#484 Explanation of Frame Types

[Nick Prosser]

The frame counter in VirtualDub2 indicates the type of frame being displayed in square brackets: the ones I see most commonly are I, P, K, D, and +

From my own knowledge/experience (and from diving through what I can of the forum here), I understand what K, I and P indicate. My colleagues and I have always assumed that D and + frames were duplicate frames, but I've been unable to find anything that backs this up.

I was wondering if anyone here had a definitive definition of each of these frame types ( and if there's any official documentation that supports this, that would be even better). And if we are right about D and + being duplicate frames, can anyone describe the difference between them?

For reference, I understand I frames to be Intra-encoded frames, where all regions of the image are fully encoded. P frames are predicted frames, which rely on previous frames as reference and only contain data that has changed since the previous reference, and K frames are IDR frames, which are key frames that specifically close the previous GOP. Please correct me if I'm wrong.