Nikon finally gets the noise down.

Started Nov 17, 2007 | Discussions thread
alanr0 Senior Member • Posts: 2,164
Re: QE and well capacity

John Sheehy wrote:

alanr0 wrote:

If we increase QE without increasing well size, there is no
improvement in SNR, but we achieve that SNR with fewer photons.

How is signal defined? Ultimately, the signal we are interested in
is the light going through the lens. In that sense, the SNR is
higher with higher quantum efficiencies, even with the same full-well

As long as the sensor does not saturate, increasing quantum efficiency will always increase SNR.

I was responding to PIXmantra's comments about the supposed link between well capacity and quantum efficiency, and his interpretation of Roger Clark's comments. In that context, maximum SNR is achieved close to saturation, and an increase in electrons converted will lead to highlight clipping.

My point was that if we double QE, but reduce exposure by one stop to avoid clipping, then (to first order) the SNR is unchanged. This is the case whether we measure (signal in highlights) (noise in highlights), (signal in highlights) (noise in shadows), or some other measure of SNR. Definition of SNR from a saturated sensor is left as an exercise for the reader.

However, I have dipped into a couple of PIXmantra's earlier posts, and I am now rather less certain what he was banging on about.

Alan Robinson

Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow