I can't see clearly now
September 23, 2010 4:02 PM Subscribe
Given a known myopia prescription, and distance from monitor, could a filter be applied to an image so that it would be clear to me, without my glasses?
I was at home last week with a headache, and had taken my glasses off to avoid pressure on my temples, but had to keep putting them back on to check a slowly updated logfile, which got me wondering if there wasn't some sort of image processing magic I could get my computer to perform, such that I would be able to see something 8 feet away, as if I was wearing my glasses. Since my eyes are not evenly bad, OD 11 in one eye, and 13 in the other, it would presumably be only partially effective. Of course then I have astigmatism too.
To be clear, I'm talking about a gimp/photoshop style filter, not a physical filter I'd put in from of the monitor.
I was at home last week with a headache, and had taken my glasses off to avoid pressure on my temples, but had to keep putting them back on to check a slowly updated logfile, which got me wondering if there wasn't some sort of image processing magic I could get my computer to perform, such that I would be able to see something 8 feet away, as if I was wearing my glasses. Since my eyes are not evenly bad, OD 11 in one eye, and 13 in the other, it would presumably be only partially effective. Of course then I have astigmatism too.
To be clear, I'm talking about a gimp/photoshop style filter, not a physical filter I'd put in from of the monitor.
In theory, this could be done. Software could apply an equal and opposite "blurring" for each and every pixel on the screen.
In practice, this would not work very well because there are limits to how light or dark a pixel can be. For example, suppose you had a single black dot on a white screen. Your bad vision will blur the black dot over the neighboring pixels, darkening them. Therefore they must be brightened to counteract that darkening. But they are already white, which is as bright as they can get.
(Even if you could brighten them, of course, that brightening would also be blurred over the neighboring pixels, including the black dot, which would then need to be made blacker as a result... pretty soon, even if you start with two tones of medium gray, you reach the limits of the display.)
posted by kindall at 4:27 PM on September 23, 2010
In practice, this would not work very well because there are limits to how light or dark a pixel can be. For example, suppose you had a single black dot on a white screen. Your bad vision will blur the black dot over the neighboring pixels, darkening them. Therefore they must be brightened to counteract that darkening. But they are already white, which is as bright as they can get.
(Even if you could brighten them, of course, that brightening would also be blurred over the neighboring pixels, including the black dot, which would then need to be made blacker as a result... pretty soon, even if you start with two tones of medium gray, you reach the limits of the display.)
posted by kindall at 4:27 PM on September 23, 2010
Probably not, but on a related tangent, have you considered getting a pair of glasses that are only for computer use? You can have them optimized for the focal range you use when computing, which isn't necessarly as general purpose as you have for every day stuff.
You could also pick a more comfortable/less durable pair, as you'll only use them at the desk.
posted by TravellingDen at 4:44 PM on September 23, 2010 [2 favorites]
You could also pick a more comfortable/less durable pair, as you'll only use them at the desk.
posted by TravellingDen at 4:44 PM on September 23, 2010 [2 favorites]
I don't think this would work for the same reason that you can't ever determine the constant C when you get the antiderivative of a function. You can't rebuild data after its been discarded.
posted by Civil_Disobedient at 4:52 PM on September 23, 2010
posted by Civil_Disobedient at 4:52 PM on September 23, 2010
This is not possible. Eyeglasses work by bending the rays of light outward before they hit the lens of your eye. In other words, your eyeglass lenses together with the lens in your eye are equivalent to one lens with the proper focal length for your eye. It is not possible to rearrange objects on the plane of your screen to achieve the same effect.
posted by ssg at 5:19 PM on September 23, 2010 [2 favorites]
posted by ssg at 5:19 PM on September 23, 2010 [2 favorites]
You could simply make the charts much larger.
There are programs designed to do this for visually impaired people, so you can easily try it out. For instance in Windows, look under all programs/accessories/accessibility/magnifier.
posted by flug at 8:58 PM on September 23, 2010
There are programs designed to do this for visually impaired people, so you can easily try it out. For instance in Windows, look under all programs/accessories/accessibility/magnifier.
posted by flug at 8:58 PM on September 23, 2010
Best answer: It might be possible, but not with a traditional monitor, and probably not with any monitor that actually exists.
