[awscreo]

So after messing around with astro this summer I want to improve my practices for astro, and have couple of (somewhat scattered) questions.

I've successfully used astrotracer for 90s exposures at iso 800 that gave me nicely exposed images to play with, and K-1's sensor allows me to easily push +1 or +2 if I need to.

I'm trying to figure out if I need the dark frame suppression after every exposure (it's on now, I'm dreading the hot pixel stuff that seems to be an issue with a lot of K-1's), but reading this article here: (Astrophotography Image Processing, Clarkvision.com) seems to suggest that maybe it's not necessary, and just processing a single (or sequence) exposure and then suppressing noise seems to work just as well, if not better? I haven't tried stacking exposures to get more star data, with 90s exposures and dark frames would I be able to stack images? Stars seem to move quite a bit after 180s intervals, would stacking work in this instance?

Also, has anyone tried the Samyang 24 1.4 for astro? I've read few reports suggesting it's one of the better options for astro currently (on the budget side), and I'll be able to stack images into panoramas with a ton of star data due to 1.4 aperture.

[VoiceOfReason]

As long as you keep framing then stacking is not a problem. Also, you don't have to use dark frame subtraction if stacking. Do however many shots you want to stack, then with the camera at the same ISO at the end of your session take 10 or so shots of the same length with the lens cap on and the camera under a jacket to make sure no stray light gets in. Use those as your dark frames (noise subtraction), then find a nice evenly lit wall and expose somewhere in the middle of the histogram to take your flat frames, 10 or so, (removes vignetting and dust specs from final image) and then while at the same ISO take 10 shots at 1/8000 second for bias frames (removes sensor noise that can't be calibrated out with darks) and stack away. Also, always stack raw files, .dng is preferred.

If only doing single frames, then yes, use dark frame subtraction.

As far as using that lens, it would be insanely wide on the K1. I generally do deeper objects, so i go longer.

[awscreo]

As long as you keep framing then stacking is not a problem. Also, you don't have to use dark frame subtraction if stacking. Do however many shots you want to stack, then with the camera at the same ISO at the end of your session take 10 or so shots of the same length with the lens cap on and the camera under a jacket to make sure no stray light gets in. Use those as your dark frames (noise subtraction), then find a nice evenly lit wall and expose somewhere in the middle of the histogram to take your flat frames, 10 or so, (removes vignetting and dust specs from final image) and then while at the same ISO take 10 shots at 1/8000 second for bias frames (removes sensor noise that can't be calibrated out with darks) and stack away. Also, always stack raw files, .dng is preferred.

If only doing single frames, then yes, use dark frame subtraction.

As far as using that lens, it would be insanely wide on the K1. I generally do deeper objects, so i go longer.

I need to try stacking, I think i'll need to find a detailed guide or buy a book) sounds like a fairly complex process

Hmm, I'm using a 14mm lens at the moment, so 24mm would be more narrow. I think I can achieve same width as 14mm if I stitch photos, but i'll have higher resolution and more detail with that large 1.4 opening, that's why I was curious. As for long telephoto - I have the 70-200 2.8, keep want to try it for astro, but not sure what settings would work best, and if the astrotracer is usable at 200mm?

---------- Post added 08-04-17 at 04:43 PM ----------

As long as you keep framing then stacking is not a problem. Also, you don't have to use dark frame subtraction if stacking. Do however many shots you want to stack, then with the camera at the same ISO at the end of your session take 10 or so shots of the same length with the lens cap on and the camera under a jacket to make sure no stray light gets in. Use those as your dark frames (noise subtraction), then find a nice evenly lit wall and expose somewhere in the middle of the histogram to take your flat frames, 10 or so, (removes vignetting and dust specs from final image) and then while at the same ISO take 10 shots at 1/8000 second for bias frames (removes sensor noise that can't be calibrated out with darks) and stack away. Also, always stack raw files, .dng is preferred.

If only doing single frames, then yes, use dark frame subtraction.

As far as using that lens, it would be insanely wide on the K1. I generally do deeper objects, so i go longer.

So do I move the frame with the stars, or just the camera still and take more photos? If so, wouldn't the app then convert the stars into streaks since they don't stay still?

[VoiceOfReason]

I need to try stacking, I think i'll need to find a detailed guide or buy a book) sounds like a fairly complex process

Hmm, I'm using a 14mm lens at the moment, so 24mm would be more narrow. I think I can achieve same width as 14mm if I stitch photos, but i'll have higher resolution and more detail with that large 1.4 opening, that's why I was curious. As for long telephoto - I have the 70-200 2.8, keep want to try it for astro, but not sure what settings would work best, and if the astrotracer is usable at 200mm?

I've used Astrotracer at 500mm before. Now I use actual tracking mounts for my stuff though, the iOptron zeq25 (updated to the cem25) and the iOptron SmartEQ Pro 3200. Here's what I do with my stacking stuff:

After you have all your subs you will want to take matching darks (these are your noise subtraction frames). The way you do that is to put the lens cap on, then cover it all, and take 10 or 20 exposures of the same length as your subs. When viewing these they must be absolutely dark. Next is to take flats (these account for any dust on the sensor and vignetting) by finding a nice wall at your house that is well lit and an even shade. Take those with the histogram 40-50% from the left, and do 10 of them. Next take bias frames by keeping the lens cap on, make sure that you’re in a dark place, and do 10 shots as fast as the camera can do, which is 1/8000 for the K3 and K1. That subtracts sensor read noise. Do all of these at the same ISO. Flats must be at the same focal length and the lens/camera in the same orientation (same is true for telescopes). The darks and bias should be taken at the end of your imaging session so the sensor is the same temperature as what your subs were (subs = sub exposures, ie the light frames you will stack).

Now open DSS, load your lights (that’s the sub frames you want to stack), darks, flats, and bias. Save the file list and name it what you want. Close DSS. Open DSS up again and load the file list, check all, then register checked pictures. After that finishes then save the changes you made to the file list and close. Open DSS again, load the file list, check above a threshold (see if there are any low outliers and set the threshold above that), then stack. After you’re done save picture to file. Do not use the adjustments in DSS or you’re going to get a monochrome image, so embed but do not apply adjustments.

You’re going to get a really dark image tif file. Open it in Photoshop, mess with curves etc. but do not use auto tone, auto color, and auto contrast or else your image will look like crap. Do smaller adjustments at a time and keep working at it. You will get decent and will figure out the best workflow for you. I like to bump the saturation some, mess with the curves and levels to stretch the data, and then try to add dynamic range back in by faking it a little on HDR, but to each their own.

[VoiceOfReason]

So do I move the frame with the stars, or just the camera still and take more photos? If so, wouldn't the app then convert the stars into streaks since they don't stay still?

Move to reframe the stars. Of course any parts of the frame that aren't overlapping will be at the edges of the stacked image so yo can crop them out.

[awscreo]

I've used Astrotracer at 500mm before. Now I use actual tracking mounts for my stuff though, the iOptron zeq25 (updated to the cem25) and the iOptron SmartEQ Pro 3200. Here's what I do with my stacking stuff:

After you have all your subs you will want to take matching darks (these are your noise subtraction frames). The way you do that is to put the lens cap on, then cover it all, and take 10 or 20 exposures of the same length as your subs. When viewing these they must be absolutely dark. Next is to take flats (these account for any dust on the sensor and vignetting) by finding a nice wall at your house that is well lit and an even shade. Take those with the histogram 40-50% from the left, and do 10 of them. Next take bias frames by keeping the lens cap on, make sure that you’re in a dark place, and do 10 shots as fast as the camera can do, which is 1/8000 for the K3 and K1. That subtracts sensor read noise. Do all of these at the same ISO. Flats must be at the same focal length and the lens/camera in the same orientation (same is true for telescopes). The darks and bias should be taken at the end of your imaging session so the sensor is the same temperature as what your subs were (subs = sub exposures, ie the light frames you will stack).

Now open DSS, load your lights (that’s the sub frames you want to stack), darks, flats, and bias. Save the file list and name it what you want. Close DSS. Open DSS up again and load the file list, check all, then register checked pictures. After that finishes then save the changes you made to the file list and close. Open DSS again, load the file list, check above a threshold (see if there are any low outliers and set the threshold above that), then stack. After you’re done save picture to file. Do not use the adjustments in DSS or you’re going to get a monochrome image, so embed but do not apply adjustments.

You’re going to get a really dark image tif file. Open it in Photoshop, mess with curves etc. but do not use auto tone, auto color, and auto contrast or else your image will look like crap. Do smaller adjustments at a time and keep working at it. You will get decent and will figure out the best workflow for you. I like to bump the saturation some, mess with the curves and levels to stretch the data, and then try to add dynamic range back in by faking it a little on HDR, but to each their own.

Thank you for taking time to write a detailed answer one question - what is dss?

---------- Post added 08-04-17 at 06:41 PM ----------

Move to reframe the stars. Of course any parts of the frame that aren't overlapping will be at the edges of the stacked image so yo can crop them out.

So if I want a foreground object in the picture, I'll need a separate exposure for that one, and then mask sky into if?

[VoiceOfReason]

Thank you for taking time to write a detailed answer one question - what is dss?

So if I want a foreground object in the picture, I'll need a separate exposure for that one, and then mask sky into if?

Yes, you will need a separate exposure for a static foreground. DSS is Deep Sky Stacker, a free program for stacking and grading images. Here's a link to DSS 3.3.6 which is the latest version, though not officially released. http://deepskystacker.free.fr/download/DeepSkyStacker336.rar

For

[awscreo]

Yes, you will need a separate exposure for a static foreground. DSS is Deep Sky Stacker, a free program for stacking and grading images. Here's a link to DSS 3.3.6 which is the latest version, though not officially released. http://deepskystacker.free.fr/download/DeepSkyStacker336.rar

For

OK I'm on board next time I'm out at night I'll try to shoot stacks and see where it takes me. Thanks!

[VoiceOfReason]

You're welcome!

[DeadJohn]

...if I need the dark frame suppression after every exposure...

Dark frame suppression is IMO not needed with the K-1. This is of course based on my camera, where I have yet to spot many hot pixels, and no sign of the "white dots" issue that some people complained about earlier. I shoot raw DNG with good results at 1600 ISO.

Here's a sample with the astrotracer. Single frame, no drk frame. 100mm, f3.5, ISO 1600, 60 seconds. Processed solely with LR, reduce black point to cut light pollution, increase white point for good exposure, tweak white balance, some sharpening and noise reduction. Contrast or Clarity (Contrast in this sample, boosting both often looks overdone to me so no Clarity here).


If you do get into stacking, a general rule of thumb is twice as many images = half as much noise. 2 shots half the noise, 4 shots 1/4, 8 shots 1/8 the noise. A dark frame after each shot wastes a lot of time. Let's say you plan to stack 16 images. Turn off the dark frame subtraction. Take 16 light frames. Then take 4-8 manual dark frames with the lens cap on. You can do something else while collecting the dark data. Stack your lights and darks with Deep Sky Stacker or other software.

[alpheios]

Hmm, I'm using a 14mm lens at the moment, so 24mm would be more narrow. I think I can achieve same width as 14mm if I stitch photos, but i'll have higher resolution and more detail with that large 1.4 opening, that's why I was curious.

Using less wide lenses and stitching photos together will always result in better astro pictures. Not only because of gain in resolution but mostly because you will be able to gather much more light. I also wanted to get an ultrawide lens for astro until I learned about how focal length, aperture value and aperture size are related.

The formula is: focal length/aperture value = aperture diameter size. That means a 14 mm F2.8 has an aperture diameter of 5 mm whereas a 24 mm F1.4 has a diameter of 17,14 mm. That means the 24 mm will be able to gather a huge amount more light than the 14 mm. Even if they had the same aperture value the longer focal length lens would be able to gather more light.

I don't own the Samyang 24 mm but like you said it indeed has a very good reputation for astro photography.

