Hacker News Reader

Comment by chrisseaton | Hacker News Reader

I'm not sure it makes any sense to talk about a video being in 4K resolution without talking about the resolution of the data that went into the production of the video. I could make a 4K video of the moon right now (or when it gets dark) by sticking my iPhone out of my window. Presumably it's the source data that is high resolution in this video, not just the video itself?

View on HN

Given the zooming that takes places and the fact that the entire moon was mapped (not just the places shown in this video, except perhaps a handful of ultra hi-res pieces), it's clear that the resolution of the mapping far exceeded the resolution visible in the 4k video for all but the closest shots.

Essentially they have much better resolution source material and they scaled that into a 4K signal which allows for a viewing experience that is "better" than having it be scaled to 1080P.

As for showing it on the web? Well generally unless you have a 4K monitor you won't see it at the full video resolution. It does look pretty crisp and stunning at 4K.

> Well generally unless you have a 4K monitor

Ah this is a common misconception actually! I think that the chroma resolution in video is often less than the pixel resolution, so a pixel resolution higher than your screen resolution can still be useful to get the extra colour data.

I have noticed this with quite many video players. Downscaling from 4k to fullscreen on my not-so-4k monitor seems to provide better details than the native 1080. On YouTube for example. Do you know if this is in effect for YouTube videos?

I think it applies to almost all conventional video compression algorithms.

https://en.wikipedia.org/wiki/Chroma_subsampling#4:2:0

Big factor here is the bitrate of the video; on practically all streaming platforms 1080p video is very heavily compressed and while speculating, on sites like YT there is also encoding speed vs visual quality optimization going on.

With equal bitrates and high-quality encoders the difference should be smaller, or even turn towards preferring 1080p depending on the bitrate.

I feel like I'm missing something here. A display that has enough pixels to show the pixel resolution of the source should be able to resolve all the (subsampled) chroma information as well, should it not? (Er, at least as long as the subsampling is done at an integer divisor of the original!)

Anyway, a high-resolution source could look better than a low-resolution source when displayed on a display with the same resolution as the low-resolution source for a number of reasons, perhaps the most obvious being that the high-resolution source might be of higher bandwidth.

> I feel like I'm missing something here.

If you have a 2k screen, you may think there's no point watching a 4k video instead of a 2k video. Well there is even if you ignore bitrate - because the 4k video when resized to 2k will contain more colour information than the 2k video shown at 2k.

> A display that has enough pixels to show the pixel resolution of the source should be able to resolve all the (subsampled) chroma information as well, should it not?

Yes, but the point is that a display that does not have enough pixels to show the resolution of the source, will be able to show colour per-pixel in the 4k video, rather than colour per-every-other-pixel as it would in the 2k video.

(Chroma subsampling is not this simple in practice.)

Oh, I see now. Thanks!

True, although for this particular video I was more impressed by the high frequency edges of the mostly monochromatic moon surface. The edges of craters are very sharp on a 4K screen.

You can expect that the NASA did true 4K without blur and at a good zoom level...

You could make a 4K image with your iPhone, but it will not nearly be anything like what this video shows. Please do look at the video. I thought it was pretty amazing the level of detail available and I don't even have a proper 4K screen (2560x1600), it is much better than any other video that attempted to do something like this and it is more about the LRO mission and the data it generated than about the output resolution of the video, with the note that generating output at a higher resolution would have been pointless so they did the best they could.

I agree. Also, instead of sharing a scripted video of moon, which we already have many, I would be more interested in high resolution models that I can explore as I want.

One starting point would be http://moontrek.jpl.nasa.gov, go to "Layers" (lower right), hit "Added layer" tab, add "Freeair Gravity, colorized" (type Gravity into the dialog box and choose that one from the several alternatives), change the alpha from 100% to about 50%, and browse as you like.

I believe this should get you to Mare Orientale:

https://moontrek.jpl.nasa.gov/#v=0.1&x=-94.67723808827287&y=...

This is still a browser, but it shows the gravity field similar to the video (not as good), and it's closer to something a scientist might use.

Even if you would film it in 480p, I could re-encode it in 4k... Blu-rays used to do this, not sure if they still do (I don't get physical media anymore)...

In this case, the zoom might matter more then anything else...

It makes perfect sense to me, I read it as "Very high quality video of the moon". Just like when I can download a movie in SD or HD I know which one will almost surely look nicer even though it's possible the HD version is the same content with black bars around the sides.

> It makes perfect sense to me, I read it as "Very high quality video of the moon"

But my point is I could also film a 4K video of the moon out of my window using an iPhone, and it wouldn't actually be very high quality. Being 4K doesn't actually tell you anything about the quality of the content whatsoever.

Nor does HD, but that is also a common and useful way to describe high definition content.

Your phone has a wide FOV which would put the moon as a small circle, probably overexposed.

That's the GPs point - the term '4K' doesn't really indicate quality

It doesn’t denote quality but quality is the connotation.

In my interpretation, it indicates that the picture contains at least 4000 pixels belonging to the moon.

Pedantically this could be a "0.085k" image then.

Any reasonable reader will assume that the Moon has to fill a significant portion of the frame and be well exposed as well as sharp for this headline to be meaningful.

I agree it’s not precise and I also agree it’s not a generally very useful way to talk about resolution, but I think the authors were looking for a way to convey high resolution and quality while using terms anyone might have heard and have a rough understanding of.

Reply on news.ycombinator.com