Noise... camera ISO sweet spots?

Started Dec 28, 2012 | Discussions thread
DSPographer Senior Member • Posts: 2,464
Re: Hmm...

Jack Hogan wrote:

I prefer to use Bill Claff's read noise data for absolute comparisons because DxO's appears to be a milkshake of the four channels, possibly or possibly not scaled/balanced. Too many variables. Also it's difficult to read their SNR data in the deepest shadows because they do not display it with enough resolution (there is one stop between 0.01 and 0.02%) and I have a feeling that's one of the reasons for sometimes unreasonable sensorgen read noise values.

It doesn't matter how finely the DXOmark data is plotted when computing a fit. That is because their plot is just a mathematical curve generated from their estimates for read noise, photon gain, and PRNU for a sensor. As long as I use at least 3 plotted points when making a curve fit, I get the exact same answer for noise e-, full well e- and PRNU. (The match error residues in my overdetermined fits are consistent with double precision accuracy.

DxO data is of course fine when comparing relative data since the methodology is the same. I wish they gave us accesss to the individual channel data and told us how they actually compute SNR, though.

I really wish they just gave as access to their raw data: but since they spent the effort to get it, they can choose what they want to release.

 DSPographer's gear list:DSPographer's gear list
Canon PowerShot G7 X Canon EOS 5D Mark II Canon EF 24mm f/2.8 Canon EF 50mm f/1.8 II Canon EF 200mm f/2.8L II USM +4 more
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow