Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Why 4K?

Rob Ruffo said:
...But your eyes use a kind of super sampling to determine gradients. The more points you have to perform this psycho-visual feet, the smoother the gradient will appear, and the more 3D-like the image - this well beyond the abilities shown by the resolution-chart like tests used to determine what resolution we can see.....
Click to expand...

I agree.
NHKs research has pointed them to 7k to create a life like view of the world.
Although in the past a lifelike view has not been necessary to tell a story its use wont be excluded in the future and I wonder if the audience will vote with the feet for the immersive experience, as they seem to be doing when faced with a choice of 2D versus 3d and film versus digital projection.


Mike Brennan
 
My question is. If you current end product is 1080, will the picture quality be better if filmed at 4k and output to 1080, of is it better to film as close to the output resolution?
 
The short answer is that higher resolution acquisition and processing (like 4K) with good quality downscaling as the last step will yield the "cleanest" image in a lower rez deliverable. The longer answer is that creative intent, sense of immediacy and a host of other "conditioned" response patterns may be better served by shooting an image on camera that is very close to the desired finish. In a perfect world I would always want plenty of pixels and color depth in the grading session. You can always rough an image up, but its a neat trick to polish a turd into a diamond, even for the best of colorists :emote_whack:

Cheers - #19
 
The other thing someone pointed out to me about shooting at a higher resolution is it gives you a lot more freedom in post to use a portion of that image cropped and still retain high quality. Maybe something snuck into the frame, or it was a shot you didn't track as well as you liked.

But more so for me, when I take 1080p footage down to say 720p, on every camera I have used, I've always been more satisfied with the down converted video that 1:1 acquisition myself.

What I recognize is probably the case is many cameras of the range I have used do not do a clean down conversion in camera, or skip lines, both of which are inferior to what Red does. The only reason I will shoot lower is if I need a higher frame rate for slo mo shots. Other than that, I will shoot at highest rez.
 
Chris Clogg said:
... but with all these movies filmed at 1080P now (e.g. star wars)... they're stuck permanently unless ..."
Click to expand...


well, the example might not be best picked ...

... the elements, actors, props, part of sets .... are shot in 1080P.
the composite might be MUCH bigger ... ;-)

Making StarWars is not shooting a movie, its more like painting, layer by layer.

Matte paintings, set extensions, 3d backdrops, camera projection maps ... and so forth ... i would guess, the composite has a much higher resolution to start with ...

We are talking about ILM. I´d furthermore guess they would figure a way to re-render the whole show in 4k ... the "4k special edition" - now with extended wider sets for "red 4k digital home theater" ...

:-)

i did not cut, count and analyze the shots in StarWars I-III, (... as i did once with "the matrix"), but remember them saying something like 1700 compositing shots out of maybe 2000 that make the movie.

So those few might stick out as a little odd ... and will be reworked with some "industrial light and magic"

And then, StarWars was a first to use digital to shoot a movie and has made its mark in history and money, already.

In my opinion, since we will have it soon, 5k is the way to go if it is made today for history tomorrow, 3k if it is "short-shelf-life fast food", news, music videos, multi cam, xxx or ultra low budget indies .. blown up by style ... or(!) ... composited.
 
How about 30+ years of future proofing? Marcus Irvin

How about 30+ years of future proofing? Marcus Irvin

Jannard said:
Why shoot 4K? If the world is currently finishing in 1080P and 2K, what is the real advantage to 4K?
Click to expand...

If short term future-proofing isn't enough reason to shoot 4K, how about 40K from your existing RED one and 50K from your new EPIC. This should be enough future-proofing for this lifetime. The key is 4K of good pixels to make this possible. Read this Wired magazine article for an explanation.

http://www.wired.com/magazine/2010/02/ff_algorithm/

I'm still waiting for my full name to be added to my profile. Marcus Irvin.
 
Marcus, the Wired article was journalism, not useful. The pictures they show are "artist representations".

Graeme
 
Graeme Nattress said:
Marcus, the Wired article was journalism, not useful. The pictures they show are "artist representations".

Graeme
Click to expand...

Of course they are "artist representations". As a still shooter for 30 years, I was most interested in the science, not the illustrations.

Groundbreaking technologies(including RED), are always preceded by "journalism" and "renderings" before they actually have work samples.

This excerpt is what got me most excited, "
the beginning of compressed sensing, or CS, the paradigm-busting field in mathematics that’s reshaping the way people work with large data sets. Only six years old, CS has already inspired more than a thousand papers and pulled in millions of dollars in federal grants.
Click to expand...
"

As I'm sure you know better than I, hundreds of scientists don't often bet their careers writing papers on nonsense. If this only came from one engineer at Stanford, it would not be near as believable.

Your focus on great pixels, will benefit the most from this tech when it does arrive. IMO, RED is poised to make history in several ways and this won't hurt you guys at all.
 
Call me not convinced, but googling at the published papers on compressed sensing, I'm not seeing any reconstructed images that look any where nearly as good as the fully sampled originals...

Graeme
 
Haven't people been doing this same thing with wavelet transforms? When I was in college in the early 90's my thesis advisor was from the CS department (computer science) and he was advising someone else that was using wavelet concepts to up-rez images.

A few years later I saw some software out offering it from some unknown company for cheap, but honestly, I would have thought if the tech was any good it would have made itself into something like Photoshop by now.
 
Closest practical software app I can think of that does something similar is the fractals based photo printing software to up-rez digital stills for very large prints without losing detail.
 
Oh yeah, forgot about the fractal stuff too :) It was also popular around the same time. Played with it a little when I was doing a lot of 3D rendering with 3D studio max. I was trying to take some textures that were a little too low rez for what we were doing to see if I could at least give them a little more detail. For some things, they worked, but it was a hit or miss at the time.
 
From what I see, ProRes is based on a discrete cosine transform (DCT) compression algorithm. Cineform is based on wavelets it appears.

Although, what I am talking about is not compression, it's about generating detail from non existing detail using wavelet technology, which is what the wired story was about.

Edit: Post disappeared this is in response to, someone was asking if ProRes was based on wavelets.
 
Last edited:
David Rasberry said:
Closest practical software app I can think of that does something similar is the fractals based photo printing software to up-rez digital stills for very large prints without losing detail.
Click to expand...
Interestingly, the Genuine Fractals software recently got a mention on CSI NY in one of the lab segments, so it must still be around. Used it myself back about Windows 2000, I think. Worked really well in detailed enlargements of a 1MP photo to a large format Epson canvas print.
 
Rick Burnett said:
From what I see, ProRes is based on a discrete cosine transform (DCT) compression algorithm. Cineform is based on wavelets it appears.

Although, what I am talking about is not compression, it's about generating detail from non existing detail using wavelet technology, which is what the wired story was about.

Edit: Post disappeared this is in response to, someone was asking if ProRes was based on wavelets.
Click to expand...

Yea, when I read backward some more I pulled the post... And I had mis-spoken. It was the Pixlet codec I was thinking of.
 
ok so I've got to ask.

You guys make it seem like 4k is such a big deal and everyone who's not at 4k is going to be missing out. That being said, I want to get the 2/3" cinema. Does that mean I should work on all my projects at full 3k inside something like premiere pro and then when I got to finish it I should uprez it to 4k?
 
You must log in or register to reply here.
Back
Top