I think that the answer is yes – but there is one point that must be addressed first which is relevant to the title; the Sony ‘Star Eater’ problem. To ‘improve’ the low light performance of the Sony α7 range of full fame mirrorless cameras, Sony chose to implement a noise reduction algorithm which was applied when long exposures are taken and, annoyingly for astrophotographers, this was also applied to the raw files. Aiming to eliminate hot pixels, the algorithm tended to remove small stars as well − hence the name. There are some ‘workarounds’, in particular by using a ‘continuous’ shooting mode for which details can be found on the web. In the later Sony α7S II, the noise reduction was applied to any exposures longer than ~3 seconds but, pleasingly, in the Sony α7S it is only applied when in ‘bulb’ mode so defined exposures up to 30 seconds in manual mode are not affected. As I explain in the digest article ‘What exposure to use for Astrophotography’, I feel that the effect (not quite) of a single long exposure of, say, 30 minutes, is best achieved by aligning and stacking 60, 30 second, exposures so for the Sony α7S I do not believe that this a real problem.
So it could well be that the older Sony α7S is a better buy than the α7S II and the new price for the body has dropped to ~£900. In addition, they are now available in some numbers on the second hand market with ‘good’ copies available for ~£600. I was able to acquire one in very good condition for this sum which, pleasingly, had a very low shutter count.
An Imaging Example
This image of the Cygnus and Lyra constellations was taken using a Zeiss 45 mm Planar lens – one of the sharpest Zeiss lenses ever made.
Unlike any other modern full frame camera, the α7S series has just 12 megapixels − when 24 or greater has become the norm. As the pixels are larger, each pixel can collect more light and so the camera has excellent low light capabilities so, at first sight, making it better for astrophotography than the high resolution full frame cameras. This is not quite true. Assuming the quantum efficiency is the same, what really matters is the total size of the photon collecting area and this will remain fairly constant independent of the number of pixels on the sensor. To make a real comparison, one needs to compare the noise performance of a high resolution sensor such as that in the Sony α7 III when scaled down to the same size (in pixels) as the α7S. This will average or ‘bin’ adjacent pixels so reducing the noise. It turns out that the effective sensitivity is pretty much the same.
There is, however, a real advantage for astrophotography − that when aligning the camera on its target. By initially setting a very high ISO, the ‘live view’ enables faint objects such as, for example, the pair of galaxies M81 and M82 in Ursa Major to be visible without having to take trial exposures to make them appear in the frame. I have often spent several minutes taking a number of exposures in order to align and orientate a camera when either using a lens or when mounted on a telescope (when the problem tends to be worse as the focal ratio is typically greater). One will, of course, reduce the ISO, as discussed below, before actually taking the series of frames to image the field of view.
The α7S has a very high quantum efficiency of 62% (greater than the α7S II at 55%) and each pixel can hold ~153,000 electrons (its well depth) before saturating − higher, I believe, than any other camera save the α7S II which is fractionally greater. This is not surprising as their pixels are typically 4 times the size of other full frame cameras. This should help in giving it a high dynamic range.
The very high sensitivity should make perfect to capture meteors, either when doing a continuous sequence of, say 20 second exposures – when many more meteors will appear than seen by eye – or making a video when, with a very high ISO of say 12,800, 1/30th of a second exposures can be used to follow their path towards the ground. I doubt that this would be possible with any other camera. Alyn Williams has a Youtube video showing this in action.
All mirrorless cameras have two advantages (one great) over a DSLR.
Firstly, as the distance from the lens flange to the sensor is less than that in a DSLR, ‘non retrofocus’ lenses can be designed which may give a better image quality. (Not all use this advantage; the ‘kit’ lens for the α7S has a lot of space between the rear element and the lens mount and is almost certainly an adapted DSLR design.)
Secondly, and of real importance, is that by the use of adapters (such as those made by K&C Concept for ~£25) it is possible to mount full frame lenses from almost any full frame DSLR or Film camera system onto the A7S e-mount. They will need to used in manual focussing mode when astroimaging (as one would in any case). This makes available a vast range of legacy full frame lenses which can be used and I have adapters so that I can mount lenses from my collection of Nikon film camera lenses, some full frame lenses made for my Sony A-mount DSLR camera and also, the Zeiss lenses that I acquired for my Contax G film camera. It is great to give them a new lease of life! One example is the Tamron Adaptall 24mm f/2.5 which, when stopped down to f/4, has an excellent performance. They can be obtained for £60 to £100 along with an adapter to enable it to be mounted on the α7S. This lens is perfect for wide field imaging of the Milky Way with a full frame camera. [The K&C Concept adapters will, where possible allow the lens aperture to be set but for lenses which have an included aperture control, simpler adapters can be used such as, for example, some of my legacy Nikon lenses such as the Micro Nikkor 55mm or Nikkor 200mm lenses.]
What ISO to use for astrophotography
There is a detailed article in the digest which would be well worth reading. Many Sony camera are aid to be ‘ISO Invariant’ meaning that it is possible to use a lower ISO than one would normally use and then increase the brightness of the fainter objects (stars or nebulae) in the image in post processing. This preserves the dynamic range so that the brighter stars are not brightened as much as fainter objects. If a higher ISO is used the all parts of the image are brightened equally in camera so that the brightest parts of the image may be blown out – such as the ‘Trapezium’ region in the Orion Nebula and the heart of the Andromeda Galaxy. At low ISOs, the Sony α7S is not as ISO invariant as many other Sony cameras as its sensor has a rather different amplifier-gain architecture than most current Sony sensors. There is an unusual step in its characteristics when ISO 2,000 is reached where it is believed that the analogue to digital converter switches to a low read noise mode. Beyond this ISO, the camera is essentially ISO invariant. As increasing the ISO always reduces the dynamic range I suspect that, in general, the use of ISO 2,000 or 3,200 would be best for astroimaging. However, in order not to over expose very bright regions in an image as mentioned above it may then be better to use a lower ISO to give a greater dynamic range.
Using a manual lens (or an autofocus lens in manual focussing mode) there is a focus magnifying feature. If the ‘c3’ button (‘delete’ when viewing images) at the bottom right of the camera is pressed, an orange square appears in the centre of the frame. This can be moved across the frame by pressing on the control ring. By pressing the centre ‘OK’ button this area is magnified first by 4.2 times and, with a second press, by 8.3 times. Though slower, this enables a more accurate focus to be achieved than with the ‘focus peaking’ feature. For astrophotography use, I have often carried out this procedure in daylight and then taped up the focus ring. In fact, the older non autofocus lenses usually have an accurate end stop at infinity so that focussing is not a real problem.
The α7S has a tilting screen which makes it ideal when imaging at high elevations. There is a second advantage: when opened out, it allows air to reach the rear wall of the camera behind which lie the camera’s electronics and so helps to keep the sensor cooler than otherwise. It would even be possible to insert a small icepack between the screen and rear wall so applying some ‘active’ cooling to the camera.
Long Exposure Noise Reduction
By default, when taking long exposures most cameras employ this technique to remove hot pixels from the image. The α7S is no exception. A second exposure (a dark frame) of the same length is taken with the shutter closed and the two differenced in camera. This is fine when single exposures are taken but probably not when a long sequence of exposures is taken when astroimaging. Firstly and obviously, it halves the total time capturing photons from the sky and secondly, but not so well known, the dark frame actually adds some noise into each frame. Hot pixels can be removed in post processing and so it is almost certainly best to switch off the process in the camera’s menus.
Camera Control when astroimaging.
When astroimaging one often takes a series of short exposures, say 30 seconds, which will then be aligned and stacked in Deep Sky Stacker or Sequator. It is obviously good if these exposures can be taken automatically. Some cameras, such as my Fugi X-A10 have an internal intervalometer mode but, sadly, the Sony α7S does not. There are, however, two ways in which the taking of a sequence of short exposures can be made
- Using an external intervalometer
Costing ~£15 the control cable can be plugged into the micro USB port and set to give the desired number of exposures with its interval set to a few seconds longer than the manual exposure to allow time for the frame to be written into the SD card.
