Ideal Camera & Lenses for Wide Night Sky Shots?

[chargedmr2]

Attempting wide night sky shots is new to me, but I have been doing a bit of reading and so far, haven't come across much discussion of M43 equipment (though most of my reading has been about technique). FF cameras such as the 6D are obviously preferred along with fast lenses, but what about in the M43 world? Are there preferred cameras and lenses that usually top the list of recommended gear? Thanks for any insight.

 

[OzRay]

They work just as well as any other camera and have a good selection of lenses, if you're prepared to pay the price. A great lens for this is the likes of the 7-14mm, but a 12mm will also work well. Faster is always better, as you can use lower ISO and you should be shooting at no longer shutter speeds that 25sec, and ideally around 15 sec or so. Taking multiple shots and stacking is another option.

This was a shot I took with the E-M1 and ED 7-14mm f4 at ISO 4000 and 15 sec (plenty of room for improvement):

Subscribe to see EXIF info for this image (if available)

 

[Sammyboy]

.... E-m5 & o12/2 ....

 

[cmpatti]

The wider the lens's FOV, the longer exposure is possible without showing movement in the stars, so there can be a tradeoff between a slower, wider lens and a faster, longer one. For example, my former Pany 20mm 1.7 didn't really give me any advantage over my 7-14 f/4. I think that wider angle lenses produce better compositions (you can get more of the Milky Way in them). On a clear night at 7,500 feet altitude with a relatively dark sky, I felt that the 7-14 gathered just enough light at 7mm f/4, 30 sec, ISO 3200 to make decent astro landscapes on the Oly EM5. Here's an example:

Subscribe to see EXIF info for this image (if available)

The images require quite a lot of post processing to make them presentable (the example above is only half done--for one thing it needs some color adjustment to get rid of the green cast in the sky), and I wouldn't want to print them very large.

It might be interesting to try the Samyang/Rokinon fisheye (I don't own one) to see whether you can go longer on the shutter speed. The Oly 7-14mm f/2.8 will likely be the best lens in the system for this sort of thing when it comes out.

I understand that the Nikon 14-24 f/2.8 on a D800 is thought to be the best available combination for this.

 

[chargedmr2]

Would the E-M5 be preferred for color rendition or iso performance (or both). I believe I read that it outperforms the E-M1 in some regards (maybe high ISO). It I think the E-M10 outperforms both for ISO. Probably just splitting hairs, though.

I'm curious if the super fast Voigtlanders are useful for the night sky? Or is there nothing to gain since they don't get wider than 17.5mm, necessitating a faster shutter speed to avoid trailing (as compared to say the 7-14 as mentioned above). I guess the 4 stop advantage wouldn't be entirely lost by the change in angle of view, since that is a considerable difference. Rumor has it, Olympus patented a 12mm 1.0 lens, but who knows if that will ever be produced. If so, that could be a real winner.

 

[OzRay]

I very much doubt that there's any significant ISO difference between the various models. As has been pointed out, the wider that you can go, the better your results will be from photographic perspective. You can use narrower lenses and then stitch them together, but that will increase your workload considerably. I've tried some shots with my 14-35 f2, but they are no where near as effective as with the 7-14mm f4, when it comes to a single shot. I haven't tried the 8mm fisheye, but will give that a chance when I get the opportunity.

 

[LundyD]

I just got the Olympus 12mm f/2.0 lens, primarily for shooting meteor showers. Supposedly we're going to have a real dandy one later this month, but it's not even listed on most sites I've seen that list such things. It's the Comet 209P/LINEAR Meteor Shower that will only be visible a few hours the night / morning of Fri May 23 - Sat. May 24.

 

[bye]

Full frame is king basically for astrophotography, because with current technology, a Nikon D600 or a Canon 6D can create stunning night shots with minimal effort. You can shoot easily 6400-12800 and get great images. This is what I took as a rough shot with my Nikon Df and an UWA Nikkor lens.

Subscribe to see EXIF info for this image (if available)

The reason why full frame is easier is because, it has a 2 stops advantage. A lot of serious astrophotographers are shooting minimum ISO 6400 and sometimes going up to ISO 8000 just to get a faster shutter speed. The stars will appear sharper and the colors and look and feel are somewhat more pronounced. There are almost very little in terms of chroma noise of banding that you get from M43. In fact, I tried to get the same thing using the E-P5 with a few frustration.

Secondly, you have a lot more selection of full frame wide angle lenses that have little to no comatic aberrations. Comatic (not Chromatic) aberrations is when the lens project the image of the stars on the edges of the elements as little comets. If you are just doing it for fun and don't mind a few of those, I suppose it's fine. But if you are planning to be serious about it, then I can't find any wide angle M43 prime lens that can approach the quality of the Nikkors, Samyang or Zeiss because it's easier to get wide angle due to the non-crop nature of the full frame sensor.

