Panasonic patents sensor with white pixels to eliminate the IR cut filter

"RichA" wrote in message
...
Good idea I guess.

http://www.freepatentsonline.com/EP2434761A1.html

A filter is still needed to stop the "white" pixels from having a response
into the IR, so those pixels (at least) still need the filter, although if
they have well-designed colour filters, only the "white" pixels will need
a filter. But surely you could do that with well-designed RGB filters?

It's about improving sensitivity.

Sounds remarkably like the CMY filters which have been done before.

David

In article , David J Taylor
wrote:

http://www.freepatentsonline.com/EP2434761A1.html

A filter is still needed to stop the "white" pixels from having a response
into the IR, so those pixels (at least) still need the filter, although if
they have well-designed colour filters, only the "white" pixels will need
a filter. But surely you could do that with well-designed RGB filters?

a quick skim of the patent says that infrared is removed mathematically.

It's about improving sensitivity.

it's about removing the infrared cut filter, which they claim impacts
sensitivity. i don't think it's as big of an issue as they suggest.

Sounds remarkably like the CMY filters which have been done before.

thats what the w-r, w-g and w-b pixels are.

"nospam" wrote in message
...
In article , David J Taylor
wrote:

http://www.freepatentsonline.com/EP2434761A1.html

A filter is still needed to stop the "white" pixels from having a
response
into the IR, so those pixels (at least) still need the filter, although
if
they have well-designed colour filters, only the "white" pixels will
need
a filter. But surely you could do that with well-designed RGB filters?

a quick skim of the patent says that infrared is removed mathematically.

It's about improving sensitivity.

it's about removing the infrared cut filter, which they claim impacts
sensitivity. i don't think it's as big of an issue as they suggest.

Sounds remarkably like the CMY filters which have been done before.

thats what the w-r, w-g and w-b pixels are.

Yes, it's been done before. The problem with removing IR mathematically
could be that with the larger number of photons incident on the sensor,
the signal to noise ratio would be degraded. I agree that it's no a major
issue, and would be surprised if anything significant results. Having
said that, I would like to see what response a current IR cut-off filter
actually has, to know what its impact on sensitivity actually is.
Judgment reserved.

David

In article
,
RichA wrote:

They said this when the M8 flaw was discovered
and they had to slap those cyan filters on the front of their
cameras. A hot mirror filter (one that rejects IR) won't work, I've
tried it.

their fix was a hot mirror filter, not cyan filter.