National Geographic's top places for dark skies

Rupert. Thanks for your words of advice. I haven't had a chance to use this lens as yet. Perhaps during this weekends observing session on Blackheath. I understand Tej has also purchased a similar lens and I will be interested to hear how he's getting on with it.

Perhaps it would more appropriate to start a new thread for my following question but since I've already posted the image of the Milky Way earlier in this thread:

When I got around to processing the raw file of the image above, I noticed some odd circular patches appearing almost in a line across the image. Looking closely at the uncorrected jpeg image above, I see they are present here. I don't think its dirt on the lens, in the lens or on the camera's sensor as these patches are not present in subsequent images. However, I did move the camera after this shot. When this shot was taken Venus was very bright and someway off the frame to the lower left. Could these patches be due to lens flare created by Venus? Anyone any thoughts?

A new thread is probably wise.

I saw the circular patches. It is not dirt on the sensor as that would appear sharp. This is either something on the lens or flare. I would not have thought Venus would cause it. I assumed you used a lens hood? If not, always use a lens hood!!

The marks, whatever they may be are a classic example of why it is advisable to take flat files. If it was something on the lens then a flat would have made it possible to subtract it. You can't do much about flare.

It might be possible to rescue the image in PS by creating a layer and putting a median filter on it. If you then darken this it will make the patches stand out and look like a negative. If you then make it a 'difference' layer it can subtract this from the original. Not quite as straightforward as that but maybe if you start a new thread called 'rescue me' or something and send me the RAW, I could play with it and post some steps to show how to get rid of the bits that should not be there.

Rupert, thanks again for the advice. Thinking about it, the most likely cause of the lens flare in the image was an errant head torch! Thank you for your description on how to use Photoshop to "fix" the flare blotches. Unfortunately I've only got GIMP on my computer. If you would be prepared to have a go at improving the image I'd be pleased to let you have a copy of the raw image file.

However, using GIMP I have able to make the more prominent flare patches less apparent by selecting the circular flared region and using the levels tool to adjust its brightness to match that of the surrounding image.

Hi Clive, I know nothing about GIMP but thought it was layer based. If you select the region and change its level I think you will just end up with it looking the same as the black level will change compared to the rest of the image. You might be able to do something with curves but the transition will be the hard bit. Layers are always the best way to do this.

Happy to play if you send me the RAW. Did you only take one image or several? i.e. can a stack be made.

How much coma is too much coma?

I've been outside playing with a Sigma 20mm F/1.4 ART lens tonight... I just wanted to test it for aberrations before sending it back. I've read some mixed reviews on this lens, the majority are negative with regards to using it for astrophotography and many suggest that it suffers from bad comatic aberration... but from what I've read, all wide lenses tend to suffer from this defect (with the exception of the esteemed Zeiss Otus that Rupert mentioned in an earlier post which really is supposed to be outstanding, but scarily pricey).

Please see the single frame below with rectangular markings, where the inner rectangle depicts where stars appear more or less spherical and the outer rectangle shows where it all gets REALLY bad:
`

Looks like it's suffering from astigmatism and bad vignetting too. I'm thinking of returning this lens and getting the Samyang 24mm f/1.4, though this lens is thought to suffer from spherical aberrations and lack sharpness. Is it likely to be any better than the Sigma? Does anyone here have any experience with this lens combined with a full frame camera or simply know more about landscape astrophotography than I do (I've read that comatic aberrations worsen with full frame cameras)?

I've read so many reviews now that I'm becoming cross-eyed. Please help.

Christina,

There are a number of things to consider here. No 1 is that using ANY camera lens from ANY maker wide open will ALWAYS result in some coma and possible other aberrations if you are critical enough. I make this point so many times to people. A camera lens has so many jobs to do compared to a high end telescope lens that it just cannot be effective on point sources of light. Using one wide open will always lead to issues. They must be stopped down to at least F4, preferable F5.6 before aberrations like this will effectively disappear. You will find it much reduced from others and you may be able to use a lens with a lower F ratio but never fully open. Testing of the 20mm Art shows a massive increase in performance from centre to edge at F4.

No 2 is a wide angle lens will also suffer more from barrel and pincushion distortion, which when wide open will be accentuated. No 3 is that the bottom corners of an image can often seem more distorted due to dispersion in the atmosphere. This will be more noticeable with longer exposures but is a factor.

No 4 is that this is a 20mm lens on a full frame camera (I assumed you used you D750?). This is very wideangle, much more so than a 24mm. While I would expect better in the central part of the image I am not surprised by the edges, which if shot at F1.4 are not that bad.

Out of interest what was the exposure as I can also see some tracking error so I assume quite long?

Many errors like this can be fixed in Camera Raw if the lens is in Adobe's DB.

Re comparison of a Samyang (or whatever other name it appears under) to a Sigma Art, then there is not really one to be made. For all the hype Samyang get, they are still a cheap, manual only focus (nothing wrong with that) except with Samyang the AF focus confirmation can apparently be a bit variable (meaning focus is wrong) so you have to check focus in daylight by eye. The odd lens has punched well above its weight but generally it is the 35mm that is the secret bargain. Sigma are a premium lens maker that have developed and developed to the point where they are easily the best independent lens maker outside of Zeiss (although the new 85mm Art just beat an Otus in the lab at 1/3 the cost). They have for some time been superior to much of Canons output if not quite with Nikon but the Art models are seeing to that. They were not always this good but have progressed.

Now despite all the BS surrounding the 'Art' lenses they are still subject to variance like anyone else's lenses. I have a customer who was using a triple Canon 300mm F4L lenses on an Astro imaging rig. Generally this is considered a good lens but I got a call asking if I could collimate them. I saw images from the 3 and 2 were decidedly poor. Nothing really can be done so 2 were sold and he bought a proper scope.

The Art you have borrowed is a first rate lens. You might have borrowed a lemon, but there is also as much chance of that, if not more with the new 20mm Samyang. Personally if you are in the market for a £600 wide angle then the Art is a good choice and I would ask to borrow another if this one is suspect. I certainly would not just buy a Samyang. I would want to test it first. Although more expensive I would also keep my eyes open for a used Zeiss 21mm Distagon. This is F2.8 but will exhibit a level of consistency that is difficult to beat. A daylight image taken with a Zeiss lens is 'just right'. I had a set a few years back and sold them all when times were tough. I really wish I hadn't. A used 21mm will cost £7-800.

Failing that I would also consider a used Nikkor 20mm AI or G lens. They can be a bargain. Remember that for all the hype surrounding digital sensors, sharpness, blah, blah. B&W film has much more dynamic range than digital and the sharpest types would be the equivalent of a 500MP (that is 500 not 50) full frame sensor. A lot of old lenses had to be built to work with this level of quality. Ironically manufacturers are now having to get back up to this quality level and it costs.

So sorry for the long post, but I would try the lens again at F4 and personally would limit the exposure to about 20s (500 rule suggests 25). I would also aim at the zenith for a pure test.

Thanks very much for the detailed reply, Rupert. Your advice (and expertise) is immensely helpful.

Perhaps this Sigma lens is a lemon… The reviews I’ve read regarding Samyang suggest that the 24mm f/1.4 lens is a bit touch-and-go in this regard too. It would seem that many people have returned up to four lenses before finding one that’s not defected in some way. I believe that the same is the case for the 14mm f/2.8 within the Samyang range too.

I suppose I may have become sceptical about the performance of the Sigma lens due to the amount of negative reviews I’ve read regarding coma. It’s difficult to determine whether these are indeed BS or genuine. Review sites such as this one, often tend to link to products where they can be purchased so it’s hard to tell if there is an ulterior motive to the reviews or whether they are genuinely unbiased.

I did indeed use my D750. The exposure was 10s at f/2. Remembering your advice from before, I stopped it down in a bid to reduce aberrations.

My kit lens is a 24-120mm f/4 which I thought was quite slow for Milky Way landscape photos. I took this lens to a dark site in France last summer and found that I was struggling to pull out detail. Here is the result of one of my test shots there (which was pointing to the zenith - by coincidence):  https://twitter.com/XtinaChester/status/832936511507152896

This is my first attempt to process a Milky Way photography so please be mindful that it’s probably not amazing. It’s a 6-frame stack using a free version of Nebulosity so there are lines running through it too. I wanted a lens at f/1.4 so that I could stop it down to ~f/2 (f/3 max) to draw out more Milky Way structure. All my photos from this trip are lacking detail and I experimented with an array of ISO's each time to see what would work. I'm reluctant to stop down to f/4 as I suspect that I'll be back in this same predicament.

Hi Christina

OK well if it was just 10s then the other artifacts I saw were not tracking.

10s is too little exposure and ISO 8000 too much gain (all digital camera do is increase amplifier gain from a default. This just adds noise). Nikon cameras are better than Canons in this regard but I don't go above 1250 with a DSLR. The 500 rule is something which works well if you are a bit conservative with timings. So 500/focal length = exposure time with a full frame sensor. Therefore 500/20 = 25seconds, which is why I said go for 20 secs.

