The milky way over the Wadden Sea

This image was taken with the X-T4 and the Laowa 9 mm. The sky was a 4 minute exposure using the iOptron Skytracker Pro at 800 ISO, the foreground a 20 seconds exposure at ISO6400 without the tracker.

As an avid sailor, I try to make it a habit to make a sailing trip to the Wadden Sea, a tidal sea in the north of the Netherlands, at least once a year. Naturally I try to combine this with taking photographs, and as it offers a stunning landscape with its mudflats and wide horizons, usually I come back with a crop of images that I’m really happy with. However, one image has been eluding me for the past 4 years. I had in mind to photograph our flat bottomed ship (a Vollenhovense Bol, for those interested in such things), high and dry on the mudflats, with the milky way in the background (astrophotography being another genre which really interests me). This is not as easy as it sounds, however. The area of the Wadden Sea is generally one of the darkest in the Netherlands, so no problem there. The problem is in the tides, and the weather of course. Apart from having to be granted a clear night sky, a feat by no means a given with the weather patterns in the Netherlands, the timing with the tides must be exactly right. The ship must be on dry land somewhere between 11 PM and 1 AM, there should me no moon, and the ship should also be floating again at a not too ungodly hour to continue what is in the first place a holiday.

In april we managed to grab a weeks sailing, and on the final day I was granted a clear night sky. However, the moon was up and ruined any chance of a good milky way shot. In addition, the bottom of that part of the sea where we had parked for the night was not solid sand, but a sort of sludgy ooze. Not the best ground for a tripod. As one of our shipmates was an astronomer, we banded together with our photography and astronomy apps, and concluded that the best chance to get a nice holiday and good weather, and make my dream photograph, was August 13th. As it happened, we could rent the ship again that week and we booked it on the spot (helped also by the fact that this was a holiday that was solidly Coronoa-proof, something quite uncertain for other holidays at that time). 

So, last week was that week. How did it go? Well, the 13th of August was cloudy, so no joy. Fortunately, we saw that coming and made the most of the few nights preceding the 13th for some attempts. The best opportunity came on the 9th. We had beached our ship on the mudflats just northeast of the little harbour of Noordpolderzijl, Groningen. We knew that between 10 PM and 1 AM we would probably be dry on the flats, and hoped for solid sand. Our wish was granted, and apart from the light pollution of the large city of Groningen to the south, the conditions were perfect.

Usually I only bring my small travel tripod on holiday, but as there was so much photography at the core of this holiday, I decided to bring my large and heavy tripod. To protect it from the sand and the residual salt water the bottom part of the legs were wrapped with garbage bags and ductape. I also decided to bring my iOptron startracker, as this tripod was sturdy enough for its weight. As camera I was a bit torn between the Fujifilm X-T4 and the Fujifilm X-T2, the X-T4 winning in the end by a narrow margin (although in hindsight I think I would have preferred the X-T2, there were some weird issues with unsharp images on the tripod, but this might have been due to shifting sands). The 9 mm Laowa and 12 mm Samyang completed the kit.

Between 10:45 and 0:45 I played around on the sand next to the ship, as my shipmates were stargazing in the dark. After some preliminary shots of the milky way and deciding which angles were best for the ship – milky way combination, I tried some extra shots with the star tracker. As at 0:30 the moon rose (also a magical sight) and the water started to rise again, I got back on board. Trying to edit a milky way photo on an iPhone really doesn’t do it justice, and the tracked images could not be edited anyway (as the foreground was blurred due to the motion of the star tracker), so really checking the files had to wait until we got home. 

For the final figures I cannot admit to few edits. The best files were obviously the ones where I had used the tracker (at ISO 800, in order to keep the shutter speeds at a manageable level), but these files had to be combined with the other images which had a sharp foreground. No problem for the mostly straight horizon, but not so easy for the ship, which I had to manually remove the background from using a wacom tablet and pen. Combining the cleaned up forgrounds with the milky was was then easy, although I might take a few more tries to really clean up the files. 

‘it has been a dream photography project that has come to fruition after several years’

I hope I haven’t bored you with my ‘little’ story above, but as I said, it has been a dream photography project that has come to fruition after several years of – not exactly failure – but lack of opportunity. I’m pleased with the result, and one of the images has already been bought by the major sailing magazine of the Netherlands as its opening spread (parly financing my holiday, which is nice). I hope you like the result as well!

For those interested in some technical stuff: I used the iOptron Skytracker Pro for the sky section of these images. At ISO800 I was looking at 4 minutes exposures. In hindsight I think I could also have gotten away with 8 minute exposures, but Wadden Sea sand is not the greatest of supports for the tripod, so at that time I didn’t risk it. I was also limited in my time on the sand as the water and the moon would rise shortly after 0:30. The 12 mm Samyang was used at f/2, its brightest possible aperture, while the Laowa 9 mm was used mostly at f/2.8, and once at f/4, as the Laowa has some horrible vignetting that is most visible at f/2.8. Shutter speeds for exposures without star tracker were 15 seconds for the 12 mm and 20 seconds for the 9 mm (its wider field of view allowing for a slightly longer shutter speed).

Stargazing in Leiden

Somewhere this summer I started developing the idea for a new photography project. My attempt to photograph the Perseids in august showed that it was possible to see at least a few stars, despite the light pollution, in Leiden and surrounding area. In addition, a successful attempt to photograph the milky way in Spain also rekindled my interest in astrophotography.

I decided to see if it was possible to photograph the stars in the centre of Leiden city. The location that I wanted to try this was the Leiden Observatory, as this was a beautiful place on its own, but also closely connected to the stars. Photographing the stars at high ISO, large aperture, and long shutter speeds as I had done in Spain to capture the milky way was not possible in Leiden as it would lead to severe overexposure of the observatory in the foreground. Reducing exposure to an amount not overexposing the foreground would lead to loss of stars in the background. I decided the best course of action would be to do startrails.

The question of timing was also important. During summer and autumn sunset was late, which meant having to wait very long until it was completely dark. The warmth also kept a lot of moisture in the air, which often resulted in a haze above the city. Needing a clear and moonless night also added to the problem of finding the right night. However, last week the temperature suddenly dropped drastically and as the sun set at around 16:40 PM, the sky was sufficiently dark at around 17:30 to start with startrails.

My last startrail attempt had been quite a while ago, and I therefore assumed I would make a lot of mistakes and was glad I had a backup attempt the day after (which was also clear and cold, after that the clouds and warmth would start creeping in again). As a one-and-a-half-hour exposure would also lead to over exposure of the observatory, and in addition cause a lot of noise on the image, I decided to take individual 30s exposure images and stack these later in photoshop.

My setup was the XT-1 with Samyang 12 mm f/2.0. This meant manual focus, but as I was working on tripod and only had to focus once, this wasn’t much of a problem. After some test shots, I set the Samyang to f/4.0 and selected a shutter speed of 30 seconds at ISO 200, set the camera to continuous, attached my remote shutter release, locked the shutter button and prepared myself for 90 minutes of waiting in what turned out to be the coldest November night since 1998.

As my feet slowly froze to a state of numbness I started to see more and more stars and decided that my attempt might actually work out well. Until this moment I had seriously considered that light pollution could ruin the entire exercise. It didn’t, but I did start to notice other problems. In my enthusiasm to photograph the observatory I hadn’t really considered in what direction I was pointing the camera, I had just chosen the best viewing angle. By fortunate coincidence the direction I chose was north, meaning that Polaris was in my image and the stars started circling nicely around it. I was saved by the stars, so to speak (note to self: always check a compass). However, as the minutes on the timer creeped by, one problem that I hadn’t considered presented itself: Schiphol Amsterdam Airport is just north of Leiden, and one plane after another started to make its descent right through my frame. After the third plane went straight overhead and curved gracefully to the northwest, I decide that cursing didn’t help much, and resigned myself to an image filled with not just startrails, but planetrails..

One hour and 15 minutes later found my toes non responsive and my fingers numb, despite the double gloves, so I decided to call it a night. Getting home, my toes slowly returned to a normal temperature while I loaded all images first into Lightroom to convert the raw files and then into photoshop to create the startrail. Google was my friend and after checking up the stacking procedure I started to work. It is possible to merge all images at the same time, but this would also include all trails made by airplanes, so instead I opted for the more time and effort consuming course of merging all images manually after removing the airplanetrails. I also chose the best foreground and made sure that during merging all other foregrounds were lost. In the end I created two images: one with the planetrails, and one without. A very time consuming process, which, to be honest, I had to repeat twice to create an image I was satisfied with.

It was very satisfying to finally being able to complete an idea that had been months in the making and then to see it completed well! When I started in the cold, I had no idea if it would work, and even during the recording there is no way of knowing how the image would work out, as all you see are individual 30s exposures. The final result was worth every bit of effort and time however!

A few notes for next time: do not forget a flashlight. Do not forget to put a lens hood on the lens (I had forgotten to put it in the bag), as light may reflect on the wide angel lens, causing ugly reflections on the photo’s that are difficult to remove. Check the direction to shoot (north is of course best because you see the stars circle around Polaris). Check the lunar calendar: no moon is best. I had a lot of help planning this shoot from the TPE and PhotoPills apps. Also, wear snowboots. And take something with you to keep you occupied during the long exposure. Check the direction of the wind (windless is of course nice for reflections in water) in connection with the runway direction of Schiphol.
I did go again next day, but the image was not as nice as the first one, which is still my favorite! Less wind however, so better reflections and less airplanes. However, this time I was disturbed by a few boats full of people that sailed every 30 minutes through my frame. It can’t be 100% perfect I guess..

Yes, of course I’m going to show you the images, they’re below!

First day, planes removed
First day, planes removed

First day, with planes
First day, with planes

The second day
The second day

 

Spelen met sluitertijd: panning

Panning

Eén van de belangrijkste ingrediënten bij fotografie is de sluitertijd. Sluitertijd kun je mee spelen: een korte sluitertijd zorgt ervoor dat een bewegend voorwerp stil op de foto komt te staan, terwijl een langere sluitertijd juist voor beweging en dynamiek in de foto kan zorgen doordat je onderwerp beweegt.

Een derde manier om te spelen met sluitertijd is ‘panning’. Het principe van panning is dat je, door een niet te korte sluitertijd te kiezen en je camera mee te bewegen met een bewegend onderwerp, je dit onderwerp stil op de foto zet, terwijl de achtergrond lijkt te bewegen. Hierdoor creëer je de sfeer dat je onderwerp snel beweegt.

Gisteren besloot ik om eens te kijken wat er allemaal bij komt kijken om met deze techniek aan de slag te gaan.

Allereerst is het nodig een sluitertijd te kiezen die lang genoeg is om bij het meebewegen van je camera de achtergrond te laten bewegen, maar kort genoeg is om het hoofdonderwerp niet bewogen te laten zijn. Deze sluitertijd hangt natuurlijk ook af van je onderwerp: een auto beweegt sneller dan een fiets, en behoeft dus een andere sluitertijd.

Ik koos op goed geluk een sluitertijd van 1/30 bij het fotograferen van auto’s.

Bij een sluitertijd van 1/30 seconde komen er echter andere problemen om de hoek kijken. Gisteren was het zonnig. Zelfst met de laagste ISO waarde die mijn camera in RAW kan bereiken (ISO 200) en het kleinste diafragma van mijn 35 mm lens (f/16) was met een sluitertijd van 1/30 seconde de foto nog steeds flink overbelicht.

Gelukkig was ik hier een beetje op voorbereid en had ik een ND4 filter meegenomen, wat het licht met 2 stops vermindert. Op deze manier kon ik nét een opname maken met bovenstaande instellingen.

Na wat pogingen kon ik de onderstaande opnamen maken:

Panning
Panning: ISO 200, f/16, 1/20 sec, ND4 filter

Panning
Panning: ISO 200, f/16, 1/20 sec, ND4 filter

Voor een eerste keer best goed gelukt, vond ik, maar auto’s zijn redelijk voorspelbaar en nu is het dus zaak op zoek te gaan naar moeilijkere onderwerpen. Mijn ultieme doel is natuurlijk om dit ook te kunnen toepassen op zeilboten, maar dat zal waarschijnlijk een vastere bodem vereisen dan een op de golven dansende rubberboot!

Een aantal punten die verder nog door mijn hoofd schoten:

  • Omdat je ‘beweging’ in je foto wil creëren in de richting van je onderwerp mee, is het belangrijk beweging in alle andere richtingen te voorkomen, omdat anders je onderwerp ook bewogen is. Hoe langer de sluitertijd, hoe moeilijker dit is. Stabilisatie in je lens of camera kan hierbij een belangrijk hulpmiddel zijn. Helaas heb ik nog geen ND filter voor mijn enige lens met stabilisatie.
  • Om het principe van panning toe te passen op een zonnige dag is het noodzakelijk een ND filter te gebruiken. Helemaal omdat ik de opname nu op f/16 moest maken, op welk punt diffractie je scherpte flink verminderd. Het liefst maak je de opname denk ik rond f/8, waarvoor je dus nog een extra stop ND filter nodig hebt. Op zonnige dagen is dus een ND8 filter een must. Bij grijzer weer kan minder wellicht volstaan.
  • Wanneer je onderwerp zich in een rechte horizontale lijn beweegt zou je kunnen overwegen een statief te gebruiken om vertikale bewegingen te voorkomen.
  • Ik heb er nu voor gekozen om van te voren scherp te stellen op een punt waar de auto’s langs zouden komen, zodat ik niet last-minute de autofocus hoefde te gebruiken. Een andere methode zou bijvoorbeeld zijn om de continue focus te gebruiken.