Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

ARRI 48fps test

Marc Wielage said:
The data storage requirements for a project like this -- especially at 4K -- will kill any post house I know of.

I would strongly suggest you check with the editing, visual effects, 3D, and color correction companies you plan to use to make sure your workflow can be accomplished at this rate.

Bear in mind that many (if not most) Hollywood feature films are still being finished in 2K. 3D doubles the data storage requirements. 4K quadruples the storage. Doubling the frame rate will double it again. The cost and time required to handle this is horrific, certainly beyond any production I've ever worked on.

I think this falls into one of those "interesting ideas that's completely impractical" categories. Don't forget Doug Trumball's years of experiments with Showscan back in the 1970s, several of which were documented in the SMPTE Journal and in American Cinematgrapher. Increased frame rates have been tried before. I'll be very surprised if even Jim Cameron could pull this off for Avatar 2.
Click to expand...

Well, I was suggesting i wanted to see some footage for how it might look in order to decide if it was preferable.

But, that having been said, Moore's law is driving incessantly forward, and what is difficlut today will be common place tomorrow. After all, 4K post pipelines exist today where just a couple of years ago there were none.

A 2D 48fps feature is no more data than a 3D feature.

-sc
 
Jeff,

It would be appreciated if you would take a little more time and list those sources (Or some of them.) now. I'm sure there are some fascinating reads in there. I am also interested to understand how the conclusions were drawn.

My intellectual rationalization has been that we have come to expect 24fps as a part of the film proscenium, and it was evolution that led us there. No doubt science will help us to understand the magic of it, that understanding will make us better manipulators, and cinema will continue to evolve.

I am interested to put these theories into practice. Will audiences accept the higher rates in action sequences, inter-cut with drama of slower rates? Will they find those action sequences more taxing, and therefore more thrilling and real? If we fail to stimulate the fight or flight release of adrenaline, will our action sequences become too taxing and fail with the audience? When shooting higher rates, are there any techniques that we can employ to induce the emotional connection that may be lacking?

It is all fascinating to me.
 
Marc Wielage said:
Increased frame rates have been tried before. I'll be very surprised if even Jim Cameron could pull this off for Avatar 2.
Click to expand...

I wouldn't, not in the least. In fact, I expect that he will only go ahead with Avatar 2 once all the pieces are in place for him to shoot, post and exhibit at 48fps. Knowing Jim, he will want more, he'll say "we need an absolute minimum of 60fps" and he'll end up getting everybody else to compromise on 48. Count on it. The fact that the storage and media handling requirements for something like this would be staggering (staggering being a concept that's constantly loosing relevance these days) does not mean it is impossible, and as long as it is possible, or marginally so, Jim has the clout to demand it and he will get it done. Do you really see a cinema owner not undertaking whatever hardware upgrades are needed to show Avatar 2 (and 3!), which would actually be minimal if they already have digital projection, and miss out on the cash cows those films will be? Once the distribution people are given a deadline, directly from Jim as it has been done before, and they quietly buckle as they have before, then the studios and all production and post people will quickly follow suit. Few people could pull such a coup off, if any outside of Cameron actually, but he definitely can and you can count on him going after it. I'll take bets on this :-)
 
What about video game experiences?

What about video game experiences?

When talking about high-fps experiences I find it amusing that everybody mentions near inaccesible showscan venues and one-time-only limited-audience demos but nobody thinks about near ubiquitous videogames.
For too much time during my teenager years I was involved in high speed competitive PC games. I had a CRT monitor that with the right resolution I could get to display really high refresh rates (120hz). For a passive audience it might be harder to discern the difference between different framerates, but for a critical reflex-driven and precision hand-eye coordinated interaction, 48fps was never nearly enough. 60fps+ was the minimum for confortable play, and 120fps the desirable constant speed. Granted, that was because a physics engine glitch gave you some advantages at that speed, so 120hz just for visual info reasons may be overkill, but I can tell you from personal experience that dropping the fps under 85 became a noticeable loss.
At those speeds the visual stream becomes a window into another reality, the "motion blur" is integrated naturally in your visual system and not with conventional exposure time, and split second decisions flow organically from your conscience to your muscles to the screen. The feeling of inmersion is enough to hook you for hours.