2. Using the camera’s WiFi facilty with an app running on a smartphone or tablet
I have used two free Apps called Time lapse and DSLR Dashboard to act as intervalometers which run on my Samsung tablet. Selecting the ‘Smart Remote Embedded’ in the cameras applications menu, a password was provided that was inputted into Time Lapse to initiate the WiFi linking between them. The App first allows single exposures to be taken to enable the camera to be aligned onto the target object(s). When a single image is captured, it appears in the ‘gallery’ and can be zoomed into to check that the lens is focussed on ‘infinity’. It can then be used to take a sequence of frames.
Having linked the camera to my Samsung Tablet in the Time lapse App, I then found that I could then control the camera using a rather more versatile App, DSLR Dashboard.
My, second hand, α7S came with an external mains power supply which provides 5 volts over a micro USB cable to plug into the micro USB port when the camera is off. This charges the battery internally. (An orange LED beside the USB port shows when the battery is being charged.) This can be useful when driving as the battery can be charged using a 12 volt to USB 5 volt converter (such as the Anker Car Charger ~£6 from Amazon).
However this is not good when several batteries are needed to be charged − and they will be as the A7S battery is really rather small. (But, pleasingly, the same battery is used as in my Sony A5000 camera.) In this case, it is best to buy a dual battery charger with two batteries such as the Powerextra charger available from Amazon for ~£19.
Using a ‘dummy’ battery.
As the capacity of the batteries used by the camera is not great, for serious astrophotography use it may well be best to use an externally powered dummy battery. They can either be powered from the mains as a combined unit or with 5 volts from a USB port. I think that the latter will be best. This type could either be powered from the mains unit that was supplied with the camera to internally charge the battery. A better solution would be to use a ‘power bank’ as used with smart phones or tablets. These, such as my Grundig, 8,000mAH capacity unit, will easily power a full nights imaging.
Could the α7S be used as a general purpose camera?
In this case one would probably want to acquire lenses which are autofocussing and perhaps image stabilized. The good quality, image stabilized, Sony 28-70 mm f/3.5 to 5.6 ‘kit’ zoom can be obtained second hand for ~£140.
Does one need more than 12 megapixels? Very rarely, and it is possible to make prints of 14 inches by 10 inches at 300 dpi and even ‘competition’ prints at 16 by 11 inches would still be at ~260 dpi. A little interpolation and sharpening could provide excellent prints for any normal use.
I have to say that, though less in weight than my Nikon D610 DSLR, it is not light in comparison to my mirrorless APS-C cameras and I will continue to use these for general ‘out and about’ use. In particular, my 21 megapixel APS-C Sony A5000, which is reviewed in the digest, has a tiny extending kit lens and fits easily into a jacket pocket.
This is where a higher number of pixels could well be required. But, usually, the subject is static and so one can ‘stitch’ a number of frames (which are taken using a longer focal length lens) together in software such as the (free) Microsoft Ice program. The α7S has a 4240 x 2832 pixel sensor. If the camera is held vertically and three overlapping frames are taken and combined, a similar proportion 3×2 image will result having 6360 x 4240 pixels, giving nearly 27 megapixels! I have done some tests stitching three images taken with a focal length of 58 mm and compared the result with a single image taken at 38 mm giving the same field of view. The stitched result is obviously a lot sharper, but what I was not expecting is that, when reduced in size from 27 megapixels down to 12, the stitched result is still significantly sharper than the 12 megapixel image.
The Sony α7S was not designed for astrophotography − it was really designed to take videos in low light and can even output 4K video to an external storage device. It can produce excellent AVCHD 2.0 videos and can also use the higher quality XAVC-S codac. It has provision for an external microphone (quite common on DSLRs) but also allows the use of headphones (which most don’t). This is another of my interests and is a major reason why I have purchased both the camera and the 28-70 image stabilized lens mentioned above.
If I am honest, I do not really expect to take higher quality astrophotos than with my Nikon D610 full frame camera which has a superb 24 megapixel sensor. But I do think that it will be easier to use; the alignment problem is greatly eased and the WiFi remote control ability very useful. As its size and weight is less, I will be more inclined to take it with me by air to dark sky sites and so perhaps produce better images than with the Micro 4/3 or APS-C cameras that I have, so far, been using when travelling afar.