Herein below is attached my first official short movie presented by Gorilla Film Studios (formerly MasterPlan Productions), entitled The XmasTree Hunt. It was directed, recorded, edited, mixed and graded by yours truly, Tomas Stacewicz. Music by courtesy of Kevin MacLeod (incompetech.com) used under a Creative Commons Attribution 4.0 (creativecommons.org/licenses/by/4.0/) license, and the Jingle Punks. Drone music was provided by courtesy of December Nightskies and Mathieu Lamontagne & Emmanuel Toledo used under a Attribution-Noncommercial-NoDerivatives 4.0 International (creativecommons.org/licenses/by-nc-nd/4.0/) license, and BethMcDonald used under Attribution-Noncommercial-NoDerivatives 3.0 Unported (creativecommons.org/licenses/by-nc-nd/3.0/). The film is 15:06 minutes long and was published for the first time on YouTube on 10 January 2017. The XmasTree Hunt may best be described as a narrative documentary. It was shot with the Blackmagic Pocket Cinema Camera and captured on CinemaDNG RAW, using the Olympus M.Zuiko Digital 14-42mm 1:3.5-5.6 and LOMO Meteor 5-1 lenses. It was filmed on location in Vättlefjäll and Kungsladugård, Göteborg, Sweden, between December 18 and 24, 2016. The exterior woodland scenes were filmed on 2016-12-18, and interior scenes of the decorated Christmas Tree on Christmas Eve, 2016-10-24. Copyright © 2017 Gorilla Film Studios.
Both myself and my wife, Jessica Dahlheim, served as camera operators; the entire opening titles sequence may be credited to her, as well as a close-up during the climactic tree cutting sequence, whereas the rest of the film was shot by yours truly. The majority of the shots were filmed handheld in a run-and-gun fashion, with no external stabilisation aids, the exception being the shot taken of the decorated Christmas Tree. During the shoot I took time and care to test out a native Micro Four Thirds lens, the Olympus M.Zuiko Digital 14-42mm 1:3.5-5.6, and how well the BMPCC behaves with its electronic button focus and iris controls. I tried to observe the optical quality on the Olympus lens, in particular its sharpness, internal reflexion and scattering (i.e. lens flare), as well as its general look. I also tested out the Fancier FC-270A Tripod and FC-02H Fluid Head during the end titles sequence, as well as the zoom control of the LOMO Meteor 5-1 lens. The film was entirely shot without a script but I wanted it to have a narrative based on our plan to cut a fir-tree, bringing it home and decorate it as a Christmas Tree; the place of the scene was already chosen at a farmer’s estates in a beautiful landscape on the outskirts of North-Western Gothenburg, at Vättlefjäll, where you can pick your own tree for a small fee. Everything was filmed spontaneously while following the subject, my wife Jessica, while she were looking for the optimal Christmas Tree. So it’s a documentary, but it has its narrative which vaugly follows a three acts structure. The title grew out from the narrative and was tweeked to become a pun on the most famous pirate copy of the Tilta cage for the BMPCC, the CamTree Hunt.
General camera settings: ISO 800, Shutter Angle 180º and 24 fps. Exterior shot camera settings (with Olympus M.Zuiko Digital 14-42mm 1:3.5-5.6): White Balance 7000K, F=14, and f/3.5 to 5.6. Interior camera settings (with LOMO Meteor 5-1 17-69mm f/1.9): White Balance 3200K, F=17 to 69, and f/1.9. General audio settings: All sound were recorded on the inboard stereo microphone of the Blackmagic Pocket Cinema Camera, with microphone and channel (line) levels set to 100% (generally untouched in post, with some exceptions where sound levels either had to be lowered or raised). The RAW footage and sound was edited, mixed, graded and rendered on Blackmagic DaVinci Resolve 12.5.4. Colour correction was made manually using the Blackmagic Cinema Camera Film to Rec. 709 v2 LUT somewhat as a reference, raising Saturation to 100 and keeping Hue at level 50, adjusting the final grading with the aid of color wheels while checking the scopes to create a balanced RGB profile in the highlights. Extensive image stabilisation of the footage was performed on the majority of shots, adjusted additionally with a 2% maximum Zooming. The only footage not needing any image stabilisation, besides the shot taken on the Fancier tripod, was actually those taken by Jessica (she has proven herself to have a steadier hand than yours truly; she is henceforth to be known as the living steadicam). The final result was rendered using the DNxHR HQ (8-bit 4:2:2) video compression codec and two cannel PCM audio codec wrapped in QuickTime and Ultra HD 4K 3840 × 2160 resolution. Exported thus, it was finally uploaded to YouTube, enhancing the image with a Saturation of +1 with no additional image manipulation to retain the flavour of the original footage.
