Your response is very helpful. On the current Construct 3 release, I broke the actual snapshot into two steps, each trigger by a button. After requesting the camera source (event #1), I take a photo with User Media when the Go button is pressed (#3). The Stop turns the video into a photo. When the snapshot is ready (#4), I load it into my Me sprite (500 X 500). Or is it me My sprite, OK lost my train of thought.
I have an overlay (called Overlay) that is over Me. Basically a frame, with the photo underneath.
Next when I'm ready to download the photo and frame with the Download button (#5), I figure out the pixels on the screen using HeightFactor and WidthFactor variables. Then I take a snapshot of my canvas using the Overlay to guide my offsets and width/height of the snapshot. When the canvas snapshot is ready (#6), I download using the Browser.