PixInsight Forum (historical)
PixInsight => General => Off-topic => Topic started by: Hans Pleijsier on 2010 April 27 13:27:02
-
OK, just to let you know I did a little test to try and find an acceptable workflow for my canon 40D cr2 files.
In this testing I investigated 28 combinations of different PI-settings.
Below you find the workflow that sofar worked best for me in terms of noise, banding, bad pixels etc.
I hope that other Canon D40 users will jump in and add some finetuning or redesign on this modest attempt. :P
So, please let me know if you think I go off track in a way that this workflow is less then efficient or even destroying information.
Ok, here it is ...
Step 1.
start up the script BatchFormatConversion and import all (L+D+B+F) Canon CR2 files after adjusting DSLR_RAW settings (in the format explorer) to (1) Create_Raw_Bayer_CFA checked and (2) NoBlackPointCorrection checked as well. Leave all other options unchecked and the sliders untouched.
Comment: You will notice that this is not in line with the advice of the big YediMaster who is in favor of the Create_Raw_Bayer(RGB) option. However this leads to a lot more noise and banding problems for me then the Bayer CFA-option .
Step 2.
Produce MasterDark (and MasterBias/MasterFlat if applicable) using the tutorial of Vicent Peris (forum message of April, 8, 2010).
Comment: In my test I used 25 Lights and 7 Darks only; no Bias and no Flats.
Step 3.
Calibrate the lights with the masters using Vicents tutorial.
Comment: -
Step 4.
Debayer the lights with the script from Ken Pendlebury. I select the Superpixel option.
Comment: This requires further debate because I expect that using the Superpixel option leads to loss of information and a smaller database for processes as ACDNR. Is bilinear the better option here?
On the other hand: if you do use bilinear you will notice that you will receive a alarm notice from PI during ImageIntegration: no significant data in MRS noiseroutine ....
Step 5.
Apply StarAlignment/standard settings work fine.
Comment: -
Step 6.
Use ImageIntegration/standard settings worked fine for me (winsorised sigma clip 5/5)
Comment: -
This workflow delivered the best overall results in my little test for Canon 40D cr2 files.
regards, HP
-
Thank you: Canon EOS 40D.
-
It would help if you posted the links to Vincents and Kens workflow?
-
It would help if you posted the links to Vincents ...
Sorry, but his name is Vicent. ;-)
-
Do you know of any reason why the workflow would differ for the Canon T1i .CR2 files?
-
Do you know of any reason why the workflow would differ for the Canon T1i .CR2 files?
The above workflow seems to go only as far as integration. There's a lot more to do.
Since the above workflow was posted in 2010 prior to the release of the BatchPreprocessing script (BPP), a better workflow is described in the PixInsight video tutorials Part 1 (http://www.ip4ap.com/pixinsight.htm) using BPP, and you can find several online videos describing just that part of the process.
There is a generic workflow for any DSLR found here:
http://lightvortexastronomy.blogspot.com/2013/12/tutorial-post-processing-pixinsight.html
Keep in mind that some parameters may need tweaking for your particular set of images.
It's also very useful to know that deconvolution should be applied to the luminance before stretching. This is explained in Part 1-20 of the PixInsight tutorial.
John C.
-
Do you know of any reason why the workflow would differ for the Canon T1i .CR2 files?
Bad me. I forgot to mention that all Canon images need to be processed with the CanonBandingReduction script (CBR), and for best results this should be done on the individual Flats, Bias, Darks, and Lights prior to ANY other processing. Using the ImageContainer script in conjunction with the CBR script (and possibly the ProcessContainer script if you want to combine several processes into one automated step) to automate the process makes life a lot easier. The results of using CBR is that the CR2 images are converted to FIT images. You can then separate the RGBL as needed for various processes.
Don't think of me as a guru about any of this. I just learned about all this last week. Once I got past my issues using the BPP script, things are taking off fast.
John C.
-
I'm puzzled as to why a BandingReductionScript would be needed, as I have never seen banding in my frames.
-
Gary,
many Canon's need it after BPP. If your camera does not, you can feel lucky!
Georg
-
Well, it's very possible I just don't know what to look for, but I have seen vertical bands illustrated here and there and I have not noticed anything like those.
-
I certainly would NEVER apply the banding script to my raw calibration or light frames!!!! - it would destroy their integrity.
-
I have seen banding with early 1000D firmware versions. Canon fixed this with an update. It was exceptionally bad. Modern DSLRs are probably not as affected. Dithering will generally fix it anyway.
-
There is a generic workflow for any DSLR found here:
http://lightvortexastronomy.blogspot.com/2013/12/tutorial-post-processing-pixinsight.html
Keep in mind that some parameters may need tweaking for your particular set of images.
It's also very useful to know that deconvolution should be applied to the luminance before stretching. This is explained in Part 1-20 of the PixInsight tutorial.
John C.
The Mercieca tutorial looks very helpful.