Technical 4k-8k video question (Vidiot?)

This is one of those things that the members of the Society of Motion Picture & TV Engineers, the
Association for Moving Image Archivists, and other organizations all endlessly debate. To me, there's a point where even when you up the "K," there's just noise up there. My joke for years is, "if we do this film project in 4K, you'll be able to get an even better look at the grain!"

But for the actual image in the silver halide particles, I'm not convinced it's there. I don't think the lenses made prior to about 1980 could even resolve enough line pairs per millimeter (lp/mm). They couldn't focus well enough, and I don't think the emulsions were good enough. This is particularly true for Panavision & CinemaScope anamorphic lenses, which were notoriously soft for many years because of the extra glass elements needed for the optical squeeze.

The exception would be 65mm spherical productions like Lawrence of Arabia, My Fair Lady, and 2001, which were incredibly bright, crisp, and sharp. I think privately, Kodak admitted the best you could get from a traditional 35mm camera neg (OCN) -> interpositive (IP) -> internegative (IN) -> release print was 2K, and that's on a good day, with a well-maintained projector and perfect lenses. And there were many eras where the DPs loved the look of lots of soft focus and diffusion; Superman, Star Wars, Ryan's Daughter, Love Story, The Way We Were, lots of 1970s films were "soft & glowy," for lack of a better term. Sharpness was not what they were going for.

But films after 1980... the lenses and emulsions got a lot better. When Kodak released the Vision series negatives in the 1990s, I just about fell out of my chair by how sharp and (relatively) grain-free they looked. Right off the negative, I think you can make a good claim that this was -- best case -- 6K of resolution. But very few people ever got to see a picture right off the negative.

TL;DR: Real-world, the numbers are a lot of bullspit, and great pictures are beyond just the "K" in terms of numbers. I think the lighting and the "character" of the image means more than how sharp it is, per se. I'm not even sure that sharpness is the most important factor. You can make a similar point that frequency response alone is not the most important factor in sound quality.

 

With more and more films being shot digitally (as opposed to traditional film), how does that affect the 4K/8k/etc. scenario?

 

There are a lot of crazy audio/video people here who do qualify as being very, very critical about the quality of the experience, and I bet they could see the difference in quality.

Everything we master these days is typically at HD or 2K, but we monitor on 4K monitors that basically line-double the input signal so we're looking at faux UHD. There's not enough artifacts for me to mind it. If I flip the big switch and monitor in true UHD (near-4K), the difference in film quality for remastering is not gigantic. If it's a 2018 production shooting on very high-res cameras -- lots of them are 5K, 6K, even 8K these days -- then there is a visible difference between 4K and uprezzed 2K. And normal HD is not as sharp as 4K.

One big problem with monitoring in 4K is that you see all the focus problems that happened on set. And the bigger the chips, the more potential issues that can happen with depth-of-field and other lens-related problems. I'm reminded of Chris Nolan's Interstellar, shot in 65mm film, where a lot of people noticed that in closeups, sometimes one actor's eye was in sharper focus than the other eye, because the depth-of-field was so narrow.

 

On the size of TV I chose to own (32¨) 720 or 1080 content cannot be distinguished frrom one another. We watch it from 2.3 meters 7-8 feet away. But I have a CRT between my floorstander to watch my concert DVDs; they look so much better on it.

 

I don't think that's necessarily true. I just watched an older DVD on my 55" LG OLED 4K set, and it didn't bother me at all. If anything, I'd say the CRT is so old, it's rolling off all the noise and crap in the signal to the point where it's hiding the problems... but it's not telling you the truth. It's like using 30-year-old loudspeakers where the tweeters are blown, so now you can't hear the tape hiss.

 

God, I hate the soft focus and diffusion of the 70s. It was just overkill.

 

Silver Halide particles adhere to grain. The grain has detail that the human eye to brain resolves from frame to frame.

Having grain does not hurt the image it is a vital part of the image like a canvas is to paint. I got a big problem with this quote.

How can you say this? The halogenated silver atom is photosensitive. This makes up a part of the image. If you are not convinced the buliding blocks of the capture of the physical thing filmed is not there...then where is it?

Despite the so called limits of the lenses a soft image is an image nonetheless. Whether it is straight lines or a blurred lines the image is the image and still has physical properties that are put to physical film.

This idea that lenses pre 1980 have less resolution is flat out wrong. They are lens that capture information and yes the 8K scan of the Gone with the Wind 3 strip techncolor negative will get more infotmation then a 2K scan I don't care what the lense is. A physics class covers this.

I have never ever heard this and even if they did it disregards the fundamentals of the size of the molecules involved.

