Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
06-01-2021, 06:10 PM - 2 Likes   #16
Senior Member
Grimmus's Avatar

Join Date: Jun 2010
Location: Newcastle (Aust.)
Photos: Gallery | Albums
Posts: 186
QuoteOriginally posted by jon.partsch Quote
This is confusing to me. Are you saying that Sequator will basically shift the starts in each exposure so that they all line up, but will *not* also shift the foreground? If so this is interesting, and I may want to try this approach. It sound like it may be difficult to implement this along with additional offset frames intended for a stitched panorama. Are you suggesting this as an alternative approach to a single frame milky way scene *not* using Astrotracer?
Exactly! In using Sequator, you simply pick a shutter speed that will be short enough not to blur stars (I use about 10 seconds with an 18 mm lens on an APS-C 24 Mp sensor). Take as many shots as you like, more is better. When you process in Sequator, there is an option to "freeze ground". You will then paint over the sky and Sequator will create a mask. In processing, it will move all the stars over the top of one another from all exposures and do a simple average on the foreground with everything fixed and spit out a result. It sounds complicated but is in fact pretty foolproof and straightforward - lots ot YT videos on it. Here's one I prepared earlier (x posted from the DA*11-18 thread):



06-01-2021, 06:29 PM - 1 Like   #17
Senior Member
Grimmus's Avatar

Join Date: Jun 2010
Location: Newcastle (Aust.)
Photos: Gallery | Albums
Posts: 186
QuoteOriginally posted by jon.partsch Quote
Yes, I was wondering what is the best exposure value to use to get (1) thin lines, (2) good color in the trails, and (3) maximize number of trails. The way I see it, there is no option for trial and error, so I need to decide on the EV that I will use and hope for the best. This is one of my primary concerns. Also, I don't see a way to use the K-1ii internal intervalometer for exposures longer than 30 seconds. If there is a way to do this, please explain.
Don't know how to multi quote, so.....
I use a Kp so don't know the K-1 II controls but they came out about the same time. On the Kp, I have a user mode set for startrails. It is based on B mode with timed shutter activated opening up the use of shutter speeds up to 20 minutes. I then have interval shooting set, minimum standby interval, all NR off. I have it set for 20 exposures of 3 minutes, ISO800 and F5.6 - but if you want to adjust this on the night, select the user mode and then change whichever settings you want - it's still much easier to remember than doing it all in one hit.

(BTW I have set up 3 user modes for night stuff - one is high ISO with NR on which I use for single exposures, composition and focus checking, the second for stacking short exposures and the third for startrails. Test them in the backyard to avoid disappointment in the field).


Don't get too hung up on ISO. Pentax use ISO invariant sensors - you can select all the images you took and push them 5 stops in post processing if you want (equiv ISO100 to 3200). ISO is only making changes AFTER the light hits your sensor - what you capture is based only on aperture and shutter speed. My approach is to stay a little on the low side as keeping control of blowing highlights (stars) is the key issue. You can also get cute in post by adjusting exposure progressively on the trails shots to get tapers on your star trails - another x-post example below:

06-02-2021, 02:44 PM   #18
Site Supporter
Site Supporter
jon.partsch's Avatar

Join Date: Oct 2009
Location: Northern California
Posts: 442
Original Poster
QuoteOriginally posted by gifthorse Quote
For an initial set up for wide field Milky Way, leave the high ISO and the long exposure noise reduction on. Turn off the auto focus. I'd start with an ISO of 1600. I'm not familiar with the D-FA 24-70 so I'd set the aperture stopped down 1 or 2 clicks (not full stops, just a click or two) from wide open. With the Astrotracer, I'd start with an exposure time of about 2 minutes. If possible, aim the camera to the South. The galactic center, the brightest part of the Milky Way, will be low in the southern sky and will rise at around 10 pm. Check the exposure. If it's overexposed, I'd decrease the ISO and try again. If underexposed, I'd increase the exposure time first and then increase the ISO if needed. The Moon will rise around 3:30 am and will be about 10% to 20% illuminated. That should illuminate the foreground without washing out the Milky Way too badly.That's a start, but be prepared to make adjustments.
For star trails, I'd start with an ISO of 800. For maximum stars you'll want the lens wide open. Turn off the high ISO and the long exposure noise reduction and auto focus. Aiming the camera north, toward the North star, will give you shorter and more curved star trails. Aiming East or West will create longer and straighter trails. If you are doing 5 minute exposures, the Moon above the horizon will wash out the sky. For star trails, I use a freeware program called Startrails. It's easy to use.
What you see in the sky won't look like the images you take. The camera captures a lot more than your eyes see. The sky glow in the center of my first image is a small city some 30 miles away. The glow isn't visible with the naked eye but shows in long exposures. As you get more practice it will get easier.
Thanks again for all this additional detail. two clicks down is f/3.5. I couldn't help myself, and got the Sigma 35mm f/1.8 Art here on the forum. It should arrive tomorrow. It has a reputation of being tack sharp with no coma at f/2.8 and pretty good wide open too. I will be testing it out as well. I won't be able to get nice foregroud to the south at my first location, but this should be possible at my second location.

