5 ms·

The real paradox is that we think that we see the world as how it is, like a camera does, when in fact what we "see" is reconstruction based on surprisingly imp

The real paradox is that we think that we see the world as how it is, like a camera does, when in fact what we "see" is reconstruction based on surprisingly imperfect visual perception and equally imperfect recollection of prior experiences.

the_af 4y ago
A camera doesn't "see" the world "how it is".
- lostgame 4y ago
  Depending the sensor or array of sensors, cameras arguably see more of the world 'how it is' than humans are capable of. Night Vision, Heat Maps, LIDAR, and Telescopic lenses can capture more of the world 'as it is' than humans could ever hope to achieve. It's why we developed them! :)
  - the_af 4y ago
    Hard disagree. Cameras and their sensors are just different from the human eye. This doesn't make them capture reality "how it is", they just have different capabilities and they make different trade-offs. They are certainly different, but not necessarily "more real". Cameras don't capture "reality". An insightful link on this topic, shared previously here on HN, which is very illuminating: "Photography is not objective" [1] [1] https://aaronhertzmann.com/2022/03/17/photography-is-not-objective.html https://aaronhertzmann.com/2022/03/17/photography-is-not-obj...
capableweb 4y ago
> reconstruction based on surprisingly imperfect visual perception Unlike a camera? ;) > equally imperfect recollection of prior experiences I guess we (humans) put our understanding of the world into the mechanics of the camera, so somehow I see the mechanics of a camera like a summary of human "recollection of prior experiences" regarding visual imperfection.
- andyjohnson0 4y ago
  > Unlike a camera? ;) Cameras have higher overall resolution than the eye, as well as it being consistent across the sensor surface (unlike the eye). Eyes have pretty good dynamic range, but DSLRs are definitely catching up. Same with sensitivity. So I'm not sure whether I'd describe a camera as being all that imperfect. But the point I was making is that we mostly navigate the world unconsciously believing that we perceive it directly via the light that hits our eyes. But we actually "see" an unfolding reconstruction that our brain pieces together from evidence contributed by our eyes and memories. If your vision is good, and especially when you're young, it's easy to be completely unconscious of the actuality of vision. > I guess we (humans) put our understanding of the world into the mechanics of the camera, so somehow I see the mechanics of a camera like a summary of human "recollection of prior experiences" regarding visual imperfection. Interesting question. I wonder if photography changed the way that people think of their vision - compared to previous methods of visual capture like painting or drawing? I suspect not but I don't know.
- strogonoff 4y ago
  Camera captures light values of a scene at a specific moment in time. While it’s better at that than your eye, that is vastly different from perceiving reality over time. Which is, incidentally, the only way humans perceive reality. The way you perceive a scene to large extent depends on experiences you have had immediately before that (as well as in all of your prior life), your mood, what the air smells like, whom you are with, and so on. If you sit in a grey room for a long period of time and step into a garden, the colors will jump at you. If you grew up in a small town and stepped onto a rooftop in the middle of a modern megalopolis for the first time, you might initially be stunned by the amount of detail hitting you all at once; if you lived and seen that image every other day for years, you’d be inclined to filter out a lot and focus on what changed. You would not be able to be tricked by your mind into thinking the moon is surprisingly large when it’s near the horizon have you not had the prior experience of seeing the same moon overhead. So on, so forth. Furthermore, the “objective” light values captured by the camera are fairly useless by themselves. We have to trust bundled software, film manufacturers, or indeed our own skills to interpret those raw values in terms of display space, and hopefully produce something resembling what we have experienced in the moment. As an aside, it’s often frustrating how some believe that the output of camera’s JPEG engine at default settings is how reality is. It irks me when someone characterizes a dull[0] photo as “true to life”, and a photo interpreted to make the most use of P3 color space as “edited”, implying that it’s no longer representing reality objectively. There’s no objective representation of a single moment in time and there’s no useful reference point here; both outcomes are “edited”, but perhaps in the latter case the photographer tried to convey how they felt in the moment rather than defer to the view of whoever engineered their film or camera’s JPEG engine. [0] A film or camera engine, while designed by undoubtedly skilled people, is optimized to provide acceptable outcomes for most diverse scenes and doesn’t really know much about final output medium, and as a result the outcomes tend to make a very conservative use of available color space. Addendum: and yes, I do suspect that this aspect of photography has influenced (and not in a way I would characterize as “good”) our perception of reality and how we think about our vision.
