Live data from Hacker News

Lytro Unlock – Making a bad camera slightly better

github.com

61–70 of 74 posts

Re: Lytro Unlock – Making a bad camera slightly better

#61

Shame lytro never made a security video camera variant. Seems like the ideal application, ensuring that anything from vehicle plates in the background to faces in the foreground can be brought into focus. What sort of fps is the API capable of producing jpegs? It sounds like the larger illum can manage 3 fps natively, but couldn't find specs on the original lytro.

Ideally a surveillance camera will already be focused to have the entire scene sharp. The Lytro camera had postage stamp resolution as far as I remember so it wouldn’t solve the problem that you are thinking of.

> Ideally a surveillance camera will already be focused to have the entire scene sharp

In practice a lot of surveillance cameras seem to be focused to have the entire scene a blurred mess

Re: Lytro Unlock – Making a bad camera slightly better

#62
post #2

I love this. I collect interesting obsolete hardware; I bought a Light L16 recently for cheap but it can no longer update itself, unfortunately. Lytro is on my list as well. The lytro itself is a lightfield / plenoptic camera. It captures the angle of light coming into the camera as well, which lets you calculate focus AFTER the photo has been taken. Focus is, of course, itself just another computation. http://graphi…

-- Media -- https://en.wikipedia.org/wiki/Smell-O-Vision https://en.wikipedia.org/wiki/MiniDisc Toshiba Twin S-VHS changer https://youtu.be/aEEYuJw7vFk https://en.wikipedia.org/wiki/HitClips https://en.wikipedia.org/wiki/Pocket_Rockers 0.85" / 22 mm HDD used in MP3 players https://web.archive.org/web/20120718025547/http://www.toshib... -- Computing peripherals -- https://en.wikipedia.org/wiki/CueCat https://www.compu…

librarything still sells and works with cuecats

Re: Lytro Unlock – Making a bad camera slightly better

#63
post #33

Earlier quoted context omitted.

"just performing predictable mathematical transformations" Just linear algebra, to be precise. A lens is just a matrix and rays are simply vectors. Combining lenses is merely matrix multiplication and complex optical systems can be represented by the result, which is a plain simple matrix again. The Eigenvalues tell you interesting properties of your optical system. I've always enjoyed this part of physics because it…

Well, sort of... there goes quite a bit of materials science and engineering into making and shaping the glass of a lens to make it behave like a matrix (i.e. to extend the range in which this linear approximation to the actual behavior of a blob of glass is valid).

Blobs of glass are actually overwhelmingly linear :) Non-linear effects in optics are generally negligible unless you go to very high energies (scattering, like raman scattering) or special materials (like fluorescent ones). You need a high powered laser or certain materials for those regimes.

That said, the geometric functions (i.e. the positions of rays w.r.t. other objects) are probably non-linear in the object position parameters indeed, and intensities certainly are (example: moving an object in front of a screen by X amount changes the illumination in front of the screen non-linearly): it's important to keep in mind what we mean by linearity (it's in terms of light field intensities for a fixed geometry scene) -- the scene are the transfer function coefficients.

What really makes things complicated in real life is (1) the presence of noise; (2) imperfections (and unknowns) in your physical/geometric apparatus. Even if you know the system perfectly, noise generally disallows reverting (or easily reverting) transfer functions, i.e. undoing blurs and arbitrarily refocusing images with simple detectors. The imperfections and even lack of rigidity of lenses and your system add even more difficulty. That's why making a simple computational lens is not so easy.

Re: Lytro Unlock – Making a bad camera slightly better

#65
post #51

> Killer feature of this new technology was the ability to refocus the image after it was taken! (...), the camera was trying to solve a problem that didn’t exist A related camera technology, but which I've found great to use, is 360 cameras. The ability to re-frame (not focus) the image as I want in post is so great. No longer do I have to spend many attempts positioning a camera perfect, just shoot and later "point…

I worked on 360 cameras / software for many years. Looking forward to the day when an article like this comes out for one of the products I worked on. When you’re in the heart of it it’s so easy to take pride in the technical challenges you’ve overcome, but completely miss the realities of the marketplace.

I really love my 360 camera. Use it everyday. I fill my 128 gb memory card multiple times a week. I use it on my helmet when biking for work. On a selfie stick when skiing so that it looks like I have a personal camera man following me (example https://www.instagram.com/reel/CnFtVsMJd39/?igshid=YmMyMTA2M... )

But yeah, they haven't taken that much off compared to traditional action cams. The frame rate and resolution after reframing is the blocker for many. And hard to do that with the current available sensors it seems like, and no one right now wants to gamble on spending much on R&D for launching the next gen.

Re: Lytro Unlock – Making a bad camera slightly better

#66
post #2

I love this. I collect interesting obsolete hardware; I bought a Light L16 recently for cheap but it can no longer update itself, unfortunately. Lytro is on my list as well. The lytro itself is a lightfield / plenoptic camera. It captures the angle of light coming into the camera as well, which lets you calculate focus AFTER the photo has been taken. Focus is, of course, itself just another computation. http://graphi…

That’s a really interesting way of phrasing it I’d never fully considered: of course you are right, a physical lens is just performing predictable mathematical transformations on the light passing through it. Thank you, that was something of a revelation in understanding how these cameras actually work.

Lenses compute 2D Fourier transforms. And they do it really, really fast.

https://en.wikipedia.org/wiki/Fourier_optics#Fourier_transfo...

Re: Lytro Unlock – Making a bad camera slightly better

#67
post #48
post #35

Earlier quoted context omitted.

That is actually deliberate, we tend to organize everyday life and the devices we build around processes that are easy to model, that is why many things look like linear processes and harmonic oscillators (the first two term of the Taylor expansion of the actual behavior). We change the type of spring we use if the current one starts to wear out early under normal operational conditions.

Let me rephrase on his behalf. Isn't it curious just how unreasonably capable our math is at expressing those physical laws of the universe to which we did not invent.

I don’t know, seems a bit Anthropic Puddle to me.

Re: Lytro Unlock – Making a bad camera slightly better

#68
post #51

Earlier quoted context omitted.

I worked on 360 cameras / software for many years. Looking forward to the day when an article like this comes out for one of the products I worked on. When you’re in the heart of it it’s so easy to take pride in the technical challenges you’ve overcome, but completely miss the realities of the marketplace.

I really love my 360 camera. Use it everyday. I fill my 128 gb memory card multiple times a week. I use it on my helmet when biking for work. On a selfie stick when skiing so that it looks like I have a personal camera man following me (example https://www.instagram.com/reel/CnFtVsMJd39/?igshid=YmMyMTA2M... ) But yeah, they haven't taken that much off compared to traditional action cams. The frame rate and resolution…

The problem is that you’re always at a sensor deficit because you need so much more resolution to cover a wide field of view. When sensors and chipsets would finally catch up, user expectations would grow to match. Now we’re up against physics: Can’t add more resolution past the diffraction limit, and larger sensors are impractical for super fisheye lenses.

Re: Lytro Unlock – Making a bad camera slightly better

#69
post #61

Earlier quoted context omitted.

Ideally a surveillance camera will already be focused to have the entire scene sharp. The Lytro camera had postage stamp resolution as far as I remember so it wouldn’t solve the problem that you are thinking of.

> Ideally a surveillance camera will already be focused to have the entire scene sharp In practice a lot of surveillance cameras seem to be focused to have the entire scene a blurred mess

In practice a lot of surveillance cameras were never focused at all. They were installed that way and never configured

Re: Lytro Unlock – Making a bad camera slightly better

#70
post #2

I love this. I collect interesting obsolete hardware; I bought a Light L16 recently for cheap but it can no longer update itself, unfortunately. Lytro is on my list as well. The lytro itself is a lightfield / plenoptic camera. It captures the angle of light coming into the camera as well, which lets you calculate focus AFTER the photo has been taken. Focus is, of course, itself just another computation. http://graphi…

Was curious about a Light L19 seeing how cheap it is now and did some digging around for how to update. This Github project from an XDA thread may do the trick https://github.com/helloavo/Light-L16-Archive

thanks for this, I've been needing to factory reset my L16 but I couldn't locate any of the firmware
Post reply on HN