That's a great tip on letting the moon illuminate the foreground! I hadn't thought of that, though not sure I can stand to stay out that late. have been looking at photos online, and I can clearly tell which ones have a foreground shot during daylight and stitched in later. They look awful. I would much prefer a dark silhouette foreground to that.

QuoteOriginally posted by Grimmus Quote
Exactly! In using Sequator, you simply pick a shutter speed that will be short enough not to blur stars (I use about 10 seconds with an 18 mm lens on an APS-C 24 Mp sensor). Take as many shots as you like, more is better. When you process in Sequator, there is an option to "freeze ground". You will then paint over the sky and Sequator will create a mask. In processing, it will move all the stars over the top of one another from all exposures and do a simple average on the foreground with everything fixed and spit out a result. It sounds complicated but is in fact pretty foolproof and straightforward - lots ot YT videos on it. Here's one I prepared earlier (x posted from the DA*11-18 thread):
Grimmus, both your photos look very much like what I hope to capture, so I will take your advice seriously. As I mentioned in my reply to gifthorse above, I can clearly tell the photos online where the foreground was photographed separately under different lighting conditions and stitched in later - they look awful. I'm going to give the approach using Sequator some thought and look for some YouTube videos as you suggest.

QuoteOriginally posted by Grimmus Quote
Don't know how to multi quote, so.....
I use a Kp so don't know the K-1 II controls but they came out about the same time. On the Kp, I have a user mode set for startrails. It is based on B mode with timed shutter activated opening up the use of shutter speeds up to 20 minutes. I then have interval shooting set, minimum standby interval, all NR off. I have it set for 20 exposures of 3 minutes, ISO800 and F5.6 - but if you want to adjust this on the night, select the user mode and then change whichever settings you want - it's still much easier to remember than doing it all in one hit.

(BTW I have set up 3 user modes for night stuff - one is high ISO with NR on which I use for single exposures, composition and focus checking, the second for stacking short exposures and the third for startrails. Test them in the backyard to avoid disappointment in the field).


Don't get too hung up on ISO. Pentax use ISO invariant sensors - you can select all the images you took and push them 5 stops in post processing if you want (equiv ISO100 to 3200). ISO is only making changes AFTER the light hits your sensor - what you capture is based only on aperture and shutter speed. My approach is to stay a little on the low side as keeping control of blowing highlights (stars) is the key issue. You can also get cute in post by adjusting exposure progressively on the trails shots to get tapers on your star trails - another x-post example below:
Again, these star trails look very much like I was hoping to capture. What is the total exposure time on this one, do you remember? I was able to determine how to set a long espouse in a user mode on my K-1ii, but there is no option to use the internal intervalometer in Bulb mode (the option no longer appears in the drive mode menu). If this is different on the KP, then I am a bit envious, but I have an external intervalometer that should work.

Name:  20210601_201748.jpg
Views: 103
Size:  378.3 KB

This is a user mode I have called "milky way". I guess I should eliminate use of the 2-second delay as it probably makes no difference for long exposures.

I will plan a few milky way (not star trail) shots, since these don't take a lot of time. I will try single frame astrotracer, the stacking method you mention using Sequator, and a stitched version using three or four 35mm frames to cover the same approximate angle of view (in portrait orientation) as 24mm.

Star trails will be one attempt at each location, and I won't know my result until stacking later in software, but I will not sweat it too much on the absolute EV that I use. I suppose I could try a shorter exposure for short trails, like 10 minutes, to try and dial it in and avoid any blown highlights - that wouldn't take too much time.

