License it? Is this such a revolutionary idea that other companies can't "invent it" themselves without licensing it?

This is an obvious idea that everyone has been craving for, and I don't think such ideas deserve patents. If stock cameras adopt this idea, tough luck, they'll just have to compete like everyone else and add more differentiation.

It's not the implementation that matters here, it's the idea. And companies can do their own implementations of it. They don't have to license it.

They may not have to license it, but for a reasonable price I'd rather use a well-implemented library that excels at what I need it to do and goes beyond, than spend time hacking up something equivalent but inferior (hey, I've got other stuff to do too).

Google+ already does something something similar, but not exactly. Ideas are obvious, implementation is everything. This product may or may not be a better implementation, and worth paying for. Inventors of advanced camera algorithms tend to get bought out or hired.

First of all, I was not talking about patents at all.

And my first assumption was that they do something more than cropping, in other words that the implementation is not trivial. But I learned that might not be the case.

> Anyway, what I wanted to ask.. how do they go around the loss of quality? It's not a simple crop? Can they somehow utilize any part of the chip they want (as long as the maximum is the same?) since the chip is much bigger than the max possible video it can record?

Who says they do? The app seems to show the entire input so I don't think there could be anything tricky in terms of using only part of the sensor.

> Who says they do?

They advertise certain resolution. I assumed it's for the final video. If they just crop it and stretch it.. that's disappointing.

When Crisp mode is used (in iPhone 5S), we use the full camera resolution (2592 x 1936 pixels) and after transforming it we project it to the output resolution (1080p, 720p). So, the final video is.. crisp :)