Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
×
Input Devices Music

Ask Slashdot: How Can a Blind Singer 'See' the Choirmaster's Baton? 189

New submitter krid4 writes "Question from a blind friend: 'My ears replace my eyes. However, when it comes to the very moment of starting, or the change of tempi, my start will always come too late. Neither tuning in with the voices around me, nor listening to the moment of their breathing-in helps to solve this problem. Fancy that it might be possible to produce tactile pressure or even lines at the top of my right hand, head or body. Even pulses would do, because what finally counts is the moment of the 'beat' produced by the choirmasters baton.' What simple, possibly DIY solutions are possible? It would help many blind chorus singers."
This discussion has been archived. No new comments can be posted.

Ask Slashdot: How Can a Blind Singer 'See' the Choirmaster's Baton?

Comments Filter:
  • by femtobyte ( 710429 ) on Sunday April 07, 2013 @05:12PM (#43386419)

    Motion tracking video of the baton (cheap webcam view from the side, colored foam ball on the baton end, track up/down motion with some very simple image processing); convert to a usable signal (e.g. audible clicks through an earpiece when the baton reaches maximum/minimum positions and turns around).

    • alternate "visualization" method: have a constant stream of clicks through headphones (open-sided so as not to impede following the singing), varying the audio phase to move the stereo "image" side to side as the baton goes up and down; this should make it easy to follow the whole motion of the baton, not just the extrema, using the most cheap and available off-the-shelf hardware.

      • The most 1:1 mapping I can think of to use for this would be to assign baton movements to gestures and define some output value to it. For example, you could play specific pitch tones in a pair of headphones for the blind singer whenever the baton moves, and each tone would indicate the beat and tempo to use.
        • by femtobyte ( 710429 ) on Sunday April 07, 2013 @05:38PM (#43386627)

          The downside of using "specific pitch tones" is confusing the heck out of the subconscious of someone trying to sing at some other particular pitch (to match the voices around them). Something broad-spectrum and atonal (a click, hiss, tick, or thump) can relay timing and position information, without interfering with (competing for attention in the brain) tonal perception.

    • by newcastlejon ( 1483695 ) on Sunday April 07, 2013 @05:34PM (#43386593)

      Motion tracking seems excessive when the poster would be satisfied with a simple pulse. I would suggest an accelerometer mounted to baton/conductor and a rumble motor ripped from an old gamepad for the singer (or an earpiece).

      For a no-tech alternative try having the singer sit/stand near the conductor and have the latter tap their feet (hard) along with the baton.

      • Even if motion tracking is "excessive," it's probably easier to implement in readily-available hardware: use a camera already on a laptop/phone, rather than needing to wire up the conductor with a custom accelerometer mount. And ripping a rumble motor from a gamepad (and creating the software/hardware interface to control it) is a heck of a lot more work than just piping sound out to headphones. As I describe in another post above, the motion tracking (in a simple and controlled environment) is pretty trivi

      • by Kell Bengal ( 711123 ) on Sunday April 07, 2013 @06:20PM (#43386861)

        I would suggest an accelerometer mounted to baton/conductor and a rumble motor

        I'm a robotics researcher - some of my work includes developing aids for the blind. Of all the comments here, this is the sanest one and the one that would actually work for people with vision impairment. It's simple, it's cheap and it will WORK. We've had good success with similar systems for other tasks like navigation and playing soccer.

        • Thanks. Re your sig, guess which one I am.
    • by Anonymous Coward on Sunday April 07, 2013 @05:45PM (#43386663)

      They already have motion sensitive batons. A person I know uses one to practice his conducting. His old one connected to the computer through a MIDI interface (and his new one uses USB) so that way the music would follow his conducting. It is interesting how much information just the baton conveys, especially through the way modern conductors form the ictus. He can control the tempo, dynamics and many other more minute things.

