Author Topic: 5dm3 RAW Sensor differences  (Read 2825 times)

beauchampy

  • Senior
  • ****
  • Posts: 302
5dm3 RAW Sensor differences
« on: December 07, 2016, 04:16:35 PM »
Has anyone else noticed a slight shift on different 5dm3 sensors?

I recently shot a project with 3 different 5d3's. One is noticeably more magenta than the others.

WB and tints were all identical.

a1ex

  • Administrator
  • Hero Member
  • *****
  • Posts: 12564
Re: 5dm3 RAW Sensor differences
« Reply #1 on: December 07, 2016, 04:49:19 PM »
CR2 (still photo) and DNG (from RAW or MLV) samples?

There are some discussions regarding color matrices, so two images with obviously different color response could be helpful to confirm some hypotheses (in particular, the CameraCalibration matrices).

https://www.magiclantern.fm/forum/index.php?topic=10119.0
https://www.magiclantern.fm/forum/index.php?topic=13948.0

teatotalTED

  • New to the forum
  • *
  • Posts: 35
Re: 5dm3 RAW Sensor differences
« Reply #2 on: December 07, 2016, 05:00:37 PM »
Shane Hurlbut did a few years ago when he started with the 5D's, part of gear set up prior to shooting, calibrate the cameras to each other by tweaking WB per camera.

http://www.thehurlblog.com/film-school-online-color-correction-put-your-best-foot-forward/

From the page:

Quote
Each camera comes from the factory supposedly balanced but all of them have a bias.  Set up a white card with the correct color temp. on the camera, which it depends on the color temp of your light.  If you are shooting with a Tungsten source then you would be at 3200K, if it is daylight you would be at 5200K and so on.  Or you can auto white balance, your choice.  Some cameras come with a yellow bias, a magenta bias and or a green bias.  Sometimes you get one that is perfect from the factory but from my experience that is not the case.

And for more consistent, less painful colour correction, for raw to.



Walter Schulz

  • Contributor
  • Hero Member
  • *****
  • Posts: 8333
Re: 5dm3 RAW Sensor differences
« Reply #3 on: December 07, 2016, 05:10:34 PM »
And there are the guys at Chromoholics:

https://luminous-landscape.com/digital-camera-calibration-and-profiling/

I don't have tried that, though.
But first make sure not to be fooled by ACRs camera "calibration" process engine 2012.

Danne

  • Developer
  • Hero Member
  • *****
  • Posts: 7293
Re: 5dm3 RAW Sensor differences
« Reply #4 on: December 07, 2016, 05:54:33 PM »
Cameracalibration values are needed to meet the last wb calibration together with the AsShotneutral tag in a dng file. These numbers are revealed only when transcoded into a dng through adobe dng converter and are camera model specific. Without cameracalibration you typically get minor AsShotneutral tag shifts which will make wb shift slightly.
At least from what I,ve seen and tried I cannot find that camera calibration tag inside a CR2. As soon as you have that number through adobe dng converter it can be added into the AsSHotNeutral tag or as a stand alone cameracalibration tag. Better to include into the AsShotNeutral tag as Resolve and other NLE,s will be able to read that tag more likely than cameracalibration values. I include cameracalibration values in cr2hdr.app to meet the final wb correction in dualiso CR2 to DNG conversion.
http://www.magiclantern.fm/forum/index.php?topic=7139.msg172907;topicseen#msg172907

With that said more likely you,ll need to calibrate through profiling. Can be done in adobe dng profile editor and links are already posted.