A near unsettling moment when you realize the machine was seeing things beyond the space in the photo. Probably the most bizarre yet fascinating tech in the movie.
@@ekathe85 True, but current AI is different than whatever this Blade Runner tech is. In the movie you can see the machine moving around real space that actually exists, whereas our AI can only fill in the blanks by generating an environment it will 'think' is there, which I agree is completely useless for real world detective use. I feel like we're getting pretty close though.
People are working on this photo tech now. The system uses light bouncing from areas normally hidden and projects a very accurate 'echo' if you will, of what objects are relfecting that light - so yes this will be possible - at the moment the trade off is that the resolution is poor if you want to see 'around corners' - but they are working on that as well. And yes - the final shot is a different angle entirely - that's just bad work from the film's continuity team :) One of the best movies ever.
Haha....Actually, one can assume the Esper machine employs machine learning to infer what missing data is most likely to be based upon the existing light field already in the picture, and fills it in, in real time as Deckard navigates through the scene.
Deckard : Enhance 224 to 176. [a man's arm becomes visible] Deckard : Enhance. Stop. [the man's shoulder and wrist are visible] Deckard : Move in. Stop. [close-up of man's wrist] Deckard : Pull out, track right. Stop. [writing is visible] Deckard : Center and pull back. Stop. [arm and door are visible] Deckard : Track 45 right. Stop. Center and stop. [doorway and mirror are visible] Deckard : Enhance 34 to 36. [dresser top is visible] Deckard : Pan right or-and pull back. Stop. [mirror is visible] Deckard : Enhance 34 to 46. [blurred white object in mirror becomes visible] Deckard : Pull back. Wait a minute. Go right. Stop. [Zhora's arm becomes visible] Deckard : Enhance 57 to 19. Track 45 left. Stop. [Zhora is visible] Deckard : Enhance 15 to 23. [marks on Zhora's face become visible] Deckard : Gimme a hard copy right there.
@Deathbringer99699 Deckard was looking at a photo from Leon's apartment. In the photo, Roy Batty is in the tub to the left. But on closer inspection, Deckard finds Zhora in another room and pictures of scales. He had found a scale in Leon's apartment. Deckard was then able to tie in that where he could find the owner of the scale, he could find Zhora.
Yes the hard copy is not the same angle as what's on the screen, however if you compare the hard copy to what's on screen, it appears the Esper took the liberty of rotating her head to get a better angle of the snake tat.
The technology to see around the corner as he does here exists, and can be seen on MIT's website (note that the photo he puts in the device is not just printed, its attached to a disc or tape of some kind that could contain the info needed to accomplish this): web.mit.edu/newsoffice/2012/camera-sees-around-corners-0321.html
Nice find. This is what I was thinking when I posted this response to another person: Actually, one can assume the Esper machine employs machine learning to infer what missing data is most likely to be based upon the existing light field already in the picture, and fills it in, in real time as Deckard navigates through the scene.
Amazing camera 👌😍 I don't like when they do this in shows / movies based in current time which try to be realistic Of course it's perfectly acceptable in this scene since it's Blade Runner
Right at the start of this clip, he looks at the photograph surrounded by musical notation. Before he rights it and puts it in his mouth, the picture appears as a Cyclops.
I see it by u.. and i know it since, when in my time i just was a jounger man. But now, my life is short, but, "you make me beleave", again. Thanks from the frozen ass of this world. Las Condes, Gran Santiago de Chile. Austral América. September, 17. 2023.
It's a bow to a similar scene in a film called Blow-Up (1966 !) where a photographer keeps enlarging images to find something very different than what he thought he was photographing.
but in blow-up, the image becomes both bigger and fuzzier. The closer you get, the more uncertain you get. re-make was "Blow-Out"1981 by Brian de palma. john Travolta makes a field recording of a tire blow-out causing a fatal car accident. Or was it the sound of a rifle shot assassinating a politician?
@SocratesTheGadfly: I used to be so ticked off about the same thing, but look at today's technology... the imperceptible is now easily perceived through DSP. Everything emits/reflects light and casts a signature upon surrounding objects. Is it so hard to think that in 2019 a routine would be written to take samplings or averages of reflected light from 2D captured objects to cast a reliable representation of what is around the corner?
