New sensor tech coming from Panasonic

Started Feb 4, 2013 | Discussions thread
Franksson
Forum MemberPosts: 74Gear list
Like?
Re: Thanks! But I'm sceptical...
In reply to kenw, Feb 4, 2013

kenw wrote:

Thanks for the link, very interesting!

I'm a little sceptical about just how revolutionary it will prove to be in the end. Three red flags:

1. We've seen a simpler approach that didn't work - CMY arrays. You do notionally get more SNR at the image pixels but then you lose it in the color space transform. This could very well have similar problems. And it is a bit worse, color spatial information is sent to different pixels requiring more demosaicing work that could also reduce the theoretical benefits of more photons at the photodiode.

2. Acceptance angle could be very sensitive. The question would be is it any worse than for current micro-lens designs. If it is then look for edge and wide aperture performance problems. On the other hand modern lenses are already rather telecentric so it might be a non-issue. I don't have the experience to know.

3. If you read the press release carefully it sounds mostly like a software/algorithm design story. Algorithms that will make it feasible to efficiently design these concepts, not much word on manufacturing itself. That said, it appears entirely focused on sticking with mostly common fabrication techniques so that is a good thing.

So definitely a cool idea, and neat to see someone working on it. But I wouldn't get too excited about seeing the technology any time soon, or expecting it to work miracles, or expecting it not to have some other trade off to deal with.

Again, thanks for the link, interesting read (which is exceedingly rare around here these days).

There is a reference in the Panasonic announcement to Nature Photonics. In this month's edition there is a more detailed article about the technology here: http://www.nature.com/nphoton/journal/vaop/ncurrent/full/nphoton.2012.345.html

Unfortunately money is required to see the full article, but there are some tantalising snippets to be found in the abstract and in a Supplementary Information pdf here: http://www.nature.com/nphoton/journal/vaop/ncurrent/extref/nphoton.2012.345-s1.pdf

They state that the amount of light received at the sensor is 1.85 times that of a conventional Bayer colour filter array.

It will be interesting to see how long it takes to get this into a production camera.

Peter

 Franksson's gear list:Franksson's gear list
Canon PowerShot G9 Panasonic Lumix DMC-GH2 Panasonic Lumix G Vario HD 14-140mm F4-5.8 OIS Panasonic Lumix G Vario 100-300mm F4-5.6 OIS
Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow