If you didn’t hear the term “frame buffer” you didn’t get a correct answer. Applications don’t draw directly to the screen. Passing control of hardware to multiple applications would be disastrous for performance. So all of the screen writes go to a section of memory called the frame buffer, which contains a pixel map of the screen contents, which is rasterized to the display on each screen refresh. A screen recorder simply needs to read the frame buffer at regular intervals, and each capture becomes a frame in the video file.
In modern system, this is implemented as two frame buffers: a back-buffer for compositing the frame and a front-buffer for writing to the screen, and those are swapped on each frame to allow the display to be updated more quickly. This matters only because you have to make sure you’re reading from the active buffer, but the concept is the same.
Latest Answers