The Blackmagic Pocket Cinema Camera: I have really come to love this camera and its small form factor. I consciously wanted this shoot to be as much stripped down as possible, with only myself and the original camera body attached to a small MFT lens, with no additional accessories. I knew that the place that we were going to included other people, strangers, besides the owner (the Santa as seen in the film), and I didn’t want us to look like a film camera crew but rather to pass innocently as hobby photographers immortalizing the moment with our seemingly modest “DSLR”. True guerilla style. But I also knew that the actual tree hunt itself would entail long walks in deep forests and cruising through thick foilage; I had to keep the form factor as small and minimal as possible. During the entire shoot, I held the camera in its original camera body hand grip, without a strap; holding it firmly with my right hand it never slipped from my grip. And not was my hand nor my arm strained or fatigued at any given moment of the shoot. I have to conculde that the ergonomics of the camera is one of its great benefits; it is really comfortable to hold in this way during extensive and prolonged shootings while being on the run.
Unfortunately, the camera body being so light and combined with a no-weight type of plastic lens, such as the Olympus M.Zuiko Digital 14-42mm 1:3.5-5.6, there is nothing there in the physics that helps stabilising the image. And, unfortunately, I must confess that micro-shakes seems to be my Achilles‘ heel. My hands shake and I cannot help it; it’s a condition I have which has become worse as I have grown older. I barely notice it in everyday life’s situations, as they are unnoticeable in normal movement, but holding such a light camera and trying to fix it, it really shows. I must conclude that the Krasnogorsk-3 pistol grip and telescopic folding stock shoulder brace actually does help, a lot. I experienced lots more micro-shaky footage now compared to the Test Movie I. An image stabilised lens would really come handy for me when doing run-and-gun shooting; luckily enough, I manages to erase almost all of the micro-shakes from the footage in post (see below) as I otherwise managed to aquire quite stable footage when it comes to greater shakes, and movement of the camera, such as panning and tilting. The rolling shutter effect that I might have experienced during the opening titles sequence, with the passing by trees, was very marginal, neglectible; perhaps a trained eye may spot it.
I used up both of my SDXC flash memory cards and all three of my Li-Ion EN-EL20 batteries; as the shoot was unplanned and wholly improvised, I had to keep the camera ready to catch the fleeting moments when they arrived. Now and then, I managed to shut down the camera to spare the batteries, which saved me some battery life, but most of the time I had to have the camera powered on to catch interesting events. Even though I had it shut down, I’m still amazed how quickly the Blackmagic Pocket Cinema Camera turns on and boots up, to be able to take a shot, when pressing the power button, which this shoot proved to me on more than one occasion. However, sometimes moments arrive on only a few seconds notice, and I managed to miss several opportunites, which taught me to keep the camera running most of the time. Still, I neither ran out of memory or power during the entire days worth of shooting. So having two SDXC cards and two spare batteries is a absolute minimum. Filling both flash memory cards gave me ample of footage to cut the movie to its final form. The parts that I cut out were either repetitive, uninteresting or to shaky to be able to stablise in post. The only drawback from having to change batteries is that it required me to place the camera in my lap to do it, and in the process I managed to smear the front element of the lens, so that the entire image became blurred; I had to clean it in the field using my scarf (which is not particularly ideal thing to do to a lens).
Regarding the image quality, I again must conclude (watching the footage in RAW on the full frame DaVinci Resolve panel) that it is simply marvelously organic and very filmic, pleasently soft and detailed at the same time. I saw very little of the vertical fixed pattern noise that I stumbled upon in my first test movie Lux ex Tenebris, only marginally so and nothing that turns the audience’s attention away from the essentials in the frame; as much of the forest was quite darkened, there was quite lot of randomised noise in the background, but it was beautiful to watch and reminded me much of celluloid film grain. The greens, yellows and browns really stands out, creating a emerald forest kind of look. Also the reds pop up in the beginning and closing, on the suit of the dressed up Santa Clause. Allt this is due to my colour grading where i pushed the reds and pulled down the blues, with the greens as a balanced mean between them. Still, skin tones look natural. I cannot help falling in love with these images, especially in the dark mysterious forest during the middle and latter part of the film. However, it seems that I have found my first examples or infrared (IR) contamination, as the fatigues of the subject (which are dark grey or light black) had a slighly brownish-reddish hue, visible as noise grain (mostly seen in the original RAW footage). It seems I have to invest in a IR filter, but I kinda liked the look; it gave the shadows a warm and filmic feel.
I tested the auto-focus function extensively during the entire shoot. One press on the focus button at the back of the camera body activates a autofocus square which focuses objects within its frame. After the focus has been set within the focus frame it freezes and doesn’t change regardless of camera movements, which is a great feature; you don’t have to worry about a running and wild autofocus, as on your typical smartphone. The square is fixed and thus only engages objects in the centre of the viewfinder, contrary to your smartphone where the square (or circle) travels all over the screen; another big plus. I tested the autofocus on the evening at home before the actual shooting day, in quite low light conditions. I concluded that the autofocus works well enough and quite fast in most cases if the lighting is sufficient (very deep shadows may pose a problem). Unfortunately, I forgot to test the auto iris function, to check if its reflected light meter works properly as compared against my analog Sekonic L-398 light meter. However, I did use the manual iris control using the left and right arrows at the back of the camera, where the left arrow steps the iris down and right steps it up in the following f/ increments: 3.5, 3.7, 4.0, 4.5, 4.8, 5.2, 5.6, 6.2, 6.7, 7.3, 8.0, 8.7, 9.5, 10, 11, 12, 14, 15, 16, 17, 19, 21, and 22. I would say that this is good enough.
When it comes to low light capacity, I again must conclude that the reported bad performance of the Blackmagic Pocket Cinema Camera in this regard is highly overexaggerated; the BMPCC is highly underrated as a low light camera. A major part of the film unfolds in a dark forest covered with trees, under a heavy and thick overcast. The mysterious gloom is pleasently balanced. In the end titles sequence, I only used the natural light coming from the LED Christmas Tree lights, nothing more. The shot was taken just after midnight, on Christmas Eve. O.k. the image is dark, but the object (the Christmas Tree) it is still visible enough to enjoy the full glory of the decorations. Watching the RAW image, there is some visible grain, but mostly in the red background (which looks quite digital and ugly), otherwise the image is quite clean. However, out of sheer ignorance, I managed to disregard the rule of using a 172.8° shutter angle while shooting in natural light coming from a 220V power mains with 50Hz frequency; I learned the hard way that you cannot use the standard 180° setting of the shutter angle, which is adapted for countries with 60Hz mains power supplies. I saw the effect already when reviewing the take on camera, on the LCD screen, but thought it was some kind of interferance of the screen itself. When I imported the footage to DaVinci Resolve, I could see that the strobing was actually captured on the CinemaDNG files. See the full story of the stuttering strobe effect here which was generated from my mistake; however, I decided to keep the footage and edit out the most severe strobing.
The Olympus M.Zuiko Digital 14-42mm 1:3.5-5.6: I do not yet possess a keen eye to spot defects and distortions in a lens, or even to judge the ability to create sharpness to an image. What I can provide here in this review is the general feeling of the look of the Olympus lens. What I can say is that it won’t find itself being in my possession for the foreseeable future. I borrowed it for one day from my sister in law, who owns the Olympus PEN E-PL1 and the lens was part of a kit that came with that camera. As the E-PL1 and the Backmagic Pocket Cinema Camera share a native Micro Four Third lens mount, I decided to borrow its lens and try out the electronic controls of the BMPCC. I borrowed the Olympus camera as well and took the opportunity to compare the MFT mount of respective camera, and to my suprise the mount of the BMPCC has a better (as in tighter) fit for the MFT lens; the mount of the Olympus wiggles a lot more which makes the BMPCC into a quite decent MFT mount. The M.Zuiko Digital 14-42mm was announced in 2010 (together with the attached camera) and is a standard zoom lens. The actual lens has a telescopic feature and is folded into a compact form when not in use. When shooting, the lens has to be folded out, which is done by turning the zoom ring until it snaps out, making it 43.5 mm (1.17 in) long. Folding it back into its compact size, you have to press a button on the side of the lens to disengage a locking mechanism.
