alexm42 Posted at 2016-1-21 14:28
The question at the top of the thread is "Can you really see a difference between a video shot in  ...

I'm saying the same thing I've always been saying: 4k resolution is pointless for most users and they will see no improvement by outputting video in it. Nothing you've said has changed that, and nothing you've said has affected my original position. You're confusing what I'm saying by conflating raw resolution with changing standards and footage aquisition.

*Editing* in 4k is a different matter as it takes into account details like CS; I said that in my original post (there are a ton more changes than 4:4:4 in 4k standards, but most are completely irrelevant). However that doesn't make 4k resolution useful or different. In other words, I answered a question more completely versus just saying yes or no.

Finally, didn't we move off the 1080p versus 4k generalized debate? Weren't we talking specifically about the P3 cameras? Do the P3 cameras do 4:4:4 CS? Because although it's common in the 4k world, it's not exactly a given that something will aquire 4:4:4 (and even less that a display will output it, even at 1080p).


On a completely subjective note, I'm going to venture to say that  at the resolutions people are watching, 4:4:4 CS will not make a difference. Even at 200% crops on that video at full screen, you really have to be pixel peeping to pull out compression artifacts (at 200%!). Extrapolate that back out to the eye resolution chart, and you run into the same 1080p vs 4k resolution issue (here is an article that explains how CS resolution is derived - it only discusses up to 4:2:2, but the details are what's important at 1080p and higher).  This isn't a comparison of moving from one color space standard to the next; this is almost exactly the same as 48 khz versus 96 khz quantization; basically by the time you're comparing such miniscule specs, you should've already improved everything else you possibly could've, as those will make much, much, MUCH bigger differences: better cinemtography, better lenses, better sound acquisition, etc. etc. And while I think 4:4:4 should always be the way CS should work going forward, its practical applicability is tiny and at typical web viewed formats up to about 50" displays at coach viewing distances - negligble, beause your eye has already failed to resolve it (if this was a jump from 4:1:1 or 4:2:0 it would make a huge difference, but that's not the case).

So to answer the question even further:  the original poster should watch that video, and if the CS makes a significant difference, they should edit in 4k (or 2.7k). But they should still output in 1080p (or lower, depending on application). And if they want to do that on a P3 camera, they should verify that it actually does shoot in 4:4:4 CS.

alexm42 Posted at 2016-1-21 14:28
The question at the top of the thread is "Can you really see a difference between a video shot in  ...

I'm saying the same thing I've always been saying: 4k resolution is pointless for most users and they will see no improvement by outputting video in it. Nothing you've said has changed that, and nothing you've said has affected my original position. You're confusing what I'm saying by conflating raw resolution with changing standards and footage aquisition.

*Editing* in 4k is a different matter as it takes into account details like CS; I said that in my original post (there are a ton more changes than 4:4:4 in 4k standards, but most are completely irrelevant). However that doesn't make 4k resolution useful or different. In other words, I answered a question more completely versus just saying yes or no.

Finally, didn't we move off the 1080p versus 4k generalized debate? Weren't we talking specifically about the P3 cameras? Do the P3 cameras do 4:4:4 CS? Because although it's common in the 4k world, it's not exactly a given that something will aquire 4:4:4 (and even less that a display will output it, even at 1080p).


On a completely subjective note, I'm going to venture to say that  at the resolutions people are watching, 4:4:4 CS will not make a difference. Even at 200% crops on that video at full screen, you really have to be pixel peeping to pull out compression artifacts (at 200%!). Extrapolate that back out to the eye resolution chart, and you run into the same 1080p vs 4k resolution issue (here is an article that explains how CS resolution is derived - it only discusses up to 4:2:2, but the details are what's important at 1080p and higher).  This isn't a comparison of moving from one color space standard to the next; this is almost exactly the same as 48 khz versus 96 khz quantization; basically by the time you're comparing such miniscule specs, you should've already improved everything else you possibly could've, as those will make much, much, MUCH bigger differences: better cinemtography, better lenses, better sound acquisition, etc. etc. And while I think 4:4:4 should always be the way CS should work going forward, its practical applicability is tiny and at typical web viewed formats up to about 50" displays at coach viewing distances - negligble, beause your eye has already failed to resolve it (if this was a jump from 4:1:1 or 4:2:0 it would make a huge difference, but that's not the case).

So to answer the question even further:  the original poster should watch that video, and if the CS makes a significant difference, they should edit in 4k (or 2.7k). But they should still output in 1080p (or lower, depending on application). And if they want to do that on a P3 camera, they should verify that it actually does shoot in 4:4:4 CS.