THANK YOU ~ Since I first saw the pic at age 8 or 9, I've ALWAYS said it was a 3D image while everybody told me it was "just zooming in really far" and indeed there is a 2nd reflection of the convex mirror within the angled dressing mirror (which I had always thought was a half-open door) But I find it notable that despite there being a convex mirror, we don't really see a photographer, but we do see hints of what seem to be a remotely activated camera - it's MY feeling or belief if you will, that there are either several fixed cameras around the apartment, or the mirrors themselves either hide lenses behind 'em OR possibly contain sensors to encode the imagery - ALTHOUGH, of course, the MIT "see around corners" device would be more in keeping with a Phillip K. Dick "speculative fiction" piece. It's normal up to the point where he's close-up on her arm, but changes dramatically when he sez "track 45 left" and it seems it's not a number of pixels but rather an ANGLE of incidence and 45-degrees seems about right when you imagine the turn or is it panning I should rather say, taking place at that virtual range distance from the zoom-in enlargement - Everybody I've spoken to while watching the film focuses on the fact that the woman's appearing to the left of the obstruction (sequin dress hanging over dressing mirror - which I had thought was a door - and that to the left of the sequins we were looking into the next room obscured by shadow and poor contrast) But note that when he's looking or "tracking 45 left" the image to the right of the sequins disappears, her elbow disappears then reappears significantly further left (my reason for thinking it was a mirror on the back of a door, perhaps a folding door with two mirrors or a folding dressing screen with mirror panels etc) the doubled elbow is confusing, that if it were indeed a single mirror with an obstruction that tip of her elbow wouldn't protrude as though her arm was ten feet long - So the image to the left has got to be a 2nd generated angle of view. SO - my thoughts are that it could all be a photo COMPOSITE of sorts, where several fixed 2D cameras presumably hanging on walls or mounted atop dressers or shelves or desks against the wall - perhaps all of the walls? - SO - I'm thinking these 2D images would inform a COMPOSITE in a sense, not like a typical 3D photo wherein two lenses stand apart at a distant equivalent to the average span between a person's eyes - But RATHER, this composite could be taken from ANY of these fixed lenses, once it has first established the parameters of the room and where the lenses are shituated within that space ... ??? Just a brain~fart!
LOL I love how he gets a picture of the woman in the bath when it's not actually visible in the first place :P beep beep beep beep beep chukachukachukachuka
First, I apologize for the typo, that's very uncharacteristic - I must have been in a rush or something. Yes, I know film pictures are limited by their grain, but the resolution of a film photo is purely limited by the quality of the film. Assuming you had implausibly flawless film this would be possible (although, in reality, it clearly isn't possible).
@rchealy The look around the corner scene has bugged me too for years, but actually it seems you are right. gizmodo-com/5894972/this-futuristic-camera-can-see-around-corners-using-lasers (just replace the first - with a dot)
dont know if this is the right place to ask,, im watching blade runner right now.. but i stoped after have watched this.. and i dont get it.. who is the girl in the picture?.. whats so important about it?...
I was just saying to myself, this is only possible if the photo is literally 100 megapixels. lol. I love how auto-enhance machine scenes always skip that part, understandable, but, literally a photography hobbyist, that bothers the piss out of me
Hold up. Skipping all of the normal reasons this is preposterous, did the machine actually understand what to do when he goes "hold on a minute"??? WAT.
Couple of things I've been confused about. 1: How did he find Zhora in the photo? Did he follow a reflection around a corner or zoom in really far into the mirror and see her behind whoever took the pic? 2.: Was did the photo do for him anyway? Why did he need to see Roy and Zhora in it? What did it do for him?
1: I think he just saw a mirror and thought "Hey, this ESPER machine works really good with mirrors. Here's a chance to see into the next room." And then he saw that the wardrobe in that room also had some mirrors and thought the same thing again. What he saw when he says "Wait a minute. Go right." I have no idea. Maybe he just had a hunch. 2: I thought he didn't know what Zhora looked like until then. I may be wrong about that.
There's no way to know whether the picture was taken on film or digitally, first of all. Second, film pictures are still limited by their grain. Third, you spelled "know" incorrectly.
you know, when its science fiction, you cant complain about the realism on this. Its probably not even a photograph, as far as you know. Maybe it some silicon based reflecting substance that captured all the photons hitting the camera at the moment of capture and was then interpreted by a machine like the esper. Go whine about enhance in csi videos, THOSE are hilarious.
![เปิด อาณาจักร หมื่นล้าน "สาระตั้ม" l [Nickynachat]](http://i.ytimg.com/vi/3XE0nPF4YkQ/mqdefault.jpg)
A near unsettling moment when you realize the machine was seeing things beyond the space in the photo. Probably the most bizarre yet fascinating tech in the movie.
It wouldn't give a detective anything real to go by, but you can do this right now with AI. Crazy.
@@ekathe85 wdym? the machine straight up give him information the photo didn't initially show, a lead on the woman with the tattooed face.
@@NeonVars I mean you could do this with AI now, in the real world, even though it wouldn't give a detective any reliable clue.
@@ekathe85 True, but current AI is different than whatever this Blade Runner tech is. In the movie you can see the machine moving around real space that actually exists, whereas our AI can only fill in the blanks by generating an environment it will 'think' is there, which I agree is completely useless for real world detective use. I feel like we're getting pretty close though.
People are working on this photo tech now. The system uses light bouncing from areas normally hidden and projects a very accurate 'echo' if you will, of what objects are relfecting that light - so yes this will be possible - at the moment the trade off is that the resolution is poor if you want to see 'around corners' - but they are working on that as well. And yes - the final shot is a different angle entirely - that's just bad work from the film's continuity team :) One of the best movies ever.
That's fucking awesome.
movies have continuity teams?
It’s just odd it could be done for paper prints, not requiring the raw data from the plenoptic camera.
We can do all that photo stuff nowadays on smartphones
When Ridley Scott knew how to make movies! Look at those far-field out-of-focus lighting effects and shadow-play!
@@jcudal32 If he was still able to make good movies he would have done so, at least once, in the last 2 decades. But he has not.
Man, that photo must have amazing dpi.
Haha....Actually, one can assume the Esper machine employs machine learning to infer what missing data is most likely to be based upon the existing light field already in the picture, and fills it in, in real time as Deckard navigates through the scene.
Deckard : Enhance 224 to 176.
[a man's arm becomes visible]
Deckard : Enhance. Stop.
[the man's shoulder and wrist are visible]
Deckard : Move in. Stop.
[close-up of man's wrist]
Deckard : Pull out, track right. Stop.
[writing is visible]
Deckard : Center and pull back. Stop.
[arm and door are visible]
Deckard : Track 45 right. Stop. Center and stop.
[doorway and mirror are visible]
Deckard : Enhance 34 to 36.
[dresser top is visible]
Deckard : Pan right or-and pull back. Stop.
[mirror is visible]
Deckard : Enhance 34 to 46.
[blurred white object in mirror becomes visible]
Deckard : Pull back. Wait a minute. Go right. Stop.
[Zhora's arm becomes visible]
Deckard : Enhance 57 to 19. Track 45 left. Stop.
[Zhora is visible]
Deckard : Enhance 15 to 23.
[marks on Zhora's face become visible]
Deckard : Gimme a hard copy right there.
Plenoptic light-field cameras - look it up - very interesting
The hard copy he prints off isn't even the same angle as the final freeze frame.
Now if this were CSI the enhance feature wouldn't have any limits at all.
@Deathbringer99699 Deckard was looking at a photo from Leon's apartment. In the photo, Roy Batty is in the tub to the left. But on closer inspection, Deckard finds Zhora in another room and pictures of scales. He had found a scale in Leon's apartment. Deckard was then able to tie in that where he could find the owner of the scale, he could find Zhora.
Yes the hard copy is not the same angle as what's on the screen, however if you compare the hard copy to what's on screen, it appears the Esper took the liberty of rotating her head to get a better angle of the snake tat.