The lens lacks image stabilisation and is not macro capable. But it has a quite short focus distance of 0.25 m (9.84 in). It has a maximum aperture of f/3.5 with the focal lenght set to 14mm, which changes to a maximum aperture of f/5.6 when zoomed in to 42mm, with a 29-75° angle of view. The full frame focal length equivalent of the lens is 28-84mm and I shot the entire film with the lens set on 14mm (which for a 4/3 sensor is the equivalent of 28mm and for the BMPCC is equivalent to 40mm, which makes it a normal lens). The lens is constructed with 9 elements in 8 groups and has a maximum magnification of 0.24. Folded in it makes it almost into a pancake lens, with a diameter of 62 mm (2.44 in). It is mainly made of plastic (probably only with the exception of the MFT mount which is made of metal) and only weights 133 g (4.7 oz). Its filter diameter is Ø40.5 mm, but there was no filters in the kit. Luckily enough, the sky was totally covered with thick clouds and the atmosphere was saturated with mist, so there was no need for any ND-filters and such. (Note: There is absolutely no overexposure in any of the footage, or clipped highlights in the sky; what we see in the film is the light grey of the heavy overcast and thick mist.)
I guess the image is quite o.k. quality wise. As the sunlight was blocked out from a heavy overcast and mist, there weren’t that much light to work with optimally. So it’s a bit difficult to judge the overall look of the lens. Although, I did recognise the previous image that I captured from Lux ex Tenebris, my first test movie, which means that the sCMOS sensor of the Blackmagic Pocket Cinema Camera has its distinct look that comes through regardless of lens attached. However, I did notice that the Olympus M.Zuiko Digital 14-42mm lacks its own character, as compared to my Russian LOMO Meteor 5-1 lens which in my opinion enhances the BMPCC look even further. I couldn’t notice any visible distortion of the image, although some shots contain a strange light phenomenon which resembles a flare somewhat but much more indistinct and ugly (look at 8:20 and 11:25). I couldn’t do a forced lensflare test of the Olympus lens as the sky had a thick overcast so there was never a direct sun visible. But I managed to capture the headlights of cars during the shoot and never saw any flares, so I guess the coating works well on the lens. As I was using the camera setup strictly handheld, I never used the zoom range but retained the focal lenght at the widest setting of 14mm. This made my field of view more or less equivalent of what I saw with my eyes. The reason for using the widest focal setting was to prevent judder and micro-shakes from the lack of external stabilisation (although that did no good).
I also settled on finding a optimal hyperfocal distance so that I wouldn’t have to think about changing the focus during the shoot, and use maximum depth of field. I exclusively used autofocus to set the hyperfocal distance by focusing on a distant object. It worked quite allright in the begining, but after changing my first battery I had to set it again and failed, which ruined quite many minutes of otherwise good footage. I eventually managed to set the hyperfocal distance by pressing the autofocus button, and suddenly the focus peaking glowed across the depth of field; I know that I finally had nailed it. The conclusion is that it can be ankward and quite risky to set the hyperfocal distance using the autofocus function of the Blackmagic Pocket Cinema Camera and Micro Four Third lens combo. On the footage the sharpness is generally allright, but a bit soft. However, I have no problem with softness of the image as that enhances the filmic look and looks alot like the sharpness I experience from the Meteor 5-1. In some scenes, in which the focus distance was short (especially in the scene with the sawing of the tree), I lost focus; which means that I trusted the hyperfocal distance setting a little bit to much. (I vaugly recall that I did another autofocus setting towards the end of the day, which might have been somewhat in error, another example of why one should set the hyperfocal distance manually.) But i don’t like the manual focus ring of the camera, which works fly-by-wire (of focus-by wire), i.e. there is no physical stop to the ring which makes it ankward to orient the focus (and I guess doesn’t work well with a follow focus arrangement).
