2. Various Kinds of Realism

There are different ways the term ‘realism’ can be understood when it comes to film (keep in mind that we’re talking about fiction). Hopkins ^{Footnote 3} provides a very useful and clear list of various types of claims that one could call ‘realism’ here, and I will loosely follow a part of this list, using in some places the same terminology Hopkins does, while sometimes deviating from it, and sometimes adding to it.

First, ‘being realistic’ could mean ‘being precise.’ ^{Footnote 4} Realism understood as precision is then a claim about the amount of information the viewer gets from what she sees. A very sharp 5K, ^{Footnote 5} HFR, and 3D image will thus be more realistic than a low-resolution, 20fps, 2D, black and white one, since the former will simply make more information available. In this sense, the image is understood as an ‘information-bearer.’ The kind of ‘information’ involved here consists mainly in the level of detail the image provides: for example, a very sharp and detailed image shows us a tiny wrinkle that would not be apparent on a less sharp and less detailed image, so the more detailed image provides us with more information (i.e., information about the existence of a wrinkle). This is a kind of precision that concerns the image itself, but precision can also concern the story that is being told; for instance, a fictional claim that Bilbo Baggins is 125cm tall is more precise than the claim that he is about half the size of a tall human. Such information can be directly provided in the story, where, for instance, a character can precisely refer to Bilbo’s height, or it can be, less precisely, visually accessible to the viewer, say, by visual comparison with a human character. Thus, when it comes to precision, there are (at least) two kinds of it: first, precision that concerns the image itself (its definition, sharpness, colour, as well as the frame rate in which the movie is being shot and shown—more on the latter below), and second, precision concerning information that is conveyed by the film, where this information can be conveyed equally well by a very precise image (in the first sense) and a less precise one (in the first sense): the height of Bilbo can be equally accessible in a high-resolution, HFR, 3D projection and in the projection of a low-resolution, badly executed copy. Let us use the terms ‘image precision’ and ‘information precision’ for these two kinds of precision, respectively, and let us distinguish them, even though they significantly overlap (more image precision will often entail more information precision (but not vice versa)). If our focus is HFR, it is image precision that we should keep in mind, and as we will see below it has a very high impact on the question of realism in HFR movies.

Somewhat related to image precision, another claim of realism in film concerns the recessiveness of the medium. ^{Footnote 6} In short, this is a question about how much the cinematographic image qua image (that is, qua marks on a surface) attracts attention to itself. A typical case of non-recessiveness of the medium involves accidental imperfections: if there are scratches, or flickering bright pixels, or anything else that disturbs the way the marks on the surface are supposed to be, this will distract the viewer from what she’s supposed to see in the image (say, hobbits), and attract her attention to what she sees on the surface—namely, a distribution of patches of colour organized in a particular way. This can be of course purposefully toyed with by the movie makers, who can use various kinds of sepia, black and white, or grainy film effects. In such situations, the viewer’s attention focuses on the medium—say, on the fact that in the middle of a movie the image switches to being black and white and very grainy—to help a story to be told, for instance, by implying that the scene that is being viewed happened in the past. In such cases, the non-recessiveness of the medium helps the story to be told. But in general it is recessiveness that movie makers are after, the idea being that one should ‘forget’ that one is looking at a flat surface with marks on it, in order to see in rather than to see. Thus, exceptions set apart, the more the image itself is recessive, the more realistic the film is, in this sense of realism.

The combination of a very high definition, such as 5K and HFR, gives rise to an exceptionally realistic movie when it comes to image precision. HFR plays an important role here, more important than the definition (most HD television screens are also HFR, so the two aspects of the resulting viewing experience are often not distinguished). The image is exceptionally sharp, providing a lot of detail as well as more fluid movement. Indeed, HFR removes motion blur: each frame contains less motion blur because the shutter is open for a shorter interval of time. What about recessiveness? When experienced for the first time, a very high definition HFR image clearly attracts attention to its exceptional sharpness, etc. and so it is non-recessive: the movie is thus non-realistic in this sense. Some viewers then get quickly used to it, and see such technological improvements as simply improvements; for this category of viewers, the medium thus ceases to be non-recessive, and becomes as recessive as a standard lower-resolution, 24fps cinema image always used to be. Others, however, do not get used to it, and find the overall experience disturbing and aesthetically unpleasing. In order to understand why, we must now consider other kinds of realism.