The technology to see around the corner as he does here exists, and can be seen on MIT's website (note that the photo he puts in the device is not just printed, its attached to a disc or tape of some kind that could contain the info needed to accomplish this):
web.mit.edu/newsoffice/2012/camera-sees-around-corners-0321.html
Nice find. This is what I was thinking when I posted this response to another person:
Actually, one can assume the Esper machine employs machine learning to infer what missing data is most likely to be based upon the existing light field already in the picture, and fills it in, in real time as Deckard navigates through the scene.
We can do all that photo stuff nowadays on smartphones
Amazing camera 👌😍
I don't like when they do this in shows / movies based in current time which try to be realistic
Of course it's perfectly acceptable in this scene since it's Blade Runner
Still silly it’s done from a photo print, not from raw data from a plenoptic camera.
Right at the start of this clip, he looks at the photograph surrounded by musical notation. Before he rights it and puts it in his mouth, the picture appears as a Cyclops.
I love that someone would take the time to find this video and then dislike it.
One can only assume that future photos encode extra information.
I see it by u.. and i know it since, when in my time i just was a jounger man. But now, my life is short, but, "you make me beleave", again.
Thanks from the frozen ass of this world.
Las Condes, Gran Santiago de Chile.
Austral América.
September, 17. 2023.
This is like Siri, except Esper appears to work.
We need an updated version
Love love love. Like the Mesa scene in 2049. Love!!!
Korg Wavestation years before it was even a thought.
Thats what this soundtrack is giving me.
Amazing movie on so many levels.
It's a bow to a similar scene in a film called Blow-Up (1966 !) where a photographer keeps enlarging images to find something very different than what he thought he was photographing.
yEs.
but in blow-up, the image becomes both bigger and fuzzier. The closer you get, the more uncertain you get.
re-make was "Blow-Out"1981 by Brian de palma. john Travolta makes a field recording of a tire blow-out causing a fatal car accident. Or was it the sound of a rifle shot assassinating a politician?
@SocratesTheGadfly: I used to be so ticked off about the same thing, but look at today's technology... the imperceptible is now easily perceived through DSP. Everything emits/reflects light and casts a signature upon surrounding objects. Is it so hard to think that in 2019 a routine would be written to take samplings or averages of reflected light from 2D captured objects to cast a reliable representation of what is around the corner?
Funny you mentioned that, because Vangelis later used that exact synth on some of his albums.
@andmaketherain But more importantly than that.....he's hunting replicants. He knows she is one if she's living with Leon.
why did this give me the creeps the first time i saw it
reminds me of work aside from all the darkness
THANK YOU ~ Since I first saw the pic at age 8 or 9, I've ALWAYS said it was a 3D image while everybody told me it was "just zooming in really far" and indeed there is a 2nd reflection of the convex mirror within the angled dressing mirror (which I had always thought was a half-open door) But I find it notable that despite there being a convex mirror, we don't really see a photographer, but we do see hints of what seem to be a remotely activated camera - it's MY feeling or belief if you will, that there are either several fixed cameras around the apartment, or the mirrors themselves either hide lenses behind 'em OR possibly contain sensors to encode the imagery - ALTHOUGH, of course, the MIT "see around corners" device would be more in keeping with a Phillip K. Dick "speculative fiction" piece. It's normal up to the point where he's close-up on her arm, but changes dramatically when he sez "track 45 left" and it seems it's not a number of pixels but rather an ANGLE of incidence and 45-degrees seems about right when you imagine the turn or is it panning I should rather say, taking place at that virtual range distance from the zoom-in enlargement - Everybody I've spoken to while watching the film focuses on the fact that the woman's appearing to the left of the obstruction (sequin dress hanging over dressing mirror - which I had thought was a door - and that to the left of the sequins we were looking into the next room obscured by shadow and poor contrast) But note that when he's looking or "tracking 45 left" the image to the right of the sequins disappears, her elbow disappears then reappears significantly further left (my reason for thinking it was a mirror on the back of a door, perhaps a folding door with two mirrors or a folding dressing screen with mirror panels etc) the doubled elbow is confusing, that if it were indeed a single mirror with an obstruction that tip of her elbow wouldn't protrude as though her arm was ten feet long - So the image to the left has got to be a 2nd generated angle of view. SO - my thoughts are that it could all be a photo COMPOSITE of sorts, where several fixed 2D cameras presumably hanging on walls or mounted atop dressers or shelves or desks against the wall - perhaps all of the walls? - SO - I'm thinking these 2D images would inform a COMPOSITE in a sense, not like a typical 3D photo wherein two lenses stand apart at a distant equivalent to the average span between a person's eyes - But RATHER, this composite could be taken from ANY of these fixed lenses, once it has first established the parameters of the room and where the lenses are shituated within that space ... ??? Just a brain~fart!