The LOMO Meteor 5-1 17-69mm f/1.9: I used the Meteor 5-1 exclusively during the end titles sequence when taking a closer look on the decorated Christmas Tree that was cut down at the end of the film. First of all, I no longer had access to the Olympus M.Zuiko Digital 14-42mm 1:3.5-5.6 as it was only a loan. Secondly, I wouldn’t have used the M.Zuiko Digital anyway as it is not a particularly fast lens. I thought that the stated f/1.9 of the Meteor 5-1 would do fine to capture the tree and the natural light coming from the LED lights. And I was correct. In my opinion, the lens works well together with the Blackmagic Pocket Cinema Camera sCMOS Super 16 sensor. I didn’t have to stop the exposure up in post. But I knew that already. Having the focal lenght set on 17mm, I did notice quite severe barrel distortion, however. What I actually tested on this particular scene was the zoom control of the Meteor 5-1. I attached the rod to the zoom lever on the side of the barrel. And I did some fast zooms, and moderately fast zooms. And as you can see on the film, it worked quite well. I suspected that the rod and lever arrangement would provide for really fast zoom effets, and my suspicions were confirmed. I actually cut out the most experimental part of that shoot, which I have attached in a separate film below – the deleted scenes or extended version of the Christmas Tree. Although not being particularly well damped, with the extension rod mounted to the zoom lever, it is still quite easy to pull and draw zoom, in particular when they are quite fast. Very slow zoomings could pose problems, or at least require lots of training and experience.
Fancier FC-270A Tripod and FC-02H Fluid Head: I did a quite thorough testing of the tripod and fluid head during the end titles sequence featuring the decorated Christmas Tree. My experiences are largly positive, although there are were some problems encountered. I did various pans and tilts, variating the speed. Some pans and tilts were very fast and quite experiemental; I did a series of combinations with moving the camera with the fluid head and simultaneously doing all kinds of zoomings with the lever and rod arrangement on the Meteor 5-1. I encountered no problems whatsoever with the panning and tilting; the camera was steady as a rock, at least during the shoot. Watching the footage in post I can notice some irregular and rough movement, but most likely that has rather to do with my inexperience more than the actual fluid head. The counterweight worked well, and I used it when tilting down from the christmas tree spire, which created a calm and smooth downwards tilt. Start and stop was smooth and without any jumps. Where I did encounter micro shake problems was when doing focus pulling while tracking the different parts of the tree in zoomed in close-ups. I had to make several retakes to get the shots; unfortunately, in the majority of re-takes there was still visible judder. It is also seen in some ot the footage that didn’t found itself on the floor. Thus, I have to conclude that the head isn’t that stable, even when the tilt and pan knobs are screwed tight. In practical terms, this means that a follow focus rig is essential for getting clean shots with panning and tilting, following subjects and pulling focus simultaneously. Below I present the extended or uncut version of the end sequence with the decorated Christmas Tree, with the most severe shaky and juddery footage cut out; you will see that I still had a lot of material left that was clean and smooth.
Workstation and Blackmagic DaVinci Resolve 12.5.4: I more or less used the same editing techiques and grading as in the last movie, although I did push the colour palette towards a warmer tone (see BMPCC above). I didn’t do any further indepth study of the Reference Manual but tested out some new things nevertheless. In the Edit page I learned the basics of placing clips as superimpositions over other clips on the timeline. I needed to do this to create certain effects during the end credits, such as overlaying the Blackmagic Design, CinemaDNG and Gorilla Film Studios logos. In doing this I explored the composit function which is accessible when highlighting any clip. In superimposing the Blackmagic Design logo I found the Pin Light composit mode to work best, as the background of the logo was grey, effectively erasing it and retaining the actual logo in full colour. In superimposing the Cinema DNG symbol i didn’t have to use any composit as its background already was transparent, but the Gorilla Film Studios logo was problematic. First I tried to use the Divide composit mode as that flipped the colour to its complementary while erasing the white background entirely; i greated a clean cyan version of the Che Gorilla drawing which would render it red, or at least that was my original idea. Unfortunately, it didn’t do well with the reds, creating strange fields of darker red at the edges; when playing it on the timeline it looked good but after the rendition the exported QuickTime DNxHR file degraded it. I had to remove my first upload of the film after one day, making another version using the original red head with the Color composit mode. Unfortunately, the white background wasn’t masked entirely and created a darker and silvery field. On both the BMD and Gorilla Film Studios logo, I softened the edges which created a subtle halo effect – good enough!
