SpeechJammer Puts An End To Annoying Speakers

If you’ve ever had to deal with people disturbing your peace and quiet by yammering on with their cell phones, you might be interested in the SpeechJammer.

The idea behind the SpeechJammer is fairly simple: It’s very hard to speak if your words are recorded and played back to you a fraction of a second later. This is a real psychological phenomenon known as delayed audio feedback that also has a beneficial effect on stuttering.

According to the researcher’s writeup (PDF warning), the SpeechJammer works by measuring the distance to the ‘target’ with an ultrasonic distance sensor and records the speaker’s voice with a shotgun mic. The recording of the spearker’s voice is delayed for about a fifth of a second and then played on a speaker on the front of the gun.

The researchers tested two conditions: ‘reading news aloud’ and a ”spontaneous monologue.’ Subjects who were reading news aloud had their speech jammed more often than those with the monologue, but the results look fairly promising. There’s only one video of the SpeechJammer in action (available after the break), so we’d like to see a few Hackaday readers build their own ‘shut up gun’ and send in a demo with an annoying talker to validate the results.


67 thoughts on “SpeechJammer Puts An End To Annoying Speakers

    1. We did this to a newsreader in the radio station where I used to work. Every channel in a properly set up facility could be over patched, so we took the feed from her mic in the apps room, fed it into a quarter inch tape machine and flipped the out from the record head to the replay head before inserting the signal back into her return feed to headphones.
      Using the vary speed control we were able to vary the effect at will so that she would start out fine and slowly began to sound like she was losing brain cells.
      The effect was so subtle that she was convinced that it was her problem rather than someone was manipulating her headphone mix.
      It was a cruel thing to do I know, but she was a *&$£! forever telling novice newsreaders (ever seen a grown man cry?) that they were terrible and that with her speech and drama experience she was the best voice to model themselves on.
      Yeah right.

  1. This effect also happens on Teamspeak or Ventrilo servers if someone has a mic that is feeding back.
    Also as a note; it is possible to continue to talk with this effect in place by ignoring everything that you hear.

    1. Actually, this is an interesting project. I would venture to think that if anyone was pointing ANY unwanted device at you, you would probably stammer a bit.

      I saw this on the news last Friday and the first thing I thought of was exactly what Tom Schwarz ( btw: are you the guitarist and photographer who used to live in Savannah? ) said above… that got me thinking about how to combat this. I haven’t looked at the writeup, but it appears the speakers are an array of ultrasonic transducers – which allow for hyper-directionality. That means you would have to be in a fairly tight area to even hear the SpeechJammer. It would be possible to add the functionality to detect this and nullify their transmission by playing back their signal ( from the podium ) out of phase.

      Again, it is an interesting project, but I can see the use of this getting out of hand.

      1. doing anti-sound is very difficult if not done in a very, very controlled enviroment. A handheld device would be very limited. If it was easy to implement the technology would be all over the place. This is just a device that disorientates the speaker. Very easy to do.



      2. You are correct, they would probably hear a ton of phasing, which would / could confuse them as well… or at least make them think they are faint-ish.

        This could probably be handled best with headphones, or in ear monitors which would create the controlled environment. Think Bose noise-canceling headphones, except more expensive and specific.

      1. Pretty sure vonskippy’s just at Hack A Day to talk about how much awesome hacks suck.

        I’m considering just autoreporting any comment he makes that someone hasn’t replied to yet, so the fact that I haven’t reported this one is your fault ;p

  2. The willingness to say STFU(or word to that effect), or please STFU,to be polite is less expensive. So @ vonskippy for in Japan this is not useless technology, when there is a portion of their society will not say that. Hopefully this is the proof of concept model, and the stealth on is under development. Even here in the US one isn’t likely to tell a lecturer to STFU. Although I’m not sure how stealthy it can be when ever yon in range will hear it. I use my cell phone as little as possible, because it has a delay that I can hear when using it, but other who use it can’t hear the the delay

      1. Space? I was thinking Super Low Orbit Satellite with a nuclear powered MPD thruster to counter aerodynamic drag. And switch out the directional mic. for an LRALP unit.
        Probably easier to just block access to Facebook now I’m thinking about it.

