Friday, October 28, 2016

Analog to digital, How Sound Is Recorded.

An Example of what a waveform looks like


Sound is an integral part of our lives, whether we realize it or not, and none of the recorded sounds we hear would be possible without two devices: an ADC(Analog to digital converter) and a DAC(digital to analog converter). And ADC is responsible for taking the vibrations of sound around us and storing them in digital form. A DAC does the opposite and converts the 0s and 1s of the stored sound’s data into the vibrations that our ears translate to sound. Before you can understand how either of these devices work, you need to understand how sounds are made. When a sound is generated, air molecules compress (get closer together) and rarefact (get farther away from each other). The series of compressions and rarefactions is what makes up the waveform and what our ears pick up and translate to sound. An ADC works by taking a measurement of the voltage of a sound a certain point in the sound This measurement is called a sample. The ADC takes tens of thousands of samples per second and  the higher this sampling rate is, the more detailed and accurate the recorded information is. An example of a sampling rate is 44.1k Hz. This means that the computer takes 44,100 samples per second. For each sample, the ADC computer measure the voltage at that specific point and stores that voltage as a binary number. The millions of binary measurements which make up a 3 minute recorded sound (just an example) is given to the main hard drive to save. A DAC is used when we want to listen to the sound we recorded. It does the opposite of a ADC and reads these binary voltage samples and tells whatever speaker you use what frequency and intensity to move its drivers. The drivers move back and forth and create the compressions and rarefactions our ears can detect.


Sources:
http://cloudinary.com/blog/how_to_generate_waveform_images_from_audio_files
http://www.hardwaresecrets.com/how-analog-to-digital-converter-adc-works/2/

1 comment:

  1. Cool post, Andy. I haven't much thought about how a computer must convert sound into binary, but if that sound is stored as data, of course it must. This made me think of my own blog post, which includes something about how computers must also convert color into binary. I wonder which requires a longer string of binary--an infinitesimal bit of color or an infinitesimal blip of sound? I haven't got much of a clue, but my money's on the sound.

    ReplyDelete