    • by sribe ( 304414 )

      Motion tracking video of the baton (cheap webcam view from the side, colored foam ball on the baton end, track up/down motion with some very simple image processing); convert to a usable signal (e.g. audible clicks through an earpiece when the baton reaches maximum/minimum positions and turns around).

      I doubt that would help. He needs to know about the velocity change as it happens, not after the next maximum/minimum is reached earlier or later than he was anticipating.

      • See my own reply to my comment above, suggesting using stereo sound to "visualize" the continuous baton position. Once you've got the motion capture data, you can convert it to whatever output form you need to convey an adequate amount of information. Nonetheless, I suspect that even knowing just the endpoints would be mostly sufficient --- conductors don't generally rapidly/erratically vary the tempo from beat to beat (even the sighted singers couldn't accurately follow that), so you'll have time to feel t

    • Seldom does the exact motion of the baton matter. Different conductors have different styles. But you could modify a Wii controller to follow the motion of the baton and turn it into pressure or some other kind of signal.
      • I should clarify what I mean about the motion not mattering.

        Different conductors sweep the baton around different ways. But they will have common features related to the motions (when the upbeats and downbeats are, cues, and so on). It is the timing of the motions, rather than the gross motions, that are most important.
        • How you capture the motion --- visible light image tracking or accelerometers --- doesn't particularly matter, since it's trivial to convert position-v-time data to acceleration-v-time data. A concern I have about using many readily available off-the-shelf accelerometer devices (e.g. a game controller) is the form factor; a conductor will be pretty picky about what he/she is willing to wave around for a multi-hour performance; a Wii controller would get awfully tiring compared to a proper light and agile ba

        • by gd2shoe ( 747932 )
          Kinect is one idea, but Wii controllers are far, far too heavy for this application.
          • I didn't mean it quite literally. But I don't think the nunchuck controller is really that heavy. Especially if you removed the case and the buttons (figuratively speaking) and just used the accelerometer.
    • Motion tracking may or may not be a good idea, but if you're to try it, don't do it from the side. The side-to-side movement is just as important to a musician as the up and down, especially at the start of a piece. (For example, if a conductor wants to bring the group in on a fourth beat anacrusis by beating beats two and three, the important movement will only be sideways, or away-from/towards the sensor.)

      This is a big problem for me---I play to the conductor's side in a band I'm in and it's very easy to

      • Good point. I suggested a side view for ease of unobstructed view, assuming that the 1D information would be sufficient. I think this is still likely to be a good assumption. In a more advanced choir, every fine nuance of the conductor's instruction can be critical; however, given that the blind singer is mostly able to keep up with the choir (despite zero direct info from the conductor), and hasn't been kicked out for unacceptable performance, I'm guessing that this is a more relaxed hobbyist choir that's

        • I've been sitting at my desk for the last couple of minutes studying my own conducting style. I've always tried to maintain a very clear beat (a lot of my conducting over the years has involved beginner bands, so communicating where the beat is is far more important than it might be in more professional outfits.)

          Anyway, I've noticed that my style is very three dimensional. My upbeat, for example, starts near my sternum, goes out from my body and follows a roughly circular path back to near my forehead. The

    • Ever tried one of those batons? If you put anything of weight on the end, it will seriously impede the director in his/her movements and expression. Also, it's way more than just the up/down motion. Anyone that was ever in a decent choir/orchestra will know that the director has two hands, the baton goes left right as well (up-down is only for 2/4 beat) and the direction the director looks and stands has meaning as well. The amplitude of the baton, and lifting or pushing motions of the director indicate vol

  • Can't the choir director accommodate your disability by counting down the beginning of the song? Forcing you both to adapt some cumbersome technology seems silly.

    • Re: (Score:3, Informative)

      Can't the choir director accommodate your disability by counting down the beginning of the song? Forcing you both to adapt some cumbersome technology seems silly.

      The baton is used for more than just starting the a song.

