Problem with DSLR Workflow

[Enzo De Bernardini]

Hi Carlos,

The issue is beyond my knowledge, but try to change the 'Optimization window (px)' (Master Dark section on IC) from 1024 to higher and lower values (like 512 and 2048), and, if fails, input zero (this use the complete image). Use only the image #29 (or a few of them) to minimize the test times.

Best,

Enzo.

[Alejandro Tombolini]

Hola Carlos, esto normalmente me pasa en las fotos con Canon XS (solo en algunos objetos), y el resultado es fotos que en algunos casos quedan bien calibradas y en otros no. Si ki=0 o un valor muy diferente a los otros ki, ese canal no queda correctamente calibrado y hay hot pixeles. Lo que hago es calibrar manualmente esas fotos.
http://proxima-sur-fotos.blogspot.com/2011/04/ngc-3324-y-ngc-3293-detalles-de.html

Lo que dice Enzo suele funcionar tomando una ventana mayor o toda la imagen.


Hi Carlos, this normally happens to me in the images with Canon XS (Only in some objects), and the result are pictures that in some cases remain uncalibrated and in others are correct. If ki=0 or a value very different from other ki, this channel is not correctly calibrated and there are hot pixels. What I do is to calibrate manually these photos.
http://proxima-sur-fotos.blogspot.com/2011/04/ngc-3324-and-ngc-3293-details-of.html

What Enzo says usually works taking a major window or the whole image.

Saludos.
Alejandro.

[cdavid]

Gracias!  I am going to try that.  What is interesting...I redid the entire process again today.  I still get the errors, but if I calibrate my Light images only using a MasterDark that has not been BIAS calibrated, (No Flats and No BIAS)- and unchecked the optimize box- it worked.  The image looks good, I do not see any significant dark pixels but I do have vignetting which I am taking care of in DBE.

I will try your suggestions with the BIAS and Flats and see what happens. Thanks again!
Carlos

[astropixel]

Hi Carlos. That is very similar to the issue I had with a 1000D. The solution I came up with, was to take at least 50 bias (maybe 40 would be ok, but it was 50) frames at least 30 darks and flats and follow the expanded version of Vicent's process icon - http://pixinsight.com/forum/index.php?topic=2570.msg19019#msg19019

[cdavid]

Hmmm maybe that is my problem. I shot the flats in the morning when I woke up. I had left the scope imaging all night. I shot the darks that next night but the temp was similar, maybe 2-3 degrees F difference. I took the bias frames indoor later that day.

I am going to try all these suggestions. I don't have the time in the next couple days....have to work sometime

I'll let you know how I make out with this.
Thanks again!
Carlos

[astropixel]

Carlos. Sorry to labour the point - that was how I managed calibration frame acquisition, and it was always problematic - same bundle of errors.  If you do it that way you need lots of frames. PI handles temp deviation but needs lots of frames to produce a master representative of the light frame set.

E.G., The 5D MKII sensor temperature stabilises after 18 frames (3.5 minute exposures and a 10 second delay)  - the temp difference between the dark frames is visible. So 40 darks is quite safe (or perhaps 35 at a pinch), accounting for the 17 cooler frames in the dark and light sets.

After a lot of pain, my summary is take lots of frames if in doubt - put the camera in the fridge if need be! Better still get a Peliter cooler or a dedicated imaging cam.

[cdavid]

Thanks.....now it is clear to me what is going on.  I need to be more precise about my acquisition.
Thanks again.
Carlos

[midcon07]

Hi Guys,

My first forum post. I'm about a 1/3 of the way through my trial evaluation period.  Some interesting stuff here.  The first data set I processed was from my Canon (unmodded) T2i of M20. (The "red" came out surprisingly well.) At any rate following all the instructions under the Camera RAW work flow everything worked perfectly.
Today I tried a data set from a friends Canon Xsi. He also supplied 10 bias and 10 dark frames. With this set I've had almost exactly the same experience that Carlos has. My error message stated "D:/Jerry_JD_Astro/M33/FITS/4_Alignment/debayer_M33_43_09-14-09_c_r.fit(channel #0): Unable to compute noise estimate."   Upon clicking "ok" the Processing Console reads <* failed *> and all processing ends.
Like Carlos I have started completely from the beginning several times adhering to all the instructions as close as possible. I don't say "adhering exactly" because I have some options with v1.7 x86 that weren't available apparently under v1.6.  As with Carlos I'm still searching for the answer.

Jerry

[cdavid]

Hi Jerry, glad to know I have company.  I have not had time this week with work and all to go back to the most recent suggestions ( the last couple of posts) but I plan on it this weekend.  Interesting however is that the same data had no problems in Nebulosity but did fail in DSS....so I think it probably has to do with as suggested different temperatures and factors when shooting the lights vs. Darks vs. Bias and flats.  Stay tuned as I continue playing with this.

Cheers
Carlos

[midcon07]

Yeah I will, Carlos.  I'm having quite a time figuring out why the set from the T2i had no problems while the set from the Xsi does. I'm tempted, as a matter of fact I think I will, reprocess the set from the T2i and see if everything comes out ok again. Probably be a couple days before I get it done though.

BTW is there a DSLR section to the forum?  I really haven't had a chance to explore the whole thing yet.

Waiting for rain in Northern AZ.

Jerry

[astropixel]

http://pixinsight.com/forum/index.php?topic=3389.msg23326#msg23326

Vicent's comments.

[cdavid]

There you go....thanks Astropixel! 

[cdavid]

Well....I've tried every permutation still no success.  I guess it is the data, my darks and biases must all be at different temperatures.  I've given up on this set in pixinsight. I had success pre-processing, registering, and stacking it in Nebulosity so I was able then to bring the image into pixinsight for processing.  I will give it another try with new images another time and this time make sure my temperatures are all the same and that I shoot enough Darks and Biases.

Thanks again everyone for all your help!  If you are curious, my final image after processing is at:

http://web.me.com/carlosdavid1/ThroughSpaceAndTime/The_Wizard_Nebula.html

Cheers
Carlos

[Cleon_Wells]

Carlos, nice image.
If you have the time to revisit the PI Calibration problem, Debayer the MasterBias, the MasterDark and the MasterFlat and use the Probe set to around 9-pixels and write down the levels.
My Debayered MasterBias has a level of 0.0156xx, the Debayered MasterDark should be slightly higher. Here's a screen shoot of one  Calibrated raw dark sub Debayered  and one Calibrated raw flat sub Debayered. I used 20-bias subs iso100@1/4000 to create the MasterBias, 10-dark subs iso400/120sec for the MasterDark and 10-flat subs iso400@1/250 to create the MasterFlat. I'm interested. My images are from a Canon T1i.
Cleon

[cdavid]

Hi cleon, thanks!  I will give this a try and post my findings when I get a moment.
Cheers
Carlos