What in the world is a deep-space time-lapse? Would it be so unconceivable and unfeasible to associate longer focal lengths and a background? How about getting more colors and details into an astrolapse? Here is the recipe of a brand new field of astrophotography!
If you have been following me in the past three years or so, you might have realized a real shift in the type of astro time-lapse I've produced. As I started astrophotography about four years back, I was instantly drawn to the possibility of putting the frames I would take back to back to produce motion. I've been captivated by the mesmerizing movement of the stars, milky way and other fascinating celestial events ever since. In the beginning, I remember myself watching a lot of Youtube tutorials on how to shoot a nice and smooth time-lapse of the night sky. To me, the hardest part as a beginner was to be consistent. I acquired the technique rather fast, as I went out shooting every other night, but the most challenging was to produce a high quality, jitter-free and smooth sequence (don't get me wrong, I still struggle with it to this day, and so do a lot of time-lapsers). I have now gained a lot of useful experience and I have learned from my many mistakes to come to this point. However, I was never satisfied (I never am), and I quickly lost interest for wide-angle astro time-lapse. You know, this kind of beautiful scene where the milky way galaxy moves relative to a foreground, but at a wide angle. Of course you can change places and capture new situations with new foregrounds, but I knew it was just scratching the surface with astro-timelapse. Until our time, astrophotography has been a very restricted, inaccessible and closed market, even more so astro time-lapse because of of the existing means, finance, interest... Nowadays, the improvement of our lifestyles and the technology has enabled a veritable revolution where citizens can buy professional gear and learn how to use it. It is especially true with home telescopes and digital cameras in the field of astrophotography. Time-lapse has always been a very competitive market as well and it is now booming, but astro-timelapse has no real market yet (maybe a small one for aurora advertisement). It's merely been a few amateurs and professionals that enjoy it, explaining why this field hasn't completely been explored yet.
About three years ago in Denmark, a weird idea crossed my mind one day, and without knowing if it was going to succeed, I took my star tracker out in my backyard (it follows the stars to allow a longer exposure time, so sharper and brighter images), and started photographing one of our closest neighbor galaxies (the Andromeda galaxy) at a focal length that would allow me to start seeing nice details (about 100mm) but still see a foreground (a tree line) while still taking advantage of a wide aparture. By exposing about 20 seconds, at f/2.8 (maximum aperture on the Canon 100mm f/2.8 Macro) with an ISO of 3200, my Sony a7rII was able to not only pick up the light but also incredible details and nebulosity in the arms of the galaxy while it was moving relative to the trees, with occasional satellites and shooting stars photobombing the frame. Looking at the final sequence then, I absolutely loved what I saw and started digging some other videos/tutorials about what I would call a 'deep-sky' time-lapse. I got a shock when I found next to nothing on the internet that would ressemble what I had produced. Of course it doesn't mean no one had ever tried it, but the only photographer that experimented in this novel field at the time was Randy Halverson (Dakotalapse: http://www.dakotalapse.com). His video 'Huelux' includes some rare medium-format and deep-sky time-lapse scenes that you can view here:
From that moment on, I knew I had just scratched the surface, and that it's been right there under our nose all along, waiting to be explored. I knew that the technology that is being produced today was performant enough to show this kind of stuff. Why has no one really thought about it before? The answer might be quite simple: deep-sky photography usually requires sophisticated gear and a lot of skills, but most importantly it requires an extremely long exposure time, several stacked light frames along with calibration frames to increase the signal-to-noise ratio. Some amateurs spend multiple hours imaging just one part of the sky, sometimes over several months or years, using filters and other equipment. So the simple idea of getting a relevant and interesting result in one light frame with an exposure time around only 30 seconds has been pretty unconceivable. Besides, why would you put this kind of frame together if there is no background and nothing moving? That would be very boring to watch, wouldn't it? Well, my experience from three years ago told me that it turns out to be anything but boring, irrelevant and undoable.
Throughout those three years, I've been shooting a lot of sequences which enabled me to improve my technique and come up with some advice for someone that would like to try it out. So here are some elements of recommendation to get decent results:
1) LOCATION: Any location that's far away from light pollution will do. Try to avoid areas near oceans as the humidity from the water usually brings undesired noise, especially for medium-format. The best locations are mountains and dry areas.
