The Hobbit, the Uncanny Valley, and what it means for 3D (opens in new tab)

I agree that there is no uncanny valley, but I believe that critics' complaints (and perhaps your own) stem not from independent comparisons of both 24 and 48 FPS to some ideal format, but rather from an assumption that 24 FPS is the ideal form.

If you believe that the purpose of a film format is simply to convey moving images to the viewer with as much fidelity and control as possible, as I do, then 48 FPS is objectively better (and 96 FPS would be better still, although there are obviously diminishing returns). The other opinion, which is that 24 FPS has some inherent artistic merit that makes it the ideal format for cinema, is bizarre and incomprehensible to me, and that type of argument could be (and probably was) used to argue against audio ("talkies"), color, surround sound, digital color correction, etc.

I saw The Hobbit both in 3D HFR (digital projection) and 3D IMAX ("real" IMAX, 70mm film projection, in 24 FPS). The difference in fidelity to me was smack-in-the-face obvious. HFR just looks so much better. In 24 FPS, the strobing in any shot with camera movement is so bad that it feels more like a camera or projector malfunction than a format that some people genuinely prefer for artistic reasons.

I definitely hope HFR is the future of cinema, just like I hope high resolution LCD displays become even more widespread, cell broadband networks get faster, digital cameras get better and higher resolution sensors, etc. I believe all these things are strictly better. To me, arguing that 24 FPS is better than 48 FPS is as bizarre as arguing that everyone should still use dial-up Internet just for the experience and so they will not take the wonder of the Internet for granted.

Honestly, you get used to it.

I say this just from watching game streams at 60 FPS, at first it looks like it's faster it should be somehow, but soon enough it just becomes normal and you wonder how you put up with low FPS streams.

The adjustment period might be a bit longer with film, but it'll happen eventually.

The idea is as the graph comes 'out of the valley', things become more and more real - so the familiar scenes, which look false to us, stay in the uncanny valley, whilst the less familiar scenes have already climbed up the other side. Just my interpretation / application of the age old theory though, so it's open to question no doubt - thanks for reading :-)

I think the argument is that the valley is further left on unfamiliar scenes, as "can just barely tell it's wrong" is more wrong for those things that stuff we know really well.

Alternatively you could say "Orcs are unreal, so there isn't an uncanny valley at all for them"

The graph shows a dip on the rightmost side below the leftmost side before going right up to indistinguishable from reality, that's the basic premise of the old uncanny valley hypothesis, that things which are just short of indistinguishable from the real thing are less desirable than fairly good approximations which could not otherwise be confused, due to the feeling that something implacable about them is a little off.

The "even frames" method of comparison you describe is flawed. Motion blur is very important to our perception of motion, and is the reason that motion at 24fps can look great on exposed film (lots of blur) and terrible in video games (no blur). When you skip half of the frames you have thrown out motion blur information that could have been present had the lower-rate source been created at that rate in the first place.

But yes, It would be neat to experience some legitimate side-by-side comparison of frame rates / motion blurs. Such a comparison has subtleties that make it hard to do fairly and cheaply.

http://100fps.com/how_many_frames_can_humans_see.htm (edit to fix small typo)

If you are on Windows, you can try SmoothVideo Project:

http://www.svp-team.com/

It's just interpolated 60 fps from regular frame rate movies but it works pretty well (it needs decent GPU for high-def videos).

I use it already for quite some time. Once you get pass the initial weirdness there is kinda no way back, 24 fps movies just feel broken.

It's similar to how computer games feel bad on underpowered graphics card, you do notice choppiness of 24 fps and it is pretty distracting.

BTW I did see Hobbit in 48 fps and it just felt "normal". So I suppose most of people complaining about 48 fps Hobbit just didn't get used to high frame rates yet (I expect for many it was their first experience). For me it took me few weeks for "soap opera effect" to wear off.

I'm curious how HFR with no 3D would have fared. Most of his complaints about the 3D HFR showing align more with my experiences with 3D than anything else.

There are so many ludicrous arguments in The Verge article.

> For some [filmmakers], HFR will be a potential new tool in their arsenal for telling certain types of stories in a new and exciting way, while others will be reminded of why the 2D format at 24 fps has stood the test of time for so long.

