Overall I would call my project so far a success, certain note-colour relationships work better than others for different image types but overall the pieces each had their own character and could be considered musical. Early tests left me disappointed due to the lack of cohesive theme in any of the pieces the program created, as the application matured using the test pieces a number of ideas were tried out which led to the ability for the tempo to change throughout the piece. This really helped to add a human element to the output and with the addition of carefully selected instrumentation more notes were able to shine through in the final mix, instruments with long sustain tended to cause dissonance as previous colours overlapped the current playing colours.
An interesting surprise was the melodic nature of the piece when Scriabin’s circle of fifths was used on photographs containing lots of similar colours such as in “Portsmouth at Night”. This leads me to believe that trying out other intervals could lead to interesting tonalities that may work for other image types, for example going up in major or minor thirds for each colour.
In the future I would like to extend the application to alter mix settings such as effects processing, panning and volume as this could add yet another dimension to the final output of an image. Mixing a song is a dynamic process and subtle changes such as widening the stereo field during quieter and darker sections could help sell the mood of the image in musical terms. Additionally, I would like to implement the ability to run a live webcam while the program is playing the image. While a static image would have one definitive output a webcam image could be different every time. Placing this project in an art installation environment would add an interactive element to the system, as visitors walked past the hue of clothing would have an immediate effect on the sound of the image.