Skoda developed (with the University of Salford) an all-mechanical bicycle bell that bypasses active noise cancellation (ANC) in headphones.
https://www.skoda-storyboard.com/en/skoda-world/skoda-duobell-a-bicycle-bell-that-outsmarts-even-smart-headphones/
It's tuned to 750hz, which is a "blind spot" in most ANC headphone processing.
There is also a resonator for a higher frequency, and the hammer produces irregular strikes, confusing the ANC further, and making pedestrians look behind them faster.

Video:
https://www.youtube.com/watch?v=zDaVPfpQvPI

Research:
https://cdn.skoda-storyboard.com/2026/04/Skoda-DuoBell-Research-final_cf127752.pdf

#BikeTooter

Škoda DuoBell: A bicycle bell that outsmarts even smart headphones - Škoda Storyboard

Pedestrians wearing headphones are exposed to an increased risk of accidents. In an effort to reduce collisions with cyclists, Škoda Auto, in collaboration with scientists, introduces an innovative bicycle bell whose sound can penetrate even active noise cancellation systems. In doing so, it helps prevent injuries to both pedestrians and cyclists.

Škoda Storyboard

@dec23k that is very cool. But I can't buy one!! Hopefully people start manufacturing them because the noise cancelling problem exists in Sydney, Australia as well.

#SkodaDuoBell #DuoBell #BikeTooter

@moz
It should be possible to build a simple battery-powered electronic 'horn' that pulses a tone around 750hz, and another tone in the typical 'bell' range.

I have a few ways of generating 750hz from a portable device, but no noise cancelling headphones (of my own) for testing.

@dec23k @moz

Or a 115 decibel bike air horn will just over power their headphones, make them jump 4 ft in the air and shit themselves.

https://www.ansco.com.au/air-zound-xl-bike-horn-pet-cannister

Air Zound XL Bike Horn

AirZound is a loud 115db, the Air Zound Bike Horn refills quickly and easily with a bike pump or compressor.

www.ansco.com.au
@SuperMoosie @moz
Crikey!
But also, that does look like it's intended for use in mostly vehicle traffic.

@dec23k @moz

Yes, traffic is what I used to use it for.

But one day a guy wandered across botany road in right front of me. He thought he was about to be hit by a truck.

@SuperMoosie @moz
I like how the possibilities in the research included mechanical (possibly wheel-powered) air sirens. I have used a hand-crank air siren; there's a lot of stop and start needed!

But regular DTMF tones played loud over a Bluetooth speaker might be close enough - the number 6 includes 770hz and 1477hz.
The higher tone is just a courtesy to make it sound more like a bicycle bell; it's not critical for ANC.

666 FTW anyone?

@dec23k @SuperMoosie yes, they target motorists. But they also work on other vehicles, and on pedestrians.

Air Zounds use a standard soft drink bottle cap, so you can replace the bottle it comes with with a 2l or 3l soft drink bottle. And if you're BikeSaint, pour the soft drink down a drain :) It's worth covering the bottle with tape or something just in case it gets damaged and shatters.

3l gives you quite a lot of tooting, but takes quite a lot of bike pumping action to fill

@moz @SuperMoosie
I like that sustainable design decision, and I admire your dedication to research!
@moz @dec23k
I used to recharge mine the local servo air pump.
@dec23k uuhhhh, the new world 😱🫣
@dec23k @jwildeboer This looks very much like our old GDR bike bells. The exact same mechanics inside 😂
@maxheadroom No surprises here, knowing the history of Skoda :) And by making the research available, I expect Anti-ANC bike bells flooding the market really soon :) @dec23k

@jwildeboer
I expect anti-anti ANC bell headphones flooding the market soon enough, too.

@maxheadroom @dec23k

@datenhalde @jwildeboer @maxheadroom
The website says the "hammer mechanism produces rapid and irregular strikes".

So the extra processing power needed to decode the random timing will drain the batteries in the new headphones faster.

@dec23k If people are riding bikes in places where pedestrians are walking, then they should be prepared to dismount and wheel the bike along if they haven't got a clear path ahead, instead of expecting people to jump out of their way because they've rung a bell. It is not fun to find a cyclist zooming up behind you at a rate of knots, and it's difficult to gauge where you should move to in the short space of time you have between them ringing the bell and catching up with you.
@HollieK72 @dec23k and what about the situation, when people are walking, where (only) cyclists are expected to ride?
@x_tof @dec23k In the UK, it's often very difficult to see where the cycle lanes are supposed to be, as they switch between the road and pavement without any rhyme or reason, and are therefore confusing for cars on the road, pedestrians on the pavement, and cyclists at any point. Proper cycle lanes that are separate from the road and pavement would be the solution, but that's not coming any time soon. Vehicles and pedestrians should not be in cycle lanes.
@HollieK72 @dec23k I don't think most people who use a bell (or most people who complain when we don't use a bell!) are expecting pedestrians to "jump out of the way" as they sail past too close without slowing down. I won't deny that cyclists like that exist, but most bikes slow down and only use the bell to make sure people are aware and not surprised by their approach. (I prefer a cheery "good morning" myself, or where I do need someone to move an "excuse me, sorry!").
@HollieK72 @dec23k I'm thinking here mostly of the DNA path in Cambridge, which is a shared route for bikes and pedestrians, completely away from the roads, but which was built too small for how popular it has become.
@lnr @dec23k I know Not All Cyclists Are Like That, but I was talking about the ones that are like that.

