James Cameron champions faster film projection rates (from 24 to 60 fps)
latimesblogs.latimes.com
latimesblogs.latimes.com
Awesome!
It's finally happening. The 24 fps speed is a relic from the stone age of cinema, back when technology simply could not cope with higher frame rates. So all they could do was shoot at 24.
Then something odd happened. Everyone was shooting at 24, including the grand masters of the art. All the procedures, techniques, all the clever things they devised, took that frame rate as a given. They were not simply shooting AT 24 fps, they were shooting FOR 24 fps.
Thus the "film look", which is the way a motion picture looks if it's shot at 24 fps using the standard cinema techniques. From a technical limitation, it became a cultural given. Because Bunuel and Eisenstein and Bergman all shot at 24 fps, everyone simply assumed this is how cinema is supposed to look like. Higher speed video ended up being seen as "cheaper looking, made for TV".
But it's exclusively a habit, a purely conventional thing, imposed by culture.
I remember thinking a while ago that, what it would take to break the mold is a few big names in cinema to start purposefully shoot at higher frame rates, and at the same time try and figure out new ways to express the same things at the new speed. Some geniuses need to reinvent the art to break free from the old stereotype.
Cameron filming at 48 or even (please, please, please) at 60 fps cannot do the same things that Cecil DeMille did at 24. Or, rather, is free to not do the same things, because the lower frame rate is limiting in many ways.
And now finally it happens. This is great. I hope 24 fps finds its true place - in the history books. It's time to move on.
24FPS celluloid, with its known gamut limitations and temporal resolution creates an ambiance for the film. This is not unlike the way that canvas and oils create a certain ambiance about an oil painting. Now is the future of 'art' digital printers that can reproduce what an artist might have painted in oils, something they can paint using a Wacom tablet and Illustrator? No, not really. The medium is part of the art, its part of the palette.
Do digital films at high FPS have a place? Of course they do (art is pretty liberal in what it will accept as a medium :-) but will 24fps celluloid be consigned to the history books? Answering that question is less obvious.
The change from hand cranked to mechanically operated shutters was a clear win for cinematographers because it took some randomness out of what the audience would see, talking pictures did the same. The change from celluloid to video is however a medium change not a technique change.
I believe the comment at the particular talk I attended was "If Leonardo could have taken a picture of Mona Lisa instead of painting her, would it be in the Louvre today?" Can't easily answer that.
I was rebutting the claim that shooting film on celluloid would 'go away' because there were higher frame rate technologies available. I don't believe that conclusion is supported by either current or past experience the art world. It was not a 'lament', it merely challenged the reasoning that 24 fps celluloid movies would be replaced in their entirety by higher frame rate ones.
Yet it's factually true.
> I was rebutting the claim that shooting film on celluloid would 'go away'
Perhaps I did not choose the best words.
It will not disappear from the face of the Earth, of course. But it will be relegated to "retro" or nostalgic movies, etc.
It depends on how quickly the new, less limited technologies are adopted by the big names in cinema, and of course by how quickly the old generation, accustomed to the old way a film "is supposed to look", will retrain their taste or die off. :)
there are analogous issues between film and video, and even between older video formats and digital video. the future of movies no doubt includes a choice of frame rates, frame resolutions, interlace vs. progressive, and aspect ratios as creative choices. they will be part of the panoply of possibilities that artists will be able to use.
n.b.: in the silent era, there was no real standard film speed, and the film was often exhibited (in the days of hand crank projectors) at varying speeds to suit the content of the image and the music that was invariably played live alongside it. for example, romantic scenes might be cranked a bit more slowly; comedy scenes could be cranked a bit faster. it was only when the industry started to be professionalized that silent films were shot a 16 or 18 fps. with the sound era, the fps was upped to 24 partily to improve the fidelity of the optical sound. most projectors have for decades now multi-leaf shutters which actually interrupt the light path as many as 3 times per frame elimitate the perception of flicker.
Please, no. Interlacing doesn't have the same nostalgic effect as a 24fps frame rate, and only serves to lower video bandwidth for devices that can't handle a progressive image at the desired frame rate. It's a tradeoff between vertical and temporal resolution that works well for limited technology because it preserves the sharpness of still images. Basically everything except broadcast TV can handle 1080p60, though, so let's let the hack die.