~ Jon
06-02-2021, 09:11 PM - 1 Like   #19
Senior Member
Grimmus's Avatar

Join Date: Jun 2010
Location: Newcastle (Aust.)
Photos: Gallery | Albums
Posts: 186
QuoteOriginally posted by jon.partsch Quote
What is the total exposure time on this one, do you remember? I was able to determine how to set a long espouse in a user mode on my K-1ii, but there is no option to use the internal intervalometer in Bulb mode (the option no longer appears in the drive mode menu). If this is different on the KP, then I am a bit envious, but I have an external intervalometer that should work.
This was 72 minutes total - 36 shots of 2 minute exposure (f6.3 and ISO800). The foreground was 10 shots at 60 seconds - importantly, after waiting for the moon to rise as per @gifthorse's approach (by which time I had been locked in but that's another story ). On the Kp I was able to set this sequence up, and take a comfortable position while the camera clicked through the 36 x 2 min exposures. Bummer the K-1 II doesn't do this but no problem if you have the external intervalometer.

There is a bloke called Richard Tatti who has a youtube channel called nightscape images - it is well done and has been a great source of information and inspiration. There are several videos where he walks through the capture of stacks, panoramas and sequator processing - you'll have to tolerate the australian accent though!

06-03-2021, 11:04 AM   #20
Site Supporter
Site Supporter
jon.partsch's Avatar

Join Date: Oct 2009
Location: Northern California
Posts: 442
Original Poster
QuoteOriginally posted by Grimmus Quote
This was 72 minutes total - 36 shots of 2 minute exposure (f6.3 and ISO800). The foreground was 10 shots at 60 seconds - importantly, after waiting for the moon to rise as per @gifthorse's approach (by which time I had been locked in but that's another story ). On the Kp I was able to set this sequence up, and take a comfortable position while the camera clicked through the 36 x 2 min exposures. Bummer the K-1 II doesn't do this but no problem if you have the external intervalometer.

There is a bloke called Richard Tatti who has a youtube channel called nightscape images - it is well done and has been a great source of information and inspiration. There are several videos where he walks through the capture of stacks, panoramas and sequator processing - you'll have to tolerate the australian accent though!
Thanks again, Grimmus. This stuff is golden. I think I am fairly confident on planning my shoot now. I still have some PP questions gestating in my mind, but those can wait for later. I think I have what I need to capture the best exposures I can (for a first timer, anyway), and I can worry about PP later. I am going to make a PDF of all the valuable information in this thread and write up some notes on what I specifically want to do in the field.

I am a little miffed that the K-1ii doesn't include as robust an intervalometer function as the KP that came before it. It seem like it would be simple to impliment in firmware. I hope they do so, but I can use an wired intervalometer just fine, I suspect.

I will definitely check out Richard Tatti on YouTube!

Right now, I am planning three different approaches to the Milky Way shot, and then I can evaluate the differences between methods in practice.

I think I have chosen a location in Yosemite where I can get a good shot of the southern galactic core on night 3 or 4 of my trip. I am planning star trails shots toward the north (for circular effect) and west (for "raining stars" effect). I am confident that I will get some good images out of it all, and perhaps more importantly, will learn a lot.

Thanks, everyone, for your help in this thread!!!

~ Jon
06-03-2021, 12:49 PM - 1 Like   #21
Site Supporter
Site Supporter
jon.partsch's Avatar

Join Date: Oct 2009
Location: Northern California
Posts: 442
Original Poster
QuoteOriginally posted by Grimmus Quote
There is a bloke called Richard Tatti who has a youtube channel called nightscape images - it is well done and has been a great source of information and inspiration. There are several videos where he walks through the capture of stacks, panoramas and sequator processing - you'll have to tolerate the australian accent though!
I hadn't realized it before, but I was already inspired to try a "stitched" Milky Way image by one of his tutorials:



This seems to be a great one on the use of Sequator in the method you are suggesting:



I am now a subscriber to his channel.

~ Jon
06-15-2021, 06:50 PM - 1 Like   #22
Pentaxian




Join Date: Dec 2011
Location: Nevada
Photos: Gallery
Posts: 2,356
So how did it go?

06-23-2021, 02:57 AM   #23
Site Supporter
Site Supporter




Join Date: Sep 2018
Location: SW Bavaria
Posts: 405
QuoteOriginally posted by jon.partsch Quote
Thanks Papa_Joe! I have stellarium on my phone. Is the PC version substantially different?
~ Jon
I did not work with the phone app so much. I think there are only minor differences, but I find the PC version is much easier to use. It is free as well.
I use it to get an overview of the sky before I go into the field and plan, which constellations to take pictures of.
06-23-2021, 11:10 AM   #24
Site Supporter
Site Supporter
jon.partsch's Avatar

Join Date: Oct 2009
Location: Northern California
Posts: 442
Original Poster
QuoteOriginally posted by gifthorse Quote
So how did it go?
Yes, an update is long overdue. I'll make this update in two parts.

