Hacker News Reader

Comment by shasta | Hacker News Reader

Doesn't this assume you don't know the focal plane? If you know what the viewer is focusing on, you can compensate. This means that without any holographic projection techniques, you could fully solve this problem by presenting each viewer with images customized to his detected focus. That might require an expensive headset today, which radically changes the economics of 3D movies, but it doesn't require a breakthrough in science.

In fact, it'd probably be safe to just assume that the user is focused on the primary element in the scene. Unfortunately, in a theater, you probably can't compensate for focus in a way that works for all of the audience, so this doesn't solve much.

View on HN

You know, I think a hidden point in all this is that fundamentally, part of what movies do is try and control the audience's attention. It's the entire point behind carefully choosing camera angles and focus in the first place. The movie-maker is trying to inform the audience: "These are the people you should be listening to, this person's reaction is what matters, this object is significant to the plot..."

If you can't control focus, then you are presented with a much larger scene. You (the viewer) would be choosing who and what to pay attention to, removing control from the film maker, and unless this additional degree of freedom is compensated for, the experience would go down. The way you tell the story as a result would change dramatically. You can't just tag on "3D version available!" to a movie and expect it to work, it requires a different way of telling a story.

Not as I understand it. I think you're conflating two issues. What you're talking about deals with the issue of looking at different elements of the scene and bringing each one to clarity as you look at it.

The second, larger issue, is the physiological contraction of muscles each eye individually must do to focus at a given distance. Try this: cover one eye and look at a held up finger with the other eye. Notice the background is blurry. Now make the background clear and you'll notice your finger is blurry.

The eye adjusts its shape to focus on different distances. This is independent of the two eyes working together and seeing different images which the brain later merges into a 3D scene.

The issue with all the current 3D technologies I know about is the image, ultimately, is always on some screen at a fixed distance. You can give each eye separate images, you can track where the eye is looking and modify what image is seen, but in the end, the eye muscles are always contorting the eye to see that plane 80' away clearly.

You're right. It's not just a still image that needs to be corrected - it's the whole volume of light that needs to be distorted (the eye is not a pinhole camera). But I still think it's doable - isn't that what corrective lenses do?

Not a bad idea. You could do it with one person in the theater, but you can take it one step further.

Have a camera facing the eye that can tell what the person is focusing on. (High end video cameras already have this.)

Have an adaptive lense in the glasses to vary rapidly change the focal plane from where they eye things it is to the screen. You then build software that blurs the image outside of their focal plane. Granted there are limits on this aproach, but people get used to glasses that have a fixed shift in the focal plain fairly quickly so it should be possible.

Now I don't see how you could extrapilate that to a theater full of people, but you can build a system that works for 1 person.

> Have a camera facing the eye that can tell what the person is focusing on.

Sorry I was a bit terse, but that's what I meant by "detected focus."

> You then build software that blurs the image outside of their focal plane.

That's not it exactly. The problem you're trying to solve is that, when the viewer is focused at one depth but is viewing a flat image at another depth, it will already appear blurry to him. You can't fix that with extra blurring. The fix is to distort the original image, much like corrective lenses.

And BTW, I suspect that you can't perform this correction for even a single viewer in the audience unless you're tracking his location and orientation.

Skip the screen and go with VR glasses? Then the projectors are at a fixed distance from the eyes, and you can (in theory) go wild applying individual corrections for everyone, up to and including correcting for individual focus.

You missed one of the steps

"Have an adaptive lense in the glasses to vary rapidly change the focal plane"

AKA their eyes focus on 50', and the screen is at 100’. But the glasses change that to 100' and the software shows a clear images for 50' with the correct amount of blur for stuff outside of that range. You still need to double the frame rate etc for the 3d effect, but you can alter things so brain is less confused.

The idea is without the bluring the full immage is in focus regardless of what the users eye tryes to focus on 5' or 500' the screen is still in focus. This would work ok, but by bluring part of the immage you get an even closer simulation of the real world.

Edit: http://en.wikipedia.org/wiki/Adjustable-focus_eyeglasses

So I did!

Reply on news.ycombinator.com