Abstract: A method and apparatus for keying digital color TV signals to avoid "stairstepping" of pixels at the edges of the key while retaining a hard switch between the keyed scenes. This is achieved by comparing successive digital keying signals with a desired digital clip level. In the event that the clip level occurs between two successive digital signals an interpolation of the clip level between these is created fractionally. A digital representation of this fraction is then included in the keying signal so that keying is accomplished on a sub-pixel basis without acutally increasing the frequency of sampling.

Abstract: A method and apparatus for keying digital color TV signals without exceeding the normal sampling frequency (13.5 MHz) and without obtaining a degraded image. A digital keying signal and a digital color TV signal are supplied to upsamplers (10,12), respectively, which double the sampling frequency of the incoming signal by linear interpolation. The 27 MHz sampled outputs are keyed in multiplier (14). In one line to a switch (24) the 27 MHz, 12 MHz bandwidth keyed signal is passed through low pass filter (16) and demultiplexer (18). The latter removes each alternate sample to provide a 13.5 MHz signal of 6 MHz bandwidth. In a second line to the switch the keyed signal is demultiplexed to provide a signal which would have resulted without upsampling, which is then delayed (22). Edge detector (26) senses high frequency keying signals and controls switch (24). In the absence of such high frequency keying signals the signal to the second line to the switch is taken.

Abstract: A method and apparatus for keying TV signals. A keying signal, K, is employed to create first and second modified keying signals, K(A) and K(B), which commence at independent points in time. The modified signals are compared and, as a result, a revised keying signal, R, is produced commencing at the leading edge of one modified keying signal and terminating at the trailing edge of the other. The operator selects R according to one of the two rules (i), (ii), the selection being by trial and error:rule (i):if K(A)>K(B) then R=K(A)if K(A)<K(B) then R=K(B)rule (ii) (which is the converse of rule (i)):if K(A)>K(B) then R=K(B)if K(A)<K(B) then R=K(A).In this manner, the revised keying signal R can be made to coincide with the video signal, V, to be keyed.