Input needed

blackstar

Senior Member
Hi,

I scrapped a camping trip to Joshua Tree NP for this year's last chance of shooting the MW photos due to the pandemic and the wildfire disaster. But still can't stop taking the MW shots... So I checked up PhotoPills and Stellarium and saw the image at my place at a time within a narrower window of visibility (~8:00 to 9:30 PM in Oct and 6:00 to 7:00 PM in Nov):

Screenshot:
Screen Shot 2020-10-12 at 1.08.11 PM-s.jpg

So I grab my camera and enter my backyard, pointing to SW, and take some shots even can't see the MW with naked eyes at all. This is the original jpg:
2020-10-14 20.18.24-s.jpg

Best PP I tried: (MW is there, barely visible in the image)
.jpg

I would like to hear your opinions and comments on the possibility of shooting better MW images at this place, time, and condition, as well as improving pp. Will take either "possible and how" view or "impossible and why" view. Thanks
 

blackstar

Senior Member
Thanks, Jake.

I know light pollution in the city can kill much of your sighting and imaging at nature objects. That's one point of impossible mission to overcome in this environment, I guess. Here is the PP image (Darktable) from RAW: (not much better... )

2020-10-14 20.18.24-DT-s.jpg
 

BF Hammer

Senior Member
I have many thoughts here. I will try to keep it organized. First of all, while I have places dark enough to photograph the Milky Way within 90 minutes drive, I have known all along that trying to do that in my suburban back-yard is a fantasy. I can take photos of planets and some deep-space objects, but not the Milky Way. Andromeda is near-impossible too. Light pollution is the price we pay for being safe at night on roads and walking in urban places.

My first thought is that in the configuration options of Stellarium is a check-box to apply estimated light-pollution to your sky to filter what you are able to see. Turn it off to see what could be visible, but turn it on to see what really will be visible.

Your technical settings are close to what I would do. For a f/3.5 lens I would bump up that ISO to beyond 3200. And you are going to need to take more photos. If your D3500 has an intervalometer like my D7000 and D750 do, it should be taking a 15 second exposure every 20 seconds. And have it take 80 photos, maybe even more. I have a Sigma 18-35mm f/1.8 lens available for DX format, so I shoot with it at f/1.8. On my FX D750 I have a Sigma 20mm f/1.4 that I actually use at f/1.8. I will set to ISO 3200 at those settings and take 15 second exposures.

Then there is the reason for taking 80 photos at least. You are going to use software to stack the images and the light data will be additive to make the Milky Way pop and blend out that excess noise from the high ISO. If you are using a Windows PC, that software will be Sequator. Mac users have been using Starry Sky Stacker for even longer. The software does the masking of the foreground and handles the movement of the stars across the sky.

The key to your photo is a darker night site. You cannot overcome ambient light pollution in a wide-field photo. DarkSiteFinder is one tool to help, but it is not perfect. In my case, I found a park just outside of a small town that is considerably darker than the map would indicate. There is not a single street light in any direction for a mile. The woods around screen out the street lights of the small town nearby. To the north is all state wildlife refuge and public hunting ground. Just a few farm homes in the location. And I discovered years ago that I can actually see the Milky Way with the bare eye there. Not the colorful version in my photos, but the faint milky-white that the Milky Way gets it's name from.

Screenshot_2020-10-17 Light Pollution Map - DarkSiteFinder com.jpg

So I am blessed to have some locations. But if you look at that map on a national level, my location in southern Wisconsin is just all light pollution. I see better dark skies in pockets in California.

Edit: after finally reading what the Bortle Scale actually is, maybe that map is more accurate than I thought. My backyard would easily be class 6-7 while I think I would call my country park location a solid class 4. The map says 1-2 class difference and I observe 2-3. The visibility of the Milky Way is one of the key determining criteria for classifying on the Bortle Scale.
 
Last edited:

blackstar

Senior Member
Thanks, BF, for your insightful thoughts and great tips.