(When it comes to recessiveness, films shown in 4D are a good example of a failure of recessiveness. 4D consists in projecting a 3D and high definition film and in adding wind, water splashes, smoke, seat vibration, and strobe lights, mainly to action scenes. This does not (and is probably not meant to) help the viewer to immerse oneself in the story told; rather, it is often experienced as distracting—while sometimes these distractions are judged as being pleasant, sometimes as unpleasant, they are almost always experienced as distractions. Thus, the ‘medium’ is here very non-recessive.)

3. High Frame Rate

What do we see when we see a movie such as The Hobbit? There are at least three candidates: first, we see a flat surface with marks on it (the image); second, we see actors in hobbit costumes (this is what Hopkins calls the ‘events filmed’); and third, we see hobbits (Hopkins refers to this as the ‘story told’).

The first thing to note here is that we never see these three things at the same time—it takes some switching back-and-forth between the three kinds of experiences to have them all. Richard Wollheim, when talking about still images, says:

But this does not mean that we have two experiences at the same time; rather, Wollheim talks about one experience with two aspects. In the case of our three candidates, it is clear that we never have three such experiences at the same time. Indeed, one can either focus on the surface qua surface, or one can focus on the actors (for instance, on the good or bad way they play their parts), or one can forget about the surface and the actors and simply focus on the story and immerse oneself in it. The latter point is exactly what recessiveness is about when it comes to the surface: when a film is realistic in the sense that the medium is recessive, we typically do not consciously see it when we see a movie.

The second thing to note is that in some cases not all of the three levels of experience of a film are present (simultaneously or not). This is where HFR (coupled with high resolution, such as 5K) is a good test case. Indeed, given how sharp, detailed, and fluid the image is—that is, how realistic in the sense of image precision the film is—some viewers have trouble getting to the third level of experiencing the movie. In short, the image is so sharp and detailed that one so clearly sees the actors in costumes that this makes it hard to see them as hobbits. The high resolution alone is not the main and only culprit here, since it is HFR that is responsible for what is often referred to as the ‘soap opera effect’ (soap operas are typically shot at 50fps or 60fps ^{Footnote 8} )—one gets the experience of watching a cheap television program instead of a movie. Thus, many (but not all) viewers of The Hobbit as projected in HFR report a very disturbing experience: when watching the movie, they are continuously seeing the make-up and the costumes for what they really are—that is, make-up and costumes—which prevents them from seeing hobbits. Outstanding image precision in association with the ‘soap opera effect’ are responsible for that. The ‘events filmed’ are thus represented by the image in such a way that the ‘story told’ never really becomes available, or at least not without effort, simply because, to put it crudely, the events filmed keep attracting attention to themselves.

This has of course a lot to do with what typical viewers are or aren’t used to, and it’s a question of discrepancy between image precision and the quality of costumes, make-up, props, etc. Early movies clearly suffered from the discrepancy: the props, costumes, or, say, alien characters involved in science-fiction movies were easily seen to be props, costumes, and masks. This discrepancy was, at that time, something that viewers were well used to, and when shown at 24fps, this all became entirely acceptable as a representation of a fictional reality: such is the ‘magic’ of the ‘cinematic quality’ of a 24fps movie. In the following decades, the discrepancy faded and effectively disappeared. Movies became so well done, special effects, make-up, costumes, etc. became so well executed that—when shown at 24fps (even in HD, 2K, or 4K)—there was rarely any discrepancy at all, and the audience became progressively used to films where forgetting about the image as well as about the events filmed, and focusing only on the story told, is immediate and easy. In the case of many movies, a typical viewer is nowadays actually entirely incapable of distinguishing actors from Computer-generated imagery (CGI) characters, ^{Footnote 9} and make-up, costumes, props, and similar never appear to be what they are—as long as the film is not too realistic when it comes to image precision, which is exactly what happens in the case of HFR movies such as The Hobbit. Thus, what HFR does is bring back the old discrepancy, which typical viewers are no longer used to. (To counterbalance that effect, props, costumes, make-up, etc. would need to become much better than they are—at least, what they are in The Hobbit. Characters such as Smaug—the dragon—do not suffer of this HFR effect for precisely this reason: there is no discrepancy, because there are no dragons.)