    • Doesn't help with tempo changes in the middle of a song.
      • Isn't learning these things part of rehearsal? Maybe an extra session with the conductor to determine what will be happening beforehand.
        • by AK Marc ( 707885 )
          Tempo changes and such should be covered, but starting a song from silent room would be hard to do. They don't yell "5-6-7-8" clicking the drumsticks together before starting Ave Maria.
        • By that logic a conductor is not needed during a performance; all they would have to do is press play. Every performance is a little different as the conductor modifies volume and tempo.

    • by arth1 ( 260657 )

      The way I've seen this regularly done (in one choir and one orchestra) was the conductor tapping the music stand with one rod while making visual cues with the other.

      I though this was common practice when having blind members?

      • All I have seen tapping used for is to get a choir's attention. How does tapping convey volume change? Can the tapping be heard during loud portions of the song? Does tapping interfere with the music? Depending on where the wand is in the air also conveys which beat Conducting is much more than being a metronome.

        • by arth1 ( 260657 )

          All I have seen tapping used for is to get a choir's attention. How does tapping convey volume change?

          It doesn't, but the point here (RTFS) was to be able to start at the right time. A blind musician is expected to remember what the conductor wants.
          (So are others too - the conductor's role during a real concert is minimal; rehearsals is where you learn how to play or sing the piece the way the conductor wants>)

  • by sanpitch ( 9206 ) on Sunday April 07, 2013 @05:20PM (#43386481)

    If you're singing in a choir, then you're standing next to someone else, who is likely sighted. Just have them give you the cue. It could be that they hold your upper arm, and slide it down to the elbow it as the choirmaster's baton drops. If the choirmaster gives a four-count before starting, then the helper's signal may be four squeezes on your arm, or four taps on your shoe. I don't imagine that it would take much training for a new person to help you with this, and it's much cheaper than some high-tech solution which may not work.

    • by Cow Jones ( 615566 ) on Sunday April 07, 2013 @06:18PM (#43386849)

      FWIW, this is exactly what we do to cue a blind choir member.
      It's not a geeky solution, and it involves people touching each other, but it's very reliable.
      I can't imagine any sighted choir member refusing to do this.

      • by martin-boundary ( 547041 ) on Monday April 08, 2013 @05:52AM (#43389613)

        It's not a geeky solution, and it involves people touching each other, but it's very reliable.

        Ah, but I'm afraid that only solves 99.99% of cases. To fully solve the problem, you also have to deal with the 3 blind-choralists-in-a-row problem, which is well known to not have an analytic solution except in trivial cases. Perhaps a numerical approach using a Sobolev auditorium might work, though.

    • by bmuon ( 1814306 )

      You're assuming the tempo doesn't change throughout the piece. This isn't true for a lot of musical pieces. This really needs a technical solution.

      • by gd2shoe ( 747932 )

        No tech solution is needed, just a greater refinement of the helping friend idea.

        For instance, holding hands down low, and having the sighted friend move his hand in a very subtle beat pattern when things change.

  • I know -- it sounds like a joke in poor taste. But hear me out: By using a kinect to track the wand, you could have it control a shock collar by having it emit short pulses to create morse code. You can wear the collar as a bracelet and have it turned down super low.

  • She couldn't hear people enter the shop...

    Vibrating bracelet bluetooth bracelet and replace the baton with a wiimote :)

  • Ask someone who's encountered the problem and seen it solved... Like a choir director or one of the many organizations of choir and choral directors. Here's one. http://www.chorusamerica.org/ [chorusamerica.org] Part of the purpose of such organizations is to share information about what works for common and uncommon situations.
  • by robbak ( 775424 ) on Sunday April 07, 2013 @06:03PM (#43386769) Homepage
    Although the choir starts singing on the large downward movement of the baton, that is not the cue the choir is using - if the started singing after seeing the downward movement, they would always be late. They are actually taking their cue from the very subtle upward movement just before the downward sweep. Even detecting this would be difficult. The size of this movement, and the delay between this movement and the drop, whether a movement is the of the 'get ready' upward sweep... all very difficult and confusing things. And the nature of the movements will change depending on conductor, the nature of the music, or even the conductors mood. The human brain sorts all of these things out just fine. The best idea is one I read from another poster here - have the neighbour of blind singer give them their cue.
  • OK, so we've got posts here about wiring up a webcam to a computer to do motion tracking and such...how about build a simple pedal with a momentary push button and have the conductor tap his feet on that, then connect that to, say, a 555 timer to generate an audible tone while the button is depressed which can be routed to some sort of earpiece or headphone -- or even a piezo element which the singer should be able to feel vibrate or click. Whole solution shouldn't cost more than $20, and that's at Radiosha

  • I play tenor sax (badly ...) and practice to backing tracks provided by my ever-patient teacher. He provides some ticks (baton hitting the stand...) before the music and after that, if I drift from the backing track I know I'm out. Or it is :) It's difficult if you're a soloist, and doing an unaccompanied bit: I had "The Long and Winding road" once, which has a bit where the soloist (me) is all by himself for about ten seconds ,,, cojones of steel...

    And good luck to your blind friend. It's diffic
    • by gd2shoe ( 747932 )

      Lord knows how your friend will tell a minim from a crotchet.

      I've never had the opportunity to ask this before: Does anyone still call them that, or were you joking? If that really is what you're used to calling them, then where are you?

      • Dunno about kittenman, but we still call them that in Australia. I do speak (and prefer) the 1/(power of two) system though. If my daughter's education is anything to go by, it seems the systems are used interchangeably in schools.

  • by Tsu Dho Nimh ( 663417 ) <`abacaxi' `at' `hotmail.com'> on Sunday April 07, 2013 @06:20PM (#43386857)

    However, when it comes to the very moment of starting, or the change of tempi, my start will always come too late.

    Ah .. the trauma of remembering band practice:

    Every conductor has a different style. The signal to start your part of a song that has already begun may be a small flick or pointing of the baton in your general direction, barely interrupting the overall tempo of the conducting, or if you have a dramatic conductor it can be a two-handed "picador going over the horns" gesture ... or no gesture at all.

    Because the baton may be signalling to someone near the OP - in front or behind - but not the OP, the problem is discrimination as much as detection.

    Also, it's not always a down beat. Changes of volume, extended notes and the final cut off of a long final note may be sweeping or tiny gestures sideways or straight towards the choir or orchestra.

    Very few conductors will make big changes in tempo from what was practiced. No good will come of it.

    In short, it might be more practical to start on the second note and drop out on the next to last note, paying attention to the parts of the production that immediately precede your bits so you are ready for it.

    • by Twinbee ( 767046 )
      To make matters worse, the 'beat' for some conductors comes in on the downstroke, and for others, on the upstroke, and variations in between. It varies according to country, but even within a country, there's differences.
  • A friend's hand, tapping in rhyhm; or the baton itself tapping the podium. Or the friend's foot tapping the rhythm, which you hear.

    For solos, some creativity on the conductor's part can eliminate the problem.

    For a tech solution, I wonder if there is something from wii?

  • I sang in a semi-pro choir for a while and at one point our director had us all move to the edges of the largish church we were rehearsing, had us face the church walls (i.e. away from each other).... and start singing in unison. Believe it or not, if you know the music and the group you're singing with, it's very doable.

    Barring that, having someone who knows what they're doing holding the blind person's hand and tapping or squeezing should do the trick.

  • As an erstwhile musician, my opinion is that conductors are overrated. Once you learn the music you should be able to perform by listening to the others in the group.

    I played in a professional orchestra where the conductor gave a downbeat, a rest, and the orchestra had to play the opening note in unison without any indication from the conductor. It worked fine. There a many groups playing very complicated music, with abrupt and gradual changes in tempo, without a conductor. Jazz and jam bands do this wi

  • by bratwiz ( 635601 ) on Sunday April 07, 2013 @08:38PM (#43387529)

    Simple, replace the baton with a starter's pistol..

