jmac698
20th August 2011, 22:27
I just thought of an interesting experiment. In NTSC at least, dot crawl changes on each frame, so if you average two frames together and it was a still frame, there is no more dot crawl. The result is as good as separated chroma (s-video).
The problem is, how to get the same image for two frames, and different chroma phases. You can play a video at half normal speed, then average them in script. You could also record a TV and it's re-run. If you have a good luck, the 2nd recording will be in a different color phase.
It was amazing to me to realize that two recordings of the same video could differ by color carrier phase.
In analog systems, like VHS and LaserDisc, the chroma phase is always the same when played back. On digital systems, I assume the phase is set from the moment the transmission encoder machine is turned on. There is a random chance that any material played twice will be broadcast in a different phase.
Not only will averaging eliminate dot crawl, but reduce added noise since transmission as well. I would love to test this theory.
Meanwhile, is there anyone familiar with how TV stations originate signals - can you confirm that it is a function of the final transmission encoder (what is this box called?), or at least from the playback device (the satellite decoder box, for example)...? For cable tv, perhaps the encoder?
The problem is, how to get the same image for two frames, and different chroma phases. You can play a video at half normal speed, then average them in script. You could also record a TV and it's re-run. If you have a good luck, the 2nd recording will be in a different color phase.
It was amazing to me to realize that two recordings of the same video could differ by color carrier phase.
In analog systems, like VHS and LaserDisc, the chroma phase is always the same when played back. On digital systems, I assume the phase is set from the moment the transmission encoder machine is turned on. There is a random chance that any material played twice will be broadcast in a different phase.
Not only will averaging eliminate dot crawl, but reduce added noise since transmission as well. I would love to test this theory.
Meanwhile, is there anyone familiar with how TV stations originate signals - can you confirm that it is a function of the final transmission encoder (what is this box called?), or at least from the playback device (the satellite decoder box, for example)...? For cable tv, perhaps the encoder?