Can any one here help me I am tying to make a sox fillter tat will go between a mono mic and the program chatscript.
The idea is to take the raw audio in put from the microphone and make an inverse copy of it. Then using its built in echo function add multiple echoes to both of them, 5 echoes on each track, with values of: 11, 17, 23, 31, 47 and: 13, 19, 29, 37, 53, millisecond. these 2 tracks are then converted into a bindery file by using soxes multiplication option. This treats the tracks like bindery numbers and multiplies them together. If you played this track it would sound nothing like the original input, but it
contains the information about how the sound chained over the previous twentieth of a second.
The idea is to use this distorted file, which is rich in parametric information as the input to a speech recognition program. I have the sox pdfs but I am not clear on the order of the commands.
I’ve used SoX a bit, but I’m no expert, and (as this is the “Audacity” forum) I think you’ll be very lucky if a SoX expert happens to see your post.
You may do better asking on the SoX mailing list: https://lists.sourceforge.net/lists/listinfo/sox-users
thank you, i googled sox forum and was pointed here. i am grateful for the link