46

I happen, time and again, to read and hear people – among which people working in cinema – regretting the fact that film (in the sense of actual reels of plastic strip) is nowadays very rarely used when shooting movies. Is there some ground to such nostalgia?

Of course I am not looking for opinions for or against film vs. digital media, but sourced, objective reasons. Is there anything that could be done with film but not with digital supports? Are there measurable differences in effects on viewers?

10
  • 5
    It arguably still is better. Not sure who told you no one uses film any more, though. It's still widely used.
    – Catija
    Commented May 17, 2015 at 17:52
  • May I know why the question was downvoted, in order to try to improve it?
    – DaG
    Commented May 17, 2015 at 19:11
  • @Catija: I was under the impression that, almost always during filming and always when screening in theatres, digital media are used, but I will be ready to stand corrected.
    – DaG
    Commented May 17, 2015 at 19:14
  • 4
    Are you talking more about film as a recording medium or as a presentation format? Or both? Can you clarify your question slightly?
    – Catija
    Commented May 17, 2015 at 20:09
  • 2
    "is no more used for movies" -- says who? Christopher Nolan for instance still uses pellicule.
    – BCdotWEB
    Commented May 17, 2015 at 20:56

3 Answers 3

55
+100

I'm going to start an answer but this is by no means complete.

Whether you're talking about film as a recording medium or as a presentation format, there are people who prefer both. You can do a lot of research on why one is better than the other but a lot of it comes down to personal preference in the end.

As a Recording Medium

There are pros and cons for film and for digital. Some directors are sticklers for using film, particularly for quality/stylistic reasons, while others like the freedom allowed with digital media. Here are just a few of the major differences.

Dynamic range - or How pretty will it be?

You're going to get better whites and blacks and colors with film than with digital. The digital sensors just aren't quite sensitive enough (yet) to mimic film:

There is no limit to the number of possible levels of color on emulsion film, whereas a digital sensor stores integer numbers, producing a limited and specific possible number of colors. Banding may be visible in the unusual case that it is not obscured by noise, and detail may be lost, particularly in shadow and highlight areas.

Here's an article comparing digital and film stills to show the dynamic range differences between the two... it's pretty impressive.

Here's a summary statement on the site:

It's 2014 and when I shoot digital, I'm still using a cheap one for most things. That's because in late 2014, even the expensive ones still blow out highlights in a nasty, computerized, unacceptable-looking way. Full-frame digital gives slightly more leeway, but it's still not slide film.

At least when highlights get blown out with film, they look natural. As we've seen, Fuji Superia negative film was able to tolerate seven full stops of highlight overexposure and still yield a pleasant, smooth-looking picture. It could be ten or twenty stops overexposed, and the worst that would happen is that it would yield a smooth-edged transition to pure white.

As to your basic question, this is the immediate answer... none of the other factors listed below are important to directors who care about this one. Film was (and is) still better in a quantifiable way, than digital.

Everything below here is why so many opt to forgo what some consider a very minor difference between the two and make the switch to digital:

Cost - or How much do I have to fork over?

Film is expensive.

With film projects, a major cost of making is the cost of film stock (around $769 USD/roll), processing, digitizing (so that it can be edited digitally), preserving, and storing the film negatives. And, with the increasing use of digital, these costs are likely rising as fewer film-processing companies are around... If you're interested in the costs, there's a site that outlines some of them here, not sure how recent the info is.

That isn't to say that digital doesn't have costs of its own... Camera cards, storage media, etc ... do have expenses associated with them.

Ease of Loss - or How much can I screw up and still get a movie out of it?

I'm not going to say that it's impossible to screw up or lose digital files... it is... but it's not nearly as easy as screwing up a roll of undeveloped film. One little light leak and the last two hours of work is down the drain. With film, you have to train people how to properly load and unload film rolls and to particularly be wary of any light... and you have to protect those shot and un-shot rolls from the moment you buy them until they've been finished processing. Even only taking into account the time from when shot to developed and digitized, it can be a day or two... or a week before the film is truly "safe"... and this often requires mailing or couriering it across several states depending on if there's a processing house anywhere nearby.