@dec23k I can appreciate the engineering, it feels like sometimes my bell is inaudible to everyone but me.

On the other hand regarding the safety aspect, I've started to change my habits so that when pedestrians are around, I try to slow down enough such that my bell wouldn't be necessary to avoid an incident at all. I think the best safety action would be building more separated lanes.

Perhaps I just don't live in a city that's big enough?

@dec23k interesting.

From the paper I understand the 750Hz tone is around 3dB less attenuated, but only in quiet areas with already not much noise, like parks.

In noisy areas, any tone that is distinct from the noise is more easily heard. But everything is attenuated and the warning distance is considerably lower already.

I wonder if these ANC headphone makers are obliged to support 'emergency' vehicle tone detection and not blank it out? Does anyone know?

Or there is the air horn solution🤪

@Slash909uk
Regular DTMF tones (played loud) might be close enough - the number 6 includes 770hz and 1477hz.

The higher tone is a courtesy to make it sound more like a bicycle bell, so it's not critical for getting past ANC.

666 FTW? Anyone?

@Slash909uk
> I wonder if these ANC headphone makers are obliged to support 'emergency' vehicle tone detection and not blank it out?

I would assume so, because of the potential legal liability.
Yet 750Hz meets this only partially, as it's the lower range of the lower band of the UK's and few other European countries two-tone pattern¹.

So that would assume there are more frequencies to take into account and wanted to check in with RTINGS tests, but they »

@dec23k

Two-Tone Siren - Technical Guide & History

Complete guide to the Two-Tone (Hi-Lo) emergency vehicle siren: technical specifications, waveform analysis, historical evolution, and how it works.

@dzwiedziu @Slash909uk
Regular DTMF tones are close - the numbers 4 5 6 include 770hz.

May or may not be close enough.
I don't have any ANC headphones so I can't test that myself.

@dec23k Sounds good, but who's actually selling it?
@petererer
Nobody, yet.
They made enough to lend one each to a set of food delivery riders (in London) for road testing, so it's not vapourware.

@dec23k Yeah, let's fix all sound sources, instead of the single receiver! What about the other important sounds a pedestrian is supposed to hear?

https://society.oftrolls.com/@geert/116368228635841792

Geert Uytterhoeven (@[email protected])

@hrw Yeah, let's fix all sound sources, instead of the single receiver! What about the other important sounds a pedestrian is supposed to hear?

Society of Trolls

@dec23k
I'm sure it is a great bell, and being able to be heard even with noise cancelling headphones is great. But I'm still not sure I get the video.

The video says that the bike accidents have increased with 24% due to noise cancelling headphones, and implies that this bell would fix that.

That means that the hypothesis they had was that a person on a bike saw the pedestrian, rang its bell, got no reaction and then just ran over the pedestrian?!

@dec23k ANC does not and cannot cancel single audio events... That's not how this works. That's not how any of this works... 🙂‍↔️ANC let's you hear a bicycle bell even better because it cancels the constant traffic noise that would drown voices, horns and bells.
@MattisCB @dec23k Seems like that is how it works plus the distraction of listening to loud music: https://www.dekra.mx/en/keep-your-ears-open-in-traffic/
@EdBruce @dec23k i have and use ANC headphones. Loud music and distraction, i give you that.

@dec23k

So, in the use-case scenario here, I'm on my bike, and I see that there's a person walking/wheeling ahead of me and they're wearing noise cancelling headphones.

Question:

Why do I need a bell?

Assumptions:

(a) If they're ahead of me and in my path, I'm either on a footpath, shared path, or carriageway;

(b) I have seen them.

Answers:

(b) I have seen them;

(a) If I'm on a footpath, I need to move; if I'm on a shared path, I need to wait; if I'm on a carriageway, I need to wait.

@clickhere
In your scenario, if you don't need a bell then you also don't need a Skoda bell (which may or may not cut through ANC).

The video that I linked in the OP has a clip of a busy bridge in London, with a number of Obvious Tourist pedestrians who had stepped down from the footpath into the cycle lane.

The web site and PDF mention 2 frequencies, and don't say whether different headphones have a different "blind spot" frequency.

I'd advise field testing with tones before buying a bell.