But I agree with you -- this is a mere affectation, a slightly way of turning common images into something evocative. And this becomes much easier to achieve when starting with a high resolution image.
The same was not true for still photography. It certainly is an open question on whether a negative association with Video will go away or if it will remain.
Mostly, I don't see why my film fps preference is more culturally constructed, or constructed in some more objectionable way, than basically every single other thing having to do with film, which is nothing but an aesthetic and cultural phenomenon to begin with.
I mean, we might inquire into the habits and conventions that lead you to prefer 60-fps film, just as much as we look into those that lead me to prefer 24-fps film! Perhaps you have an aesthetic preference that films "should" be more realistic, and less stylized? Or a technological-aesthetic ideology that films should use "better" technology because it must necessarily be better aesthetically?
There are objective differences. 24 fps contains less information than 60 fps. Its temporal resolution is 2.5x worse.
Believe me, I can clearly see my own cultural bias. It keeps telling me 24 fps somehow does look okay. But we are not complete automata, we can distinguish between instinctive reactions (acquired tastes) and reality.
And the reality is, the higher the frame rate, the more information is recorded. An intentionally crippled low-framerate movie may serve certain esthetic norms, but that's pretty much all it does.
As far as 24 fps cinema is concerned, thanks to a century of conditioning, we're all pavlovian dogs.
I tend to prefer more stylized films overall, not just in temporal resolution either. If you prefer more realistic ones that seem like you're really watching a scene that's happening, that's a legitimate preference, but it's not mine. I'm not sure the 24-fps (or better, 18-fps) effect is the most important or best anti-realism aesthetic effect in a film-maker's toolbox, but it's one of them. It's probably, as you say, an accident of history that it's as widely used as it is, but it's in my view a positive accident of history. Modern Hollywood is far too realistic in its shooting style imo, and this is one of the few fortunate areas where it hasn't gone all the way off that cliff.
Which is not to say that 24p won't get largely eclipsed. I suspect it will. But not because it's universally 'better'- just more appropriate to a growing number of things.
A friend recently picked-up an LED "3D" tv. Now, I always thought the 3D is a BS gimmick, and I still do, though I will conceed that it does work far better than I ever thought it would. Which is, not at all.
But the TV -- a samsung -- had a feature that reminds me a bit of up-converting, however it's used on every input, even 1080P from the blu ray. I googled it at the time but the gist is that it, on 2D mode, it computes deltas between frames and inserts new ones to make the image seem more fluid.
IT LOOKED LIKE TRASH!
It made everything look soap-opera or like a documentary. It messed up the depth of field. Sure, things were hyper-crisp but alarmingly so. It made it look cheap. Now, there's a setting buried in the control panel to tweak or disable that, but this is how it came from the factory, and i thought it was un-watchable.
So it's not merely lighting.
One last thought which is I actually think the perception of higher quality shots in films has a lot to do with stedicam which of course will be the same at 60fps. But that's just a guess.
It may look like trash, but--then again--so do upscaled images.
Your dissection of what is wrong with Samsung's interpolation had nothing whatsoever to do with seanalltogether's comment.
Now come back at me with some word-by-word, point-by-point dissection, but i'm done, this isn't worth it.
Absolutely preposterous. Depth of field is a function of lens focal length, aperture, and sensor size. There is no way that Samsung's TV could in any way effect the depth of field of the original recording. If it could, then it would be a magical motherfucking TV.
Because as much as you clearly know about film tech more than I do, you seem to be lacking a slight amount of common sense: What's happening there is processing of a digital signal. Specifically, the TV is creating frames for you to create the illusion of smoother video. And what this effect does, is it appears to flatten the shot.
What your slashdot-worthy post overlooks is that you're speaking from your limited knowledge of film. However, in this TV, there's a feature that lets you apply this setting to just half the screen, so you can see them both, side by side, and tweak as you wish.
Just like in life, there's no amount of theoretical education than can make-up for actually being there, for actually having hands-on.
