Patent: Expanded dynamic range using DPAF sensors

Quarkcharmed

EOS 5DMkIV
Feb 14, 2018
462
310
Australia
www.michaelborisenko.com
The alignment where the left-half pixel and right-half pixel such that the left-half pixel is sky and the right-half pixel is the building and vice versa... on one row of pixels?
In your message https://www.canonrumors.com/forum/index.php?threads/patent-expanded-dynamic-range-using-dpaf-sensors.37447/page-2#post-787800 you never said it was about a whole row. You were talking about one pixel. In one row that's unlikely of course. But in two half-pixels or in several adjacent half-pixels - easily.
 

Don Haines

Beware of cats with laser eyes!
Jun 4, 2012
8,092
1,571
Canada
This is the same idea as HDR photography. Take one shot at one ISO, another at a different ISO, and merge the pictures. Now we get to take both pictures at the same time.

I believe that several forum members proposed this about a day after Canon announced their first DPAF camera.
 
Last edited:
  • Like
Reactions: nchoh

SecureGSM

EOS 6D MK II
Feb 26, 2017
1,155
189
The alignment where the left-half pixel and right-half pixel such that the left-half pixel is sky and the right-half pixel is the building and vice versa... on one row of pixels?
just thinking.. with this dual pixel raw tech.. we are essentially looking at a two subframes with a half-pixel shift each if combined in a single image.
similar to this one here:


I might consider investing some time experimenting to see if there is a detectable difference in sharpness between a single and "stacked" dual pixel raw 5d IV files.

p.s. actually, looking at this post here: there is a difference

 
Last edited:

Uneternal

EOS T7i
Jan 25, 2016
58
32
p.s. actually, looking at this post here: there is a difference

The difference there is just from different saturation and contrast. I don't see a difference in sharpness.
 

nchoh

EOS RP
Apr 3, 2018
205
109
Calgary
In your message https://www.canonrumors.com/forum/index.php?threads/patent-expanded-dynamic-range-using-dpaf-sensors.37447/page-2#post-787800 you never said it was about a whole row. You were talking about one pixel. In one row that's unlikely of course. But in two half-pixels or in several adjacent half-pixels - easily.
Yes, true. I just didn't know how to explain at that time... sometimes, taking an experiment considering just one pixel does not make sense when expanded to the whole sensor.

That said. It seems that many people are not reading the article correctly. It is about amplification during read.

The amplification of one half of the pixel could also mean de-amplification of the other half, so that the pixel as a whole would be at the "correct" exposure". Blown out pixels would still be blown out on the amplified pixel but would have information on the de-amplified half-pixel. And vice-versa for pixels that are too dark for any information. That would give expanded ISO.