Just like dial-up for Internet and horses for transportation stood the test of time for so long. This is such an unabashed argument that "it should stay this way because it has been this at for so long" that I feel sympathetic embarrassment upon reading it.

The author's failure to be immersed is clearly due to his a priori insistence that 24 FPS is the way movie should be shot and projected. He says that himself. But there is more:

> In the opening hour of The Hobbit shown in 3D HFR – I don’t recall hearing a single sigh, or laugh. Not one. When I went to see the exact same seen with an audience of the same size on a 2D projection – I heard regular chuckles and laughter… why?

Here's why: because there are really high odds that of all the people across the world that watched the movie in both formats, at least one person would experience this. Why did Bob win the lottery? Is it because he's good at guessing lottery numbers, or is it because the odds were high that someone would win the lottery, and we chose Bob for analysis after learning that he won? In case anyone wants a counter-anecdote, I watched The Hobbit in both formats and there was a lot of laughter in both theaters. I don't have any real evidence, but I suspect most showings in all formats had considerable laughter.

> It’s like being on a film set in person: all of the magic is lost. You get to see behind the curtain and you’re no longer under the spell…

I think most critics and movie buffs are well aware that "the magic" comes from the viewer's deliberate choice to suspend disbelief. If this weren't the case, then filmmakers themselves would be unable to enjoy a film, since they would fully understand most aspects of how the film was produced. Again, this just reaffirms the author's own admission that he will only choose to be immersed if a film is projected in 24 FPS.

> The makeup wasn’t as terrible as some people say, and most of the VFX were stunning but not all. When I saw them in 2D however – it was almost like seeing another film. My attention wasn’t drawn to them … As I was focusing on central action. That challenges the "Suspension of Disbelief" theory that we all need to believe what we are seeing on screen and to get lost in it…

Oh boy. The author seems unaware that all or nearly all modern theory about suspension of disbelief places most of the responsibility on the viewer, not the artist. No artist can forcefully make you forget or ignore the fact that you're sitting down in a dark room with a bunch of other people.

I suppose I cannot dispute the author's claims about experiencing physical pain from the 3D. Perhaps there truly is some sort variation in people's visual systems that causes this. Other than the slight physical discomfort from wearing 3D glasses over my corrective glasses, all I can say is I haven't experienced that. If this were his central criticism, then it would be valid, although depending on the rate of occurrence of this medical condition it may or may not be reasonable to expect a blockbuster filmmaker to accommodate those who suffer from it.

The 60fps myth should be wearing thin by now, and I'm sad to see it repeated so often on HN.

I don't have the relevant links available to me currently, but suffice to say the eye can pick up on far, far more than 60 fps.

The fast forward look is just an adjustment period, you experience the same watching videos of games at 60fps (even though they're played in 60fps normally, it's a very odd thing).

If you believe 60fps is better than 24fps, I have a really hard time believing there's something about our visual processing that makes 48fps worse than 24fps.

HFR is nothing new. The problem is it sucks. It gives everything that is not a fast-action scene a cheap soap opera/reality show look.

I predict that a film's frame rate will become simply another choice for the cinematographer to make. Just like different film stocks and processing techniques are employed for different genres (e.g. bleach bypass in war movies), perhaps the FPS will indeed become genre-specific like you suggest. Debate about whether 48 is superior to 24 is misplaced -- it's an aesthetic choice, and the fluid movement that comes with additional frames works well in some scenarios but not in others. Many shots in The Hobbit looked terrible, perhaps demonstrating that 48 is not the ideal frame rate for an epic fantasy, but for some films it would be quite suitable.

I'm thinking action movies will benefit. Comedy, less so; depending on the type you might not want it too realistic.

Because for some reason they have applied a text-shadow effect with a colour of #fff (white). So normally it is invisible against the also white background, but once you select you put the selection colour behind it causing the shadow to show up.

Well, the uncanny valley does rely on the adaptation of human brains to quickly judge familiar things. Human faces are the most obvious and common example, because the human brain and visual system has unsurprisingly adapted to be extremely fast and aggressive at recognizing facial features. The effect will never be as strong with something like, say, the face of an aardvark, because everyone sort of knows what aardvarks look like, but few people are familiar enough to instantly recognize deviations and thus interpret the image as creepy or disturbing.