So really, save the hyperbole.
This actually makes it more like a video game in which discreet images are generated from the world, instead of natural film/digital sensor in which millions of photons are captured for a short duration.
Honestly, download an original British or Canadian drama show that doesn't get carried in the US and you'll see low budget. You'll see horrific, unwatchable direction.
I'm sorry, but give Sam Raimi a video camera with any frame rate and there's a significant chance he'll produce a blockbuster with it. I mean Oliver Stone released big budget movies using Super8!
The Hollywood "look" is genericness, because like everything else about Hollywood the studios are whole-heartedly fearful of anything that might stop someone watching a film, and in doing so turn people off watching films by lacking originality from plot to cast to cameras.
Practically nothing that doesn't feature US accents and locations is picked up by the networks - sometimes period co-productions sneak in, as long as there's pretty frocks and no swearing.
You're missing Misfits, for instance, which is the best show on TV...
I'm talking, look at the first episode of Primeval. It was atrocious across the board from acting to directing and 3D graphics; especially for reaching an audience of ~6.8 million. Compare that to the outstanding production qualities of Stargate Universe that only got ~2.3 million in a 5x bigger market.
My example are shows like the Canadian Flashpoint, which despite a decent budget and getting 1 million viewers can just be awful. The stories often have little reasoning behind them (shit literally just happens), the direction can often be gaudy by trying to make grand shots and just utterly failing. It often feels cold, like they're not using close-ups right. I often find this in many British shows too, comedies and traditional dramas not so much, but especially a lot of the police shows can be awful.
And Primeval could never approach, say, the brain-rotting stench of The Cape. SGU's poor ratings are not a badge of honour, it is a cancelled show!
As for police shows - that's pretty much a US speciality, but the really good ones are under constant threat of cancellation.
I do concede though, while I've frequently thought of British TV lately "damn, there's a lot that's good right now!", this thought has never consciously crossed my mind while watching anything Canadian. Due South, anyone?
Disclaimer: I'm a layman, but that's my opinion as a South American that has watched both North American and European TV (with long stays in Canada and Austria).
Sadly we were using first generation G4s with Final Cut 1.0, so to do this for any any film over about 5 minutes meant that the render time would be measured in days, not hours.
Google for 'samsung soap opera effect'.
I do not think that is an accident. After all, once we all buy our new 3D TVs, something will have to happen for people to buy new TVs again in 3-4 years.
And before you mention bandwidth considerations, I do not have actual data, but I suspect that when using H.264 High Profile, the difference in bandwidth between 48fps and 60fps won't be that large. There is a limit to how much you can change visually in a single second: human perception.
Most theatrical DVDs released after the early 2000s were 24Hz, as far as I know and depended on the DVD player to convert to NTSC. So now in modern players, they're outputted at 24Hz for 120Hz and 240Hz TVs.
The only downside is if you get crappy TVs and/or Bluray players that do horrible pulldown, I guess. But I've gotten hardpulled down DVDs that look horrible as well (I'm looking at you, BBC).
To the best of my knowledge, H.264 only supports two GOP structures (short = 6 frames / long = 15 frames), both of which operate independently from the playback speed.
In other words, a short GOP setting produces 8 blocks per second running at 48, and 10 per second running at 60. Assuming a constant bitrate (I know, I know), the throughput of the entire system needs to increase by an additional 20% to handle 60. And that's where things fall apart.
Consider the iTunes store, where I believe HD rentals top out at 720p. Or most HD broadcasters, who skew away from the 19.39 Mbps ceiling on the ATSC MPEG-2 spec in favor of programs encoded at 13.something Mbps (the legal floor) thereby allowing an SD channel encoded at ~6Mbps to be delivered in parallel. Does this exercise in bandwidth maximization impact the quality of the HD component? You bet. Do the networks care? If an additional SD means they can dramatically increase the number of ads served, then no, absolutely not.
So regardless of how great the acquisition setup, and how awesome the playback system, there's always the network in between them to consider. And with providers moving toward bandwidth caps, I see picture quality going down, not up. So, streaming 1080p60 H.264 encoded at (say) 7-9 Mbps? Choice, yes, but total fantasy for the time being.