The first update regard my milky way image. I didn't bother to take any additional images when I was in Yosemite. I had other responsibilities, including playing dinner host at our campsite (and single malt scotch steward later in the evening) and the good locations for milky way images were far from the valley. I thought it better to evaluate what I was able to achieve at my first camping location, and plan to tweak my strategy for my trip to Lassen in July.

Here is the best I was able to achieve. It is three frames taken with the Sigma 35mm Art lens at f/2.8. The frames are overlapped by about 50%. Each frame is a composite of 4 images. Each image was acquired with a 10-second exposure at, I think, ISO 3200
Name:  Stitch1-1-Pano_Final_LR_Edit.jpg
Views: 49
Size:  108.7 KB

The stars are mostly pixel-sharp except for some coma around the edges and corners. I have a couple of problems with this image, however. There are multiple satellite trails that I don't know how to remove. Any advice on this? Also, you can see some artifacts (dark splotches across the center of the frame) where Lightroom stitched the individual frames together. Any advice on how to deal with this?

Here is the best result I was able to achieve with the D FA 24-70mm. This image is also a stack of images, but only of a single frame. I took this at f/3.5 to try and get better focus, but my lens seems that it was unable to focus all the way to infinity at the 24mm setting. I was able to get pinpoint stars with this lens only at 35mm and above. This performance seems pretty bad, even for a zoom!


Here is a 100% crop from the center of the image above, as you can see, the stars are not even close to being in focus:


...and, for comparison, here is a 100% crop from the Sigma 35mm Art lens:


QuoteOriginally posted by DeadJohn Quote
Star trails are much more forgiving exposure-wise than the Milky Way.

Milky Way tips:
[LIST][*]Don't worry too much about EV. I suggest starting the night with an educated guess at exposure, then take a test shot and examine it to see if you need to adjust settings.
  • At 24mm f2.8, I would use 3200 ISO, 20 second shutter for my first test shot. (although anywhere from 1600 to 6400 ISO can give similar results; lower ISO gives less initial noise, but you'll need to lift shadows more which adds that noise back in)
  • Zoom all the way in to the test shot to check focus, star trailing, and tripod vibration. A little bit of trailing is okay; you won't notice it at normal viewing sizes.
  • Check the histogram. Make sure it's not too far to the left with completely clipped blacks. A histogram peak approx. 1/3 of the way from the left is a common suggestion.
Regarding the histogram peak tip, I found that I required ISO 128000 to approach a histogram peak at about 1/4 the way up on the scale at f/2.8 and 24mm and 10 seconds on the D FA 24-70mm. This seems crazy! @Grimmus recommended Richard Tatti on YouTube. Richard mentioned taking 10-second exposures at f/2.8 and ISO 1600 or 3200 (don't remember precisely, but I *think* he said 1600) with his Sigma 35mm Art lens on his FF Nikon for his milky way panoramas. This doesn't seem possible given my experience. At those same settings, my image above from the Sigma 35mm Art had the histogram pinned up against the left hand side of the scale (i.e. there was no "peak"), and the image on the rear of the camera looked almost completely black. I was able to extract the image above from the images I collected at ISO 3200, but this doesn't seem consistent with Richards results.
06-26-2021, 05:24 AM - 1 Like   #25
Senior Member
Grimmus's Avatar

Join Date: Jun 2010
Location: Newcastle (Aust.)
Photos: Gallery | Albums
Posts: 186
I'll pre-empt this by saying my comments are based on limited practice on this nightscape caper, so feel free to ignore.......still learning lots........

