I tried to take a picture of the milky way, but as you can see on crop of the picture I've uploaded (100% size), there's a lot of purple patches on the black areas.
Anyone knows why this happened (I've never seen than on any other milky way shots I've taken, except some chromatic aberration around the shiniest stars ofc)?
And more importantly, anyone knows how to treat this? I tried a hue/saturation filter in photoshop but the patches still show (they just change color)
Thanks!
Yup I see Magenta and Teal patches. Noise, there are several types of noise, but I'm thinking sensor?
What settings did you use for the shot, and how many shots were taken (if stacked)? I could be mistaken, but it feels like the ISO could have been bumped a pinch higher.
I'd try making a duplicate layer and load the new layer into camera raw filter. In there make a new mask for the entire image, a new slider will appear called moire reduction. Play with that slider till the magenta/teal is gone... and hit okay going back into photoshop. Now switch that layer to Darken blend mode. See if that helps.
Thanks for the reply, didn't know that moire function! It did help a little bit but it's still quite visible.
Only 1 shot, f/4, 25sec, iso3200 at 10mm (on A6500 so 15mm FF equivalent). I know that's not how you get great shots, but I've gotten decent pictures with these same settings (without these magenta patches), so I was wondering what I did wrong this time.
Sorry to hear it didn't help much... Hum.. it's puzzling that this doesn't happen all the time.
I can only suggest instead of taking a single 25s shot, take five 6s shots at the same settings and either manually rack 'n stack the images - or use a free software like "Sequator". This should reduce this type of noise by quite a lot.
Thanks, I'll definitely try that next time. Is that really how it works? I can take 5 very short shots (they'd be almost blacked out with the same settings at 6seconds) and stacking them would give me enough brightness to see the milky was as I see it on that shot?
I thought the way to use it would be to take 5 (or more) shots with the same settings (so 25 sec exp) and then staking them would reduce noise but still give me the same sort of exposure. Is that not a good way to do it? shorter exposure time is better?
I also heard astro is a field where you can over expose by a stop then reduce the exposure in post. It's about saturation of the sensor to help with noise reduction as well.
With stacking, noise is reduced as it's the variable that will change between shots, and it's "averaged" away.
Note: I can only share what I've learned with practice - I only started shooting astro 18 months ago.
The other way is to get one of those tracking mounts that go on-top a standard tripod. This way you can bring the iso way down and shutters open for longer. I hear even simple tracking is a game-changer for astro.
Thanks again for the tips. I also saw these tracking mounts, has to be game changing if the tracking is done properly, you can shoot at iso 100 and expose for however many seconds you need :) I wish I had that during my travels
The tracking mounts are definitely a game changer. I have been able to get 60 seconds at 135 mm with no star trails. That was with a rough polar alignment
How much have you brightened the image in post? It’s common if you don’t leave enough light in that when you increase brightness and or shadows you’ll see purple patches at the bottom of the shot. That may be the reason. It’s happened to me before. Although your settings seem ok.
You're right, I'm at +0.85 on exposure. But even when I bring the exposure back to default settings, they're still visible (see attachement, this is sooc)
This is another one, sooc, same body, same night, still iso 3200 but at 8 sec / F1.4 (different lens, 30mm)
Loads of magenta around the bright stars but no magenta patches (or very dim at least) like the other picture.
Yeah, it's basically not leaving enough light hit the sensor. I had issues before with some time lapses where I wanted them to be done so I used 10 sec exposures in pretty dark areas. When I brought up the exposure there was so much purple pixelation at the bottom of the frame.
This one here looks great! The magenta may just be a white balance change. Also, you're getting some coma there so stop to about f2 or f2.8.
I haven't done much astro (good astro anyways). Do you know the light pollution of the area you are shooting? In the little bit of experience I've accumulated shooting astro I've noticed that light pollution is the biggest factor affecting color of astro photos.
I ask this because the area you are concerned about is within the light band of the Milk Way, so it should be white-ish/brighter. The sensor is definitely picking up that there is some amount of light there and that there is more light there than in areas just surrounding it like in the top right half of the 1:1 magnification you posted so I don't think it is noise (not that it can't be removed with noise reduction). I think that if it was true noise you would see it across the entire photo. Also if you look carefully, the "noise" is the same color as the center of your milk way just not as bright.
The a6300 has very little noise at iso3200 .
I'd love to see the entire RAW photo just to compare to my RAWs
Here is a shot from deep in the Wyoming back country. This is shot with a 20mm. The milk way is going to be a tad wider in my picture but the ratios will still be the same. Notice that the darkest part of the band only makes up less than 1/3rd of the whole band. There is easily a third of a "white wash" or more on either side that is still part of the Milk Way.
Thanks for your reply.
There was some light pollution as there were lights right in front of me (a camping area) but not too much (that camping area was in the middle of the desert in Namibia).
but to be fair, I think my problem happens all across the picture, not just one specific area. And I actually find it most disturbing in those dark areas on the corners, that should be black and are half magenta instead.
