My colleague and I set out to further explore the sketches and possible movements. The sketches offers multiple body parts to be tracked which opens up the field for use to use any body part we desire to use. There is even possibility to track two bodies in one sketch but we won’t do that given the social distancing but also the fact that we work on two separate locations.
Something that lies within my core interest when it comes to designing is efficiency but I also have a great admiration towards precisions and accuracy. How can precision and accuracy be represented in a movement? That would be interesting to find out, so still on the basic sketches I coded a quick console.log command to show the value of where my wrists are (the x and y-coordinates). What we found was that it was it was a challenge to get a exact value. We tried different distance and angle to the (built-in) webcam over three different computers. Depending on the camera the value was always moving even though the wrists was resting on a table being completely still. With that in mind creating an interaction where a huge demand is on precisions and accuracy was out of the question at least for the accuracy I had in mind.
We tried if the drunk rudolph sketch was any more stable but it exhibited the same problems of inaccuracy and moving graphics even though the body part is completely still. At that point we guess it has to do with the library that tracks the motion.
A complete neglect of precise movements is not what we want to do, so we are going to ask during the coaching session if there is any workarounds to the issue or a different approach take to accuracy and precision as its too important to just neglect. If we did that we would only have gross motor skills to explore which is not necessarily the path we want to take.
Micro blogg 