Processing Raw images for scientific use

Started 9 months ago | Questions thread
Dave_J_E Regular Member • Posts: 139
Re: Processing Raw images for scientific use

amirber wrote:


We have recently purchased a Sony a7R IV camera and intend to use it for scientific data analysis. For this reason we collect image in manual mode and only use the raw images. the processing is rather straightforward:
* capture dark image (with closed lens cap in a dark room)
* capture an image of the test object.

While we were experimenting with different settings of the camera I made sure to collect dark images under the same mode, ISO and exposure time and got very consistent values there, however when I processed the target images parts of the frame (about 1%) of the pixels were of lower values than those of the dark images. This came as somewhat of a surprise to me, is it possible that some unaccounted parameter caused the black level / pedestal to change automatically?

I process the raw images using a standard python library (rawpy) and would like to verify if I'm missing something in the process. If there is a whitepaper relevant for this I would very much appreciate it if you could point me in the right direction. I usually use computer vision cameras (Basler) which use Sony sensors, This is not an issue there so I'm guessing theres some free parameter hiding from me.


I'm wondering whether these are masked pixels. Sensors typically have a small group of masked pixels that can be used as a black level reference. My Sony sensors have a few columns on the right side of the frame masked. You can view and measure them in RawDigger.


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