That's why more often than not the results from those research papers seem insufficient to me. They always place the top speed of human perception somewhere between 48-60fps whereas the gaming community I frequented ten years ago considered as common empirical knowledge for this perceivable limit to be 85-100 fps. The lack of motion blur in those games makes it difficult to compare the effect with traditional cinematography, surely the intrinsic motion blur might make it feasible to achieve the same perception of movement with less discrete frames per second.

It's a pity that TFTs became a step backwards on refresh rates, because now all of this is a strange memory of a past life, lacking any scientific rigor.

Although nowadays I'm a firm supporter of traditional 24hz cinematography, those memories still make me very curious about how a fast-paced action scene might look shot and projected at high fps today. Think about car stunts, kung-fu style fights, fighter plane dogfights, or just sweeping camera moves with lots of parallax, etc benefiting from this especially when talking about 3D presentation, where 24hz has shown to be insufficient. (The hardest shots to integrate in Avatar where not because of convergence/interaxial problems but because of the 24hz stutter IMHO). I see the use of this as an extension of the logic that led Ridley Scott to shoot the Gladiator fight scenes with faster shutter speeds. A nearer-to-reality sharper-blade feel for a scene that needs it, only without stutter.

I'd just like it to be easier to experiment with this without access to DCI projectors.
 
It's not that Douglas Trumbull didn't find that audiences could perceive smoother motion above 60 fps when he developed Showscan, it was just that the audience perception of motion improvement and clarity slowed down above 60 fps, peaked at 72 fps, after which any minor improvements weren't offset by the increase in difficulty to shoot and project film at those frame rates. 60 fps was settled on as being more practical than 72 fps I believe.
 
Is Trumbull's research available on the web?
Nice how 72 fps is exactly half of the 144 fps of the DCI projectors, allowing for 3D multiplexing.

Also Jeff, I second Scott's request for those research papers if possible.

Thanks.
 
Daniel,

Good points, really. You don't need DCI compliant projectors for such tests, I've conducted a few myself using 720p and 1080p DLP and LCD projectors, most of which nowadays are 120 hz, and where projecting at 60 fps is very easy. I, for one, LOVE and HATE 24 fps cinematography, I adore the dreamy quality it lends to static images and abhor the stutter and strobing it produces when in motion. However, irrespective of that, 3D is one field where higher frame rates are not just an aesthetic choice but a necessity. Yes, you can make 3D at 24 fps but it looks at its worst, whereas 48 fps and higher really add to the realism of the effect. People forget, or don't know, that our eyes construct depth around objects based on textures, and how the same textures are perceived from two different angles/perspectives. The easier those textures are to see and blend together, the better the stereo effect. Which is why shallow DOf is bad for 3D because whatever is blurred lacks enough textural information to yield a strong depth effect, and why the strobing of 24 fps also destroys the clarity of those textures. Showscan failed not because it wasn't sharper and better than regular 35 mm at 24 fps, but because the public and executives failed to see enough of a difference, and studio execs have always been on the lookout to decrease production costs, not to increase them. IMAX on the other hand succeeded because it added an element that set it apart from regular projection: the size of its screen. Of course the image is better, look at how big it is! Then again, IMAX has not exactly been what I would call a sound success either, more existing screens are being closed down than new ones opened, and the IMAX Lite concept is nothing but a capitulation. Cameron is going to push REALLY hard for 48 fps projection, and he is VERY hard to deny. Then again, will the public see a difference between the first Avatar at 24 fps and the second and third at 48 fps or higher? Probably not, and if they do, it will most likely be attributed to the fact that the technology has gotten much better in the intervening 4 or 5 years before the sequels hit...
 
