Wide Field Processing of Cygnus

[msmythers]

With many very wide field images being captured lately I thought I'd share a video of a different set of processing steps then most would use. Yes there are the standard DBE, BN and ColorCal, even TGVDenoise along with ColorSaturation, SCNR and MorphologicalTransformation for star brightness reduction. I use DBE a little different because of the Milky Way running through a large portion of the image, more like what you might do with a large nebula.

What makes most of this different is I do all of the major work while linear. I only do a final tweak of cropping and lower the overall brightness with Pixelmath after stretching the image.

My goal is not to show that this is how you must process your image but to make you think about other possibilities with all the tools and scripts PI has to offer.

Here are the steps used in the image:
DBE
BN
CC
PixelMath    making a Gray mask
TGVDenoise   using the gray mask, local support and Statistics tool for setting.
StarMask   from linear image
MT   for star brigthness reduction using star mask
ColorSaturation
DarkStructureEnhance Script
SCNR
ColorSaturation
HistogramTransformation    stretch to non-linear using STF as input
Crop   remove bad edges
Pixelmath   final brightness reduction

This is what I consider a somewhat simple processing. Certainly the image could be tweaked in many ways. I could do a better job maybe with the color cal or DBE?? I wanted to show the subtle changes in this region. I have 2 images in the folder, one is with the star reduction and one without, just to show the impact. The video has no audio and the encoding smoothed the image to the point where it's not easy to tell that it needed a little help with the noise. It did.

The image is from last October. It was 23 exposures of 120 seconds each at ISO800. The camera is a DIY modified Sony Nex-5 with a Minolta 45mm pancake lens at f4.5. The only filter was a IR/UV cut filter, no LP filter. This was taken from my red/white zone front yard in Florida.

https://drive.google.com/folderview?id=0B13r3kYqTf8NUy1lemJleFRfYnc&usp=sharing

If the video starts out blurry, pause and then play.



Mike

[Jason Tackett]

Thanks for sharing this instructive video. I like the simplicity of the workflow and I learned some subtleties of a couple of the tools that I wasn't aware of. Very nice.

[mmirot]

Is there supposed to be sound?

Max

[bmhjr]

Thanks Mike.  This is good timing because I have a 35mm shot of the same area I need to work on.

Bill

[msmythers]

Max

No audio, I mentioned that in the posting. I have speech issues so I don't have a microphone on my computer. Someday I hope to figure out how to add text into videos.


Mike

[bmhjr]

Mike,

There is a lot of great info in there, thanks.  One part I am unsure of is the TGVDenoise settings.  You showed getting the values for edge protection from the statistics.  But, the TGV setting shows 0.000186 and the Stats for StdDev show 0.0186.  What am I missing on the order of magnitude setting?

Thanks
Bill

[msmythers]

Bill

With the gray mask I need to change the setting from 0.0186 to 0.000186. I find I get better results with the gray mask. With the new, not officially released version of TGVDenoise there is a section called Image Model. With the use of the additive noise setting you can use the StdDev value. I like the new module much better but I didn't want to show it since there is no version yet for people with OSX on a Mac. I know Carlos is working on a whole new TGV module which it sounds like these settings or the way we use the module is going to change. I'm looking forward to it and the new DBE when it's released. What other software do you get major changes for the price of the original admission.


Mike

Oh yeah, thanks for asking.

[bmhjr]

Bill

With the gray mask I need to change the setting from 0.0186 to 0.000186. I find I get better results with the gray mask. With the new, not officially released version of TGVDenoise there is a section called Image Model. With the use of the additive noise setting you can use the StdDev value. I like the new module much better but I didn't want to show it since there is no version yet for people with OSX on a Mac. I know Carlos is working on a whole new TGV module which it sounds like these settings or the way we use the module is going to change. I'm looking forward to it and the new DBE when it's released. What other software do you get major changes for the price of the original admission.


Mike

Oh yeah, thanks for asking.

Thanks for the explanation, I forgot about the mask.  I agree that steady updates are a huge plus with Pixsinsight.  That, and the fact you can do what ever you want to solve a problem (for good or bad in my case) keeps interest high and results always improving.

[Alejandro Tombolini]

Hi Mike, nice image! and thank you for sharing the video, it is very clear. Only a comment about Color Calibration, I would have used the average of more stars as a white reference, may be the all picture in this case. More stars will ensure you that a sufficiently complete set of spectral types is being sampled.
Saludos, Alejandro. 

[msmythers]

Alejandro

Thanks. I tried the entire image but the color was more magenta. I probably needed to lower the structure and noise settings with that lens. Maybe on the next revision, haha.

Thanks again.


Mike

[mmirot]

Max

No audio, I mentioned that in the posting. I have speech issues so I don't have a microphone on my computer. Someday I hope to figure out how to add text into videos.


Mike

Clearly, I have reading issues so were even.

[biosif]

Hi.

I have done something, but i think if you want to share original image, to try.

[msmythers]

The images in the folder are the finished images. One is with star brightness reduction and one is without. So many people with wide field images don't reduce the brightness of the stars which can make it difficult to see the gas, dust and nebula patterns in an image.

The video is just to show some techniques to try when working on a wide field image of the Milky Way.


Mike

[wimvb]

Great video, Mike. I will give your workflow a go on an image I have had problems with.
Thanks for sharing.

Wim

[kolec]

Very nice video

kolec