Panorama photography with the DJI Mavic Air 2

Since I carry a DJI Inspire 2 Quadcopter in a large wheeled hardshell suitcase on most of my photographic trips, my standard luggage has grown to a weight of more than 65 kilos. Although the image quality of the APS-C sized sensor of the DJI Zenmuse X7 is in a class of its own, for the reasons of space and weight I'm always interested in the smaller options available on the market. In terms of packing size and weight, the first edition of the DJI Mavic Air was nothing but sensational. For my trip to Greenland in September 2020 the copter nerds from Globe Flight have given me the successor, called Mavic Air 2, for a thorough test.

View fullsize

The DJI Mavic Air 2, ready for takeoff (Photo: DJI)

First of all, with the tiny sensors of the small drones you of course have to be prepared for higher noise levels, but at least the Mavic Air 2 has a slightly larger sensor than the previous model (1/2" instead of 1/2.3") - and you don't have to take every picture with the intention of a large format fine art print. On the other hand, one thing in which the first Mavic Air already excelled, was automatic panorama shooting, which I think is an excellent way to get the most out of a small 12 megapixel sensor. This applies to both classic panoramic images (which simply show a larger section of the landscape than the angle of view of the lens allows) and digital image presentation forms such as interactive 360° spherical panoramas.
 By the way, Mavic Air 2's lens focal length is 24mm (in relation on 35mm "Full Frame") - in my opinion a very good all-round focal length for a drone. It is perfectly suited for creating panoramic shots.

There are four different types of panoramas that can be shot with the Mavic Air: Horizontal, vertical, 180° and the 360° sphere. You first determine the initial angle of view and then start the panorama shooting. The drone does the rest, and it is extremely quick and easy - hardly a minute passes from the pictures taken to the montage of the panoramic image as a JPEG file in the drone. The results of in-camera stitching are not bad, but not outstandingly good either. There occasionally are stitching errors - especially if you have moving elements in the overlapping field of view between the shots.

However, if you're willing to go few steps manually, you can achieve much better results. The real highlight of the drone's automatic panorama function is that it can save the RAW files (DNG) of the individual images parallel to the montaged panorama (must be activated first).

This results in the following workflow for me (here the example of a 360° panorama):

1. Recording of the panorama using the automatic function of the copter

2. Development of RAW files in Adobe Camera Raw (or Lightroom) with the focus on optimal technical reproduction of the tonal values (shadows, highlights)

3. Assembly of the panorama from the developed TIFF files in PTGui (using masks where necessary)

4. Editing of the mounted picture in Adobe Photoshop with regard to the aesthetics of the tonal values and colours, insertion of a "cap" (the copter can't photograph steeply upwards, so this gap in the picture can be closed in this step)

5. Creation of the interactive web version of the panorama using a suitable tool/viewer, currently my favourite tool is marzipano.

Another big advantage of the self-assembled images is that they use the maximum number of available pixels, thus offering a visibly higher resolution. A corresponding 360° panorama from PTGui is 18,486 x 9,243 pixels (W x H), whereas a file mounted in the drone has only 8,192 x 4096 pixels. In addition, DJI applies an unsolicited HDR function to the images during assembly to compensate for the contrasts. You my like this look - but it is definitely not my cup of tea.

Here are two examples of 360° spheres I photographed with the Mavic Air 2. I think the technical quality is impressive. With the Inspire 2, of course, I achieve even more detail resolution as well as more dynamic range in the shadows and highlights. Sensor size and resolution are irreplaceable. But compared to the size and weight of an Inspire 2, the performance of the Mavic Air 2 is simply amazing.

The completely mounted and edited image file of the spherical panorama "Glacier Milk Lake" - please click on the image to get to the panorama viewer …

The completely mounted and edited image file of the spherical panorama "Paradise Lake" - please click on the image to get to the panorama viewer …

And while I'm about to praise this little drone - here are my 4 personal highlights from its use in Greenland:

• Up to 35 minutes flight time per battery feels somehow endless.

• The camera delivers excellent material for her size, up to 4K at 60 FPS.