First, I use ClearDarkSky to seek places and spots of low light pollution -- real dark sky. And like you, found a local nearby old recreation area that is polluted much less than in the city (orange-yellow zone instead of red). Explored it in early summer and ran away by being attacked from a heavy swamp of deadly poisonous mosquitoes near a river. The people in the area mentioned good time to visit and avoid mosq is winter... Nothing's perfect!

Appreciate your tip about Stellarium. I checked it and found it defaultly set light pollution scale = 2 that shows the MW image visible (see my original image #1). Turn local data on and scale is up to 7! MW image no longer visible! Exactly the real scene like image #2. So I was cheated to run out to shoot some hopeless photos...

Your idea and insight of stacking night sky images get me into serious consideration on putting some effort on the processing tool. I now only use mac, so the free Sequator isn't a choice. SSS is a paid app for mac user. Don't know which one you use and wonder if you would share with some examples, especially of MW images? I have got a free small app: Lynkeos. Haven't got a chance using yet, so don't know if it works.

I do see some light and make some clear points out from your post. Really appreciate.
 

BF Hammer

Senior Member
I told this story in the "post your Milky Way" thread. This is my very first attempt to photograph the Milky Way in 2017. I was not at the "dark park" purposely to do this, I was trying to capture aurora that night. I had no idea how to do it. My camera body then was a D7000 and I did (still do) have a nice lens in the Sigma 18-35mm f/1.8 Art.

nvV13l4.jpg

Single, non-stacked image with too-low ISO and too-long exposure time. I processed the crap out of that thing to get that photo.

But one year later I did some research. I used google and the dark skies map to locate a county park in Amish land to my north. I mis-used Google Earth to estimate where the Milky Way core would be at what time. Don't use Google Earth that way, it will steer you wrong. I thought it would be above a picnic shelter, instead it was prominently above the outhouse. I kept hearing coyotes so I was unwilling to explore much in the dark. Important part was I began to learn stacking with an early version of Sequator and what I needed to refine.

lIk8ftV.jpg

This was the shot I wanted.
jCohD7p.jpg

Fast forward to 2019 and I have many hours of watching pros on youtube. I study that same Amish-country park from Google Earth and decide I may have a south-view possibility at the boat dock for the lake. I thought the year before that it may be too wooded. I was armed with a new D750 but did not have a good full-frame wide-angle lens to use. So I just used the Sigma 18-35mm f/1.8 and tried. I made technical mistakes, including mistaking the zoom for the focus and not noticing in the dark. This was the one image I salvaged before the rest were taken out-of-focus.

KCdHBLQ.jpg

But 1 month later I had purchased my Sigma 20mm f/1.4 and returned determined to not screw-up.

ViKJbfm.jpg

7xAif9h.jpg

So then this past summer I tried some impromptu photos at that same park as the original 2017 photo. I was there photographing Neowise. I used the 20mm f/1.4 and I was also armed with a new 150mm square filter holder and light-pollution filter to fit. It helped a bit with the sky contrast. But I did not get as many good photos as needed because of dew and fogging. Also you see the reflection of the lens hood in the back of the filter. I should crop it, but the photo loses something when I try. So I leave it.

HyP9r92.jpg

I credit the improvements to Sequator over the past 3 years partly for my better photos. I did invest in some better equipment and put in personal experience time. I'm still learning. I do use a Windows PC so no help with Mac software. Some things I did learn and should try to do better is that you can do your own layer-mask with a photo of the foreground you take before or after and try to make a nice long-exposure at lower ISO. I did that with the outhouse photo and it worked well. I forgot to do it with the lake photos, and instead had to rely on Sequator giving me a foreground with noisy data. It worked OK, but it could have been better with the manual touch.
 

blackstar

Senior Member
BF, Wonderful photography story and examples.

Like to ask clarification and confirmation with each photo image you post:

#1 -- single shot, post processed, no stacking
#2 -- stacked image (can you share one individual image b4 stacking for comparison and how many single shots?)
#3 -- "the shot I wanted" -- ???
#4 -- single shot, post processed, no stacking
#5 -- single shot, new FF lens
#6 -- stacked image from #5 and others (wonder how you stacked out the nice color cloud below the MW? how many single shots?)
#7 -- Stacked image (how many single shots and one example?)

