Skip to content

A python project that shows static noise, but reveals a hidden message when the right sound is played.

Notifications You must be signed in to change notification settings

b-destoop/hidden_noise_message

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

25 Commits
 
 
 
 
 
 
 
 

Repository files navigation

denoising

principle

The app reveals a hidden message when the right sound is played in front of it, but only for the duration of the sound.

This is done by showing static noise (like on old tv's) when no sound is registered. When a sound is picked up by the system's microphone, the noise shifts into the hidden message, but only by the amount of similarity of the sound that is picked up and the secret sound that is to be played.

optional improvement : have the code show when no noise is made!

noise visibility tactic

Create noise using a gaussian curve. For the values which coincide with the code image, this gaussian curve should have a smaller standard deviation.

note : the code that is to be shown should be white on a black background.

sound

Sound is captured using the default microphone of the system.

The volume is calculated using volume = 20 * log10(avg(sample)). This parameter is very volatile, therefore the following helper parameters are also calculated:

  • max volume: the highest value of volume that was measured with the currently running application
  • min volume: the lowest value of volume that was measured with the currently running application
  • relative volume: a number indicating where the current volume sits between max volume and min volume. 1 means the current value of volume is the same as max volume. The same applies for 0 being represent of min volume
  • average relative volume: a time averaged version of relative volume. This gives a slower changing parameter. This parameter is to be used for the visibility application.

todo

  • create a window with static noise
  • receive sound
  • find comparison metric between sound and preferred sound (fourier? convolution? machine learning?)
  • load in image with a message
  • distort image according to the comparison metric
  • create an extra window with tuning specifics
    • load in image
    • choose a microphone source
    • sensitivity dialling
    • goal sample loading
  • create transfer function for the similarity to the amount of revealing of the hidden message.

About

A python project that shows static noise, but reveals a hidden message when the right sound is played.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages