RED as a FILM LOOK

[tomridlle]

thank you Axel,really taught me a lot
can we add some fillter in the compression for mpeg？
the grain in the 3d production is also a big problem,if we don't add some grain the animation looks fake， especially composite it with the live action

 

[Axel Mertes]

I think that it all depends upon the story.

The story dictates the visuals.

For that super-clean grainless look to work for the story, it has to be a situated in the present, or in the last... 20 years.

It might be used in a futuristic story or in a story about a totally imaginary world.

But before say... 1984 .... it wouldn't look right, IMHO.

For example, would you use this kind of visuals in a WWII movie... in the trenches... super clean trenches footage?... wouldn't work for me...

And would you give me a super clean and pure, grainless close-up on Julius Caesar, while giving a speech to his seasoned soldiers.

There's 5000 years of human history that kinda suggests grain ... and only a mere 20-30 years that would allow for a super clean picture to be a good thing for the story.

For me, grain is here to stay.

Anyway, juz my 2c.

Different people, different ideals.

Sya, gotta watch Munich now...

Spielberg is a grainy guy.

Hi Karapetkov,

I know what you mean. Right now poeple use the grain to position the material in that time frames, totally right. I am not sure how I'd react when seeing a sandals movie in total clean appearance. Maybe its not that bad, but unusual?

With the extremely noise free recording of future camera systems, we may end up with a new look. Only then we can do the comparison which does affect us more, the version with or without grain. And then grain seems to become an purely artificial thing, as we have no real grain anymore.

Imagine:

For compression quality and efficiency we may produce perfect clean footage and some metadata is kept within the data stream to add the according "film grain" we want, live during playback. Patend pending...

So you can look 300 on 1950's film stock, or 2008's filmstock, at 500 ASA or 250 or 1600...

So we have a new control on the monitor: Brightness, Contrast, Grain...

Who knows...
Axel

 

[Axel Mertes]

thank you Axel,really taught me a lot
can we add some fillter in the compression for mpeg？
the grain in the 3d production is also a big problem,if we don't add some grain the animation looks fake， especially composite it with the live action

Composites are always an issue, when combining 3d animation to live action you need to add similar grain to make it feel looking "right".

Sometimes people degrain the live footage prior to combining the layers, add grain later.

Or you add grain just to the layer you created artificially.

You must be very careful and have an eye for this, otherwise you will run into problems. There are methods for extracting grain / noise from images to get a mask layer that can be used to add the same grain / noise to "noisefree" layers.

You may e.g. shoot a clean plate, so a stable image from the scene, just some seconds. Then you make a binary XOR extraction of the difference between each subsequent images (delta data). That should be much of your noise level already. But there are many many methods to solve the task.

The good composites are those you don't see...

Axel

 

[Radoslav Karapetkov]

On the other hand, we have the "whole" future as a possible "place" to put a story.

Which is pretty much unlimited. And for the future, a grainy look wouldn't be always appropriate.

Would a grainy vision work for a mission-to-mars kind of movie? Not always.

Probably if a RED camera was available in the time when Kubrick was making "2001", he would have chosen it... or at least the 65mm 8K version of it .

Or "Solyaris", "Stalker"... I bet that Tarkovsky would also love a super clean footage from a RED for his sci-fi films.

Alles ist gut.

We have a super-clean RED for the "future" and a grainy, magic-bullet-ed RED for "Julius Caesar".

Great.

 

[Bing Bailey]

its not the mileage baby its the grain LOL I like grain. I agree with spielberg's comments about grain giving film life when nothing else in the frame is moving. course we can add digitally if we really want it.

human history has never been pristine. it has texture and so should the films we make. but there are ways to do this in costume design , makeup effects , production design and in post. we only want the source superclean so we can make these decisions in post later and then change them if we don't like what we get. if I wanted the feeling of looking through a window I'd go look through a window. I like options and thats what RED ONE gives us.

 

[Sanjin Jukic]

Grain stays even if one day film or digital would die.

It's a standalone aesthetic image category or an option.

Just have a look at the advertising in magazines.

When you have a grain it is about a nostalgy/nostalgie sign

and also it goes beyond that.

Reality is not only a present time.

The film or movies knows that the best.

Hollywood a factory of dreams... stories are involved here...

... past, present, future... and beyond that is eternity (in our human minds)...

and then beyond that is a STARDUST ****************************

LINK>>>

Just because to get a real grainy STARDUST must see this picture below here:

 

[Radoslav Karapetkov]

we only want the source superclean so we can make these decisions in post later and then change them if we don't like what we get.

Agreed 100 %.

 

[Kreisky]

With the extremely noise free recording of future camera systems, we may end up with a new look. Only then we can do the comparison which does affect us more, the version with or without grain. And then grain seems to become an purely artificial thing, as we have no real grain anymore.

I like that !

 

[Sidney L. Plaut]

we like dogme films where I work

 

[Christoffer Glans]

Screens might not even be present in the future. We might have biodigital chips in our heads that generate threedimensional screens and holographic images. It might also be possible to use these in a lucid dream state to create objects we can touch, smell and experience in what we define as "reality".

And these things you can mock, but that might be because you aren't currently updated in biotech/biodigital science.
In a hundred years this might be possible and with the current accelerated development of better computers and soon also quantum computers with the ability to make two opposing choices at the same time... well, let's say I'm sure we will see major breakthroughs in biodigital science even before a hundred years have passed...

All in all, photography hasn't been around for such a long time... who do we try to cheat by saying that we as photographers will be around a hundred years from now? Many of us who are photographers might be the same ones who would have been interested in painting before photography was invented.
So the future of photographers will be the creators of these holographic simulations and dreamscapes... we might do what we are supposed to be; painting with light, but this time we can bend it by pure will inside these created realities.

Catch my drift? "Noise free future cameras" in what timespan are we talking?

 

[Axel Mertes]

Glazarus,

I guess its time to make a break and go back into the Matrix, do you?



I am there already...
Axel

 

[Christoffer Glans]

I guess its time to make a break and go back into the Matrix, do you?

I see you have been living two lives... Mr. Mertes

 

[conrad gaunt]

In my experience a little bit of grain can help distribute quantization, and help remove banding too, especially if fading to black. My understanding of mpeg is that high frequency blocks of pixels don`t show the effects of banding as much as smooth gradients do, so more quantization is acceptable when allocating data for such a high freq. pixel block? I never seem to get banding on smooth areas of eg, sky , when data rates are good anyway, but sometimes when natural scenes are complex, even with the DC precision set high etc, I see nasty artifacts in trees etc, whilst the sky is usually near perfect. I think there are different scales of noise, and types, some maybe useful, although s16mm film grain, it appears, is not one of those. Any unnatural noise added to motion images will be at a risk of jittering, but this isn`t always the case, or a problem. I can`t imagine wanting to add noise to red footage often personally, if ever. If I did it would have to be brown in origin, and grow all over the image as part of some , er , sci-fi experiment.
I guess some people might want to match film stock, with recode, but I`d rather intercut scarlet footage personally, or just use film totally. I wouldn`t rule out noise completely as a useful, digital tool, if you`re into vfx, but definately do target test encodings.
That said, I haven`t liked many, if any, digital grain effects, or audio distortion effects for that matter, that I`ve come across over the years. They always leave me a little cold, like using a transistor amp straight after using a mates 100W valve amp with the same guitar.
More research into simulating analog signals using discrete digital samples, for artistic purposes (rather than scientific) definately ticks an interesting alchemy box for me, and IFS/noise systems are still open areas for research papers, in the 3D software realms.
I have been privately experimenting with simulated guitar feedback effects for years (the kind hendrix could make,with a crybaby wah wah, a fuzz box and valve amp.) , using only clean guitar sounds as inputs, just for the hell of it really, and because it seemed impossible at first. the valve amps warmth always leaves me lusting a better digital alternative, sometimes I think such endevours are futile, but possibly why they draw me in.
The psuedo subjective nature of de-Bayering CFAs interests me in the same, but slighly strange way, always tempting me to add one or more iterated function systems into the mix, and brown noise, just to see what resolution the image stops growing at. That would be interesting noise to me, but probably for still images more than motion ones. Similar to software that attempts to use several images to create a "super-resolution" image, but using only one image, and an IFS. Like a cancer? Ooo, I`m bamboozled!