Pardon me for so many requests. Thanks
 

BF Hammer

Senior Member
BF, Wonderful photography story and examples.

Like to ask clarification and confirmation with each photo image you post:

#1 -- single shot, post processed, no stacking
#2 -- stacked image (can you share one individual image b4 stacking for comparison and how many single shots?)
#3 -- "the shot I wanted" -- ???
#4 -- single shot, post processed, no stacking
#5 -- single shot, new FF lens
#6 -- stacked image from #5 and others (wonder how you stacked out the nice color cloud below the MW? how many single shots?)
#7 -- Stacked image (how many single shots and one example?)

Pardon me for so many requests. Thanks

Shot #2 was the portrait mode photo I took that night, I also composed in landscape to cut out the outhouse. But it looks more amateurish with just some tree-tops at bottom frame. I did not know to take several dozen photos at that time, so I only had 11 images in the portrait pose. Not even sure if I used all 11 in the stack, might have had to toss some with things like satellites showing. But I do have the Jpegs still saved, here are 2 outtakes. A sky exposure and a darker exposure for the foreground. A simple layer masking allowed me to fix that blown-out foreground because there was an actual light on a pole just out of frame.

DSC_4968.jpgDSC_4984.jpg

#3 is the shot I wanted because I wanted to have the Milky Way at that location, considerably farther to my right. I estimated it would not line up for another 3-4 hours and I was not waiting that long. I was already past closing time for the park, so any sheriff deputy cruising by and noticing me likely would have asked me to vacate. At the least.

#5 is really a stacked photo. It was my second series of images because I moved me and my chair off of the boat dock. The clouds and twilight were right for me to begin serious work. Single frame outtake below.

DSC_0385.jpg

But #6 was actually me goofing around as I was waiting for the clouds to move out and the twilight to darken. Turns out I liked the "selfie" shots and Sequator made the cloud motion blend into something dramatic. That bright area right side on horizon is light pollution from Portage, WI and some sunset twilight. Nearest community in that general direction. There were about 35 photos taken in that set. I think that is what I programmed into the camera intervalometer for all 3 sets that night (there is also a portrait mode set). An outtake below, but I think I had to throw out a good 10 photos from the beginning of the set that had clouds over most of the scene.

DSC_0347.jpg

#7: I took more than 40 images for that particular set. However I only could use 20 images max to stack from the set. Lens and light-polution filter were fogging up badly as they were being taken, and I did not notice until near the end. 2 outtakes below.

DSC_1435.jpgDSC_1461.jpg

I have since purchased a USB-powered lens dew heater wrap, which helps that fogging on lens, but not if it happens on the filter.

So it should be plain to see that while I have learned much on this subject, I am still learning and gaining experience. I will take progressively better photos in the coming years. Unless I veer off into another photography subject and devote myself to that for a while. :D
 

Attachments

  • DSC_0350.jpg
    DSC_0350.jpg
    28.4 KB · Views: 47

blackstar

Senior Member
Thank you so much, BF.

After all your explanation and sharing studied, still have more inquiries for you to help: Do you just throw raw files into Sequator to run stacking? or convert raw into TIFF and pre-process them b4 stacking with Squator? or just use jpg's to run stacking? The fact that the final stacked images stand out as brilliantly beautiful from the seemingly crapy single exposures really makes the (stacking) process impressive. (But I realize your effort and investment in equipment to improve single exposures also make a great difference. And post process also a big factor.)

Viewing from your outtakes, it seems the MW images are visible in them, at least faintly though, if not very clear. So, I would assume for all four sets of single exposures, you were there able to see the MW in the sky with naked eyes (at least faint scene). If I'm correct, the place you took shots must have a light pollution scale = 3 or 4. So it's not comparable (aside from your investment and efforts) to the place I took shots -- backyard in town (scale=7). And was not able to see even faint MW in the sky.

I think I'm going to get a new Rokinon 10mm f2.8 wide angle CS lens for my D3500 after failed to try two new Rokinon 12mm f2.8 fisheye FF lenses: incompatible lens! I figure that using Rokinon 14mm f2.8 FF with d3500, it will come down to ~18mm wide which overlap my kit lens. OTOH, when upgrade to FF body, Rokinon 10mm f2.8 CS lens will come down to ~15mm wide, which is close to 14mm. Am I on the right path? Afterward, I may try to take shots at other darker places and hope to get better result.
 

BF Hammer

Senior Member
Thank you so much, BF.

After all your explanation and sharing studied, still have more inquiries for you to help: Do you just throw raw files into Sequator to run stacking? or convert raw into TIFF and pre-process them b4 stacking with Squator? or just use jpg's to run stacking? The fact that the final stacked images stand out as brilliantly beautiful from the seemingly crapy single exposures really makes the (stacking) process impressive. (But I realize your effort and investment in equipment to improve single exposures also make a great difference. And post process also a big factor.)

Viewing from your outtakes, it seems the MW images are visible in them, at least faintly though, if not very clear. So, I would assume for all four sets of single exposures, you were there able to see the MW in the sky with naked eyes (at least faint scene). If I'm correct, the place you took shots must have a light pollution scale = 3 or 4. So it's not comparable (aside from your investment and efforts) to the place I took shots -- backyard in town (scale=7). And was not able to see even faint MW in the sky.

I think I'm going to get a new Rokinon 10mm f2.8 wide angle CS lens for my D3500 after failed to try two new Rokinon 12mm f2.8 fisheye FF lenses: incompatible lens! I figure that using Rokinon 14mm f2.8 FF with d3500, it will come down to ~18mm wide which overlap my kit lens. OTOH, when upgrade to FF body, Rokinon 10mm f2.8 CS lens will come down to ~15mm wide, which is close to 14mm. Am I on the right path? Afterward, I may try to take shots at other darker places and hope to get better result.

When I use Sequator I use the Raw files. I have learned to help things out a bit by viewing each file and tossing out the ones that have visible blur or streaks from satellites, fireflies, aircraft. But Sequator has a checkbox in the processing menu to help fix those if those photos are included. There are HDR and auto brightness options as well as an adjustable light pollution filter option. I may run the stack 3 or 4 times with the different options to get a pleasing TIFF. Then I do the fine-tuning with RawTherapee and GIMP.

My workflow changes here if I am taking photos of a planet, comet, or a galaxy. I am using Deep Sky Stacker for those larger objects in the frame where extra stars appear. I still use Raw images with DSS, but I will do even more runs and try different options. Stacking is a lot slower in that software, but I will get better results if there is no stationary foreground. Planets with no stars in the frame normally requires I go to RegiStax which is what that software is designed to do. My TIF exports actually crash the program so I export .PNG to make it work. I avoid working with JPG where I can in post.

Yes I can actually see the Milky Way at the locations I take the photos at. Faintly, about Bortle class 4 as I read the link to Wikipedia right.

As for your lens choice, I am not much help. I have always had at least a D80 or newer DSLR of similar class. My lens options were never limited much with F mount. I never had anything wider than 18mm with a crop sensor body. But I can say I love the Sigma 18-35mm f/1.8 Art. It should work with your body, and I think you might be surprised what they sell for used. I remember buying mine sort of gray-market for under $600, and that was 4.5 years ago. Of course now that I have a FX format D750, my widest lens is now 20mm which is about 14mm equivalent to a crop body. I'm OK with that.
 

blackstar

Senior Member
Thanks, Needa.

The discussion of MW processing from the link is more about how to process an under-exposed MW image. In my case, it's more about heavy light pollutions making MW invisible and so the stacking process could reduce noises and bring out enhanced (visible) MW image (assuming each single shot already in normal or over-exposure).
 

blackstar

Senior Member
Thanks again, BF.