Returns for the Samyang are no surprise. No different to cheap, badged Chinese telescopes. The odd one can be brilliant but on the whole, far from it. Not that £400 is cheap for a lens but for something exotic (24mm F1.4 is) then if you pay Peanuts, you get Monkeys. BTW, if you want a wide angle then the one lens you should save up for is not really the ones I mentioned, its the Nikon 14-24mm F2.8. This is used by a lot of Astro Photographers with Canon, Nikon and CCD alike. A grand used but worth every penny as its a sought after pro lens that is like a bag full of primes. I'm saving up for one.

If people are returning lenses for coma then you have to ask how are they using them? The only way they can determine coma is with something like a star test and that's not really fair because if used in daylight on a terrestrial subject they would never see that. Distortion, Chromatic Aberration, Longitudinal Aberration and plain old Sharpness (but in the eye of the beholder) yes. Coma, no.

F2 is not really any different from wide open. Only at F5.6 does a lens really deliver. Most lenses sweet spots from F5.6-F8. Always have been because this is the aperture which uses the best part of the lens. A caveat needs to be mentioned here because I should really say approx 2 stops from fully open. Your F4 kit lens in theory would be best at F8. Having said this the faster the native F Ratio the smaller the radius of the curve in the lens and the more distortion because that is harder to make so the edge of the lens suffers. If you want to see how lenses perform then I suggest looking at either DXO or Photozone (just google with the lens) as they offer tests.

F4-F5.6 aperture range most people imaging with telescopes would kill for. The key to seeing anything is not to end up with a picture you can see. It is to end up with anything at all! The Rosette image I took at E-Eye was 4 images (about 20 less than ideal) of 5 minutes each and this was the actual result after stacking that. ISO would have been equivalent to about 200.

RGB Stack by Rupert Smith, on Flickr

As you can see there is not a lot there. In Photoshop, if you go to Image/Equalize then this will show you image with every level maxed out so you can see whats in there. This is waht that 'nothing' image reveals.

RGB Stack Equalize by Rupert Smith, on Flickr

I use this as a guide and also a challenge. I am by no means a talented processor but I got the nothing to this so far. Getting there but really more data for longer is the way to go.

Rosette 210117 v3 by Rupert Smith, on Flickr

Your image as shot is fine for enhancement. If you shot with less ISO, longer exposure and at F5.6-8 you would be able to bring more detail out because there will be less noise or aberrations in it. Slowly, slowly, catchy monkey as they say.

You got image 3 and 2 out of image 1??? Wow!!!

Rupert, could I ask you to send those to me please and I'll figure out a way of incorporating them into the video in an effective way? Thanks! Wow.

You got image 3 and 2 out of image 1??? Wow!!!

I second that! Your Rosette Nebula image is amazing, Rupert. It really does depict what can be achieved with longer exposures at lower ISO’s. Thanks for sharing your working images as examples - it's incredibly helpful to actually see what you mean.

I was really struggling to pull out detail from any of the low ISO’s used with the 24-120mm kit lens. At the time, the 8000 one seemed to balance noise and stars best but it’s still not the beautiful Milky Way shot that I’m looking to achieve (and still rather noisy, I’d agree). Having now followed your advice and pulled both these ISO 8000 and the images captured at lower ISO’s that contain close to ‘nothing’ into Photoshop and Equalised, a whole new world has been revealed! Amazing! Thought there was little there in those low ISO images. I definitely should have taken more shots. I’ve since been advised that I should have taken about 20 more photos to stack with at a lower ISO. Though the D750 is supposed to cope with noise at high ISO's well… At least that’s what I’d read before purchasing the camera.

10s is really all that the 24-120mm lens can handle before stars start trailing. My understanding is that a wider lens will cope better on that front and enable my exposures to be longer. I’ll try 20s as you’ve suggested next time on a wider lens and see how that fares. I think the only way that I can avoid trailing otherwise is with use of a tracking mount... I simply can’t afford one of those at the moment so I'll need to find a happy compromise somewhere and try to find a way to glean the best possible results from a cheap lens (sadly, a 1K lens is but a dream too). I’m just keen to get out there and to give it a go and way too impatient to wait until I can afford quality kit.

