Aperture related scaling

All things embedded.
*NO FEATURE REQUESTS*

Aperture related scaling

Postby astronomer » Tue Apr 29, 2014 11:11 pm

After lain's discovery that nothing changes the register setting when aperture is altered, I decided to do a test to measure how the scaling factor is applied after the sensor read out. It could be helpful for locating the code.

Referring to my old post for test pattern and procedure.
The image was taken in dark room with only a screen turn on showing that test pattern, the goal is to make as uniform as possible in the linear histogram so the entire dynamic range is captured with greater than 10 pixel at each ADU bin. A 50 1.8D lens is used with 1/3EV aperture increment from 1.8 to 5.6

The RAW image is then converted to tiff using "dcraw -T -4 -D *.nef" which is then imported into R-studio for histogram generation.

Test result:

Tried 2 different lens and gaps counts are same, so it's not lens dependent but purely aperture value dependent. When lens are detached with Non-CPU entry, no gaps were observed in green channel.

The following graphs illustrates the gaps and calculated scaling factor for each channel.

F1.8
1 .png


F2.0
2 .png


F2.2
3 .png
User avatar
astronomer
Developer
 
Posts: 722
Joined: Fri Apr 06, 2012 10:06 am
Location: US
Been thanked: 293 times

Re: Aperture related scaling

Postby astronomer » Tue Apr 29, 2014 11:15 pm

F2.5
4 .png


F2.8
5 .png


F3.3
6 .png
User avatar
astronomer
Developer
 
Posts: 722
Joined: Fri Apr 06, 2012 10:06 am
Location: US
Been thanked: 293 times

Re: Aperture related scaling

Postby astronomer » Tue Apr 29, 2014 11:19 pm

F3.5
7 .png


F4
8 .png


F4.5
9 .png
User avatar
astronomer
Developer
 
Posts: 722
Joined: Fri Apr 06, 2012 10:06 am
Location: US
Been thanked: 293 times

Re: Aperture related scaling

Postby astronomer » Tue Apr 29, 2014 11:23 pm

F5
10 .png


F5.6
11 .png



I had also zoomed in the 1-500 ADU region and observed the original gaps in red channels are shifted to right when more gaps are added in due to aperture related scaling. This means the aperture related scaling should happen after the red and blue scaling.
I will post the script later. Really need to get some sleep.
User avatar
astronomer
Developer
 
Posts: 722
Joined: Fri Apr 06, 2012 10:06 am
Location: US
Been thanked: 293 times

Re: Aperture related scaling

Postby astronomer » Mon May 05, 2014 5:55 pm

Attachments
gaps_R.gif
gaps_B.gif
User avatar
astronomer
Developer
 
Posts: 722
Joined: Fri Apr 06, 2012 10:06 am
Location: US
Been thanked: 293 times


Return to Firmware

Who is online

Users browsing this forum: No registered users and 13 guests