The shot with the Sigma 35 looks to have a lot of potential - the stars look really good (as the reviews of this lens suggest)! I'm interpreting that you have taken 4 frames of 10 seconds at f2.8 in landscape orientation then panned vertically to take the next 4 and again for the last 4 then stacked the sets of 4 before stitching the three stacks? For the satellite trails, I don't know of any approach other than cloning/or healing it out before processing. Some software claims to have options to remove it but it seems to leave faint traces behind is my experience. With the dark bands, this can be caused if you don't correct for vignetting before stacking and stitching. Need to examine what lens corrections or vignettes are being applied. I usually adjust the exposure before stacking and stitching so that you have the histogram about right (1/4 to 1/3 from left without clipping highlights or blacks) and this gives you something with a bit of room to work with on the final image. I wouldn't get too involved in it while taking the photos - shutter, ISO and aperture are all fixed by other factors. Apart from the problems you identified with bands and trails, this image should be capable of being pushed a bit to bring out more detail. You can also have a go at fixing the trails and banding by doing local adjustments on the final image - if you don't want to go back to the start images. The big downside of the night stuff is there is nearly as much or more to learn about the processing when you are sitting at the computer (yuk), and this is needed to provide the feedback loop to the field work (which is the fun bit in my opinion)
06-26-2021, 10:48 AM   #26
Site Supporter
Site Supporter
jon.partsch's Avatar

Join Date: Oct 2009
Location: Northern California
Posts: 442
Original Poster
QuoteOriginally posted by Grimmus Quote
I'm interpreting that you have taken 4 frames of 10 seconds at f2.8 in landscape orientation then panned vertically to take the next 4 and again for the last 4 then stacked the sets of 4 before stitching the three stacks?
Yes, that is exactly correct. I also have a series I took that is 4x4 exposures, so that image will be a bit taller, but I haven't processed it yet.

QuoteOriginally posted by Grimmus Quote
For the satellite trails, I don't know of any approach other than cloning/or healing it out before processing. Some software claims to have options to remove it but it seems to leave faint traces behind is my experience.
That's a bummer, as I imagine this problem is only going to get worse in the future.

QuoteOriginally posted by Grimmus Quote
With the dark bands, this can be caused if you don't correct for vignetting before stacking and stitching. Need to examine what lens corrections or vignettes are being applied. I usually adjust the exposure before stacking and stitching so that you have the histogram about right (1/4 to 1/3 from left without clipping highlights or blacks) and this gives you something with a bit of room to work with on the final image.
Thanks. I'm pretty sure I did nothing to the images before stacking. I'll try the approach you suggest and see if I can get a better result.

QuoteOriginally posted by Grimmus Quote
I wouldn't get too involved in it while taking the photos - shutter, ISO and aperture are all fixed by other factors. Apart from the problems you identified with bands and trails, this image should be capable of being pushed a bit to bring out more detail. You can also have a go at fixing the trails and banding by doing local adjustments on the final image - if you don't want to go back to the start images. The big downside of the night stuff is there is nearly as much or more to learn about the processing when you are sitting at the computer (yuk), and this is needed to provide the feedback loop to the field work (which is the fun bit in my opinion)
I think I will try to fix the trails and vignetting in the original images. Many hors have been spent in front of the computer in the last couple of months trying to learn this type of *highly technical* photography. It has been frustrating, but also rewarding. If I want to be more than a snapshot photographer, then I have to put in the work, and that's okay.
06-26-2021, 08:28 PM - 1 Like   #27
Senior Member
Grimmus's Avatar

Join Date: Jun 2010
Location: Newcastle (Aust.)
Photos: Gallery | Albums
Posts: 186
I can't see the dark bands at PF image resolution but to clarify my comment, the vignette might be due to lens light transmission towards the edges however with an f1.4 lens at f2.8 I would think this unlikely.

If you have good star shape at f2.8, open up a bit and try f2. Might even be OK for a panorama at less as the centre of one image will sit over the edge of another and you'll increase the data collected by 2 to 4 times for a given shutter speed.

Not related I don't think but I had banding (concentric rings) when processing star trails and the solution was to uncheck lens corrections. No idea why, just is.....good luck and yes, the processing effort is high with not always equal measures of frustration and reward.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
astro, astrotracer, attempts, camera, ev, exposure, exposures, location, photography, pixel, post, shot, star, start trails, technique, trails, wide field
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Landscape First attempt @ star trails rtmarwitz Post Your Photos! 12 05-07-2021 04:22 AM
Thematic Post & Discuss Your Astro Tracer Star Trails & Star Shots wtlwdwgn Mini-Challenges, Games, and Photo Stories 48 02-22-2021 01:47 PM
Advice on Star/Milky Way photogrphy Ace General Photography 8 07-14-2015 09:00 AM
First attempt at star trails, failed. Need suggestions? Ajdcadet Troubleshooting and Beginner Help 12 04-15-2015 01:47 PM
Night Star Trails and Meteor - First time Kendigitize Photo Critique 5 08-18-2013 05:50 AM



All times are GMT -7. The time now is 05:27 AM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top