English English

Applies to:
All 1T-C2-XXX units that support scaling, plus C2-1000, C2-2000, C2-2000A, C2-3000, C2-4000, C2-5000, C2-6000, C2-7000 series

1. If input and output are progressive (not interlaced) and at similar frame rates:

From the above diagrams, showing Input and Output frames in time, it can be seen that the ‘phase’ relationship between input and output will affect how long it takes for an input frame to propagate to the output.

Since all CORIO2 units feature double-buffering, the scaling logic will ensure that a frame will only be read from memory once it has been completely written into it. Also, the ‘read’ logic will only move to the next frame in memory at each vertical sync (i.e. at the boundary between frames).

Example 1: the Output (read) frame boundary occurs just after a new frame has been written into memory, therefore it can immediately start reading it out and it will have just over 1 frame’s delay from input to output.

Example 2: the Output frame boundary occurs much later (just before the 2nd frame has finished writing into memory), therefore the frame delay is nearly 2 frames from input to output.

Thus the average delay is 1.5 frames.

2. What happens when frame rates are not the same?

A similar thing happens, with the frame delay being between 1 and 2 frames – measured in terms of the input frame rate (e.g. 16ms = 60Hz). However, if the output frame rate is lower, then it will take longer to display (since it updates less often) and this may add to the frame delay. In other words, it will take 1 to 2 frames before the input frame starts being output, and then (since the frame takes longer to output) it will be on the screen for longer.

This means that if the output frame rate is quicker than the first, it may give the impression that there is less delay from input to output  (since the time between the middle of the image on the input and the output has decreased). If the output frame rate is slower, then the delay could be perceived to increase – since the time between the middle of the input frame and the middle of the output frame has increased.

3. What happens if temporal interpolation is turned on? (C2-2000A & 1T-C2-760 units)

This works by merging frames together in order to smooth output any difference between input and output frame rates. Thus 2 frames from memory are read together and an ‘in-between’ one is created. The amount by which these 2 frames are merged depends on the phase of the input and output frames at that point in time. This means that the overall frame delay may be between 1 and 3 frames, with an average of 2 frames.

4. What happens if the inputs or outputs are interlaced?

Things get very complicated, depending on de-interlacing mechanism used. At best, no further delay is added. At worst, for NTSC sources using 3:2 conversion, this can add 2 fields of delay in processing the image. Interlaced outputs tend not to affect frame delay – but since it takes 2 fields to re-create the whole frame, the display may add its own frame processing delay.