2) DATE/TIME: Depending on what deep-sky object you are shooting, the time of year will also vary, so I really advise you to use an app that will enable you to plan ahead the position of the stars at a certain location and time (I personally use StarWalk 2, but there are a bunch of others). You absolutely need a moonless night, as even the slightest moonlight will fade deep-sky objects.
3) PLANNING: As you will have a bit more gear to take with you, you will need to plan your shot very carefully. Allow a lot of setup time prior to shooting, especially if you don't know your gear very well. It's a good idea to try out a scene during the day to see if it matches your taste, and if it's really feasible. You can always improvise at night, but it's way more complicated and can be rather dangerous...
4) CAMERA: Even more than for regular astrolapse, you really need the most low-light performant sensors on the market, e.g. the ones the will give you the lowest read noise, so the best final signal-to-noise ratio (also the best dynamic range, but I would say read noise performance is of utmost importance). To find the market's current cameras that pack the best low-light sensors, I invite you to watch my tutorial here. For your reference, I would advise the Sony a7s series, the Canon 6D, 5D mk IV, Nikon 810a...). With these cameras, you will be able to go to ISO 6400 without having noise problems.
5) LENSES: This is practically the most important point here. Since you will increase your focal length, you will also decrease your exposure time if you are not using a tracker, and even if you are, longer focal length lenses often have a narrower aperture. After having run a lot of tests on focal lengths from 50 to 135mm, I am 100% positive that the brands Sigma (50mm) and Samyang (50mm, 85mm, 135mm) produce as of today the sharpest, cheapest, brightest lenses for astrophotography. It is probably because of them that I am now able to get so much details and brightness in my pictures! At these focal lengths, you need very wide apertures (from f/1.4 to f/2). Narrower apertures will considerably worsen your ability to capture light and eventually details. I am currently evaluating lenses with a focal length longer than 135mm, as I am trying to zoom in even more and see where the interesting limit is.
6) PORTABLE TRACKER: I have personally succeeded in taking deep-sky time-lapses at a maximum of 135mm without a tracker, but since you'd have to keep your exposure time under 4-5 seconds (unless you want blurry objects), you would rapidly be limited to capturing very bright areas of the night sky (aka: the milky way) in pristine conditions. Since these conditions don't happen too often (especially if you live in humid areas) and since you also want to capture fainter objects (nebulae), or even since you might want to try even longer focal lengths, I highly recommend getting your hands on a portable tracker. A tracker compensates for the rotation of the Earth and enables your camera to follow the stars so you don't get star trailing and can expose longer. You will then get sharper images, more light, less noise, more details, contrasts and colors. It literally changes the quality of your shots. Remember that the sky will be fixed, while the foreground (if you have one) will be 'moving', hence the importance of planning ahead! Now, since you might want to shoot different scenes, you will need a tracker that travels well and that's easy to set up. I would recommend the Vixen Polarie or the Star adventurer mini which can easily be mounted on any travel tripod. With more experience, you can also use a motion control system that can also follow the stars. It is less precise than star trackers, but it gives you more freedom to control every component of the motion.
7) FILTERS AND ASTRO-MODIFICATION: An extra piece of gear to consider is a light pollution filter. Most light pollution filters (Clip-in Astronomik or Optolong) won't do, as they introduce too much coma, color shift, distortion, light fall-off, and sometimes sharpness. On top of that, they usually cut off the light too much, meaning you would have to bump up all your settings to get the same results. However, some new external light pollution filters have appeared on the market in the recent years (Lonely Speck's Pure Night filter, Optolong Clear Sky Filter, Nisi Natural Night Filter ...). While these filters seem to work less for wider angles, they seem make a difference using focal lengths higher than 85mm, even when the light pollution is weak, and seem to increase the contrasts neatly without compromising the direct quality of the picture (good sharpness, no distortion, coma or fall-off...). You will still need to bump up your settings to compensate for the loss of light though!
Another important operation to consider is to astro-modify your camera (read my article about astro-modification here). As you were probably able to see in my videos, I wouldn't have been able to pick up so much red nebulosity in my deep-sky time lapses if it weren't for my astro-modded Canon 6D. It really nicely picks up all the H-alpha emissions in the sky, even the faintest! Since Hydrogen gas is probably the most abundant in quantity in our universe, allowing your camera to pick up one of its most common emission bands (H-alpha, 656.28nm) will tremendously improve the level of relevance and interest of your time-lapse.