4. Realism

What does this tell us about realism? What is at play here is a kind of realism that is close to what Hopkins calls ‘Collapse.’ Hopkins defines Collapse as follows: “a film is realistic when we see it as the photographic record, not of the events filmed, but of the events that make up the story told.” ^{Footnote 10} The idea is that the three levels of experiencing a movie Collapse in the sense that the viewer gets directly from the first level (where she sees an image on a flat surface) to the third level (where she sees hobbits), without ever noticing the second level (where there are actors in hobbit costumes). In the sense of Collapse, following what we have seen above, The Hobbit in HFR is then a very unrealistic movie. The situation is then the following: an HFR movie such as The Hobbit is more realistic when it comes to image precision and because of that it is less realistic when it comes to Collapse.

Let us avoid a possible misunderstanding: Collapse is not the same thing as transparency. Transparency is a notion coming from the debate about photographic representation, famously championed by Kendall Walton, ^{Footnote 11} who claims that photographs allow us to literally see the objects they represent—in this sense, they are transparent. In Walton’s view, photographs are produced in a mechanical way and as a consequence they are akin to aids to vision: they are prosthetic devices like telescopes or surveillance cameras that help us to literally see through them. ^{Footnote 12} In this view, when I see a photograph of my deceased grandmother, I do actually see her. Transparency, when applied to the case of cinema, is then the claim that when we see a movie—that is, a series of static photographs shown in rapid succession—we literally see through the resulting (‘moving’) image what was there when the series of photographs was shot, namely, the actors in costumes (i.e., the events filmed). If the transparency thesis is true, ^{Footnote 13} we can then articulate the relationships between the first, second, and third level of experiencing a movie by saying that, by seeing the (‘moving’) image, we see the events filmed, and we see in them the story told.

Transparency is a fourth kind of realism (until now we have seen three: precision, recessiveness of the medium, and Collapse). In this sense of realism, a film is realistic if one literally sees through it real events that happened out there in the world. If the transparency claim is correct, this is then true of all standard photography and of all standard film—except of course when CGI and otherwise added visual effects come to supplement the photographic images, since such modified images are to be understood as paintings rather than as photographs, at least partly. Transparency is a somewhat orthogonal issue to the first three brands of realism we have seen. It is an orthogonal issue because it is a claim about the nature of photography and film (in the sense of a projection of a rapid succession of photographic images), rather than a claim about our experience of film. One way to see the independence of the transparency thesis is to see that it has no implications with respect to the other three kinds of realism we have considered above. If Walton is right and the transparency claim is true, this does not imply nor preclude Collapse, precision, or recessiveness. Transparency is simply a claim about the process of production of photographic images and about a consequence of the nature of this process, namely the fact that such images can act as aids to vision. On the other hand, we have seen the relationship there is between the other three claims: more precision entails less Collapse and less recessiveness of the medium (at least to some extent), as illustrated by the case of HFR.

To come back to Collapse, one way to understand it better is to formulate it as involving a kind of illusion. ^{Footnote 14} When Collapse is at work, we are not having an experience of the events filmed but rather we have an experience of hobbits. Of course, since there are no hobbits, we are in this sense having a kind of an illusory experience. It is a voluntarily endorsed illusion, which is actually a result of a controlled use of our cognitive and imaginative faculties in order to get an experience of the story told. When nothing prevents us from doing this—unlike in the case of distractions that can lead to non-recessiveness of the medium or that prevent Collapse, as we have seen in the case of HFR movies—we use our cognitive and imaginative faculties to voluntarily fool ourselves, and the movie we are experiencing thus becomes for us realistic in the sense of Collapse, i.e., in the sense that we see it as a photographic representation of fictional events as if those were real events.