leobg 4y ago
Which is probably why art exists. Sometimes a painting seems to capture reality better than a photograph.
- lostgame 4y ago
  Still frames of UE5 are certainly getting there. (Put it in motion, and that's a different story for sure...)
bsenftner 4y ago
Not only that, but human vision has a variable zoom - objects we focus upon receive a zoom, an enlargement distorting physical size. The human vision variable zoom is similar to the "3D hat" of a sine curve. The Fovo Renderer people discuss this here: https://www.fovotec.com/the-fovorender-story-part-1 https://www.fovotec.com/the-fovorender-story-part-1
smokel 4y ago
I sometimes play with a simple experiment that I think illustrates this quite nicely: Cover one eye and look at something, then uncover the eye, and notice the added "depth" information. Personally, I don't observe the depth information from stereo vision in a way that I can easily describe in terms of a (single) photographic image. Apparently the images from the two separate eyes are combined after processing some of the depth information. My consciousness does not seem to have direct access to the entire pipeline. (I made some goggles to experiment for longer durations. Great fun.) Similar things apply to face recognition, instantly comprehending words from letters, mathematical formulas (possibly only for trained mathematicians), and perhaps even fear of spidery images.
- nowahe 4y ago
  You might be interested by the book "Consciousness and the Brain" by Stanislas Dehaene. In it, Dehaene tries to define Consciousness, based on a series of reproducible experiments designed to discriminate between conscious and unconscious processes. For example, he is able to show that our unconscious processes that are responsable for parsing text and numbers are able to assign conceptual meaning to words/numbers (ie "five" and "5" lead to the same meaning), before any conscious awareness. Although my summary hardly does it justice (as I'm still half-way through reading it), I highly recommend Scott Alexender's book review on Astral Codex Ten[0], it gives an exhaustive overview of the book and the subjects it covers. [0]: https://astralcodexten.substack.com/p/your-book-review-consciousness-and https://astralcodexten.substack.com/p/your-book-review-consc...
- xamuel 4y ago
  Your covering-one-eye thing works particularly nicely if you use it in a place where you can see the tops of lots of trees at various distances overlapping each other. Cover one eye and suddenly all those canopies become one depthless blob, like it's one gigantic treetop. It sounds like you might be interested in experiments with deliberately redirecting your eyes. There's lots of interesting phenomena you can achieve by pointing one eye at one thing and another eye at another thing. Here's a paper of mine which can serve as a sort of stepping stone into this wide unexplored field: https://philpapers.org/archive/ALEDSK.pdf https://philpapers.org/archive/ALEDSK.pdf
hprotagonist 4y ago
> when in fact what we "see" is reconstruction based on surprisingly imperfect visual perception … like a camera does :)
- andyjohnson0 4y ago
  No. Nothing like a camera does. https://www.cambridgeincolour.com/tutorials/cameras-vs-human-eye.htm https://www.cambridgeincolour.com/tutorials/cameras-vs-human...
  - IncRnd 4y ago
    Yes, like a camera. You are confusing what we "see", with interpretations that don't occur in the eyes. That's even why you put the word "see" in quotes. You used the wrong word. The word "see" explicity refers to what is observed by the eyes, not our later interpretations.
    - mtVessel 4y ago
      No, it doesn't. "See" refers to our visual perception, whether it occurs in the eyes, the brain, or anywhere in between.
      - IncRnd 4y ago
        From the standpoint of dictionaries (and history, usage, and etymology), "see" refers to the eye. Below are the first entries in three different dictionaries and an etymological entry. All refer to the eye and not the brain or other intermediate locations. Dictionaries: merriam webster: to perceive by the eye [1] dictionary.com: to perceive with the eyes; look at [2] cambridge: to be conscious of what is around you by using your eyes: [3] Etymology: eymonline: Middle English sēn, from Old English seon (Anglian sean) "be or become aware of by means of the eye; look, behold;" and It is attested by late Old English as "be able to see with the eyes, have the faculty of sight, not be blind." [4] [1] https://www.merriam-webster.com/dictionary/see https://www.merriam-webster.com/dictionary/see [2] https://www.dictionary.com/browse/see https://www.dictionary.com/browse/see [3] https://dictionary.cambridge.org/us/dictionary/english/see https://dictionary.cambridge.org/us/dictionary/english/see [4] https://www.etymonline.com/word/see https://www.etymonline.com/word/see

the_af 4y ago

A camera doesn't "see" the world "how it is".