So having said that, the best M43 ideal camera platform would be a 16MP sensor, preferably a Sony to get less noise and better color look and limit your ISO up to 3200. As for lenses, use either a 17mm f/1.8 which I used for night sky or the 12mm f/2 which I heard to be the best choice for the platform.

Now, you can use f/3.5 and f/4 lenses, but the results aren't going to look as spectacular like the sample photo I posted. It might look great if you haven't seen the photos made by a full frame camera and a skilled astrophotographer, but there is a limitation of what you can do and produce with a M43 camera. The smaller sensor is the weakness in this type of photography.

However, I've done quite well with my E-P5 in this area, but I reserve my Nikon Df if I want the best quality I can get doing astrophotography. Just don't expect the same kind of look you'll get from ISO 8000 done with a Nikon Df or Canon 6D.

Below is a shot taken in the same area with the E-P5 and 17mm f/1.8 @ 1.8 with hyperfocal distance. This allows me to extend the depth of field @ 1.8 so I can get the stars sharp as well as the foreground! If you don't shoot hyperfocal, the foreground will not look sharp. Hyperfocal distance was used with the Nikon Df shot as well.

Subscribe to see EXIF info for this image (if available)

Hope this helps.

 

[OzRay]

Funny that you should mention coma, as I was reading about that on Lens Rentals a few days ago and checked out my shots taken with the ED 7-14mm f4 and not one shot displays coma at any of the edges or corners, all the stars come out as full circles. I think that you can find m4/3 compatible lenses that are as good, if not better than the competition. I only used ISO 4000 for my shot, but I thought that noise was eminently manageable and acceptable. I'm going to see what else I can do and push the limits of the lens and camera, but I doubt that I'll be overly disappointed.

 

[PMCC]

If weight and money is not an issue, you may also consider a small equatorial tracking mount.

 

[OzRay]

I have one and I really need to learn to use it again. It is very large and heavy with the associated tripod, but it was actually reasonably inexpensive.

 

[penWS]

I tried astronography with my E-PL5 on a couple occasions and got decent results using the Panasonic 14mm 2.5 wide open with a Sony wide angle adapter and iso 1600. Now I have the Oly 12mm 2.0 which is possibly the best m43 lens for this.. or perhaps the SLR Magic 12mm 1.6. Have also got surprisingly good results with the Rokinon 7.5mm fisheye, stars remain sharp without visible trailing at exposures around 30-40 secs.

 

[chargedmr2]

Thanks for the feedback everyone. This is all really helpful. I'll try not to derail my own thread too much, but reading the responses raised a question about technique. Is it best to use hyperfocal distance, or is this simply for achieving greatest DOF for the foreground as in any other situation? What's the most foolproof way of ensuring crisp stars in the photo, from a focus standpoint only?

Although it was not about astrophotography, I've read that the hyperfocal distance doesn't produce maximum sharpness at infinity. So would it be best to focus out a but further than the hyperfocal distance? I've been using manual focus at 14x zoom with my E-M10 to satisfaction so far, but foregrounds have not worked there way into my practice shots much at this point.

 

[bye]

Funny that you should mention coma, as I was reading about that on Lens Rentals a few days ago and checked out my shots taken with the ED 7-14mm f4 and not one shot displays coma at any of the edges or corners, all the stars come out as full circles. I think that you can find m4/3 compatible lenses that are as good, if not better than the competition. I only used ISO 4000 for my shot, but I thought that noise was eminently manageable and acceptable. I'm going to see what else I can do and push the limits of the lens and camera, but I doubt that I'll be overly disappointed.

The 7-14 is not really the ideal lens for it because it is slow. If you want to achieve the best look and feel, you will need something a little faster than f/4. The ideal lens is a 12mm f/2, either from Olympus or the new Samyang. The SLRMagic 12mm Hyperprime Cine T1.6 is a lens I'm considering buying not for astro but for a unique landscape look. But since it's rated in T-stops, the f-stop would be @ f/1.4. That's even better for astro especially you can keep the ISO low.

In the ideal world, you want ultra wide angle, fast, sharp optically, superb Hi-ISO performance and very low coma aberrations. The Samyang/Rokinon 14mm f/2.8, which is pretty inexpensive for a UWA is available for $299 in EOS mount or $329 or so in Nikon F mount is an awesome price with little to no coma aberrations. We do not have something like a 7mm f/2.8 yet. The closest is the Rokinon fisheye 7.5mm @ f/3.5.

For me, the 7-14 is a too expensive approach for astro. The 12mm f/2 is better or the SLRMagic @ T1.6.