Gimme a hard copy right there
polaroid metadata. totally doable
Yeah, apparently polaroids are in 3D as well...
stable diffusion in action ;-)
if you hadn't noticed all frame references are the same
Does anyone know the piece of music at the start of the clip?
We can do all that photo stuff nowadays on smartphones
7 replicants downvoted this video.
CSI has nothing on this. Nothing.
LOL I love how he gets a picture of the woman in the bath when it's not actually visible in the first place :P beep beep beep beep beep chukachukachukachuka
“Just print the damn thing!”
Just print the damn thing!
First, I apologize for the typo, that's very uncharacteristic - I must have been in a rush or something.
Yes, I know film pictures are limited by their grain, but the resolution of a film photo is purely limited by the quality of the film. Assuming you had implausibly flawless film this would be possible (although, in reality, it clearly isn't possible).
Eight years late but wouldn't it also be limited by the lens too?
But why is his TV so small?
well it is TEH FEWTCHARR
This is child's play to me. I have a naturally suspicious eye.
@CaptainJen Because he's a lowly detective.
@rchealy The look around the corner scene has bugged me too for years, but actually it seems you are right.
gizmodo-com/5894972/this-futuristic-camera-can-see-around-corners-using-lasers
(just replace the first - with a dot)
@SocratesTheGadfly Check out Jean-Francois Rauzier's website
dont know if this is the right place to ask,,
im watching blade runner right now.. but i stoped after have watched this..
and i dont get it..
who is the girl in the picture?..
whats so important about it?...
That photo must have been shot with a 200 megapixel camera.
I was just saying to myself, this is only possible if the photo is literally 100 megapixels. lol. I love how auto-enhance machine scenes always skip that part, understandable, but, literally a photography hobbyist, that bothers the piss out of me
2:27
Hold up. Skipping all of the normal reasons this is preposterous, did the machine actually understand what to do when he goes "hold on a minute"??? WAT.
Yes. It probably understands plain speaking.
3D picture
thought you might like a spoof i did of this scene with sound effect. thanks for posting your vid of this scene & than you windows moviemaker, phew
im not familiar with the movie, but is he simply scanning an old photograph?
look this
photosynth by microsoft
it's the same!
???
How was such a picture taken? With a drone flying through the entire apartment making google-streetview snapshots of everything?
my question was not who, but how.
That's how they actually created this behind the scenes- only it was with a bunch of 35mm still photos that they pieced together.
@CaptainJen Smaller Faster Stronger Better!!!
#Google #Assistant
....😐.... - ......🙂.......
Couple of things I've been confused about. 1: How did he find Zhora in the photo? Did he follow a reflection around a corner or zoom in really far into the mirror and see her behind whoever took the pic? 2.: Was did the photo do for him anyway? Why did he need to see Roy and Zhora in it? What did it do for him?
1: I think he just saw a mirror and thought "Hey, this ESPER machine works really good with mirrors. Here's a chance to see into the next room." And then he saw that the wardrobe in that room also had some mirrors and thought the same thing again. What he saw when he says "Wait a minute. Go right." I have no idea. Maybe he just had a hunch.
2: I thought he didn't know what Zhora looked like until then. I may be wrong about that.
what bugs me the most is how he could take a picture of so many pixels
why is he using a smal low definition CRT display?
There's no way to know whether the picture was taken on film or digitally, first of all. Second, film pictures are still limited by their grain. Third, you spelled "know" incorrectly.
But, ENHANCE!!!
you know, when its science fiction, you cant complain about the realism on this. Its probably not even a photograph, as far as you know. Maybe it some silicon based reflecting substance that captured all the photons hitting the camera at the moment of capture and was then interpreted by a machine like the esper.
Go whine about enhance in csi videos, THOSE are hilarious.
“Just print the damn thing!”