lostgame 4y ago
Depending the sensor or array of sensors, cameras arguably see more of the world 'how it is' than humans are capable of. Night Vision, Heat Maps, LIDAR, and Telescopic lenses can capture more of the world 'as it is' than humans could ever hope to achieve. It's why we developed them! :)
- the_af 4y ago
  Hard disagree. Cameras and their sensors are just different from the human eye. This doesn't make them capture reality "how it is", they just have different capabilities and they make different trade-offs. They are certainly different, but not necessarily "more real". Cameras don't capture "reality". An insightful link on this topic, shared previously here on HN, which is very illuminating: "Photography is not objective" [1] [1] https://aaronhertzmann.com/2022/03/17/photography-is-not-objective.html https://aaronhertzmann.com/2022/03/17/photography-is-not-obj...

capableweb 4y ago

> reconstruction based on surprisingly imperfect visual perception Unlike a camera? ;) > equally imperfect recollection of prior experiences I guess we (humans) put our understanding of the world into the mechanics of the camera, so somehow I see the mechanics of a camera like a summary of human "recollection of prior experiences" regarding visual imperfection.

andyjohnson0 4y ago
> Unlike a camera? ;) Cameras have higher overall resolution than the eye, as well as it being consistent across the sensor surface (unlike the eye). Eyes have pretty good dynamic range, but DSLRs are definitely catching up. Same with sensitivity. So I'm not sure whether I'd describe a camera as being all that imperfect. But the point I was making is that we mostly navigate the world unconsciously believing that we perceive it directly via the light that hits our eyes. But we actually "see" an unfolding reconstruction that our brain pieces together from evidence contributed by our eyes and memories. If your vision is good, and especially when you're young, it's easy to be completely unconscious of the actuality of vision. > I guess we (humans) put our understanding of the world into the mechanics of the camera, so somehow I see the mechanics of a camera like a summary of human "recollection of prior experiences" regarding visual imperfection. Interesting question. I wonder if photography changed the way that people think of their vision - compared to previous methods of visual capture like painting or drawing? I suspect not but I don't know.
strogonoff 4y ago
Camera captures light values of a scene at a specific moment in time. While it’s better at that than your eye, that is vastly different from perceiving reality over time. Which is, incidentally, the only way humans perceive reality. The way you perceive a scene to large extent depends on experiences you have had immediately before that (as well as in all of your prior life), your mood, what the air smells like, whom you are with, and so on. If you sit in a grey room for a long period of time and step into a garden, the colors will jump at you. If you grew up in a small town and stepped onto a rooftop in the middle of a modern megalopolis for the first time, you might initially be stunned by the amount of detail hitting you all at once; if you lived and seen that image every other day for years, you’d be inclined to filter out a lot and focus on what changed. You would not be able to be tricked by your mind into thinking the moon is surprisingly large when it’s near the horizon have you not had the prior experience of seeing the same moon overhead. So on, so forth. Furthermore, the “objective” light values captured by the camera are fairly useless by themselves. We have to trust bundled software, film manufacturers, or indeed our own skills to interpret those raw values in terms of display space, and hopefully produce something resembling what we have experienced in the moment. As an aside, it’s often frustrating how some believe that the output of camera’s JPEG engine at default settings is how reality is. It irks me when someone characterizes a dull[0] photo as “true to life”, and a photo interpreted to make the most use of P3 color space as “edited”, implying that it’s no longer representing reality objectively. There’s no objective representation of a single moment in time and there’s no useful reference point here; both outcomes are “edited”, but perhaps in the latter case the photographer tried to convey how they felt in the moment rather than defer to the view of whoever engineered their film or camera’s JPEG engine. [0] A film or camera engine, while designed by undoubtedly skilled people, is optimized to provide acceptable outcomes for most diverse scenes and doesn’t really know much about final output medium, and as a result the outcomes tend to make a very conservative use of available color space. Addendum: and yes, I do suspect that this aspect of photography has influenced (and not in a way I would characterize as “good”) our perception of reality and how we think about our vision.