  • by girlinatrainingbra ( 2738457 ) on Sunday April 07, 2013 @09:00PM (#43387599)
    Multiple steps:
    1 - modify baton to include an IR-led (infrared wavelength Light Emitting Diode) so that motion tracking of the baton's moving tip can be done easily without bothering other people in the orchestra or the audience with a visible or flashing or distracting red or green LED
    2 - set up some sort of motion tracking system that can track the IR led and come up with X-position and Y-position and possibly also X-velocity and Y-velocity
    3 - calculate X-velocity as the derivative of the X-position, calculate the X-acceleration as the derivative of the X-velocity with respect to time; do the same for Y-position to calculate Y-velocity and Y-acceleration
    4 - when you hit zero-crossings for X-velocity (e.g. X-velocity goes from positive [right to left perhaps] to negative [left to right], then the X-clicker is activated
    5 - when you hit zero-crossings for X-velocity (e.g. Y-velocity goes from positive [down to up perhaps] to negative [up to down], then the Y-clicker is activated
    6 - hide/place X-clicker in the right-foot, maybe at the heel-pad/ankle region or right under the big-toe, whichever the user likes best
    7 - hide/place Y-clicker in the left-foot
    .
    alternate 6 - X-clicker-A goes under little toe of right foot, X-click-A is activated when the baton goes from (left--right) to (right-to-left), which means it hit the right-extent of travel and reversed; X-click-B goes under the big toe of right foot and it clicks when the baton stops going (from right-to-left) and reverses direction to go (from left-to-right), which is the left-most extent of travel.
    alternate 7 - do like alternate 6 but place one clicker at the back of the heel Y-click-min which clicks when the baton changes from traveling downwards to traveling back upwards, and tape Y-click-MAX along the calf, maybe 6 inches up or so, and Y-click-MAX clicks when the baton stops traveling up and changes direction to go down. This is an intuitive mapping of what the baton is doing.
    :>) If this works, please send me royalty or idea money if you're grateful. JK. No, maybe if you do make money, gimme! (alternate-6 and alternate-7 from brother on phone. Thanks!) Note that the alternate clickings will match what the baton is doing in real-geometric space!
  • Putting a bluetooth accelerometer on the baton would be nice.

    The least invasive solution would involve image processing, but I expect the specialized algorithm be finicky and I worry about frame rates.

    One thing to consider is that people actually have fairly low voluntary reaction times, so a lot of the coordination may come from viewing the physical preparation for the stroke and not the stroke itself.

  • I seem to remember that when Penn and Teller were guests at the Philadelphia Philharmonic, they had a great randomizer as to what song they would play last. The entire orchestra was blindfolded. The song was selected, and shown to the audience. Then Penn stepped up to the box, tapped his conductor's wand three times, and the orchestra playedthe correct final song, without a hitch.

  • take a kinect and write a program to track the baton.

    transmit that in ultra low latency (wired) to piezo thumpers to touch your friend on a spot on his chest or leg or wherever based on the downbeat or whichever per timesignature.

    systems needed:

    kinect software to track motion of the baton
    audio beat matching software to interface with motion tracking
    hardware interface for beat toucher.

    • or ultra low tech, strap a tube with airbag onto the conductor's armpit.
      have the other end of the tube connected to another airbag strapped to your friend's body part.

      He will feel the change in air pressure on the downbeat.

  • Install an IR LED on the baton. Google "LED throwies" for what you need. Most IR LEDs have a fairly narrow beam so sand down the lens to scatter the light more.

    Then get some IR cameras to track them. Any cheap webcam will do. Add an IR filter by having an UNUSED roll of film developed. Install one camera watching from your position, and possibly one to the side, above, or under the conductor if Z-motion is important.

    Then get some motion tracking software like this: https://en.wikipedia.org/wiki/FreeTra [wikipedia.org]

  • Gonna take a couple minor hacks, but the technology is already present.