I would like a source and that means a scientific study not a bunch of peoole trusting their eyes alone without instrumentation.

Something not "sharp" does not mean it is less resolution.

While I know that you don't like grain this does not change the fact that it cannot be separated from the chemicals anymore than the supposed "grain free" look you think is better.

Of course what is scary is one seems to forget that the artists involved chose that film stock and developed with those so called limitations in mind. You may have your preference but a technician is not the artist and should not be rubbing out vital information with a digital eraser.

This one kills me. In the studies I have read the comparison is always with the negative and how a 4K scan gives a more accurate representation to that source then an optical print does.

Yet what is ignored is that a print didn't all of a sudden have less silver halide particles anymore then a photocopy has less carbon.

A film print is a unique footprint of the artists intended vision which includes color timing and varying degrees of development requests that were made by the director.

Lately there has been a push for OCN only with obvious changes made that had no input from the directors (or rather director's oroginal notes because I swear a good deal of them have lost their minds) at all and you see it advertised everywhere.

Now I am supposed to believe that not only does the grain not matter and it should be squished like a bug but somehow the silver halide particles that adhere to it not only don't carry the image, but somhow can be separated?

Wouldn't this be like taking Bob Rosses canvas away from his paintings? I would think he would have a problem with this and I think many great directors feel the same about their work being bastardized.

DePalma movies are awesome and one of the reasons I feel is the unique dreamy look that he chose for some films like Obsession.

Whether one likes that look or not, I would hope that there would be agreement that the look would be preserved and not tweaked to add "punch" because of technician preference.

It seems the theme to this post above that if the lines aren't straight on an outline because of a blurred lense artistic choice this somehow means less resolution.

Nothing could be further from the truth.

P.S. Penthouse magazine smoky pictorials ain't half bad either.

 

Here's the thing - the projectors in cinemas are 4K. I see more than a hundred films in the theater every year, my wife and I sit pretty darn close to the screen, and on a 30-foot wide screen we can't see the pixels.

They're making 8K TVs for the same reason they made 4K TVs - because they can and they want to sell more TVs. which is not to say there is not good reasons for 4K TVs, improve color range and contrast ratio are two good reasons, but actual number of pixels is not a good one.

I'd love to have a 1080p television with HDR, because 1920 x 1080 can be delivered to my home with much less compression than 4K can, and the compression artifacts are more noticeable to me then the supposedly greater resolution.

 

There was a lot of it. I recently worked on a 1983 film that still had that "soft/glowy diffusion" crap. It's fatiguing after awhile.

I will admit to using a glow/mist effect on highlights during love scenes and stuff like that on modern films... but not the whole damned film.

 

Totally agree with you Vidiot. I have a 65" 4k tv with a viewing distance of 11', and do watch some content in native 4k. Interstellar as an example was highly enjoyable, and certainly there was a smoothness and richness to the picture you don't get at 1080. I'm buying 4k content where the reviews (or the material) seem to warrant it however, as I am planning to go to 85" at some point, and I'm pretty sure it'll be more noticeable at that size.

 

This is probably going to be a silly question, so forgive me in advance... but can you personally discern what film stock was used on a particular film when watching it? It's easy to spot films (primarily made between 1977 and 1984) shot on 5247 for example - that had a very distinct and recognizable aesthetic - but if pressed, could you tell which film was shot primarily on which stock? I would assume being a digital colorist by profession, you'd recognize certain stocks or even digital capture cameras by sight now...

One of the reasons I'm asking is you mentioned the Kodak Vision series released in the 1990's, and I too noticed just how sharp some films of that era were... micro-grain, high-contrast, wonderful color saturation etc. It may not be a great film overall, but Star Trek: Generations is one gorgeously-photographed and lit film, courtesy of the late and great John A. Alonzo... would that have been shot on a Vision series (if so, which one?), because it looks remarkably sharp and detailed even now...!?

I'm coming around more to digital capture cinematography... I recently saw footage of Solo and was sure it was shot on 35mm... stunned when I found out it was digital... the times they are a-changin'...

 

I wonder how much of that is due to the egos of aging stars... and perhaps Lola and other digital "beauty work" companies have eliminated the need to have the rest of the film softened.

 

Star Wars lighting was aging stars? Vidiot has made it clear how he detests the look of the film which is fine, but to go a step further is another thing all together.

Gil Taylor made the decision based on his artistic choice the same as he did the Omen and some of the others I listed.

I was talking about this just this morning with a film friend and they were like what??? I like how those films look why is this a problem?

The quest for super clear "HD" look only has robbed art of personality and worse it has come to this foolishness that a soft focus or diffusing light presence somehow means less detail.