Of course, Cameron is talking about making theaters competitive, so maybe that's his point. Given that BluRay has effectively failed as a universal home standard, and that instant-availability is hugely favored over order-and-wait programs, image quality of filmed entertainment at home is now tied to network limits.
The real question, then, is "are there enough cinephiles who actually care?"
Not exactly - while your 80% approximation is true for uncompressed footage, h.264 encoded video and essentially every modern day video compression schema will inherit data from the frame(s) before it. The higher the framerate, the less (generally speaking) the images will change between frames - leading to a reduction in data per frame.
Here's an example:
File 1: http://files.jjcm.org/60fps.mp4 - 127KB
File 2: http://files.jjcm.org/48fps.mp4 - 119KB
As seen here, the 48fps video 94% of the size of the 60 fps video. Granted, this is a singular example, but at least it supports the concepts that I described above.
I have to admit, I couldn't see any difference in the quality of the motion between the two clips, but thanks for posting them.
For what it's worth, I suspect what you describe (720p60 @ 6Mbps) is more than fine for most people. In other words, Cameron's idea that people will be going to theaters because of magical picture quality seems like wishful thinking. I suspect people go to theaters because they like going to theaters. It's social. It's big. It's immersive and (relatively) distraction-free - all part of a fun night out.
Are they going to go home to 720p60 @ 6Mbps and think 'gah!'. Unlikely.
I'm sorry, but it needs to be stated: the above isn't true and the calculations have nothing to do with reality. H.264 supports pretty much any GOP structure and you don't get from frames to final bitrate by dividing frame rates by GOP size.
Accepting that I was mistaken about the rigidity of h.264 (a possibility I acknowledged up front), and assuming that you DID establish a GOP structure of 6 frames (at a constant bit rate) shooting for one minute at 48 fps would produce 80% of the data generated by shooting at 60fps.
Indeed, this ratio would be true if you used no compression at all. The point was that in a fixed GOP scheme (again, at a constant bitrate), the difference in FILE SIZE would increase with the difference in frame rate, so the idea that file sizes for 60 fps material would only be marginally larger than those for 48 fps seemed mistaken.
My only mistake was in assuming that the GOP is fixed in h.264. I agree that "you don't get from frames to final bitrate by dividing frame rates by GOP size." But then again, I never said you did.
That's why a game at 24fps looks bad, but an (action) movie at 24fps looks good.
"Looks good"? Maybe passable. And that's only because of a century of experimentation with camera techniques allowing to eschew the limitations of the format.
E.g. try giving an amateur a 24 fps camera and ask them to shoot a sequence that involves a lot of panning. The result is going to look like junk.
Even in the hands of a professional it still doesn't look good. But we've all trained our brains to shut up and accept it, because we've all watched so much 24 fps content. It's 100% pavlovian conditioning, nothing else.
To someone coming from a universe where all motion pictures are shot at 60 fps or above, our movies would look excruciatingly bad.
Specifically, cone response time seems to be on the order of 15-20ms if I read http://www.ncbi.nlm.nih.gov/pubmed/14990682 correctly. And that's just to go from initial stimulation to peak response; presumably the response also dies off in nonzero time, so the full width of the response in time is more than 20ms. Rods respond slower than cones according to http://en.wikipedia.org/wiki/Photoreceptor_cell though I can't find numbers.
Cameron's proposal will introduce similar perceptual problems to HD video - fake stuff will look more fake, every tiny detail of a performer's face will be apparent, and scenes that play a certain way at 24fps might fall flat at 60. One study showed that physical comedy was funnier in low res/low frame rate, and the same action looked painful to the audience in greater definition and detail.
I'll watch a good story in any format, thanks.
Trumbull's research found that 60fps somehow resonated the most with the brain -- no slower, no faster.
Check out this infographic on Roger Ebert's blog comparing the different formats:
http://blogs.suntimes.com/ebert/assets_c/2011/01/resolution%...
The full blog post is well worth a read, especially given the story of a single passionate entrepreneur trying to change Hollywood and his story of trying to get Christopher Nolan to champion the format:
http://blogs.suntimes.com/ebert/2011/01/more_than_ever_the_f...