I’ve returned the Sigma. There were just far too many aberrations and I wondered whether I got a dud lens so will try again at f/5-f/6 as you’ve suggested with a new wide lens and see whether that works out any better. Before I replace the Sigma for another of the same, I’m tempted to try the Samyang just to see the results. I know, I know, I may get monkeys but I want to see these primates for myself. Might try out the 14mm f/2.8 instead of the 24mm f/1.4… Thinking that this will then enable me to take longer exposures at a low ISO than I’d be able to achieve with the 24mm (my concern with the 24mm is that I’d end up in the same situation as with the kit lens discussed above - that image stack was taken at f/5).

.. Upon closer inspection of the equalized photos...

This is ISO 800:

24mm 10s f/5 ISO 800 by Christina Chester, on Flickr

And the is one of the images that created that stacked image at ISO 8000:

24mm 10s f/5 ISO 8000 by Christina Chester, on Flickr

Looks like there is a great deal of noise in the one at a lower ISO too 😟 This is a bit confusing now.

Cristina,

The 500 rule is a usable rule for imaging without visible star trailing so you should be able to get more than 10s with a wide angle. Understand that star trailing is still there but the reduced star size and wider view mean you can't see it easily.

Stacking is a method used to increase signal to noise ratio. Every image stacked has only one constant, the stars. Sigma Kappa type stacking will reject data that is not duplicated, so as noise is random, it gets rejected while the stars are kept. Up to 30 frames this means the S/N becomes much greater. Therefore you can then stretch to amplify the image without amplifying the noise as much. You can't fake it by taking one picture and duplicating it because the picture is identical so nothing is rejected and the S/N stays the same.

Taking more shots on a fixed tripod is OK if you don't have a foreground subject. If you do then you have to fake it and take a foreground image to place over the sky (that you have lowered).

Your camera has low noise due to on board handling of the image but it is not magic. Go above ISO1250 and it will become visible. The Sony sensors Nikon work with are getting close to ISO invariant. This means that if you took a picture at ISO 100 and one at ISO1600, then increased the exposure in processing on the ISO100 to be the equivalent of ISO1600 it would look the same. The advantage being that software is a bit more subtle doing this than the camera. In reality, its not really the ISO we need for astrophotography, its exposure, we need to capture more photons.

Why the fascination with getting a super wide angle lens? You will always get more distortion and more problems. Leaving lens aberrations aside a super wide means you will;

- Have a greater chance of a light pollution gradient

- If you put a light pollution filter on the front of the lens it will not work as well at the edges because the lens is 'seeing' light through the filter at an angle which is not filtered as effectively as light passing through perpendicular to the filters surface.

- Have to pay more relatively due to the complexity of the lens.

The wide angle I currently use is a Nikkor 24mm F2.8 AIS. Good, reliable, pro grade manual lens that new was over £500. It has a lens profile built into your Nikon and also in Camera Raw. Cost me £99 off ebay. No brainer. Why spend more until you feel the lens is your limit?

You really do need to get this 'I can't take pictures at F5' nonsense out of your head. If you limit yourself like that you will end up only taking pictures full of artifacts. As I said, most imagers are using scopes with f ratios ranging from F5-F10. When I do solar system imaging it is at F32+! Yes they are tracked but its photons you must capture which require extra time.

Tracking mounts are not expensive. You can make one for £50-£100. You can buy a reasonable one for not much more and a really good one for less than that Sigma. Depends what you want to do.

You will have to come along to the deep sky talk in March.

Cristina

The equalize function is just to show what is there. There will be lots of nasty's. Don't worry. Can you dropbox the original RAW file for the ISO 800 one so I can play?

Why the fascination with getting a super wide angle lens?

I’m not immensely keen on the appearance of super-wide photos, to be honest... I’m simply looking to find a means of getting around the problem of being able to take longer exposures without stars trailing and without spending money on a tracking mount. I wasn’t aware of the light pollution gradient issue that wide lenses experience, though. There you go! Another new lesson (amongst the many others in this thread)!

Thanks for offering to take a look at the RAW file for the ISO 800 image. I have uploaded the five photos that I took at ISO 800 in that series for you to enjoy…  Or shake your head in dismay at 🙂

I look forward to hearing what you think of those. A link will be in your inbox very shortly.

Why spend more until you feel the lens is your limit?

All for that! Just think that I've reached the limit with this particular kit lens.

Hi Christina,

I am a new member (you may remember my twins from the Venus observation last year).

Instead of a tracking mount for the camera alone, I just got an adapter to use with my existing Celestron mount - 20 pounds.

In terms of wide angle - happy to bring my Nikon 14-24 along for one of the next member observation evenings if you want to try that. Can't make Blackheath this month though as I am travelling.

Best

Thomas