1 person found this helpful
The whole point of limiting the chroma is that you can get away with it. It's not supposed to be noticeable, and without an identical 4:4:4 copy to compare with, you're unlikely to spot it. The only time it's really a problem is when you need to process the image, for example if you are doing green screen compositing.
There's no reliable way to detect color sampling. If the source clip had a lower color sampling rate than the codec it was transcoded to, you may possibly be able to detect the missing values (I haven't actually tried this though). You won't be able to detect if the values have been clipped in the colorspace conversion though.
I think this is one of those things that people worry far too much about. There are situations where it's important but I always favor a lower compression ratio over a higher color sampling rate.
See this article:
> The only time it's really a problem is when you need to process the image, for example if you are doing green screen compositing.
And then it matters enormously.
So how much of this information is lost in final compression? If you have 4:4:4: and compress out to h.264 mp4 for HD boradcast or internet do you still have that much colorspace?
Thanks for the reply,it was very helpful. It does seem that this is the latest obsession with video. And out of all of the obsessions, HD vs full HD, 720p vs 1080i vs 1080p, this one takes the cake because there is no measure.