With digital, when you pop the card out of the camera it goes to your data wrangler who will download the data and likely make multiple copies of it and verify that the files are good. Within an hour, your data is "safe". If you're smart and do make multiple copies of the data and keep them in different places, the chance you'll lose everything is miniscule.

Reviewing Shots - or How quickly can I see the mess the actors made of my dialogue?

As stated above, you don't know what you've got on film until it's developed... which can take a day to a week. Some sets may do a simultaneous recording on a digital camera but it may not record an identical image to the 35mm camera.

Digital media can be instantly reviewed by simply watching the digital files on the camera.

Length of Shots - or How often do I have to stop and change my media?

A standard roll of film has a run time of about 11 minutes when run at 24 frames per second... regardless of the type of film, actually. It's the same running time for 35 mm and 16 mm, just smaller (shorter) rolls for the 16 mm. Note, this also means you have to change rolls of film every 11 minutes of shooting time!

Depending on the digital camera used, recording quality and the size of the memory cards, you can get much more time on a digital camera. Here's some info from the Arri Alexa's recording info page:

Alexa record times

To paraphrase, with a 64 GB card, you'll get between 24 and 210 minutes of recording time depending on the codec you use.

There are other differences, obviously, but these are the big ones.

As a Presentation Format

The dynamic range issues exist here, too but the production companies have a bit more control and they're a bit more finicky about price.

Here's some numbers from an undated article on the University of California Institute for Research in the Arts:

Today, the driving force isn’t so much a single movie as it is the studios’ bottom line — they no longer want to pay to physically print and ship movies. It costs about $1,500 to print one copy of a movie on 35 mm film and ship it to theaters in its heavy metal canister. Multiply that by 4,000 copies — one for each movie on each screen in each multiplex around the country — and the numbers start to get ugly. By comparison, putting out a digital copy costs a mere $150.

A similar amount is noted on the Wikipedia page about release prints.

Release prints are generally expensive. For example, in the United States, it is not unusual for each one to cost around $1,500 to print and ship to theaters around the country. The cost of a release print is determined primarily by its length in feet, the type of print stock used and the number of prints being struck in a given run.

6
  • 7
    There's also the risk from the theaters perspective of damaging the film. The theater I used to work at as a projectionist would occasionally have what was refered to as a 'brain wrap'. This would happen when the film would get tangled around the projector and could permanently damage the film. In addition with film, someone has to take the film out of the canisters and splice it together.
    – kuhl
    Commented May 18, 2015 at 2:24
  • If I remember right from the commentary, the cost of all of the digital media used for shooting Spy Kids 2 was about the same as the cost of all the film used for El Mariachi.
    – supercat
    Commented Jun 28, 2016 at 18:41
  • When using a quality sensor with adequate light, could one avoid the "digital look" of blown highlights by under-exposing everything a stop and then doubling the intensity of almost everything, but with a shallower clipping function? Under-exposing and boosting would double the noise, but using a quality sensor with adequate light doubling the noise could be less objectionable than blown highlights.
    – supercat
    Commented Mar 22, 2017 at 21:53
  • @supercat I'm not sure... I feel like I've heard of people using 3D rigs set up to record identical images where the cameras are set at different stops to get a wider range and then they composite the images in post... The problem with your method is that if it's all under-exposed, you'll crush your blacks while (possibly) rescuing your whites... so you've only made a bigger problem.
    – Catija
    Commented Mar 22, 2017 at 21:58
  • @Catija: Dynamic range at the bottom can be a problem too, but depending upon the sensor it may be less severe than at the top. Using other capture approaches to improve the dynamic range is better if one can do it, but would of course require more expensive camera gear. Instead of using two sets of lenses and optics, though, perhaps it might be better to use two sensors and a beam splitter that transmits 90% of light to one sensor and reflects 10% to the other. That would cost less than 1/3 of a stop of light, but allow a ten-fold boost in highlight range. If the camera is built well...
    – supercat
    Commented Mar 22, 2017 at 22:30
3

