by the lack of search results on Google, I’d say they gam­bled right.

    Either way, your time on this was appre­ci­ated. It’s always nice to know your part of a crowd, even if it is small. ;)

  24. Chris says: Jul 22, 2009 @ 2:25 am

    that was really inter­est­ing. when i first saw the com­mer­cial i had no idea what melody they were try­ing to play with the car… it sounded like they just made it up

  25. Trevor Cox says: Mar 01, 2010 @ 11:51 am

    This is a great post and I’ve just used it on Sonic Won­ders, my new web­site about Sound Tourism.

  26. Peter says: Sep 16, 2010 @ 7:53 am

    Bravo! Ter­rific analy­sis. Must have been thrilling to find that the num­bers and pitches worked out when you pre­dicted their mistake.

    Too bad Honda got filched by the peo­ple they con­tracted. The whole con­cept is a cool idea – hope­fully some­one else will do it bet­ter… and maybe include a longer and more inter­est­ing melody.

  27. Pingback: Honda Needs a Tune-Up | Maszman Speaks! Jan 06, 2011 @ 9:11 am

    […] I’m sure you’ve seen the Honda com­mer­cial where they cut grooves in a road so that a Honda Civic plays a tune when it dri­ves over it at 55 mph. What you might not real­ize is that Honda made a mis­take in cut­ting the grooves and edited the audio to make it sound sort of corr.… […]

  28. David Edwards says: May 10, 2011 @ 12:13 am

    Hi David.

    Jeremy Heil­man raised the idea of using diag­o­nal grooves, or cut­ting dif­fer­ent grooves for each side of the car. That could be tried on the exist­ing road, by using a Renault 16 as the test car. This had two wheel­bases which dif­fered by 2.76 inches, so you would get a dif­fer­ent effect from each side — kind of like dri­ving at two dif­fer­ent speeds at the same time.

    Wikipedia has an arti­cle about this car; http://en.wikipedia.org/wiki/Renault_16

    David Edwards

  29. jeff says: May 12, 2011 @ 6:13 pm

    I have no idea what any of you peo­ple are talk­ing about.

  30. jeff says: May 12, 2011 @ 6:18 pm

    My friend James has no idea either. You peo­ple give me hope for the future.

  31. jeff says: May 12, 2011 @ 6:19 pm

    I didn’t mean the last com­ment sarcastically.

  32. Mike J. says: Aug 21, 2011 @ 3:39 am

    Late to the party but unfazed…

    How come the tire man­u­fac­tur­ers haven’t caught on? Your tires could play a wake-up mes­sage at all times! My tires already have a dis­tinct growl on hard turns, too, which could be adapted into a crit­i­cal warning.

    On the other hand, the reduc­tion in crashes from dri­vers falling asleep might be off­set by the num­ber of inten­tional ditch­ings by dri­vers dri­ven to insan­ity by the same three notes over and over and over and over.

  33. Larry Winiarski says: Aug 04, 2012 @ 7:06 am


    It’s obvi­ously not the William Tell over­ture.
    Maybe it’s the “Will you promise not to Tell overture”

  34. Pingback: Musical Road Oct 17, 2012 @ 1:19 pm

    […] you prob­a­bly noticed, the pitches are off.  Some­one wrote an excel­lent arti­cle explain­ing why that’s the case. Some of the grooves cut out to cre­ate the […]

  35. Trevor Cox says: Nov 18, 2012 @ 5:09 am

    I vis­ited the re-made road in Lan­caster last sum­mer which is badly out of tune as well, if any­thing the high note is even flat­ter than the orig­i­nal. I took along a tape mea­sure and an audio recorder as well. I drove at at about 60 mph using cruise con­trol to main­tain con­stant speed.

    The grooves are 1″.

    For the low­est note there is a spac­ing of 3.846″ between groves, mak­ing a dis­tance between iden­ti­cal parts on the rum­ble strip 4.846″. This gave me 217 Hz, a note just below A3.

    For the high­est note of the melody the fre­quency was 333 Hz (clos­est note E4) when it should be an octave higher than the low­est note (434 Hz, A4). The dis­tance between iden­ti­cal parts of the rum­ble strip was 3.152″ when it should have been 2.423″.

    It seems that they didn’t half the spac­ing between the grooves, and if they had done that it would have been less flat (but still awful).


  36. Pingback: Joel Eaton » Joel Eaton Sep 14, 2013 @ 10:51 am

    […] Singing Road! If you drift off the side of a road and hit a rum­ble strip, you’ll get a dis­tinc­tive sound intended to alert you and pre­vent an acci­dent. The pitch of the sound you get depends on the spac­ing between the bumps or ridges. So if you make lots of ridges like a rum­ble strip and vary the spac­ing between the ridges cor­rectly, then dif­fer­ent musi­cal notes can be made. Close together ridges (say 6mm apart) give high notes, and far apart ridges (say 12mm apart) give low notes. Make the right pat­tern of ridges then as a car dri­ves over them, a tune is played. Here’s a story of how Honda tried this for an advert then totally fudged it up. […]

  37. Dave Scheivert says: Oct 07, 2014 @ 1:19 am

    David, seems to me you just have to change the notes that put this in a Minor key, and it would be pass­able. (and not make me crazy IF I have to drive over it)

  38. Pingback: Syngende veier i susende fart | ballade.no Oct 07, 2014 @ 3:13 am

    […] aller første gang, om syn­gende veier da Cox spilte en smakebit av bil­gi­gan­ten Hon­das forsøk på å få en vei i Cal­i­for­nia til å spille Rossi­nis William Tell […]

  39. Pooriya says: Apr 28, 2015 @ 11:35 am

    Hi, my name is pooriya and im try­ing to build amu­si­cal road in my coun­try. I live in iran and my equip­ments are low, i actu­ally cut grooves of 12,6 mm in width and depth with a 12 mm spac­ing and another 24,12mm in width snd depth with 24 mm spac­ing grooves on the ground but when i go over them its mute ‚there is no sound . And i dont know why! Tell me whats wrong please help me thanks

  40. davidsd says: Apr 28, 2015 @ 12:10 pm

    Hi Pooriya, What a great project! My guess is that your grooves might not be spaced far enough apart from each other. You want to make sure that when wheels pass over the grooves, they oscil­late up and down a lit­tle bit. If the grooves are too close to each other, and the wheels are big, then the wheels won’t be able to tell the dif­fer­ence between them. So my sug­ges­tion is to try mak­ing the grooves far­ther apart. (Note that this means you’ll have to move faster over them to make the same sound.) 12mm sounds like a rea­son­able depth. But I’m not an expert, so you’ll have to experiment!

  41. Pooriya says: Jun 14, 2015 @ 6:56 am

    Hi david thanks for the advice. But i have a major prob­lem with what you basicly say here and that is based on my exper­i­ment which i cut grooves of 24,12 mm in width and depth with 36 mm spac­ing and 48,12 mm in width and depth with 72 mm spac­ing and another 24,12 mm in width and depth with 96 mm spac­ing then used sen­sors to deter­mine the fre­quen­cies and found out that the fre­quen­cies does not depend on the spac­ings becuase the first and last grooves which are the same in width but have diffrent spac­ings, had the same fre­quen­cies at a given speed! And i think its the way it should be because we have two fre­quen­cies here one is the impact of the tire and the groove which i think should deter­min the note and the sec­ond fre­quency is the repeati­tion of the first. To bet­ter under­stand this you can think of a sound of a fork hit­ting on a table, that has a spe­cific fre­quency but reapet­ing it could be 10 times per secon witch is a fre­quncy but doesnt deter­mins the note am i wrong??

Leave a Reply