Patent: Quad Pixel autofocus image sensor

[Mt Spokane Photography]

there is an extra stop of DR that lives in the second sub frame of the DPRAW file. I do use it when multiple exposure are not an option - e.g. slow / fast moving subject.

I've tested it, it appears to be a fraction of a stop, but does help. Not enough for me to bother using it.

I've heard from only a very few who use the DP Raw feature. It will be interesting if they come out with a quad pixel to see if there is a application.

 

[Mt Spokane Photography]

For real? That's awesome! Do you need to use Canon's DPP to process? And would this also benefit max usable ISO in low light, or is it just for low ISO DR?

You need to separate the DP RAW images into two files as well as the merged file from the camera. There is software called DPR Split that does this. Then, you stack the images to obtain extra DR.

The top image is the original, the bottom is the processed one.

 

[SecureGSM]

For real? That's awesome! Do you need to use Canon's DPP to process? And would this also benefit max usable ISO in low light, or is it just for low ISO DR?

yup, here:

DPRSplit will help pull more dynamic range from Canon 5D Mark IV Dual Pixel Raw files

DPRSplit can help you pull about +1EV of usable dynamic range from Canon's 5D Mark IV Dual Pixel Raw files—as long as you don't mind a bit of parallax error, that is.

www.dpreview.com

this won't affect max usable iso in low light. quite opposite, it affects ability to recover clipped highlight.
an extra stop of DR (on demand) is a wonderful feature to have.

 

[CJudge]

Yeah. I think people underestimate the DPP factor on the lack of uptake for Dual Pixel features. That is precisely and only what has kept me from exploiting the focus-after-exposure feature. Even though its effects are slight, I'd still go out of my way to get the extra millimeter of sharpness if I could do it in Lightroom or, really, anything else. That software is truly dreadful.

Considering that Canon's new c-RAW format is much friendlier on storage space (as well as to my eyes having practically zero effect on image quality compared to standard CR3 files), I would agree with you that I would love to take advantage of what DPRAW offers, but there's no way I'm switching my whole workflow just to do it. If QPRAW is indeed put into use, and Lightroom enables access to these post processing options, I'd be all over that!

 

[CJudge]

yup, here:

DPRSplit will help pull more dynamic range from Canon 5D Mark IV Dual Pixel Raw files

DPRSplit can help you pull about +1EV of usable dynamic range from Canon's 5D Mark IV Dual Pixel Raw files—as long as you don't mind a bit of parallax error, that is.

www.dpreview.com

this won't affect max usable iso in low light. quite opposite, it affects ability to recover clipped highlight.
an extra stop of DR (on demand) is a wonderful feature to have.

Thanks for sharing that! I wonder though, if you can separate and then merge the two image files (as described by Mt Spokane Photography above) , would you not effectively cut the noise in an image by averaging it?

 

[Mt Spokane Photography]

Thanks for sharing that! I wonder though, if you can separate and then merge the two image files (as described by Mt Spokane Photography above) , would you not effectively cut the noise in an image by averaging it?

Canon merges the two images in the camera, which presumably reduces the noise. You gain more DR by a small but visible amount, and hopefully reduce noise by merging three images. Its somewhat klunky, the files are .dng. You should read about it at their site.

 

[uri.raz]

If Canon makes a QPAF sensor, would it be debayered same as DPAF (= sum four sub pixels of same color, and interpolate colors same as non *PAF sensor), or would it gain by adding four adjacent sub pixels as in the orange squares, and not interpolating color at all?