HN Theater @HNTheaterMonth

> It does not produce a focused image

I don't think this is technically correct. There are many focused images being produced: https://www.youtube.com/watch?v=rEMP3XEgnws

> but it's not at all like a conventional camera lens

The physics is exactly the same as a conventional lens, because each micro lens is conventional lenses.

It's all conventional, up to the point of having to separate all of the real, focused, images to make one clear image.

For proof, if you block all the micro lenses, except one, you would see a boring, focused, real image projected onto the sensor. In fact, the original light field cameras, that you could buy off the shelf, had a single, separate, image per microlens. Looking at the image in this press release, it looks the same.

⬐ eloeffler

So it's also not a single exposure because each microlense-sensor team has its own exposure.

At this point we could connect each sensor to its own circuitry and storage and then say it's a cameraless photo altogether :)

My go to for light field cameras would probably be this video:

https://www.youtube.com/watch?v=rEMP3XEgnws

It explains the specifics of how light field cameras work quite well, but doesn't go too deep into light fields. For more of an overview of light fields in general, I can recommend this video:

https://www.youtube.com/watch?v=BXdKVisWAco

My answer to your question would be that light field cameras sample discrete perspectives from a 4D light field. These samples can either be 1) combined directly (requires a very high view density) 2) interpolated to try to recover the continuous 4D light field function (this is an active area of research), 3) downsampled to volumetric 3D (Google's approach with "Welcome to light field" etc.), or 4) downsampled to 2D + depth (a depth map)

Each of these use different methods.

⬐ yomism

In that channel there are some videos of Brian Kernighan talking about his Bell Labs years. Pure gold.