5. In Defence of Collapse

We have seen four kinds of realism in film, and we have seen how three of them interact. Having these kinds of realism in mind, what should we conclude? Should we say that The Hobbit is a realistic film? Or should we refrain from such an unqualified general statement and only always talk more precisely about realism of such-and-such a kind (precision, recessiveness, Collapse, transparency)? Indeed, we can say that The Hobbit as shown ^{Footnote 15} in HFR is very realistic in the sense of image precision, that it is not realistic when it comes to Collapse, and that it is not realistic when it comes to recessiveness of the medium, although this latter kind of realism is very much subject to how much typical viewers are used to such type of images. But, then again, is The Hobbit a realistic movie, or is it not? Asked in this way, the question does not perhaps bear much importance—it may merely be a terminological question. But the question that lies behind it is perhaps more interesting: is one of the brands of realism we have seen above more important than the others? Is there one kind of realism that is to be privileged and to be considered as being the realism in film? ^{Footnote 16}

I believe this is indeed the case and that it is Collapse that we should take to be the most important form of realism, when it comes to fiction. Bearing in mind our test case of an HFR movie, the failure of Collapse explains here very well the overall failure of the movie, in the sense that it gave rise to many reactions of disappointment. Indeed, the film simply misses its target: the whole point of a movie such as The Hobbit is to tell a story about hobbits. If it doesn’t, it simply fails, no matter how skilfully done it is or how extraordinary the quality of the image is. Telling a story is what movies are for (keep in mind that I am focusing on fiction here). It’s their main purpose and their very reason for existence. ^{Footnote 17} James Conant ^{Footnote 18} put it nicely: “for something to be a movie it must offer us glimpses of a world. [I]t must visually present a world into which it can invite us and in which we can become absorbed.” ^{Footnote 19}

This is why, in a nutshell, I submit that Collapse is the kind of realism that is more important than the others. Take, for instance, image precision. If a movie fails in this respect, for instance, if the image is not very sharp, as long as Collapse is preserved—that is, as long as one is able to see hobbits in the movie—it can still be experienced as a movie, and serve its main purpose perfectly well. This is not so if Collapse fails. If Collapse fails, the whole point of the movie is lost. We have seen that Collapse is based in a kind of voluntary illusion and in the use of our imaginative faculties. This kind of illusion and use of imagination is indeed what storytelling is all about. This is why an old black and white, low resolution, 2D movie can be better, as a means to tell a story, than an HFR movie with outstanding image quality. More often than not, less is simply more: a black and white image with a shallow depth of field can leave more room for imagination and more room for stories to be told/imagined than a very sharp, perfectly exposed, high-resolution image of the same reality. Accuracy understood as fidelity to reality is not what counts. Contrary to documentaries, fictional films are not about recording what exists in the world. To stick to our main test case involving HFR, we thus see that motion blur—which is exactly what, on purpose, HFR prevents—is a good thing, similar to how a shallow depth of field is a good thing, or to the way an underexposed scene is a good thing, in order to leave some elements unshown and simply imagined.

The whole point of a fictional film is to (i) tell a story, and (ii) allow viewers to immerse themselves in the story told. If Collapse fails, (ii) is lost, and so is one of the very reasons for the existence of the movie qua work of fiction. Granted, a story can be conveyed by a movie where Collapse fails—after all, we do get the story about Bilbo when we watch The Hobbit—but not in the immersive sense. ^{Footnote 20} Collapse is necessary in order to have an experience of only the story told. When Collapse fails, distractions that prevent full immersion are in place (similarly to what happens in the case of 4D movies; see above). The other forms of realism are not of such importance and do not have such a consequence if they fail. This shows, I submit, that Collapse is the main kind of realism, when it comes to fiction.