And they seem to suggest that RED made the entire chart, which they didn't. RED made the accurate part of the chart, but MaxiVision48 added their own nonsense.
The impression I get about this company is that they tried to sell a product to a bunch of people who didn't want or need it. The reason RED is so successful is that they actually made a camera directors wanted to use. Directors want to be able to do long takes, they want to see instant footage, they want to be able to go directly into the editing room after shooting, and they want a cheap medium. MaxiVision doesn't offer any of those advantages. In fact, if it ends up being harder to process, it's actually worse on all those points.
MaxiVision48 seems a little like BluRay. It's the next generation of a dead technology. Film will always be useful for certain films, but the benefits of digital go way beyond high resolution.
Anyway, interesting story, thanks for the link. Gives me some food for thought about my own business.
Slightly off topic, but you just nailed it, sir. Upvoted accordingly.
1) according to the Ebert article, it's not twice as much film, it "requires 50% more film and one stop more light (a trivial issue)"
http://blogs.suntimes.com/ebert/assets_c/2011/01/MaxiVision%...
Notice what's implied in the diagram is that the MV48 format would use existing 35mm film and would modify existing cameras for it.
2) it's way less expensive than 3D, both to shoot and display.
3D is just as much a competitor as the Red camera. It's what the movie studios and theaters have been banking on to compete with the quality of home theater setups, not an increase in 2D image quality. Caring about the unique look of film compared to digital while trying to increase the quality is why directors like Christopher Nolan like the IMAX format, but shooting with 70mm film is not only expensive but unwieldy because the cameras are huge. Sort of like what bulky 3D cameras were like before James Cameron forced the invention of a new one for Avatar.
No one shoots with 70 mm film. 65 mm film is the capture size, 70 mm is the print size (added 5 mm for audio tracks).
And what about the second point about 3D being an expensive sham that movie studios are using to milk more money out of the audience? And this isn't the first time they've tried that - we're going through the same fad that hit in the 60's and it'll die for similar reasons.
Regardless of whether it's Maxvision or RED, an increase in picture quality way beyond 1080p is possible and being held up by bean counters, not the directors. It's not like RED's native high resolution is translating directly to the screen - it's being downgraded for the projectors at movie theaters. I'm getting this from the FAQ page on the RED website. Scroll down to the last one - "So, where can I see 4K?" and you'll see promises of 4k projectors and displays coming soon.
The Maxivision goal was a noble one, but it was just too expensive to work. Increasing the framerate of content playback on digital projectors will be a far less costly initiative than the Maxivision proposal would have been. Unfortunately I'm not sure 48 fps vs. 24 fps has enough "wow" factor to convince the average movie goer to go to the 48 fps screening over a 24 fps screening, but I really hope it will.
Either way, I'm looking forward to any improvement that doesn't involve 3D. Ebert's other guest author the week before the Maxivision one was an anti-3D article by Walter Murch, probably the most famous film editor and sound designer ever:
This is watching the downloaded avi.
http://www.100fps.com/how_many_frames_can_humans_see.htm
Humans can notice stuff that hapens in 1/200th of a second.
His other pages are worth a look too, you can find them in the upper-left corner under FAQ.
It is nice to see that I'm not the only hacker with an interest in video technology.
(And there's plenty of content that's 60fps, just not a lot of American content that's 60fps.)
Doubling the framerate will not double the bitrate. (Though it will increase it somewhat, probably not as much as you'd expect) Modern video codecs are getting pretty smart.
In my opinion, higher framerates just end up making everything look like a sitcom.
I suppose I am just wary of change simply for the sake of change. I would like to be able to see the same 24fps and 60fps scene side by side and evaluate the advantages myself.
I assume if they have a modern digital 2D playback system this may not be a problem. However, what if they have just good ol' projectors that require film reels?
Questions on my mind:
- Can we roll film at a higher speed on the projector? Can we double the reel intake rate it if we go with 48 fps? Is it tougher to do fractional speed increases (like 60 which is 2.5 x 24)
- Are there any refresh issues - I assume no since the light is always turned on.