@clickhere @dec23k My use case for bells is that there are people ahead of me walking in the road, and I want to let them know I'm there when I'm far enough away that I won't startle them when passing politely (assuming it is safe to do so). Some people prefer a bell to a call -- sometimes a bell announces "bicycle" rather than "runner" in a way people prefer.

I don't cycle on paths or the pavement.

@ravenbait @dec23k Oh yeah, in general, I use a bell now and then, in a fairly similar way.

But if I see that someone has noise cancelling headphones - or otherwise doesn't seem to have reacted to the sound of the bell (they may be deaf or hard of hearing) - then I'll adjust what I'm doing. (Like, I won't barge past, usually just take my time and wait for an opportunity to pass safely.)

But this over-engineered, highly specific use-case? Nah.

@clickhere @ravenbait
I just saw something vaguely interesting in a video, and saw that nobody on Fedi had written about it. So I wrote something on Fedi about it, and then more than 100 accounts liked or boosted it.

The video has quotes about "working with partners" in London, and "sharing this with the world", but Skoda seem to have only made enough bells for the field test, and aren't trying to sell them.

Maybe they just want more people to be aware of Škoda's bicycle-manufacturer origins?

@dec23k @clickhere It's weird, isn't it? It's basically stealth marketing. I wonder who the audience is.

@dec23k @clickhere Or who the *intended* audience is, I should say.

Full disclosure: we have a car chez Ravenbait, and it is a Skoda. Other than it being a reliable workhorse that is relatively cheap to repair, I have no allegiance.

@ravenbait @clickhere
Full disclosure: I have never owned a car, and I no longer own a bicycle (was a daily rider until I had 4 stolen in the same 12 month period when I was living between the canals). Getting around mostly on foot and on public transport. Never had the car brain, but the rothar brain is still strong.

I also don't own any ANC headphones, so I can't test for any possible "blind spot" frequencies.

@dec23k Full disclosure: I'm multi-modal.

(oo-er)

Although, tbf, my motorised mode was years ago and a Vespa, but it counts. (I *can* also drive a car, just not well or legally. Maybe I should get one!)

@ravenbait

@ravenbait @clickhere
And now there's a yt-short video about it from renowned mathematician Hannah Fry.

@dec23k Ker-ching!

Yeah, Skoda have been leaning into the cycling buzz a fair bit over the years, so this could be more marketing, all right.

I mean, I think it is interesting as an exploration of the physics, etc., but not so sure about the bike bell application - maybe emergency warnings in public places, or something?

I must take a look at Fry's video, though, love her. She's such a nerd, and such a great communicator for normies like me.

@ravenbait

@clickhere @dec23k I'm certainly not the right market for their bell -- I have a couple of very loud, sustained, but discreet bells that I bought for their minimal handlebar real estate and delightful sound. That thing looks like you could use it as a life preserver if you fell in a canal! I will give them points for trying to build a bridge between motorists and cyclists, though, because we sorely need bridges.
@dec23k
750Hz seems weird, as there is no one emergency signal frequency, yet better than my idea of blasting directional infrasounds, to make pedestrians viscerally afraid of crossing cyclists' paths ^_^

@dzwiedziu
That 750Hz could be an artifact of the DSP in the headphone design, or even something as simple as the casing around the microphone causing a notch filter (not as dramatic as what they show in the video). I would expect it to be different across different manufacturers.

If 770Hz is close enough, then playing the DTMF for 666 loud over a Bluetooth speaker should also work!

@dec23k That 750 Hz bell that uses the 750 Hz gap in noise cancellation headphones seems to be all nice and good, until you look at the actual frequencies that the bell emits:

- 1980 Hz (missing sometimes?)
- 5040 Hz
- 9000 Hz
- 13300 Hz

The image below was created using Specdroid while playing the video.

@lenaschimmel
So there isn't any 750Hz audible in the video?

The 1980 Hz is probably the secondary resonator, missing sometimes maybe because the rider's glove was muting it?

The higher frequencies could be harmonics or other resonances.

@dec23k This is rubbish. 750Hz is around F5#, the F# in the octave above middle C.

So the claim is that if you put on your ANC headphones and run your fingers up the white and black keys, then a few notes will be particularly louder.

If you have a piano or a keyboard, give it a go now. See for yourself. One of the cases where a cheaper instrument is better.

If this really existed, audiophiles would be besieging the HQs of Sony, Bose and Apple over poor headphone performance whenever near the huge amount of popular music in F# minor.

---

So of course I had to read the paper. Turns out the claim isn't that 750Hz is louder. The claim is that 700-800Hz is psychometrically more noticed. So this need not be a feature of ANC at all. It could be a feature of humans.

And why are these frequencies more noticed as bicycle bells by human listeners. That would be because these frequencies are often used by bicycle bells. People pay more attention to the things they are listening for. Shocked I am.