1080p or 4K resized down to 1080p?
Did your mother drop you on your head as a baby?
>>54319984
rude
>>54319970
Native 1080p looks better!
4k at 4k
>>54319970
Probably the latter because 4k supersampled to 1080p would be less artifacted.
>>54319970
If you're asking this for a TV, resolution is the least important aspect. I would focus on whether it's LCD/LED/OLED/Plasma, and its refresh rate. 4k is immaterial unless you're sitting closer than 6 feet (or more than 40 feet) away, which most people aren't. I'd rather have a 1080p OLED 120hz TV than an 4k LED 60hz TV.
To answer your question, though, it really doesn't matter.
>>54319970
It should be around the same as 4k is an even multiple of 1080p
>>54319970
It's the same because most movies are shot at 4k (or if filmed on 35mm/70mm it's scanned/tc'd at 4k or 8k) and then sampled down to 1080p for home video. Only difference here: When the production team samples down a film from 4k or 8k they know what they are doing and use high quality filters like bicubic sharpener or maybe Lanczos (or others, who knows). In any event, it's going to look a lot better than the image your on-the-fly scaler produces.
>>54322406
>it's the same
>(something something)
>which makes it actually not the same
Fuck I'm dumb. What I meant to say is - it depends on the content and for movies, the native 1080p version is going to look better (at 1080p!)
interesting thread
/g/ settle something that my housemate and I were arguing about
if viewing a 480i/480p dvd on a ps4 or similar bdp with upscaling capabilities on a 1080p hdtv is it better to display the original 480i/480p signal or upscale it
personally I prefer the original
>>54319970
4k is always better.
>Say you have, in 1080p, a single pixel of noise.
And in a 4k video, you also have a single pixel of noise
>The pixel in the 4k, is 1/4th as significant, and as such, doesn't have as great an effect.
>So when you downsize (average), those 4 pixels become one, and that pixel is 1/4th noise, instead of entirely noise
Also youtube specifically makes 4k source better at 1080p
>For youtube specifically, if you select the 4K video stream and play it on a 1080p monitor, you're getting a higher bitrate than the 1080p stream from youtube.
>>54322738
It has to be upscaled (and possibly deinterlaced) either way. The difference is between either the ps4/bdp or the TV. I would assume that neither is as good as mpc-hc + madVR.
>>54322779
>>So when you downsize (average), those 4 pixels become one, and that pixel is 1/4th noise, instead of entirely noise
That assumes that the source of the 1080p is not 4k (see >>54322406). If the source is 1080p, then there is nothing to downscale/compare it to. If the source is 4k, then the 1080p is already what you described - a version where 1 noise pixel is now 1/4th of a noise pixel. Your argument really doesn't make any practical sense at all (unless you are talking about comparing a 4k bluray encode to a 1080p encode where the 1080p encode came from the 1080p remux instead of the 4k one, in which case that comparison is unfair).
As for youtube, the reason it looks better is, as you said, not the resolution but the bitrate, making that point irrelevant too.
4k is better. Although the most huge improvement comes for actual cinematography. If you take 1080p source and do a digital 4x zoom in post, now it looks like ass.
If you shot in 4k and need a digital 4x zoom for that section in post, since 4k has four times the resolution of 1080p, that 4x digital zoom has only reduced the total resolution to 1080p exactly instead of to a quarter of 1080p (in the case of an original 1080p source)
>>54322830
>really doesn't make any practical sense at all
It does, in terms of content creation.
>>54320158
This.
on a LCD native always looks better, downsampled just becomes a blurry mess.
its easier and less expensive to just use AA be it full screen, temporal or whatever.
>>54322830
>obviously all source is shot in 4k first
No, no it isn't, you fucking idiot.
Lots of youtube vids were shot with cameras that had a max res of 720p or 1080p and never touched 4k at all.
Obviously, your idiot idea of his question (does 4k downsampled to 1080 or another 4k camera downsampled to 1080 look better?) doesn't make any sense to ask.
>>54322920
>Lots of youtube vids were shot with cameras that had a max res of 720p or 1080p and never touched 4k at all.
But then you can't really ask the question because it depends on the camera itself. 4k video usually means bitrate starvation, which makes the whole pixel/noise argument irrelevant because the 4k source looks worse than the 1080p source.
>>54319970
To begin with, 1080p and 4K are almost the same. You need a good-ass monitor or tv to tell the difference, so it's a meme resolution.
Even 720p is fucking good.
>don't fall for the K meme
>>54323250
:^)
>>54323266
>using the smiley with a carat nose
All other things being equal, 1080p from a 4K camera should look better than 1080p from a 1080p camera.
This is because each RGB pixel is created from just one sensor pixel.
The remaining two colors are interpolated from surrounding pixels.
But if you down-sample from a 4K sensor to a 1080p file, you can use 1 red, 1 blue and 2 green pixels for each RGB pixel with no need to interpolate.