- How about sound? [The assumption here is sound is on the same reel - but I could be wrong]
If there are any projectionists in the HN house, I would appreciate you filling us in...
- sound is either on the same reel or on a DTS disc synced to the reel. - I was surprised that a rather small theater in my metro area (no stadium seating, out in the boonies) has digital projectors in all of its screening auditoriums.
For modern movies, I would imagine that in 10 years or less there will be no more acetate prints and all distribution will be digital. For digital prints framerate is solely limited by the response rate of the projector and the throughput available in the playback system.
Another case in point: people still shoot and love photos on Tri-X, a grainy B&W film, even tho' you can buy cameras that do a dozen or more full-colour megapixels off the shelf...
It'd be awesome, but it's a long, long way away.
As Jim says, this is low-hanging fruit. Easily done with most everything that's on the shelf today.
For fully rendered content, there is likely no additional man hour time, but for any frame by frame work involving humans, post processing could take a lot longer.
My point was for archival purposes. If you record data at a degree or two higher level than human perception, you can always add whatever post-processing you want to get any kind of artistic effect desired. You want that old jerky 24fps stuff? Fine. Post-process it to get it. People 200 years from now will be able to watch it in 24fps, in black-and-white, in 2-D, or whatever the initial artistic intent. But future audiences and artists might also choose to remix or to see it with more data.
We are currently in a situation akin to knowing how to shoot color movies but refusing to because everybody liked black and white so much. Or being able to record audio but being afraid of putting all those movie pianists out of work.
This kind of thing can be framed up as artistic all day long, but in reality it's just about fear of change -- more to the point, fear of mucking around with your business model too much.
So having learnt that 120FPS is the limit, I'd shoot for 240 or 480FPS.
Also, the days of watching just the raw footage are long over. These days, almost every frame you see in a movie has been heavily post-processed. Quadrupling the frame rate for no reason other than to be safe or to allow others to remix your work in the future means quadrupling render times and, for jobs like rotoscoping, quadrupling people's workload. It just doesn't make any economic sense.
Unless you were talking about shooting just the raw footage with higher resolution and frequency (and downsampling it prior to post-processing), but then what people are remixing isn't your film but your footage. Plus, you still have increased cost for more sophisticated equipment and for storage.
But this issue is not unique to frame rate, or even cinema. We ran into this same problem when HDR took off in still photography. How much of the art is the medium, and how much is the processing?
I think you can argue it either way, but my point was that additional data can always lead to various interpretations later on, while less data always leads to a less-varied range of future possibilities. Speculating on what kinds of post-production work might be done, who might do it, or whether it's a good thing or not misses the point. What used to be part of the medium is now part of the process.
Leonardo would have taken the picture, then painted the painting. What hung in the Louvre would have been up to him as to what he decided to release. Would the painting be worth any less if we had a fully-rendered, highly-detailed model of the studio, subject, and the artist? Not at all, but there are many derivative works we could create with that kind of data that we could never do from just a painting. Think of it in a silly way: if I see a man and draw a stick man, am I locked in forever for only remembering that man in such simple terms? Or might I want to come back and paint him? Why make me choose when the tech lets me have it both ways?
I won't go into the technology/cost issues, as these things have a way of changing dramatically from year to year.
All I was saying is that realistically, and unfortunately, the economic realities of film production make it unlikely that anything will be shot with quality much higher than what the stakeholders can make use of in the short to medium term.
I respect your point and I mostly share the sentiment. Even so; no, da Vinci wouldn't choose what to put in the Louvre. That's the point. We chose what to put in the Louvre, and if da Vinci would have chosen another medium he would have created something else, which might not hang there or anywhere, today.
I'm doing a program where I have great use of Redis for some things, while I could use MongoDB for others. Even so, I choose to use Redis exclusively because I have a feeling that the constraint will help me create a better program.
I really believe in constraints in creation. But of course, if Cameron needs 60 fps he should shot in that!
The individual eye's components do have a frame rate, but they are not synchronized, so the eye as a whole can detect changes smaller than that.
48 fps might be a fine thing, but there are other factors to enjoyment that should be looked at.