Your picture looks great, I definitely don't get results like that! Or maybe I just have no clue how to edit..
Here are the raws of the 2 magnifications I posted, I'd definitely love to hear your thoughts if you have time to look at them
https://we.tl/t-YyjGPhiL6W
Thanks
Just checked out the photos you sent. The one with the campsite in view is certainly being affected by noise and likely the light pollution as well. The green and red pixels are a dead give away. The second picture however looks fairly clean. I might be wrong but the picture looks white washed probably by the ambient light from the campsite but there isn't as much noise. If you ever get to shoot somewhere like that again I would definitely suggest getting as far as possible from the campsite. Even when I was living in a small town in Wyoming I had to drive out of town a couple miles. The porch lights and couple neon signs were enough to affect the picture quality. In Florida, I've made several attempts to take some astro shots and they all look like the white washed picture. I always thought it was just because the ambient light pollution in Florida.
As far as color, the RAW files from Wyoming and WV have color from the start. Those places have almost no light pollution. They are both in grey or black areas on a light pollution map. The ones from Florida have little to no color with the exception of a handful. Most of Florida is yellow or brighter and you really need to get into the dark blue or grey category to get good pictures.
Thanks for the explanation and taking a look at the pictures! I didn't realize that the slightest light pollution could have a meaningful impact (I just imagined you'd need a massive city around to have an impact on the sky).
But no regrets on that one anyways, I was definitely not going to go outside the camp, as there were lions, hyaenas, leopards, and all sorts of stuff that could interrupt my long exposure shots :D
Most of the time I shoot from Bortle 8 skies which = light pollution filters, and double the amount of time in exposures. Im still learning how to so post processing. Try taking 10 to 20 shots and stack them in Sequator. It has a function where you can freeze the foreground. Using the 500 rule you might try increasing your exposure time to 15 seconds.
To me it seems like high sensor noise (high exposure time and high iso may be contributing), combined with high exposure correction. I'm assuming your lens won't shoot much faster than f4.0. There is a really good channel on YouTube called the lonely speck that will help you with editing astro photos as well as teach you techniques to stack the images. I would say probably shoot at around 8-12 seconds as you're most likely to minimize a lot of star trailing and it'll help reduce some noise that you're seeing. Maybe also click the moire function at the bottom of lightroom in case no one has suggested that. (try and adjust your white balance and green and pink balance (they go over the best method to do that in some videos on lonely speck, that change alone I've found is enough to bring some noise down on it's own).
Thanks for the reply.
You're right, I'm using the Sony 10-18mm f/4. My problem with 8-12 second shots is that I'd need at least iso 6400 if not more to have a picture that is not too dark.
High exposure time contributes to high sensor noise? So even if I had a motorized tracking head, I shouldn't do 30sec+ exposures at low iso?
Moire function does help a little bit, with noise reduction too.
I'll check that youtube channel, thanks for the recommendation!
Yes the longer your sensor is exposed the longer it is running and putting out heat (which is exponential, not linear) it's technically not noise but it will do things like cause odd hot pixels etc (every sensor behaves differently), newer cameras generally don't have any issues. I just assumed that could be an issue (went back and looked at your photo and don't see that issue, sorry for that confusion).
I'm no expert about this but astro is all about signal to noise ratio. By increasing your iso you are increasing your sensitivity (issue is that it will also pick up more noise). normally at lower iso's there is more actual light vs random light as you raise your iso you're increasing your ability to pick up that noise until it drowns out the actual light. It's not as simple as lower iso automatically means better quality image quality (in most cases lower iso shots just don't produce enough noise to drown out the actual signal.
Imagine standing next to one person and you try talking (they then talk as well with someone listening 5 feet away from you). They'll be able to hear you no issue , now redo the same experiment with 500 other people (highly unlikely the observer will hear you, you'll be muffled out, you can think of iso in a similar way).
To overcome that you can always stack your photo's. So I wouldn't be so scared to shoot at 6400 and expose a bit less, then stack about 5-6 images you'll notice that you'll have much less noise. I wish I had an example but my lens is an f1.4 so I always run iso's around 3200 and below but even when shooting I'll stack and notice a difference when zoomed in. What I've found is that it's easy to correct for noise, but much harder to turn star streaks into actual dots, prints as large as 5x7 will show streaks if you give them a close look). You can also do noise reduction in your camera as well but the issue with that is that it will double your shooting time per each shot, not a great option as stacking is just easier once you learn it). In camera exposure compensation (with iso etc generally produces a cleaner result than increasing exposure using a slider, iso invariance cameras change that and that's much more complicated and there are debates about if it does something etc, ignore that part just for simplicity sake here.
If I'm wrong about any of this I hope someone reads my essay and can correct me as well.
Thanks for the detailed explanation, it makes sense.
I didn't know much about astro when I shot these pictures (not that I'm an expert now lol). Next time I will definitely do multiple shots + probably buy a faster lens + tracking mount, hopefully I won't have to deal with issues like that anymore :)