Fixing one’s legacy

With Maschine 2.0 I had to define new classes of assets for hardware as well as software representation of content. This had been done in a way that was not really so forward thinking. After many years it was generating too much manual work and made the content production more complex than it should have been.

Reverse engineering designing

As I saw the opportunity of a refresh of the pipeline with Kontakt 8 and new hardware, I started to reflect on how the rules that I had specified could be automated. After a few attempts I decided that :

• full automation - detecting the logos “presence” beyond bounding box for exemple - would be huge work and ultimately feel limiting for users (third party publishers and inhouse designers).
• the app goal was to give the publisher the guarantee that the output was technically clear
• the result should be directly ingestable by the inhouse pipeline therefore removing any QA work on those.

Pragmatic execution

The execution was done in Juce with my former colleague and jack of all trades Mickael le Goff @SYNESTHESIS. Was that the best choice? in retrospect I think we should have used webtech instead, but Juce seemed like the lowest hanging fruit for us in our context at the time.