    1. Naw, Rush has been doing radio for too long. Whether you think good or bad of him, when you speak for your job, I imagine he would have no or very little trouble ignoring it. Also helps that he is hard of hearing.

  3. Now instead of playing back what they have said intercept the signals that control speech, sense what they are about to say then fire that at them. The brain should pick it up as if they have said it and not trigger the action :)

    1. I’d be more worried about stuff like ADS, acoustic weapons, and taser drones since they’re a bit harder to counter using methods that won’t be used by riot cops as an excuse to brutalize protesters even further.

      With this all you need is some mental training or just stick your fingers in your ears.

    1. would that matter if it picked it self up? It would end up adding to the jamming effect, or creating feedback which would be ear splitting for the person speaking… which would probably also cause them to stop talking. Not that I agree with that, but it would do the job.

  4. I read that Bob Pease, whose obituary was posted here last year, built something similar. He put it in his office to annoy bosses/coworkers that came in to needlessly interrupt his work. It shifted the audio to a higher frequency, which neatly solves the potential for it to record itself; since frequencies outside normal human speech can be filtered out. Though I don’t recall a delay or inducing stuttering, it was reported it had the intended effect of making such interruptions shorter.

    The British “Mosquito” uses a 17.5khz tone to annoy teens. Shifting to a frequency like that would allow you to use a piezo, which makes it smaller, louder, and more directional. And if the frequency is chosen correctly, more selective to the targeted individual.

    1. That’s pretty funny ( what Bob Pease did )… One of my old co-workers ( we were both audio engineers ) would do something pretty mean when a vocalist did something that he didn’t like. He would screw with them by pitch shifting their vocals through their monitor by just enough to get them to correct so they are off in the house. Pretty mean, but a funny idea.

      1. I was recently at a filk music convention, and the sound guy decided to prank the MC. He shifted her voice up by a third and played it back to her in the monitors. She stopped talking out loud and shifted to ASL signing, which was a bit garbled because she had a mike in one hand, but some gestures were very clear even to those of us who don’t sign or were at the wrong angle to lipread.

  5. This may be more effective with a changing delay. For each pause (as in each word) increase the delay or lessen it.. I think it would cause more confusion. Also dont forget random animal noises.

    “So I was like talking to *Quack* the other *Woof* when all of the *Mooooo* a *Ca KAW*”

  6. #1, if the lecture is bothering or boring precious you, leave.
    if you paid to be there, you would not use this.
    if you did use this once, you would not be welcome to attend again. This is rude.

    #2, if someone pointed it at me, I would consider it a weapon and act accordingly.

    #3 yappy cellphones??.. buy a jammer.
    Discreet is elite, and effective.

    If “my friend” built one “my friend” would include scaleable deterents, sonic phasor to vibrate skull bones, high pressure ultrasonics, shimmer strobe (8Hz option), tasor, (–classified–), (–classified–), on-off switch, remote functions.

    My backpack doesn’t have room to tote one of these, pepper spray is smaller and really gets to the point.

    I expect this would earn you bracelets for possesion of a device [blahh-blahh] causing a public disturbance.
    Getting charged, even with charges dropped = no visit to US ever without harassment.

  7. Easy to demonstrate yourself: Get some headphones and a random microphone, plug them into your sound card, dial in some delay (what? of course your modern discrete sound card has a delay function!), and try talking.

    It’s hard to do.

    This is why performers at full-scale shows either have in-ear monitors, or huge near-field wedges on the floor pointed at their head: The sound from the front-of-house PA is too late to hear yourself in a timely fashion, while also too loud to hear yourself in your own head.

  8. John Boy & Billy, a radio show in North Carolina, did this several years ago with a regular guest. From what I understand, they just delayed the audio playing into his headphones while he sang “Silver Bells”. They had a lot of calls about making fun of disabled people, etc, so they described how they had set up the gag.

  9. I had trouble with this same effect when I was working with audio distortion. I was using VSTHost to filter my voice but it was really hard to continue talking due to a half-second delay. I couldn’t test like that so I had to record it and play back when I was done.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.