cyan and magenta dots/noise in RAW results - RAW Shooting and Processing - CHDK Forum

cyan and magenta dots/noise in RAW results

  • 5 Replies
  • 2613 Views
cyan and magenta dots/noise in RAW results
« on: 23 / April / 2009, 03:31:28 »
Advertisements
I have purchased two A590is at different times, and to my surprise the results are extremely different. (see http://www.flickr.com/search/?q=test%20canon&w=87084332@N00). I thought it was probably just a lens issue, even though the poorer quality one had a smoother operating and more solid feeling lens barrel.

Upon closer examination, however, it appears that the sensor also responds differently. The newer A590is had scattered cyan and magenta dots across the entire image, especially in highlights. I tried to change the "bad pixel" correction in the CHDK menu, but those dots remain.

I tried switching off DNG format, and those dots are gone in the raw .CR2 files. However, the exposure information is gone and the colors just look way off. Is there any way to keep exposure information and remove those dot/artifacts?

Is this a common problem? What can I do to remove it in my RAW results? Thanks.

(let me know if I should provide an image sample of this problem. Thanks.)
« Last Edit: 23 / April / 2009, 05:58:02 by ciaconne »

Re: cyan and magenta dots/noise in RAW results
« Reply #1 on: 23 / April / 2009, 16:49:21 »
I solved this problem by re-doing the badpixel.lua script which is supposed to be a required step for saving in DNG format. I suppose the older test results doesn't apply to the new camera. By running the test again it generated a new profile for the new camera and that solved the DNG noise problem.

I wonder if that is a good indication that they had put in a different sensor for the A590 generations though. Interesting.

*

Offline PhyrePhoX

  • *****
  • 2254
  • make RAW not WAR
    • PhyreWorX
Re: cyan and magenta dots/noise in RAW results
« Reply #2 on: 23 / April / 2009, 17:05:57 »
ah, now i get it.

well, no, not different generations. just two sensors. there are no sensors that 100% alike. every sensor in every camera produced has badpixels. some more, some less. but they all do. and they all have them at different places. thats why you have to redo the badpixel.lua script on each camera seperately.

Re: cyan and magenta dots/noise in RAW results
« Reply #3 on: 23 / April / 2009, 18:01:07 »
yeah, that makes sense. though I am not sure how to explain the dramatic different results from the two A590is samples. It seems to be too dramatic to be just optics alone. Who knows. It's end of life anyway. No wonder I hear inconsistent reports of A590is image quality.

I'm just glad I can shoot RAW without the noise, and with less blurriness!


*

Offline PhyrePhoX

  • *****
  • 2254
  • make RAW not WAR
    • PhyreWorX
Re: cyan and magenta dots/noise in RAW results
« Reply #4 on: 23 / April / 2009, 18:38:38 »
it's not optics, it's physics ;)
try to use the badpixel file from the "bad camera" in the good camera. suddently the dngs from this one will also look bad.

Re: cyan and magenta dots/noise in RAW results
« Reply #5 on: 23 / April / 2009, 19:59:58 »
eh.. i was referring to the low quality image outputs from the older camera.. not sure what's wrong with that one.

 

Related Topics