 

[bye]

One must understand what photography is all about. Photography is all about recording an image in 2 dimensions; not 3 dimensions. Depth is a created illusion due to the circle of confusion. Therefore, in photography, there is ONLY ONE PLANE on the subject matter that is truly sharp. Everything else in front or rear aren't technically sharp. They are sharp ONLY due to our vision perception. Since a human vision has its own rated MTF (resolving resolution), what one sees as sharp may not necessarily mean that another sees exactly with the same acuity as you do.

Infinity is infinity. Infinity is not a set distance, because every star you photograph don't have the same exact distances to the camera's sensor plane. So in theory, perhaps only one star in your camera vision falls in one of the ideal focus plane is sharp and every other stars are only sharp because they fall within the DOF. If we simply extend the front DOF to include the foreground, then you are simply moving the one focus plane slightly forward while maintaining infinity DOF. This only works well with sharp lenses, and not all ultra-wide angle lenses are sharp when they are wide open. So when people say hyperfocal distance don't work for them, I suspect it is because they are not using a good sharp prime lens that are both razor sharp and exhibit little to no coma aberrations. You have little selection in the M43, but plenty in the full frame world. The acuity does not drop with a good sharp lens if you miss focus, because with the extra resolving power, it still gives you good perceived sharpness because it made up by being able to capture more of the detail. This is especially true with a 16MP sensor and up. You get color tonal changes that are smooth and pleasing. Whereas an already not so great lens which performs well if you focus correctly start to rear its ugly head when you start to mis-focus.

Another potential challenge to experiencing poor success with hyperfocal photography is the fact that most modern lenses do not have printed hyperfocal distance chart like the good old school lenses do. So you guess. Fortunately, my 17mm f/1.8 has this and so I assume the 12mm f/2 has the same thing. I find it pretty accurate. I use DOFmaster.com to calculate my hyperfocal setup and it works really well. In both of my photos, the large prints are sharp on both the stars and the pier.

Therefore, if you have a so so prime lens that is somewhat sharp, stay with infinity. If you have a better sharper lens, you can experiment with hyperfocal.

 

[LovinTheEP2]

I very much doubt that there's any significant ISO difference between the various models. As has been pointed out, the wider that you can go, the better your results will be from photographic perspective. You can use narrower lenses and then stitch them together, but that will increase your workload considerably. I've tried some shots with my 14-35 f2, but they are no where near as effective as with the 7-14mm f4, when it comes to a single shot. I haven't tried the 8mm fisheye, but will give that a chance when I get the opportunity.

Stay away from the EM-1 for Long Exposure work like you would be doing for nighttime star photography. The PDAF array on the Panasonic sensor of the EM)1 add's a crazy amount of noise. I haven't been able to see any reports of the EM-10 if it suffers the same as it uses the same Image Processer but the sensor from the EM-5 which doesn't exhibit the long exposure issue. So it's an ISO issue but a exposure length issue. Yes it can be addressed with dark frame substraction (even though in my opinion there is a tradeoff in IQ as a result) but for long exposure night sky photography that isn't really an option though as stars move etc. and try to clean up the noise with dark frame substraction would take some serious serious work.

 

[LovinTheEP2]

Thanks for the feedback everyone. This is all really helpful. I'll try not to derail my own thread too much, but reading the responses raised a question about technique. Is it best to use hyperfocal distance, or is this simply for achieving greatest DOF for the foreground as in any other situation? What's the most foolproof way of ensuring crisp stars in the photo, from a focus standpoint only?

Although it was not about astrophotography, I've read that the hyperfocal distance doesn't produce maximum sharpness at infinity. So would it be best to focus out a but further than the hyperfocal distance? I've been using manual focus at 14x zoom with my E-M10 to satisfaction so far, but foregrounds have not worked there way into my practice shots much at this point.

The great images I have seen with respect to foregrounds I believe are processed basically as seperate focus exposures and blended into 1 image with Photoshop mask layers after stacking each part seperately (above horizon, below horizon stack sets) - I think most are illusians that they are all 1 focus point images.

Any chance you could repeat some of the work people have done with their EM-5 and EM-1 with resprect to long exposure dark frames to see which profile the EM-10 exhibits?

 

[chargedmr2]

One must understand what photography is all about. Photography is all about recording an image in 2 dimensions; not 3 dimensions. Depth is a created illusion due to the circle of confusion. Therefore, in photography, there is ONLY ONE PLANE on the subject matter that is truly sharp. Everything else in front or rear aren't technically sharp. They are sharp ONLY due to our vision perception. Since a human vision has its own rated MTF (resolving resolution), what one sees as sharp may not necessarily mean that another sees exactly with the same acuity as you do.

Infinity is infinity. Infinity is not a set distance, because every star you photograph don't have the same exact distances to the camera's sensor plane. So in theory, perhaps only one star in your camera vision falls in one of the ideal focus plane is sharp and every other stars are only sharp because they fall within the DOF. If we simply extend the front DOF to include the foreground, then you are simply moving the one focus plane slightly forward while maintaining infinity DOF. This only works well with sharp lenses, and not all ultra-wide angle lenses are sharp when they are wide open. So when people say hyperfocal distance don't work for them, I suspect it is because they are not using a good sharp prime lens that are both razor sharp and exhibit little to no coma aberrations. You have little selection in the M43, but plenty in the full frame world. The acuity does not drop with a good sharp lens if you miss focus, because with the extra resolving power, it still gives you good perceived sharpness because it made up by being able to capture more of the detail. This is especially true with a 16MP sensor and up. You get color tonal changes that are smooth and pleasing. Whereas an already not so great lens which performs well if you focus correctly start to rear its ugly head when you start to mis-focus.

Another potential challenge to experiencing poor success with hyperfocal photography is the fact that most modern lenses do not have printed hyperfocal distance chart like the good old school lenses do. So you guess. Fortunately, my 17mm f/1.8 has this and so I assume the 12mm f/2 has the same thing. I find it pretty accurate. I use DOFmaster.com to calculate my hyperfocal setup and it works really well. In both of my photos, the large prints are sharp on both the stars and the pier.

Therefore, if you have a so so prime lens that is somewhat sharp, stay with infinity. If you have a better sharper lens, you can experiment with hyperfocal.

Thanks again for taking the time to put together these responses--it's very helpful for my understanding of some of the basics.

One day I may look into getting a good FF camera in addition to my travel friendly M43 system, but for now, my skill levels simply can't justify it. I suppose when I get to the point that my limitations are gear related, it will be time to upgrade.

 

[chargedmr2]

The great images I have seen with respect to foregrounds I believe are processed basically as seperate focus exposures and blended into 1 image with Photoshop mask layers after stacking each part seperately (above horizon, below horizon stack sets) - I think most are illusians that they are all 1 focus point images.

Any chance you could repeat some of the work people have done with their EM-5 and EM-1 with resprect to long exposure dark frames to see which profile the EM-10 exhibits?

My wife and I are leaving on Sunday to travel Turkey and Greece for two months, so I don't think I can get to this before our trip. But I am very curious as to what the results will be and will definitely do some testing as soon as I can. Hoping for E-M5 performance in this regard!

 

[bye]

Thanks again for taking the time to put together these responses--it's very helpful for my understanding of some of the basics.

One day I may look into getting a good FF camera in addition to my travel friendly M43 system, but for now, my skill levels simply can't justify it. I suppose when I get to the point that my limitations are gear related, it will be time to upgrade.

Personally, I would not recommend getting a good FF unless you are serious about this. It's a significant investment. If you must carry one on a trip, some people buy a used Canon 5D Mark 1 12MP full frame or a Nikon D700 and pair it with a Rokinon/Samyang 14mm along with their APS-C Fuji/Sony or M43 kit. However, any M43 body based on a Sony sensor will perform well including your E-M10 and the only thing you need is a fast sharp prime lens, which is the O12mm, but I see the Samyang is not too far behind. It would be a waste of money simply to get FF and a lens just for one specific type of photography, whereas M43 with the Sony sensor excels pretty much in most things like my Nikon Df (the Lord of Darkness) against my E-P5 whereby the fast primes and zooms I have for my MFT kit and IBIS pretty much cancels out the shortcomings. In fact, my E-P5 outperforms the Df because it has IBIS in regards to low light photography. The only thing that it does not do well is astrophotography because of the smaller photo sites. IBIS in this case can not overcome this limitation. Don't worry it. Just bring a GOOD tripod and shoot. That has more bearing than a good full frame sensor. If your tripod is flimsy and weak, any breeze and blowing wind make for unsharp night photos. No ultra-light tripods; minimum is a Manfrotto 190 or heavier because of the longer exposure.

In regards to focusing a star in live view; try turning Live View Boost to on. This will help you manually focus on the brightest star. Get an app by the name of Stellarium for your smart phone or tablet (both iOS and Android) and it will tell you the stars alignment, constellation etc which will become helpful to you to get the best star shots. Also check clearskycharts website for the clearest night sky. All of this preparation will make for a better star photo then the best FF equipment. And then when you take more photos, you will decide if you really need FF, but I still strongly suggest going the used route while keeping your E-M10 kit.
I only use FF because the Nikon Df is one of my work cameras.