Ahh, yes. I totally missed those details. Maybe it's just that the following got misinterpreted then:
For DreamWorks, the amount of data needed to create a stereoscopic film leaped by 30%."
After giving it some thought, it could be that their new stereo pipeline consumes more disk space than their old one... That would be expected if they increased their output resolution or used more passes in their comp process. Either way it just sounds more like a "bigger and better!" type of number thrown out there for PR reasons rather than a clever optimization.
Glad to hear it. I mostly use multi-layer EXR files for my work and those are disk hungry as well, as you probably know (up to ~250 MB per frame in my case). But since there's no need to have 1 image set per eye, I was somehow expecting deep images to consume less space... Thanks for the tip!
Instead of rendering images for both eyes, you render a "deep" image that contains depth information. The images for each eye are then written out of post production rather than out of the CG software.
Stereoscopic vision effortlessly focuses on objects we look at with both eyes; it takes a conscious effort to focus on an arbitrary point that isn't seen by both eyes. As an example, put a finger in front of one eye in a way that the other eye doesn't see it, just as is the case with Glass (close the "finger" eye to check). Now of you try to focus on your finger, you'll notice that your vision doesn't naturally do so ; the natural reflex is to move your head back. Since you can't move your head in this case, focus needs to be forced in place and has a hell of time getting a clear view of the finger.
The Apache systems completely replaces the field of view of the targeting eye and is designed to work alongside binocular vision, overlaying data atop what is seen by both eyes ; albeit in different colors (augmented reality). The perspective remains the same for both eyes though.
The problem with Glass seems to be in forcing a spatially unrelated image onto one eye forcing the focus to shift from from the environment to the Glass display, the strain coming from the other eye having to focus somewhere in mid-air. That's unnatural and needs to be forced without a distinct object to look at.
Producers do own the copyright if they put out novel content, although that rarely happens outside of megacorps for the following reasons:
It's useless to be based outside of major metro areas as all the crew talent is there. Major US cities and surrounding ares cost a lot to live in compared to the mid-west or anywhere else except western Europe ; that implies that crews have to get paid a minimum of $30k/year. As a US business, one MUST have all software licenses to operate without getting squashed by a lawsuit from media software companies, and depending on the type of artist and production a SINGLE licensed computer can cost anywhere from $10k to $25k+. Add network and server costs on top of that as the volume and size of the databases prevent them from being decentralized (see here: http://slashdot.org/comments.pl?sid=3355921&cid=42465107). Compared to the rest of the world, cost of living and licenses here make production costs the highest of everywhere else.
So, in conclusion, piracy has two effects on my line of work : artist and technical crews get paid shit, have no benefits and will be quickly replaced by younger personnel in time because income gets eroded by downloads. And, on top of that, no-one else in the world has to worry about running unlicensed software, reducing costs by at least 50%. As a small upstarting digital media producer, I can't compete with that. Now, if fellow./ers don't have a problem with megacorps and lawyers being the only people who can afford a living in the US, then by all means, keep pirating your content and software!
Measuring the pendulum's motion in 3D implies having at least stereoscopic vision. While I agree that specific hardware can be made to compute the proper flight once the position and velocity vectors of the pendulum are determined, the quadrucopter would still need to carry the gear required to acquire the pendulum.
As far as I can tell, that would at least require a 360 radar for object position detection and a couple of high speed, high resolution cameras mounted on 360 degree capable, fast moving tracking system. That part is what seems to be the most difficult, and I don't see that amount of gear mounted on this lightweight model without severe speed and agility penalties.
This looks like it was achieved using motion capture equipment, as seen by the usage of retro-reflective balls (see them shine at 1:44). If that's the case, it would imply that the computing was done remotely. Motion capture gear typically works at 120 frames per second, which would give the system enough time resolution to figure this out on the fly with a decent PC on the back-end.
It'll be a while before the quadrucopters can do this in their own but the program figuring out the catching move is impressive nonetheless.
What is more important, in the LONG haul? The space 'race' will be more like a long drawn out marathon punctuated with the occasional leap forward, which would be better news than what we're hearing on a daily basis...
Some will argue that the science or the technology isn't there yet. Well, we still need to get started somewhere. Look at the discovery and settlement of the american continent: from the day of first setting foot in 1492 to having an independent government established, how much time has passed? Using 1789 as a reference, that's nearly 300 years. Look at the numerous innovations that happened in between ; not to mention the numerous failures, human or otherwise. But all things considered, it was a worthwhile fight and we're all much better off today thanks to the efforts of our predecessors.
Achieving the conquest of space is the greatest challenge of humanity has yet to face. It will take a change of perspective as the planning timescales will shift from years to decades, if not centuries. But look at it this way, will we all not be better off in the end? Won't the effort bring many new discoveries and economic stability in the mean time..?
It is said that the first step in any endeavor is the most difficult. I, for one, vote to take it. Enough with the political bickering already! We've got better things to do as an intelligent species.
It might even be enough to prompt a major "recall" of US companies, in which case it might become interesting. On top of that, bringing HQs back would also help employment as well.
