Design sounds

[Brian Higgins]

Any of this makes ?

Design Implications for Your Project

• Pitch mapping: Use logarithmic mapping from distance to frequency, since human pitch perception is logarithmic.

• Tone quality: A pure square-wave buzzer may fatigue users—consider adding simple patterns (pulse trains, vibrato, duty-cycle modulation).

• Frequency range: Stay in ~300–3000 Hz, the sweet spot for human hearing.

• Redundancy: Combine pitch + pulse rate for a robust, intuitive signal.

• Avoid tiny changes: The “just noticeable difference” (JND) for pitch is ~0.5–1% in midrange. Changes smaller than that won’t be heard.

Brian Higgins

VA Researcher for blind mobility “Laser enhanced Echolocation” ClearPath Navigation 

Sent from my iPhone

[Chris Albertson]

Yes, makes perfect sense.  Think about the piano keyboard.  Each “C” is an active part and is TWICE the frequency of the previous C.  It is non-linear.  Someone figured this out 10,000 years ago.

And yes, those electronic beep sounds on devices like microwave ovens and oven timers are horrible.    I’d use sound samples from musical instruments, like a flute sound or whatever.   

Not everyone is good at hearing pitch.  It is like eyesight; there is a wide range of ability.  Make your device adjustable so that those with excellent pitch hearing can take advantage of it.   But those who can’t hear a sharp or a flat can get a stronger cue.

--
You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/hbrobotics/973B7D19-564F-4123-BFEC-FBA76BE4A76C%40gmail.com.

[Craig Austin]

Brian,

I had a little fun with this and wrote a Python script (actually Google's Gemini wrote the script under my direction) to demonstrate all the elements called out in your Design Implications. The attached script is written for a windows machine and will need Python with tkinter, numpy, and sounddevice installed to run it. If you can manage that, I think it will do a lot to help understand those implications and make it easier to select good parameters for your cane. You'll probably need to change the piezo speaker in your cane to a regular speaker to implement the variable frequencies on it. But you can try these parameters out on a windows machine before deciding the parameter limits... And, before I get yelled at by everyone, yes I should be using GIT, I've just got more to learn before doing so and didn't want to delay this any longer.

--