When you talk about how well you can see with a particular optical system, you talk about its "resolution." One way to define the resolution is the smallest distance there can be between to point light sources while you can still tell there are two of them. I think for a "typical" human eye the resolution is a few minutes of arc --- maybe a twentieth of a degree. The full moon that's out tonight is about half a degree across, so that resolution suggests that with your corrected vision you could look at the moon and imagine dividing it into a ten-by-ten grid of fairly distinct regions, but not much better than that. If you look tonight, the moon is in the sky near Jupiter, which is essentially a point of light and should have the smallest possible apparent size. (If you don't look until tomorrow, they'll still be fairly close.)
Now what happens to me when I take my glasses off is that the moon and Jupiter turn into round shapes that are about the same size (with very different brightnesses). So I think the resolution of my uncorrected vision is about half a degree.
As others have already said, if you take a regular old image on static display and convolve it with the lousy resolution function of your eyes, then the information in that image is basically destroyed. If I understand what you want, you're wondering whether there is some way of encoding the same information at multiple places on the display, so that light from different blurred parts of the screen interacts somehow to cancel out the blur and deliver you a sharp image. That's basically what a hologram does. The surface of a regular photograph has information about how much light to reflect of each color from each part of the surface. The surface of a holograph film (like the one on a credit card) has information about how much light to reflect, but also information about what phase of light to reflect. Since fixing the phase of light allows you to arrange for constructive and destructive interference, light corresponding to a particular part of the image (say the far wing of the bird on your credit card) reaches each eye from a different part of the film. That's what gives the hologram a 3-dimensional appearance, even though it's an image on a two-dimensional film: there is extra information in the image, so light from the same part of the image can reach your eyes from multiple places on the film.
But that also limits the usefulness of what you can do with a hologram. In order for light to interfere with itself, it has to have a fairly narrow wavelength range. This is why each part of the image on the hologram has a fairly strong color, and why the colors change as you move the film around --- the "rainbow" effect. I can imagine constructing a hologram where the image appears "above" the film rather than "inside" the film, and where the image is far enough above the film that it's within the range of focus of your uncorrected eyes. But you couldn't do that with an LED monitor --- LEDs don't allow you to control the phase of light. You'd have to have some elaborate laser setup, and what you got out would look like a hologram.
The only non-hologram method I can think of for encoding information about the same part of an image in multiple places are the old random-dot stereograms that have finally mostly disappeared. Those have obvious limitations as well. An important limitation for your question is that if you can't resolve the dots, you can't see the image.
So: probably not, but interesting question. I think a holographically assisted display would be an interesting senior project for a upper-level physics undergraduate, but I bet the second pair of computer-only glasses would end up cheaper.
posted by fantabulous timewaster at 9:48 PM on September 23, 2010 [1 favorite]
When you talk about how well you can see with a particular optical system, you talk about its "resolution." One way to define the resolution is the smallest distance there can be between to point light sources while you can still tell there are two of them. I think for a "typical" human eye the resolution is a few minutes of arc --- maybe a twentieth of a degree. The full moon that's out tonight is about half a degree across, so that resolution suggests that with your corrected vision you could look at the moon and imagine dividing it into a ten-by-ten grid of fairly distinct regions, but not much better than that. If you look tonight, the moon is in the sky near Jupiter, which is essentially a point of light and should have the smallest possible apparent size. (If you don't look until tomorrow, they'll still be fairly close.)
Now what happens to me when I take my glasses off is that the moon and Jupiter turn into round shapes that are about the same size (with very different brightnesses). So I think the resolution of my uncorrected vision is about half a degree.
As others have already said, if you take a regular old image on static display and convolve it with the lousy resolution function of your eyes, then the information in that image is basically destroyed. If I understand what you want, you're wondering whether there is some way of encoding the same information at multiple places on the display, so that light from different blurred parts of the screen interacts somehow to cancel out the blur and deliver you a sharp image. That's basically what a hologram does. The surface of a regular photograph has information about how much light to reflect of each color from each part of the surface. The surface of a holograph film (like the one on a credit card) has information about how much light to reflect, but also information about what phase of light to reflect. Since fixing the phase of light allows you to arrange for constructive and destructive interference, light corresponding to a particular part of the image (say the far wing of the bird on your credit card) reaches each eye from a different part of the film. That's what gives the hologram a 3-dimensional appearance, even though it's an image on a two-dimensional film: there is extra information in the image, so light from the same part of the image can reach your eyes from multiple places on the film.
But that also limits the usefulness of what you can do with a hologram. In order for light to interfere with itself, it has to have a fairly narrow wavelength range. This is why each part of the image on the hologram has a fairly strong color, and why the colors change as you move the film around --- the "rainbow" effect. I can imagine constructing a hologram where the image appears "above" the film rather than "inside" the film, and where the image is far enough above the film that it's within the range of focus of your uncorrected eyes. But you couldn't do that with an LED monitor --- LEDs don't allow you to control the phase of light. You'd have to have some elaborate laser setup, and what you got out would look like a hologram.
The only non-hologram method I can think of for encoding information about the same part of an image in multiple places are the old random-dot stereograms that have finally mostly disappeared. Those have obvious limitations as well. An important limitation for your question is that if you can't resolve the dots, you can't see the image.
So: probably not, but interesting question. I think a holographically assisted display would be an interesting senior project for a upper-level physics undergraduate, but I bet the second pair of computer-only glasses would end up cheaper.
posted by fantabulous timewaster at 9:48 PM on September 23, 2010 [1 favorite]
Response by poster: Whatever image you're looking at—whether sharp, blurry, or covered with baby sea otters—is going through those eyes and thus getting blurred.
As George Costanza references, I can use the pinhole effect via squinting, and get a sharp picture, at a drastic reduction in available light, so it's not as clear cut as you make it out, I think.
You could also pick a more comfortable/less durable pair
I've never been offered a choice in frames based on durability versus comfort. I always pick frames based on how they feel, then how they look. I get the lightest possible lenses and frames already.
You could simply make the charts much larger.
On my 27" monitor, if I expand text large enough to be able to reliably guess what it says, I can fit at most 6 letters on screen.
If I understand what you want, you're wondering whether there is some way of encoding the same information at multiple places on the display, so that light from different blurred parts of the screen interacts somehow to cancel out the blur and deliver you a sharp image.
Exactly the lines I was wondering along; interference patterns and whatnot. I wasn't envisioning the filter being able to deliver 100% clarity to an image the same size as the monitor displaying it, more like a 27" monitor could create a faux-clear image 10" across.
Guess I'm still waiting for the bionic eye. Thanx all for responses.
posted by nomisxid at 7:57 AM on September 27, 2010
As George Costanza references, I can use the pinhole effect via squinting, and get a sharp picture, at a drastic reduction in available light, so it's not as clear cut as you make it out, I think.
You could also pick a more comfortable/less durable pair
I've never been offered a choice in frames based on durability versus comfort. I always pick frames based on how they feel, then how they look. I get the lightest possible lenses and frames already.
You could simply make the charts much larger.
On my 27" monitor, if I expand text large enough to be able to reliably guess what it says, I can fit at most 6 letters on screen.
If I understand what you want, you're wondering whether there is some way of encoding the same information at multiple places on the display, so that light from different blurred parts of the screen interacts somehow to cancel out the blur and deliver you a sharp image.
Exactly the lines I was wondering along; interference patterns and whatnot. I wasn't envisioning the filter being able to deliver 100% clarity to an image the same size as the monitor displaying it, more like a 27" monitor could create a faux-clear image 10" across.
Guess I'm still waiting for the bionic eye. Thanx all for responses.
posted by nomisxid at 7:57 AM on September 27, 2010
« Older How do I download an entire website from Google's... | Windows XP Profile Won't Load Newer »
This thread is closed to new comments.
Whatever image you're looking at—whether sharp, blurry, or covered with baby sea otters—is going through those eyes and thus getting blurred.
That said, you can use something like the Mac's Universal Access control panel to dramatically change the contrast, allowing for better discern-ability of text.
posted by blueberry at 4:22 PM on September 23, 2010