Since these companies are based in Ireland for tax purposes, that might indeed end up happening...
On the other hand, that could also force said companies' tax dollars back into the US which wouldn't hurt at this point...
This kind of reminds me Half-life 2 where the L.E. to civilian population ratio is something like 50%.
If that's the direction we're heading , I wonder who'll be left to grow food, build houses, make furniture... Who'll take care of the monster once the monster has gotten rid of it's "foes".
I certainly wish that was the case. A lot of Sci-Fi masterpieces (Kubrik, Star Trek, Tron...) have encouraged me to care and learn more about science and can't say how proud I am of what human ingenuity can achieve.
However, to my great dismay I've noticed that a love sci-fi requires both a good imagination and a decent education ; which often come as a pair but are hardly a standard feature of the majority of peoples' lives. If it weren't for all the dick waving contests, I'm sure we'd be much further along by now and all that needs to be done is to fix education (I have high hopes for web-based schooling). Better education leads to better cultural advancement and overall human progress.
-- Nothing is enough for whom enough is too little - Confucius
I can't be sure but I'd wager that it has something to do with the camera technology. See here:
http://www.red.com/products/scarlet#tech-specs
The more pixels you have in the live footage, the easier it is to get clean results when adding the VFX. Think of it this way: it's easy to scale down a large digital frame, but it'll look like crap if you do the opposite since we're dealing with a pixel grid rather than analog film stock.
Actually, it's all the data required as input to make the *final* frames. We're talking many layers of video at 32 bit per channel (128bit images), VFX cache data which can be GBs per second of footage, thousands of textures that are also GBs in size, point clouds... All of that is meant to retain a maximum amount of flexibility before finalizing the footage. Read up on the REYES pipeline for detailed info.
Disclaimer: Film and animation professional and professor.
Slashdot Mirror
Ahh, yes. I totally missed those details. Maybe it's just that the following got misinterpreted then:
After giving it some thought, it could be that their new stereo pipeline consumes more disk space than their old one... That would be expected if they increased their output resolution or used more passes in their comp process. Either way it just sounds more like a "bigger and better!" type of number thrown out there for PR reasons rather than a clever optimization.
Glad to hear it. I mostly use multi-layer EXR files for my work and those are disk hungry as well, as you probably know (up to ~250 MB per frame in my case). But since there's no need to have 1 image set per eye, I was somehow expecting deep images to consume less space... Thanks for the tip!
I can't be certain that this is the technique Dreamworks used, but it makes sense and would save disk space:
http://en.wikipedia.org/wiki/Deep_image_compositing
Instead of rendering images for both eyes, you render a "deep" image that contains depth information. The images for each eye are then written out of post production rather than out of the CG software.
Wouldn't that just push the problem further by a few degrees?
I am gross and perverted, obsessed and deranged
I've existed for years but very little has changed
I'm the tool of the government and industry too
For I'm destined to rule and regulate you
I may be vile and pernicious, but you can't look away
I make you think I'm delicious with the stuff that I say
I am the best you can get, have you guessed me , yet?
Well I'm the slime oozing out of your TV set!
You will obey me while I lead you and eat the garbage that I feed you
Until the day that we don't need you, don't got for help, no one will heed you
Your mind is totally controlled, it has been stuffed into my mold
And you will do as you are told until the rights to you are sold...
Stereoscopic vision effortlessly focuses on objects we look at with both eyes; it takes a conscious effort to focus on an arbitrary point that isn't seen by both eyes. As an example, put a finger in front of one eye in a way that the other eye doesn't see it, just as is the case with Glass (close the "finger" eye to check). Now of you try to focus on your finger, you'll notice that your vision doesn't naturally do so ; the natural reflex is to move your head back. Since you can't move your head in this case, focus needs to be forced in place and has a hell of time getting a clear view of the finger.
The Apache systems completely replaces the field of view of the targeting eye and is designed to work alongside binocular vision, overlaying data atop what is seen by both eyes ; albeit in different colors (augmented reality). The perspective remains the same for both eyes though.
The problem with Glass seems to be in forcing a spatially unrelated image onto one eye forcing the focus to shift from from the environment to the Glass display, the strain coming from the other eye having to focus somewhere in mid-air. That's unnatural and needs to be forced without a distinct object to look at.
I'm all for it. Either that or at least remove the regional monopolies these guys abuse all too often.
Producers do own the copyright if they put out novel content, although that rarely happens outside of megacorps for the following reasons :
It's useless to be based outside of major metro areas as all the crew talent is there. Major US cities and surrounding ares cost a lot to live in compared to the mid-west or anywhere else except western Europe ; that implies that crews have to get paid a minimum of $30k/year. As a US business, one MUST have all software licenses to operate without getting squashed by a lawsuit from media software companies, and depending on the type of artist and production a SINGLE licensed computer can cost anywhere from $10k to $25k+. Add network and server costs on top of that as the volume and size of the databases prevent them from being decentralized (see here: http://slashdot.org/comments.pl?sid=3355921&cid=42465107). Compared to the rest of the world, cost of living and licenses here make production costs the highest of everywhere else.
So, in conclusion, piracy has two effects on my line of work : artist and technical crews get paid shit, have no benefits and will be quickly replaced by younger personnel in time because income gets eroded by downloads. And, on top of that, no-one else in the world has to worry about running unlicensed software, reducing costs by at least 50%. As a small upstarting digital media producer, I can't compete with that. Now, if fellow ./ers don't have a problem with megacorps and lawyers being the only people who can afford a living in the US, then by all means, keep pirating your content and software!
Measuring the pendulum's motion in 3D implies having at least stereoscopic vision. While I agree that specific hardware can be made to compute the proper flight once the position and velocity vectors of the pendulum are determined, the quadrucopter would still need to carry the gear required to acquire the pendulum.
As far as I can tell, that would at least require a 360 radar for object position detection and a couple of high speed, high resolution cameras mounted on 360 degree capable, fast moving tracking system. That part is what seems to be the most difficult, and I don't see that amount of gear mounted on this lightweight model without severe speed and agility penalties.
This looks like it was achieved using motion capture equipment, as seen by the usage of retro-reflective balls (see them shine at 1:44). If that's the case, it would imply that the computing was done remotely. Motion capture gear typically works at 120 frames per second, which would give the system enough time resolution to figure this out on the fly with a decent PC on the back-end.
It'll be a while before the quadrucopters can do this in their own but the program figuring out the catching move is impressive nonetheless.
Lie detector tests aren't considered valid evidence in France.
https://fr.wikipedia.org/wiki/D%C3%A9tecteur_de_mensonge
What is more important, in the LONG haul? The space 'race' will be more like a long drawn out marathon punctuated with the occasional leap forward, which would be better news than what we're hearing on a daily basis...
Some will argue that the science or the technology isn't there yet. Well, we still need to get started somewhere. Look at the discovery and settlement of the american continent: from the day of first setting foot in 1492 to having an independent government established, how much time has passed? Using 1789 as a reference, that's nearly 300 years. Look at the numerous innovations that happened in between ; not to mention the numerous failures, human or otherwise. But all things considered, it was a worthwhile fight and we're all much better off today thanks to the efforts of our predecessors.
Achieving the conquest of space is the greatest challenge of humanity has yet to face. It will take a change of perspective as the planning timescales will shift from years to decades, if not centuries. But look at it this way, will we all not be better off in the end? Won't the effort bring many new discoveries and economic stability in the mean time..?
It is said that the first step in any endeavor is the most difficult. I, for one, vote to take it. Enough with the political bickering already! We've got better things to do as an intelligent species.
That 40 year old movie predicted this problem spot on:
https://www.youtube.com/watch?v=HFvT_qEZJf8
http://arstechnica.com/security/2008/10/aclu-23-of-us-population-lives-in-constitution-free-zone/
It might even be enough to prompt a major "recall" of US companies, in which case it might become interesting. On top of that, bringing HQs back would also help employment as well.
A bucket is filled with many drops, is it not?
Since these companies are based in Ireland for tax purposes, that might indeed end up happening...
On the other hand, that could also force said companies' tax dollars back into the US which wouldn't hurt at this point...
The hell it's not!
Malware developers...
That sounds like the Randian utopia, but with the bad stuff swept under the carpet. Progress indeed.
This kind of reminds me Half-life 2 where the L.E. to civilian population ratio is something like 50%.
If that's the direction we're heading , I wonder who'll be left to grow food, build houses, make furniture... Who'll take care of the monster once the monster has gotten rid of it's "foes".
I certainly wish that was the case. A lot of Sci-Fi masterpieces (Kubrik, Star Trek, Tron...) have encouraged me to care and learn more about science and can't say how proud I am of what human ingenuity can achieve.
However, to my great dismay I've noticed that a love sci-fi requires both a good imagination and a decent education ; which often come as a pair but are hardly a standard feature of the majority of peoples' lives. If it weren't for all the dick waving contests, I'm sure we'd be much further along by now and all that needs to be done is to fix education (I have high hopes for web-based schooling). Better education leads to better cultural advancement and overall human progress.
--
Nothing is enough for whom enough is too little - Confucius
I can't be sure but I'd wager that it has something to do with the camera technology. See here:
http://www.red.com/products/scarlet#tech-specs
The more pixels you have in the live footage, the easier it is to get clean results when adding the VFX. Think of it this way: it's easy to scale down a large digital frame, but it'll look like crap if you do the opposite since we're dealing with a pixel grid rather than analog film stock.
Actually, it's all the data required as input to make the *final* frames. We're talking many layers of video at 32 bit per channel (128bit images), VFX cache data which can be GBs per second of footage, thousands of textures that are also GBs in size, point clouds... All of that is meant to retain a maximum amount of flexibility before finalizing the footage. Read up on the REYES pipeline for detailed info.
Disclaimer: Film and animation professional and professor.
Not to mention crowd-sourcing the story development thanks to slashdoters' broad imaginations...