In this "your wrong" "you don't know science" blah blah it masks the real problem. It is ridiculing someone...anyone that doesn't drink the Koolaid.

You can shoot film your entire life, but it does not change the physical properties that allow you to develop your art to begin with.

If aging stars is why people like Gil Taylor made their decisions what about the DNR processed crap that gives a blemish free make over?

I am betting I am just a wee bit closer to the truth in that capacity. Of course in the "I am always wrong" camp this will most certainly fly by as if it has no merit.

 

No, we have the ability to keep the entire image soft except their skin if we choose to do that. Eyes, lips, teeth, hair, costume... all sharp. Even different parts of their faces and necks can be softened in a different way from other parts. So the filtration is a question of style, not to flatter the actors.

 

i read once that HDTV really placed a heavier burden on make up artists for TV shows, news, etc. I don’t know how much if that is true, however I could see that happening.

 

I think we've already hit the limits of extracting useful pixel data for most B-movies produced before this decade. Film stock plays a large role in how much information was originally captured.

The differences between a 4K and 2K image capture are vanishingly small for mediocre and/or inexpensive cinematography.

 

Howard Stern had a porn actress on some years ago, and she said, "oh, we're not looking forward to high definition in our industry."

Noted cinematographer Steve Yedlin (who recently shot the blockbuster Star Wars: The Last Jedi) is very outspoken about his belief that higher resolutions do not matter above a certain point:

DP Steve Yedlin Blows the Lid Off Camera Resolution Myths

A Clear Look at the Issue of Resolution - The American Society of Cinematographers

Steve Yedlin, ASC - ResDemo

Mr. Yedlin is extremely technical and really understands this area in extraordinary detail (no pun intended), and I tend to side with him on the subject.

 

When the Tonight Show went HD, Leno's makeup artist created airbrush makeup because the powder style makeup made people look like they had powder on their faces. I think they got a technical Emmy for it.

 

Vinegar Syndrome has proven all this wrong with HD scans of original negatives of more adult content.

People have seen beyond HD in the movie theaters for over a century. No attempts to rewrite history will change this fact.

If anything has been true from the recent 4K scans of older material is that improvement is obvious.

The pixel is the limit and it has quite a way to go to match analog material that has far more complicated information to mine then a even a 2D scan can produce.

 

The "Golden Age" films were shot to the same standard as mainstream ones for the most part, because a much of the crew were mainstream techs moonlighting.

Not sure what you mean. I just got out of First Man, seated in the second row of a theater with a huge screen - and the projector was a SONY 4K unit, and there was no hint of the pixel structure.

 

I was responding to this comment:

Yes I agree and this idea that HD scan is going to unearth more then what was originally shown does not make sense. HD at its best is usually filmed 6K. There is no question that HD filming is not there yet by a long shot.

Now all that said, did you really think I was unaware that adult films of years past were put to analog FILM medium?

The comment made by the actresses were obviously ones that have just been exposed to high quality because they have basically been crappy SD content for years. VHS was king for how long?

If we are ever going to have a sensible discussion we have to move past the basics.

And? It is easy to fool the human eye/brain because it wants to be fooled.

If you think you need 4K to hide a pixel structure what do you make of photographs that have far less information and you can't see it there either?

I have never got your point on this one. No offense. It doesn't prove that analog film has been matched.

When I say beyond HD, it is because obviously analog film doesn't adhere to a pixel count. But the pixel equivalent is what people debate, and I do not get this notion that film all of a sudden had less quality before the 90's.

Though it makes some people mad, if you study the basics of what was put on that film print, we are talking detail that can't even be seen to the maximum without a microscopes and I am supposed to think that a pixel that can be seen by the human eye (when you get up really close that is), matches it? Sorry, it does not.

Let me say this again. A pixel even in the smallest amount can absolutely be seen because humans are more than capable of discerning that kind of detail if you walk close enough to the TV/movie theater screen. You could actually focus on one pixel if you were crazy enough to do that.

Focus all you want on a silver halide cluster and it is IMPOSSIBLE for the human eye to discern one molecule.

The size of a 4K pixel on an projected screen can be measured in inches (depending on size of the largest screen). The size of a molecule on a film print even when projected is in the millionth of a millimeter at the largest.

So in the end, what makes more scientific sense?

 

You're off by a factor of ten. I have the pixels of a 4k image on a 30' wide screen at just over a tenth of an inch each.

 

Penthouse" there's a grain of truth there.

 

It depends on size of screen. I used Imax that can be 110 feet. So put it down to centimeters, my point remains unchanged except for the slight error.