Consider the following setup, where the incoming image has the same resolution as the DMD device.
Image D–1
The image is sent to the projector, pixel by pixel, line by line from top left to bottom right. For 60Hz, it takes 16.6667ms for all the pixels to arrive at the projector (disregarding blanking pixels).
Theoretically, as soon as the first pixel arrives, it can be transmitted directly to the DMD. Since incoming and output pixels follow a one-to-one relation, when the last pixel arrives it is sent to the DMD and at that time the DMD can immediately start to divide the image into bitplanes and display the image. This means that the latency never can go below 16.6667ms.
There is more processing being done to the pixels (color analysis, memory handling etc.), so in real life, latency will never be this low.
Non-native image
Image D–2
Consider this setup where the incoming image is smaller vertically than the DMD resolution. To maintain correct aspect, the image will be displayed by filling in black pixels above and below on the DMD. The DMD device needs to receive all the pixels it shall display, including the black ones.
In this case, the Pulse image processing will start to prefill the DMD with black pixels before the actual pixels arrive on the input. Then it follows a one-to-one pixel copy until the last incoming pixel has arrived. After that, the additional black pixels are transmitted until the DMD has received all the pixels for a full frame. It is not until then that the image can be displayed.
Therefore, the black pixels at the bottom of the image directly adds to latency since the image cannot be display until all the extra black pixels have been transmitted to the DMD.