leobg 4y ago

Which is probably why art exists. Sometimes a painting seems to capture reality better than a photograph.

lostgame 4y ago
Still frames of UE5 are certainly getting there. (Put it in motion, and that's a different story for sure...)

bsenftner 4y ago

Not only that, but human vision has a variable zoom - objects we focus upon receive a zoom, an enlargement distorting physical size. The human vision variable zoom is similar to the "3D hat" of a sine curve. The Fovo Renderer people discuss this here: https://www.fovotec.com/the-fovorender-story-part-1 https://www.fovotec.com/the-fovorender-story-part-1

smokel 4y ago

I sometimes play with a simple experiment that I think illustrates this quite nicely: Cover one eye and look at something, then uncover the eye, and notice the added "depth" information. Personally, I don't observe the depth information from stereo vision in a way that I can easily describe in terms of a (single) photographic image. Apparently the images from the two separate eyes are combined after processing some of the depth information. My consciousness does not seem to have direct access to the entire pipeline. (I made some goggles to experiment for longer durations. Great fun.) Similar things apply to face recognition, instantly comprehending words from letters, mathematical formulas (possibly only for trained mathematicians), and perhaps even fear of spidery images.

nowahe 4y ago
You might be interested by the book "Consciousness and the Brain" by Stanislas Dehaene. In it, Dehaene tries to define Consciousness, based on a series of reproducible experiments designed to discriminate between conscious and unconscious processes. For example, he is able to show that our unconscious processes that are responsable for parsing text and numbers are able to assign conceptual meaning to words/numbers (ie "five" and "5" lead to the same meaning), before any conscious awareness. Although my summary hardly does it justice (as I'm still half-way through reading it), I highly recommend Scott Alexender's book review on Astral Codex Ten[0], it gives an exhaustive overview of the book and the subjects it covers. [0]: https://astralcodexten.substack.com/p/your-book-review-consciousness-and https://astralcodexten.substack.com/p/your-book-review-consc...
xamuel 4y ago
Your covering-one-eye thing works particularly nicely if you use it in a place where you can see the tops of lots of trees at various distances overlapping each other. Cover one eye and suddenly all those canopies become one depthless blob, like it's one gigantic treetop. It sounds like you might be interested in experiments with deliberately redirecting your eyes. There's lots of interesting phenomena you can achieve by pointing one eye at one thing and another eye at another thing. Here's a paper of mine which can serve as a sort of stepping stone into this wide unexplored field: https://philpapers.org/archive/ALEDSK.pdf https://philpapers.org/archive/ALEDSK.pdf

hprotagonist 4y ago

> when in fact what we "see" is reconstruction based on surprisingly imperfect visual perception … like a camera does :)

andyjohnson0 4y ago
No. Nothing like a camera does. https://www.cambridgeincolour.com/tutorials/cameras-vs-human-eye.htm https://www.cambridgeincolour.com/tutorials/cameras-vs-human...
- IncRnd 4y ago
  Yes, like a camera. You are confusing what we "see", with interpretations that don't occur in the eyes. That's even why you put the word "see" in quotes. You used the wrong word. The word "see" explicity refers to what is observed by the eyes, not our later interpretations.
  - mtVessel 4y ago
    No, it doesn't. "See" refers to our visual perception, whether it occurs in the eyes, the brain, or anywhere in between.
    - IncRnd 4y ago
      From the standpoint of dictionaries (and history, usage, and etymology), "see" refers to the eye. Below are the first entries in three different dictionaries and an etymological entry. All refer to the eye and not the brain or other intermediate locations. Dictionaries: merriam webster: to perceive by the eye [1] dictionary.com: to perceive with the eyes; look at [2] cambridge: to be conscious of what is around you by using your eyes: [3] Etymology: eymonline: Middle English sēn, from Old English seon (Anglian sean) "be or become aware of by means of the eye; look, behold;" and It is attested by late Old English as "be able to see with the eyes, have the faculty of sight, not be blind." [4] [1] https://www.merriam-webster.com/dictionary/see https://www.merriam-webster.com/dictionary/see [2] https://www.dictionary.com/browse/see https://www.dictionary.com/browse/see [3] https://dictionary.cambridge.org/us/dictionary/english/see https://dictionary.cambridge.org/us/dictionary/english/see [4] https://www.etymonline.com/word/see https://www.etymonline.com/word/see