• The Copter flies absolute precisely and seems very responsive and controllable.

• The whole system works absolutely smooth and trouble-free, there was no signal interference or other unexpected behaviour.

Traces of climate change (1)

Time and again I am asked whether I already have noticed anything of this much discussed climate change during my work in the Arctic. Sometimes I hear the desire for appeasement, reassurance - "no, don't worry, you can't see a thing of it yet" - but unfortunately I can't help you out with that. In fact, I see significant changes from year to year, although these can of course be caused by short-term fluctuations. Unfortunately, the tendency over the last few years goes very clearly in one direction: mass (height) losses of glaciers, retreat of glacier fronts wherever you look. But much more frightening than the current changes are the traces of what has been going on for decades.

On this picture taken in summer 2018 more than 500km north of the arctic circle, we see a glacier that extends from a height of approximately 1500 meters down to the sea. However, I should say "had extended", because there is practically nothing left of the once large glacier. The non-eroded, semi-circularly pushed up lateral moraines look as if they were left by the ice in a great hurry. Of course this did not happen overnight, but also not in the usual glacial periods.
From exactly this area I was told that a few decades ago, such glaciers were still used in winter from the frozen sea for dog sledding across the mountain ranges into the neighbouring fjord. But without ice this does not work any more.
This is how strikingly such a completely disappeared glacier first caught my eye in 2018 - but since I know what to look for, I have found a lot of it and many more drastic signs of the Arctic in a state of flux.

It is quite obvious: we have been in the middle of the climate change for a long time. And our remaining time is running out. Every ton of carbon dioxide that we blow into the atmosphere today will remain there for a very long time. There is only one really viable way to escape the catastrophe - to reduce our carbon dioxide emissions sustainably and massively.
We (the industrial nations) have been living on credit for a long time, which we have granted ourselves arbitrarily at the expense of future generations. And we have concluded a Paris Climate Agreement, the honourable goals of which cannot even come close to being achieved with the "climate package" recently adopted in Germany. Not even theoretically and with a huge portion of super-optimism. Just the good old "keep it up", adorned with a little greenery.
This is what I call malignant acting - in view of what is at stake for us all! And I think to myself every day: How can someone call himself "conservative" or even "Christian", who is so indifferent to our planet? After all, it is still the only one we have.

4 x Light

Once and again people ask me the question "what is the most important thing in landscape photography?". Of course, this question cannot be answered unequivocally due to the many ingredients of a good picture - but there is no doubt that light is one of the essentials of many good pictures. Light widely determines our technical approach and puts it's basic mood over the scenes. It defines intensity, models the subjects and is often the main motif by itself.

I took this collection of four pictures of the same iceberg around the end of May 2013 on four different occasions over several days (altogether there were even 5 variants, but 4 are better arrangeable in one picture). 
It was only at home that I discovered, that the pictures showed almost exactly the same section of an iceberg in different lighting situations - so diverse was the effect of the position of the sun and the weather as well as the rotation and position of the slowly drifting iceberg relative to the sun.

On the upper left you can see afternoon backlighting with amber-coloured ice fragments on the water surface, on the upper right there is the typical mercury light of the Arctic when the sky is overcast, on the lower left the night sun hides behind a cloud bank and finally on the lower right there is the unbelievable light of the midnight sun shining below a dense cloud cover, which makes the ice shine like gold when the sky is dark.

It is, by the way, not the first time this has happened to me - from such series I have concluded that my inner compass for image aesthetics seems to provide reproducible results in a certain way, since in different lighting situations I unconsciously chose the same characteristic image framing several times. On the one hand, this sounds good to me - as I can assume that my compositions are not purely random products. On the other hand, it makes me ask myself the question whether I did not reproduce myself by selecting a section, just like a predictable photo robot. However, I have to postpone the clarification of this question, as I haven't given it enough thought yet.

After all, I am satisfied with the result of the comparable image frames inasmuch as that they look somewhat aesthetic in each of the light situations and on the other hand clearly show the different effects of the image component "light".