Drawing sounds

Would it be possible to create Draw sound feature? When user enters the draw mode and with mouse draw the curve which can be used to generate sound.

For example drawing this, processor can transform into a sound

Using it as a waveform or a spectrogram ?.

There is a [u]Draw Tool[/u] but it’s really indented for “little repairs”.

The real problem (besides accuracy!) is the amount of information required… For example, if you draw 10 cycles at 1kHz that’s 1/10th of a second of sound, and if you play it that’s not enough time to perceive pitch and it will sound like a “click”. And, it’s just not practical to draw thousands of cycles to represent anything useful or interesting.

Would it be possible even if you draw that high amount of information, Audacity would discretize it?

That spectrogram could be very useful in Audacity.

It’s not as fertile a technique as first appears:
#1. You can only draw in one direction, so most of the English alphabet is not possible as a waveform.
e.g. “M” “N” “V” “U” are possible, but “A” “X” “E” “K” aren’t.
#2. completely different looking waveforms can sound the same.

It’s a very niche interest, & there’s already free software which does that,
e.g. … https://www.youtube.com/results?search_query=AudioPaint

You can create a waveform with the Draw tool like this:

Second Track000.png
Then select the drawn waveform like this:

Second Track001.png
Then repeat it lots of times with the “Repeat” effect, like this:

Second Track003.png
To get a sound that is long enough to play, you will need to repeat the sound hundreds of times.
The resulting sound will probably just sound like “beeep” or “buzzzz” (not very interesting sounds). The interesting character of musical sounds mostly comes from the way the sound changes over time. In this case, the sound is not changing over time, so is not very interesting, though these steps do (literally) what you requested.

Audacity would discretize it?

Yes. Digital audio is always digitized/quantized in amplitude and time. With CD audio there are 44,100 samples per second where each sample represents the positive or negative amplitude (wave 'height") for one instant in time. See [u]Digital Audio Fundamentals[/u].

When you play the sound, the digital-to-analog converter in your soundcard “connects the dots” and smooths/filters to create (or re-create) a continuous waveform.