 

[Dj Joofa]

And these things you can mock, but that might be because you aren't currently updated in biotech/biodigital science.
In a hundred years this might be possible and with the current accelerated development of better computers and soon also quantum computers with the ability to make two opposing choices at the same time... well, let's say I'm sure we will see major breakthroughs in biodigital science even before a hundred years have passed...

Glazarus, it is easy to believe the "miracles" of modern science these days. Your usage of "two opposing choices at the same time" reminds me of Minsky's famous book "Perceptrons", where he showed that a simple neuron can't even distinguish between a XOR function and single-handedly tried to discredit the fledgling field of neural networks. However, that effort was shown later to not hold true in the case of multi-layer neural networks.

Artificial Intelligence, computer vision, and pattern recognition are collectively at least close to 50 years old field, however, they have produced little technology that is worthwhile. And, before you start coming after me, I must say that I was trained in pattern recognition and computer vision, which is my primary background. So, I have heard such arguments regarding biology-based machines for a long time. I do believe that something will eventually emerge from biology-based technology, however, it may not be what we think currently given the popular culture.

Kubrick came up with a film entitled "2001: A Space Odyssey" in late 1960's, but 2001 came and passed by, but we still do not have a computer even remotely as intelligent as HAL.

 

[Christoffer Glans]

Kubrick came up with a film entitled "2001: A Space Odyssey" in late 1960's, but 2001 came and passed by, but we still do not have a computer even remotely as intelligent as HAL

But the cartridges that they store digital information, the "cells" of the HAL computer closely resemble holographic storage disc prototypes that are researched today as the future storing media.
It's not all wrong, just because the date is...

As for quantum computers, they are in development today as we speak.
They are just not cheap or working in such a stable way that we can use it to work with.

I can not predict the time it will happen but biodigital entities like advanced nanites and so on isn't the science fiction we used to think.
We have simplified nanites for creating windows that never attract liquid like water (just push it of the windshield of a car as en example).

And as Einstein said: Science can never develop new theories or technology without imagination.
Which is what I believe is the case with some scientists today; they just don't have imagination.

 

[Radoslav Karapetkov]

we like dogme films where I work

I must also admit that I'm a bit dogme-atic myself.

But, just a little bit.

 

[Dj Joofa]

But the cartridges that they store digital information, the "cells" of the HAL computer closely resemble holographic storage disc prototypes that are researched today as the future storing media.
It's not all wrong, just because the date is...

Well, I was not even born in 1968 when Kubrick's "2001: A Space Odyssey" came out so I am not responsible for any information contained in it.

And as Einstein said: Science can never develop new theories or technology without imagination.
Which is what I believe is the case with some scientists today; they just don't have imagination.

Yep, that part I agree.

 

[Nova Invicta]

Mpeg 2 & 4

Mpeg 2 & 4

The BBC have officially stated by 2010 they will no longer accept 16mm film because of the grain when applying 10 - 18Mbps for HD broadcast. OFCOM the UK broadcast regulator wants to see compression further reduced to compact more channels into the limited UK bandwidth so effectively compounding the problem further.
The way in which Mpeg compresses is the issue NOT film but as someone else has commented the engineers & IT guys have won the day artistic choice is effectively dead for the BBC. Mpeg roots are in broadcast not in cinema but with digital projection and Blu-ray were seeing broadcast standards dictating the standards for all display going forwards.

 

[Christoffer Glans]

Broadcasters just try to push the bandwith over the edge to get supermany channels instead of quality.
Channel 4 in Sweden used to be 1 channel when we went into digital broadcasting. The quality of series like Lost and so on broadcast in SD to a HDTV with perfect reception almost looked like HD television because the signal was so perfect.

I have witnessed how the channel quality went from being perfect to shait after they pushed in alot of unnecessary channels that are supposed to cover different genres like documentary, film, sport, classics, children, you name it. Now almost all the major Swedish channels have subchannels that nobody (nobody normal ) watches but who takes up bandwith.

I wonder... is there ANY high quality HDTV broadcasters that broadcast 1080p with all the bandwith possible? I'm sure that when enough people have HDTV's in their home a channel like that will be on everyone's favorites just because it offers perfect quality. They wouldn't need to market it more then "highest quality HD direct to your home".