In film, the dLogE curve, the characteristic curve of film emulsion is non-linear. With exposure time on the x-axis and negative density on the y, at very low levels, increasing light/exposure time, has little effect on film chemistry. It's asymptotic. Then, the curve, curves and starts to rise, linearly for much of the film's exposure range. That's where cameramen like to work: where you double the light (open up a stop) and you get twice the negative density. Then, at the top of the curve, it levels off to a asymptotically horizontal line. Adding more light provides no new negative density -- there is no more to be had at that level. Makeup men use to use that, in film, to hide wrinkles in the eyes of actresses of 'a certain age'. Film couldn't resolve the detail of the fine wrinkles: they were lost in the highlights.

Digital doesn't behave that way. It's dLogE curve is linear. Cameramen need to know this and be able to work with it. As in Photoshop's Curves tool, adjustments can be made, but at 24 frames per second, that, too, becomes a cost issue. So the 'look' of film changes in digital, or someone spends some more money, after arguing that only cameramen will see the difference.

7
  • 4
    There is a lot of jargon in this answer that makes it hard for someone without a lot technical knowledge about film to understand. For instance, I don't know what film emulsion is or why it is important. Your answer may be wonderful and worthy of many upvotes, but without understanding it, I can't say. Perhaps you could explain some of your terms? Commented May 25, 2016 at 2:18
  • 1
    I wrote a lengthy response to you, but do not see it. I wrote another response after, to another person, and saw that. So if you saw no response from me, hit me up again and I'll give you an answer. We could try email, but I'm willing to go at this again. Issue may be my own. I don't know. Commented May 25, 2016 at 17:21
  • 6
    You can always just try editing your answer with more information. I'd rather have you do that than send an e-mail, since then anybody viewing this question can benefit from what you have to say. Commented May 25, 2016 at 17:23
  • Faster, shorter answer to give you something. Emulsion is the chemistry that Kodak makes and puts on the long 35mm strip of plastic we all call 'film'. The chemistry responds to light: brighter image with more light, darker with less. Film of a give emulsion doesn't respond to light at very low or very high levels in a linear way: a little light doesn't give you a little brightness; it gives almost none at all. This is the asymptote, the non-linear part of the curve that describes how the film chemistry, the emulsion responds to light. Commented May 25, 2016 at 17:27
  • Digital is linear all the way: a very little light will give you a PROPORTIONALLY very little exposure. The relationship between light and exposure for digital is linear from dark to bright. Film is different at both darkest and lightest levels: less exposure variation for light input variation. Ask for more if you like. Commented May 25, 2016 at 17:27
0

Are there measurable differences in effects on viewers?

With film there is flickering of the image and your brain engages the persistence of vision in a very different way than while watching streaming content - whether interlaced or progressive. The difference being that with film, for some small period of time the audience is in the dark.

Also, generally with film, you are watching reflected light filtered through a photochemical medium; with digital video you are generally looking directly into the light source, and the streams of pixels are electromagnetic.

Other than that, it's just aesthetic preference.

2
  • I do not have a source link for this as it was an old local tv news article but no more than about five years ago but it stuck in my mind. Anyway it was about a few people who -preferred- watching their TVs in black and white (TV's were mono ones rather than just the colour turned down) due to an apparently sharper or better picture. It was not as far as I am aware down to any financial reasons even though (in theory) over here in the U.K you can obtain a black/white licence for much less than a colour one but it's complex and off topic to go in that as it involves recording capabilities etc.
    – AndyF
    Commented Jan 10, 2020 at 23:33
  • 1
    This is because the chroma or color information has only a fraction of the bandwidth of the luma (brightness) component, so a color tv broadcast signal is actually a high resolution black and white signal with a low resolution hue signal overlaid on it. This aligns with how we process visual information, our brains dedicate much more power to interpreting brightness than color, so this reflects how we see and was a handy way to add color to a black and white signal in a backwards compatible way. It was definitely a noticible effect if you compared black and white to color tvs side by side. Commented Aug 22, 2020 at 8:55

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .