Your high-converting reaction video composite and performance marketing summary brief are generated below.
The reaction studio stores percentages for picture-in-picture placement and converts them into canvas coordinates. In its default vertical canvas, width is one thousand eighty and height one thousand nine hundred twenty. A thirty five percent floating bubble is three hundred seventy eight pixels wide and tall. At x seventy percent it begins at seventy percent of remaining width seven hundred two, or four hundred ninety one point four pixels. At y seventy five percent it begins at one thousand one hundred fifty six point five. These two bars compare bubble width three hundred seventy eight with canvas width one thousand eighty at point three pixels per canvas unit, measuring one hundred thirteen point four and three hundred twenty four. The percentage locates the box within remaining space, not its center. Corner overlay keeps width equal to canvas width times scale, but sets height to width times sixteen ninths. For the same three hundred seventy eight width, height becomes six hundred seventy two. Floating bubble instead keeps both dimensions equal. The two bars share point five pixels per dimension unit, showing bubble height one hundred eighty nine and corner height three hundred thirty six. Split frame ignores the scale slider for box dimensions and uses half canvas width with full canvas height. Top-bottom uses full width and half height. Layout choice therefore changes which state fields have geometric effects. These are source compositor rules rather than proof of any real camera capture. The source records canvas captureStream at a requested thirty frames per second. It does not add an audio track to that stream. The main volume, microphone volume and ducking settings are stored in a campaign brief, but this code does not implement an audio mixer connecting them to the recording. The two bars depict requested rates thirty and fifteen at ten pixels per frame-per-second unit, three hundred and one hundred fifty, as an illustration of rate settings rather than measured delivery. If no recorded chunks exist, the fallback labels serialized state bytes as video Web M, which does not make them decodable video. Inspect actual media and its audio tracks before calling an export complete. No live microphone, sound mix or real presenter is assumed.