Buried in a project right now, but will post what I found as soon as I can. But searches on "Visual Processing in the Brain", "Darpa Studies: Tactical Task Saturation", and others may yield further reading. Try: "Managing Tactical Information and Soldier Cognition in the modern Battlespace". I think that was a good one.
 
This is a power-point that I thought was interesting. Kinda touches on the concepts presented earlier. But emerging Mil-Tech that we will see in our more sophisticated products downstream.

Imagine if more of your cognitive bandwidth could be harnessed passively. For example, if you were using a camera and the camera could deliver to you information on its operating parameters in an intuitive data stream (non-conscious) and allowed you a control interface that gave you a high degree of input without interrupting your task concentration.

For example: What if you could smell bad focus? When a subject you are shooting moves outside their pre-determined focus zone, the camera control interface emits a bad smell. Your natural reaction to a foul odor could be the control stimuli that corrects it-- without your even being consciously aware of the problem. ("Your footage stinks..." -- sorry I could not resist.)

That type of thing. Anyway-- fun stuff.

http://www.au.af.mil/au/awc/awcgate/darpa/augmented_cognition.pdf
 
Just to add my two cents about games: as a game developer I was taught to do things by The Book. And The Book so amply says that all games should attempt to run at 60hz to 120hz with a solid, sustained, 60fps. This, however, doesn't always work. So, the settled "minimum" requirement for smooth motion has always been 30fps sustained. This isn't based on any actual research into the subject - it's an empirical estimation that is derived from how things were done in other games before them.

Above 60fps perception of smoothness does, indeed, drop of noticeably. There is no real "need" to run a game at 400fps or something that ridiculous - the eye isn't catching all those extra frames anyway! There have been claims that you can see the difference going from 60 to 100fps - even though in general the law of diminishing returns kicks in violently here and few can see the difference, if any. Going from 24fps to 30 to 60 however...has empirical evidence that the movement is "choppy" and not as smooth until you reach for the higher framerates.

Now, some can argue that this proves that, in the end, 24fps is the ultimate standard - that the FPS wars are doomed just as the resolution wars are. Improvements beyond a certain point "aren't needed".
I, for one, disagree. I think 24p definitely has a legacy going for it and a certain oddball charm. However, I would argue that similar to sensor size, choosing a framerate is an aesthetic decision and not, NOT, because you can chose between "video-y" (which has a bad connotation) or "filmic" (which has a good connotation) image but rather because you can choose whether your movie requires the extra frames or not.
I would also argue that 50fps, 60fps, 30fps etc have a place in the viewers mind. Doing away with "stuttery motion" might actually help quite a few viewers enjoy the film more! I think that viewing a film shot and distributed with a 1000fps would be pointlessly ridiculous - but I would still say that we have a way to go before we see where each framerate has it's place.

In the end, I wish to hell cinema(s) would do away with a "standard". Let the filmmakers decide how they want their viewers to experience the footage and let the creative juices flow. Then when I see the next craptastic mediocre Hollywood film, I can enjoy it in blazing 60fps and pretend it makes a difference :closedeyes:
 
Anybody feel that a small jump would be interesting? Like 26-27-28fps? Many at least like the look of 30fps, and many consider 24 part of the film look, so isn't anyone considering splitting the difference between 24 and 30?

But I agree with Erik. Let the artist decide what the hell he feels is best! And in a digital world that should also allow for variation up to whatever the generally available max framerate projectors would handle! Let me shoot intimate dialogue at 24 then gradually work my way up, or slam right into, whatever I felt like using for and action and the like.
 
Jeff Coatney said:
Buried in a project right now, but will post what I found as soon as I can. But searches on "Visual Processing in the Brain", "Darpa Studies: Tactical Task Saturation", and others may yield further reading. Try: "Managing Tactical Information and Soldier Cognition in the modern Battlespace". I think that was a good one.
Click to expand...

