We were inspired to do this because of our shared love for music and our newly found interest of image processing.
What it does
You give the program an image and it returns to you a .wav file based on the colors in the picture.
How we built it
We used Java to do most of the image processing because it is what Hunter knew the best. We used C++ to convert the sound to a more natural sounding guitar based on a project Matthew had made before. We used the CLI program SOX to convert from a text file of data to an actual .wav file. Finally, Python was used to connect all the pieces together and create one coherent project.
Challenges we ran into
We were having an issue making the music sound natural so Matthew remembered an old project and we integrated that single component into it. We also had a huge issue with size. We had audio files that would be gigabytes large because too big were used. We had to create a compression algorithm to keep the musical integrity of our product.
Accomplishments that we're proud of
We got the project to work way faster then we thought. It gave us a lot of time to fine tune the controls and make the music better.
What we learned
We learned to do some basic image processing and some complicated debugging from using a multi-step process.
What's next for Image2Sound
We were talking about eventually making a sound to image program that would do the opposite of what we made.