Page 1 of 2

Rho Ophiuchi Cloud with Black Point hack

PostPosted: Sun Aug 17, 2014 9:30 pm
by astronomer
I did a road trip at the end of June with my camera patched to output raw sensor data instead of pre-processed one by Expeed 0x67000000. Here are the result.

Rho Ophiuchi Cloud complex near star Antares. Total exposure is 100 minutes in 50 2-min sub-frames at F3.5 180mm. Tracking is done with AstroTrac TT-320X.

Image

Original Post here.

Now people have been asking if they could see the difference visually. Now the answer clearly is Yes!

Example:
DSS.jpg

This is occasionally the case when you use typical DeepSkyStacker pipeline with stock firmware output. And in fact outlined here by Wei-hao. It due to the temperature variation inside DSLR that cause dark current to vary frame by frame. But for perfect sensor calibration, it requires matching the black point after a complete dark subtraction with that of flat frames. When the 2 don't match, you have annoying color vignetting when the Y-intercepts of pixel intensity are not brought together to the origin before multiplying the sensitivity coefficient.

Now the problem raise when Nikon stock firmware simply clip off the dark current with some value greater than the actual one. This means more than 50% of pixels in the dark frame are zero making it impossible to do a proper calibration. With the hack, I could develop a new pipeline to normalize the dark current and extract the correct value before applying flat fielding.

The result under the same level, stretch looks like this:
custom_pipeline.jpg


Now you could use dark frame at different temperature or exposure time, or even taken at a different day, making it possible to collect more photons from the deep space.

astronomer

Re: Rho Ophiuchi Cloud with Black Point hack

PostPosted: Sun Oct 26, 2014 6:58 am
by SK8R5100
this is really cool and looks great
you are very talented

Re: Rho Ophiuchi Cloud with Black Point hack

PostPosted: Mon Nov 17, 2014 5:52 am
by aliensporebomb
Beautiful images. Amazing! I would love to try something like that but don't even know where to start.

Re: Rho Ophiuchi Cloud with Black Point hack

PostPosted: Mon Nov 17, 2014 7:35 am
by astronomer

Re: Rho Ophiuchi Cloud with Black Point hack

PostPosted: Sat Sep 05, 2015 8:41 am
by Silentg
Hi Astronomer,

Can you share that workflow? One of my biggest frustrations is not having enough dark frames after the fact (I know, I should take more at the time, but I don't have the discipline yet because I haven't been doing deep sky stuff for a while). Being able to take them after the fact would be a godsend from a convenience and accuracy standpoint, especially considering that my D800 is not cooling modified and never will be because it's my main body. The absolute most expensive software I'm willing to buy for now is probably Pixinsight, but only if it's worth it. DSS is doing the job just fine, even with the black point set to zero. I have no idea why it's working for me most of the time.

Re: Rho Ophiuchi Cloud with Black Point hack

PostPosted: Sat Sep 05, 2015 10:40 pm
by astronomer

Re: Rho Ophiuchi Cloud with Black Point hack

PostPosted: Mon Sep 07, 2015 8:12 pm
by Silentg

Re: Rho Ophiuchi Cloud with Black Point hack

PostPosted: Thu Sep 24, 2015 6:52 am
by astronomer

Re: Rho Ophiuchi Cloud with Black Point hack

PostPosted: Wed Oct 21, 2015 9:20 pm
by SXBB
I have an astro modified D5100 with hacked firmware and DIGICAMCONTROL for capture. I'd really love to be able to implement this. I haven't done any programming for for a while but I'm game to try R. How difficult would it be for me to use this script to pre-process my files? I've been shooting lights and darks, dithering and stacking them using my D5100 but haven't used the dark current mod yet. The darks have helped with hot pixels but I know I've been losing data from dark clipping. I have 5 nights of dark sky shooting coming up in a few weeks. I'd shoot everything on my trip using the true dark current hack if I can figure out how to do the pre-processing.

Thanks,

Bruce

Re: Rho Ophiuchi Cloud with Black Point hack

PostPosted: Thu Oct 22, 2015 9:59 am
by SXBB
Astronomer,

I found in another thread that you use MaximDL for calibration. Is it possible to use Photoshop or Lightroom instead? I'm very new at processing images and at this point I use Deep Sky Stacker and Photoshop. I understand that DSS doesn't process these files so I'm looking for possible alternatives.

Thank You for working on this project and making your process public. It's really an amazing advancement for DSLR Astrophotography.

Best,

Bruce