Ok, so RAW is the way to go! Now it seems to be that Sequator has everything you need plus many extra features for stacking images and it's free! But I am not so lucky because I use mac. Wonder does SSS also have the same features and functions as Sequator? or even better than Sequator? :) (I want to make sure b4 paying for it)

For the moment, I am keeping the new Rokinon 12mm f2.8 fisheye FF and to practice using it (it only works in M mode with d3500).
 

BF Hammer

Senior Member
Thanks again, BF.

Ok, so RAW is the way to go! Now it seems to be that Sequator has everything you need plus many extra features for stacking images and it's free! But I am not so lucky because I use mac. Wonder does SSS also have the same features and functions as Sequator? or even better than Sequator? :) (I want to make sure b4 paying for it)

For the moment, I am keeping the new Rokinon 12mm f2.8 fisheye FF and to practice using it (it only works in M mode with d3500).

I would say start searching on youtube for SSS tutorials. I know I watched at least one a couple of years ago. I watch a lot of youtube stuff on my smart TV and at least 1/3 of it tends to be photography-related. I do remember that Raw files are recommended and part of the workflow is masking the sky portion of the base image. So it really was the model for Sequator. I really should take dark reference frames and the hot-pixel frames as part of my workflow, but I don't think I have really needed it yet. I lack the patience to do those boring parts. Maybe I would get even better results if I did.
 

blackstar

Senior Member
Below is a link to a darktable image edit. If you download the original image and place it in several folders you can copy a sidecar file from each edit to one of the folders, import the image from the folders. Then you be able to see what steps were taken to achieve the results.

https://discuss.pixls.us/t/milky-way-processing-in-darktable-looking-for-feedback/19702

Thanks, Needa. If I understand, in order to see what steps were taken to achieve the results, I have to copy the original raw file (e.g., ori.NEF), one processed jpg file (ori.jpg) and one xmp file (ori.NEF.xmp) and place them in a folder, then open the files in DT? But how? I tried importing the images from the holder in Lighttable and both NEF and jpg files show up. Then I choose either one of them in Darkroom. It just show some preset steps in the history stack without any editing steps from the xmp file. Wonder what I did wrong? Thanks for help.


 

Needa

Senior Member
Challenge Team
Thanks, Needa. If I understand, in order to see what steps were taken to achieve the results, I have to copy the original raw file (e.g., ori.NEF), one processed jpg file (ori.jpg) and one xmp file (ori.NEF.xmp) and place them in a folder, then open the files in DT? But how? I tried importing the images from the holder in Lighttable and both NEF and jpg files show up. Then I choose either one of them in Darkroom. It just show some preset steps in the history stack without any editing steps from the xmp file. Wonder what I did wrong? Thanks for help.



blackstar

No need to import JPG, just NEF and XMP, requires a different folder for each XMP plus the NEF. You can also set your options (gear icon then import tab) to ignore jpg but if you normally import jpgs remember to set it back. Just tested downloaded the NEF and the first XMP to a folder named test, right clicked the image and selected open with dt, this is also how I open jpgs in dt to edit, as I don't import jpgs. The right click method for me opens it directly in darktable mode.The image opened in dt and the history stack shows 19 edits. Then by going to any of the listed modules you can see that modules settings. If you want to output it as a jpg just export as normal.
 

blackstar

Senior Member
Thanks, Needa.

I did exactly as you expressed, but when the image opened in Darkroom the history stack shows only 9 default pre-steps (everytime open a new image file, DT system pre-edits before you make your own editing) and no change on the NEF image. I went back to check the xmp file in the folder and see it's size has changed from 10 KB to 5KB. So I think everytime I open an image file, DT would write a new xmp which would overwrite the old xmp file if one exists (which does in this case). I am stuck here! No idea how you manage to see others' editing steps... Help!
 

blackstar

Senior Member
Gee, How you get that? I did replace xmp with downloaded one, but same thing happens again and again. Can you check your DT preference setting? In "core options" panel, there is "xmp" section which has 3 options: Write sidecar file for each image, store xmp tags in compressed format, and Look for updated xmp files on startup. What are your settings on these options?
 
Last edited:
Top