Thanks Jeff.
 
Imagine you shoot 48fps for acquisition, like Cameron wants to. You've got a tool like Graeme's Magic Motion that would allow you to dial in the motion blur for a shot or scene that makes it more pleasing, and avoids the whole pan speed jitter issue. I know Graeme's position here is artistic options, and some people have a negative view on the issue apparently... but we should also think of uses for this idea beyond the initial kneejerk. If Magic Motion allows for motion blur control, 48fps acquisition could be useful in many applications- even if the endgame is 24fps playback,or 30 or 36, or 42...
 
Creative options are one thing, photographic reality is another. The truth is that very few people can really "dial in" an accurate amount of motion blur without an actual photographic reference, and dialing in an incorrect amount usually results in an image that is recognizable by most viewers as being "wrong" in ways they cannot pinpoint. The same thing is true of CG facial simulations in that humans are quick to notice anomalies in things they are used to seeing on an everyday basis, thus they notice "dead eyes" on CG characters, lack of subtle facial muscle movement, and other cues that what they're watching is not "real." This happens very quickly with things like user controlled depth of field simulations and user controlled motion blur. So while it may sound good to have "creative options" for these things, it is not something that should be done by the casual shooter or editor.

Not to mention that I still believe that anything that can be photographed should be photographed, not simulated, as by definition "simulating" is attempting to create the real thing. What's wrong with actually photographing the real thing if that's what you want in the first place?
 
I think since we're talking about ARRI and Cameron that we're talking about professional end users. I know I am.
 
XKCD said it best: "We're also stuck with blurry, juddery, slow-panning 24fps movies forever because (thanks to 60fps home video) people associate high framerates with camcorders and cheap sitcoms, and thus think good framerates look 'fake'."

Every reason anyone gives anywhere ever against higher framerates comes from associations people have become accustomed to, not any kind of objective aesthetic judgement.

Personally I don't have a problem with 24, but I don't see any point in pretending there are any real reasons to stick to it. If we changed all of a sudden, people would complain for a couple months and then get used to it. Just like every time anyone tries to change anything.
 
I don't think it's a standard most will like. I'm not saying we shouldn't be open to it, but a test is different from a feature length film. Like 3D it's a tool and it can and would be abused. I think there are technical benefits that could arise from 48fps acquisition- HDRx and Magic Motion are great sort of back room examples because the intent is always a 24 end product- but putting 48fps into projection is a different beast. I think it might create an uncanny valley effect and backfire. Luckily for Cameron, he doesn't really have to *shoot* most of AVATAR 2 in 48fps because of the nature of the film. If he changes his mind later, he can extrapolate 24fps from the 48fps footage, dial in more standard motion blur and then back off the virtual cameras to 24 rather than 48. Clicks of a mouse. Shooting 48 instead of 60 would also build him a better back door mathematically, but we'll see.
 
I have no problem with higher frame rates for theatrical presentation as long as I am not forced to shoot it. 48 fps is fine but if I wish to make a film shot at 24 fps, I should still be able to do so. I hope these projector upgrades will not lock in the frame rates should theaters and studios decide to adopt higher frame rates as an OPTION, not a standard.
 
ChrisLyon said:
I think since we're talking about ARRI and Cameron that we're talking about professional end users. I know I am.
Click to expand...

So am I. The best animators in the world aree still not able to crack the "uncanny valley" wall with simulated human characters, as evidenced by recent projects like "Tron." I wasn't referring only to indie do it yourselfers in my comment. Motion blur is not something to be played with except when needed for a specific effect. It is rather a natural consequence of shuttered photography and something that "works" for most viewers when it's right, and doesn't when it's not. Simulating it is not the same as capturing the real thing, and there will likely be many projects in which this is obvious if things like Magic Motion are widely adopted and abused.

And I haven't even mentioned the potential motion artifacts that have been known to appear in motion blur simulations....
 
You must log in or register to reply here.
Back
Top