[awscreo]

Dark frame suppression is IMO not needed with the K-1. This is of course based on my camera, where I have yet to spot many hot pixels, and no sign of the "white dots" issue that some people complained about earlier. I shoot raw DNG with good results at 1600 ISO.

Here's a sample with the astrotracer. Single frame, no drk frame. 100mm, f3.5, ISO 1600, 60 seconds. Processed solely with LR, reduce black point to cut light pollution, increase white point for good exposure, tweak white balance, some sharpening and noise reduction. Contrast or Clarity (Contrast in this sample, boosting both often looks overdone to me so no Clarity here).


If you do get into stacking, a general rule of thumb is twice as many images = half as much noise. 2 shots half the noise, 4 shots 1/4, 8 shots 1/8 the noise. A dark frame after each shot wastes a lot of time. Let's say you plan to stack 16 images. Turn off the dark frame subtraction. Take 16 light frames. Then take 4-8 manual dark frames with the lens cap on. You can do something else while collecting the dark data. Stack your lights and darks with Deep Sky Stacker or other software.

I will try my longer lens for astro, just need another day with clear skies (a hard thing to come by this summer unfortunately, it's been raining a lot in NY state). I'm planning on doing astro as much as I can before winter comes back, and maybe even then)

[slartibartfast01]

Using less wide lenses and stitching photos together will always result in better astro pictures. Not only because of gain in resolution but mostly because you will be able to gather much more light. I also wanted to get an ultrawide lens for astro until I learned about how focal length, aperture value and aperture size are related.

The formula is: focal length/aperture value = aperture diameter size. That means a 14 mm F2.8 has an aperture diameter of 5 mm whereas a 24 mm F1.4 has a diameter of 17,14 mm. That means the 24 mm will be able to gather a huge amount more light than the 14 mm. Even if they had the same aperture value the longer focal length lens would be able to gather more light.

I don't own the Samyang 24 mm but like you said it indeed has a very good reputation for astro photography.

Does the longer focal length lens really gather more light at the same f-number?
I would have have thought that if the aperture diameter is larger then the correspondingly smaller field of view would result in exactly the same total light gathering.

[Bob 256]

Lenses work slightly differently for stars (which are point objects) and galaxies (which are "extended" objects).

For stars, the actual aperture is what counts regardless of f stop. The larger the aperture, the more light is gathered for each star. If the focal length were increased (decreasing the f number), the same amount of light energy is still present in a star image because it doesn't get any larger (in practice this holds pretty true though it's a simplification).

Objects like the glow of galaxies, benefit from lower f numbers. The reason is that as focal length is increased (assuming a fixed aperture) the f number increases and the area that the light energy is spread over on the sensor increases (longer focal length = higher magnification). If the aperture were increased, decreasing the f number and brightening the image, it will balance out when the same f number is reached as it was before the focal length was increased.

This means that lower f numbers (for a given focal length) will brighten extended objects and stars while going to a larger aperture (as is the case with longer focal length lenses) will increase the brightness of just the stars and might even dim extended objects if a higher f number results.

We can beat the Hale telescope at Palomar (which is f 3.3) when shooting galaxies, but we can't come anywhere close to capturing faint stars with our piddly lenses compared to the Hale 200 inch aperture (actually the Hale isn't used too much for the galaxies we take photos of because its field of view is too small).

[slartibartfast01]

Lenses work slightly differently for stars (which are point objects) and galaxies (which are "extended" objects).

For stars, the actual aperture is what counts regardless of f stop. The larger the aperture, the more light is gathered for each star. If the focal length were increased (decreasing the f number), the same amount of light energy is still present in a star image because it doesn't get any larger (in practice this holds pretty true though it's a simplification).

Objects like the glow of galaxies, benefit from lower f numbers. The reason is that as focal length is increased (assuming a fixed aperture) the f number increases and the area that the light energy is spread over on the sensor increases (longer focal length = higher magnification). If the aperture were increased, decreasing the f number and brightening the image, it will balance out when the same f number is reached as it was before the focal length was increased.

I see. That makes sense. Thanks.