    Use an air drum stick (or similar, say, wimote?) for the baton. This needs to output to a device that can communicate to other devices.

    Built an app for android cell phones (everybody has one now, Iphone is another option of course) so the output from the baton is translated to vibration at the phone via wifi.

    While certain aspects are going to be lost, timing and emphasis can be retained.

    Phil

  • I almost had an answer but then I realized there was a big flaw, but here goes anyways. 1. As others here suggested, place infra-red LEDs on the tips of the batons. 2. Use a video camera focused on where the batons will be throughout the performance and set the contrast really high. The LEDs will appear very brightly on any video monitor. 3a. (flawed part) Use a blind tongue display (TDU) which enables the singer to "see" the batons, except now the singer is no longer a singer, they are a hummer because t
  • by b4upoo ( 166390 ) on Sunday April 07, 2013 @11:14PM (#43388061)

    In concert bands and orchestras frankly the director is not in control. The tubas and Sousaphones have enough tube length that the player must lead the beat by a fraction of a second. The reality is that the director's baton is actually following the big brass in timing. Since a chorus is often without the deep brass sections you need a device that will send you an audio cue slightly ahead of the choral director. I suspect that a portable or hand held PC like device could be programmed to measure the tempo and report the beat slightly ahead. It also would need to vary the intensity of the cue so that you would no when to play forte or pianissimo.
                        You can get what I am describing by looking at marching bands. The last row in the band will be the Sousaphones. The Drum Major is pretty much invisible to most of the band members so it really is not the drum major controlling either the pace of the march nor the pace of the music. It is the sousaphones and they are pointed right over the heads of the band members. And it is most likely feeling the sound and not hearing the sound that cues in the players. Think about all the noise in a stadium and the cheering etc.. If you are playing a clarinet or other instrument you won't be aware of the Sousaphones much or at all. But they are controlling the pace at all times. The drum major blows his whistle when the piece is finished and when a piece is to be started and puts on quite a show. The low brass knows what pace to set from rehearsals. And the bulk and weight of the Sousaphones comes into play as it is obvious that if the pace were too quick the Sousaphones could not keep up with the marches at all. Very few conductors want to admit that the low brass is running the show. Large drums can also do the same sort of function at times.

  • Instead of having a specially-made or retrofitted baton, how about having the area in front of the podium encased in a magnetic field like that of a theremin, such that the baton would provide data on horizontal position by pitch and vertical position by volume. That would give you all the data you need to feed to whatever peripheral signals the musician (come to think of it, an earbud with the output would help for the beginning of the piece, but it might be too distracting for things like tempo changes.
  • My ears replace my eyes

    Argh. What an image for a Monday morning. Good title for a Harlan Ellison story though.

  • You might want to talk to Shelley Katz of Symphonova [http://www.symphonova.com/]. He's been working on integrating digitally produced music into an orchestra. The system uses an instrumented baton, and it should be possible to output a click track, or something more theremin-like, so that you can hear using an earbud what the conductor is indicating.

  • If you don't want to use a human standing next to the blind singer then it could be solved with some nice cheap modern technology.

    If you could practically do it, attach a MEMS accelerometer (or gyro) to the baton and track the velocity of the tip (or equivalent reference point). Either wired or wirelessly transmit that velocity data to small processing box that drives a haptic device to alert the singer. Unfortunately I'm not entirely familiar with the visual clues of conducting, I've had a look at the pat

  • Well, maybe.
    I've been seeing a children's toy on the shelves recently known as V-drums, basically it's a pair of "drum sticks" that contain simple gyroscopic motion sensors - and produce drum sounds as you swing them around, the programming being made to associate various movements with the locations of your non-existent drums.

    I bet that if the choirmaster uses one of these, instead of a baton, with some open-ended headphones for the singer - they'd end up with different drum sounds for the most crucial mov

How many hardware guys does it take to change a light bulb? "Well the diagnostics say it's fine buddy, so it's a software problem."

Working...