8) SET-UP and ON-LOCATION: it might not seem like it, but shooting at longer focal lengths is more difficult than it sounds at night. Supposing that you checked all of the above points, you will realize that practice is always different from theory. On-location, the humidity and most importantly the wind will be your worst enemy. What is more distracting and annoying to look at than a shaky video? The more you increase your focal length, the more your lens and so your frame will be sensitive to even the slightest breeze. Moreover, longer focal lengths usually mean longer lenses which makes your rig more exposed to the wind, introducing jitter, blurry stars and produce shaky videos. Since you want to avoid that, you need to find a spot or a day without wind.
One good recommendation I have as well is not to shoot too close to the horizon, as the starlight has to travel through more of the atmosphere, creating more noise and less clarity/light. If you want to include a background, you will not only need to be far away from it (long focal lengths usually have extremely long focusing distances, even for slightly higher f-stops) to have background and foreground in focus, but you will also prefer a foreground that's higher in the sky, for example a mountain, say at least 15 degrees above the horizon, otherwise you will get disappointing shots. In my last movie Galaxies, you can see that I tried some close-ups of the milky way core setting on the La Palma horizon, and it was decent, but it could have been much better: even if I was higher in altitude (so better conditions), the distortion of the atmosphere is still too important to really get the contrasts and colors right. I really had to push the edits in post-process, creating more noise. However, if you look at the sequence showing the Scutum area sliding above the mountains at 85mm, it's much higher in the sky, thus significantly clearer and lighter (this area is much fainter than the core though!). I didn't have to edit these frames very much to get to those pictures, and they are not even tracked (Sony a7s, Samyang 85mm f/1.4 @ f/2, ISO 6400, 6'') :
9) POST-PROCESSING: Post-processing and your workflow is extremely important to produce top-notch sequences. You want to lose as little quality as possible during all your process workflow. Here are a few editing recommendations I can give out for Adobe Lightroom. The first thing you need to keep in mind is that noise in a still picture ins't the same than noise in a film. Grain tends to move around and it can be very distracting. One will generally be accepting a higher level of noise in a still picture than in a movie, so you will have to keep noise at its lowest while bringing out contrasts and colors. There are a few buttons I wouldn't touch too much. For example, I generally don't touch the contrast button, as keeping the middle-tones at a high level can help drown part of the noise. Also, there is a big misconception about the clarity button in astrophotography, even for stills. A lot of astrophotographers tend to bump up the clarity to try and reveal more details in the milky way. In reality, all it does is make the stars too bright, so much that when you export your picture and view it in full quality, the nebulosity and deep-sky objects are smothered by an ocean of white dots... I for one actually reduce the clarity (especially if you are using a tracker) by more than 50%, sometime more than 80%. You will lose just a bit of details in the milky way (it's actually insignificant), but this step allows you to reduce star brightness while enhancing the background sky (nebulosity). My sequences would not look like that (especially the Orion area filled with tiny stars since it is close to the milky way) if I hadn't reduced the clarity.
Since our atmosphere also alters colors, I would advise adjusting them using the dials at the very bottom. Try and google pictures of the object you are interested in, and pay attention to its 'natural colors': that will help you adjust them in post-process.
One recommendation when you are editing your pictures is to avoid using too much noise reduction right away. Since noise is better seen when it's moving around, I would apply about 20-30% in editing and more later in my video-editing tool if need be. I actually prefer a bit more noisy picture (because the nebulosity is somehow grainy too) than one that has been smoothed out too much, it looks less natural to me.
As you export, assemble and store your individual sequences, make sure you use a lossless format: you will have to make a compromise between quality and space on your computer. I personally use the codec Apple ProRes 422, 24fps, and original camera dimensions, which I can cut down to 4K or even less later. Afterwards in film-editing, there are also a number of tools that I use to get rid of the noise, apply a movement, effects...
In my new film 'Galaxies Vol II: Wonders of the winter night skies' that you can watch above, I really wanted to show the product of these 3 years of experimenting in the domain of medium-format and deep-sky time-lapse using the techniques stated. I did it for several reasons. The first one being that it really is possible to produce such single frames that pack so much details, colors, and contrasts with the right gear, skills and under the right conditions. The second reason is that I wanted to show that this type of sequence can be very relevant for recreational, educational and even for scientific purposes. After a lot of feedback and according to my personal thought, I believe these