The method moves the foveated region to the actual gaze position only after a delay equal to the system latency (d).Our technique (bottom row) predicts the ending position of the saccade at its early stage and updates the image according to the new prediction as soon as it is available (b).
Also, foveation encoding may be applied to the image before other types of image compression are applied and therefore can result in a multiplicative reduction.
Standard gaze-contingent rendering (top row) updates the image according to the current gaze prediction.
Due to the system latency, during a saccade, there is a significant mismatch between the rendering and the actual gaze position (b, c).
So Mode 0x13 was a great addition, but it had one downside – it was slow.
Mode 0x13 only allowed one byte of display memory – one pixel – to be modified per write access; even if you did 16-bit writes, they got broken into two 8-bit writes.