Making Audio Captcha using audacity

[AFJ]

Hello everyone, I have two folders. One of them contains A-Z and 0-9 audio clips of 1.5 seconds and in another folder I have different kinds of noises. I want to take 5 characters from first folder (randomly) and stitch them. Then randomly select a noise from the noise folder and add that to the already 5 characters audio that has been stitched. Can I do this using audacity?

[steve]

How many times will you need to do this? Once, twice, 10 times, 50000 times, more?

[AFJ]

I want this 1000 times first, then 5000 times and if needed then 10,000 times. I am making a dataset for training a neural network.

[steve]

If you know Python, then probably your best option would be to automate Audacity with Python. See: https://manual.audacityteam.org/man/scripting.html

(Note that the scripting API is not yet frozen, so there may be changes in later versions of Audacity, though such changes are likely to be small).

[Trebor]

... I want to take 5 characters from first folder (randomly) and stitch them. Then randomly select a noise from the noise folder and add that to the already 5 characters audio that has been stitched ...

Maybe possible with SoX ... https://en.wikipedia.org/wiki/SoX

[stitch = concatenate ]

... if needed then 10,000 times. I am making a dataset for training a neural network.

unCaptcha-ing ?

[AFJ]

Yes, I am trying to break audio captcha for which I need a dataset and I am trying to make my own dataset.