On the Edit page I also noticed that I could individually adjust the volume setting of each clip by highligting its audio track on the timeline; a very good feature! Going over to the Color page I deepened my understanding of the image stabilization feature using the Tracking Window. I found a working formula of the setting of Strong and Smooth to effectively eradicade micro-shakes to the camera: Always begin with setting Strong to 90 and Smooth to 10. When the resulting image gives a exaggerated “rubbery” or “wavy” characteristic, lower Strong to 80 or 70 until the image stabilises and looks more natural, but don’t change Smooth, with a few exceptions. However, this doesn’t work if the micro-shakes are to strong or if the movements of the image otherwise are to heavy. If so, I started off with a Strong and Smooth of 10 each, and raising them with increments of 10, sometimes (and oftentimes) mostly Strong and sometimes mostly Smooth. At last I have found a pattern and if you look at the images, there are very few which exhibit micro-shakes; and you should be aware that the greater majority had very visible micro-shaking in them, although I didn’t manage to stabilise all to my satisfaction. So I must conclude that DaVinci Resolve handles stabilization quite well, after all. Unfortunately, DaVinci Resolve 12.5.4 chrashed on me three times when doing stabilisations, two of them on one and the same day. Perhaps 12.5.4 is a bit more buggy as compared to 12.5.3 (the version that I used previously with stabilisation), either that or I managed to create a larger workload in comparison; it seems that DaVinci Resolve 12.5.4 in particular is more sensitive to working with tracking and power windows.
Lastly, finding myself on the Deliver page, I used the same formula as the one that I eventually developed to render files that are optimized for best image quality on YouTube and Vimeo, namely uploading my BMPCC 1080p footage in 3840 × 2160 Ultra HD (4K), using QuickTime as a wrapper and DNxHR HQ as a codec. When doing that render in Resolve 12.5.3, although it took all day to do the upload, I succeded using Firefox (my preferred browser up until now). But after upgrading to 12.5.4 YouTube crashed on me after less than 30 minutes on each try, regardless of the size of the file. It only accepted 1080p HD. I have made several attempts for the past weeks to no avail. (I experienced it the first time when uploading my Strobing Test footage on YouTube.) My brilliant son came up with the idea to switch to Google Chrome, thinking that as YouTube nowadays is an extension of Google it could perhaps be more compatible. I tried it out starting on January 9, 2017, which marked the end of two weeks of daily postproduction work. And it didn’t crash, taking approximately 17 hours to complete. So the conclusion is that, regardless if Resolve 12.5.4 has some issues with uploading Ultra HD 4K to YouTube or not, or if something has happened on the receiving end (i.e. YouTube) during the past month, it is obvious that Google Chrome is a more stable browser compared to Firefox when you want to upload Ultra HD to YouTube. I believe that Chrome also has a faster upload ratio (taking 17 hours to upload 15 minutes of footage) compared to my previous (successful) upload of 17 minutes footage that took more than 30 hours with Firefox. So it seems that Google Chrome is faster as well. Don’t know how Explorer works out but I generally dislike it as a universal browser. So from now on, I will switch to Google Chrome as the main browser of my workstation.
In conclusion, I am quite satisfied overall with the end result. I had a vauge vision which found itself onto the screen. It’s not a perfect piece of film, far from it. But considering the footage that I grabbed into the CinemaDNG files, I’m happy with what I eventually made out of it. The more I delve into the filmmaking process, the more I realise that postproduction is really a creative process, perhaps the most creative part of filmmaking, in wich the film actually develops and grows. I’m particularily satisfied, and quite proud, of my editing. I have yet to learn the art of editing from books, but I do feel that I have a kind of natural sense for it, a brewing instrinct of how to combine different scenes and various footage into a coherent and fluid whole, although I still have much to learn. During the process of editing The XmasTree Hunt, which took me one week to complete, (with the exception of some conscious jump cuts towards the end) I did my very best to create a seamless transition between takes, trying to mask the cuts with movement of camera, with the position and movement of legs, with the movement and position of body in the composited frame, etc. And I believe I succeded in most cases. Where I couldn’t find a seamless flow between two clips, I used a cross dissolve. I happen to like cross dissolves, however I try to use them sparingly and only when I must; I prefer to make clean cuts if I can succed in making them transit well enough and unnoticed. I’m also quite happy with several of the takes and their composition, although there are some obivious mistakes as well (such as the subject dissapearing out of frame and reentering); but these shots were unplanned and just happened, grabbed with my gut instinct in most cases. Some of them I really nailed, both wide shots and medium shots. Music wise, I’m also quite satisfied, blending traditional Christmas music of a melancholic touch with various drone pieces, although some of the latter may perhaps feel somewhat intrusive. So I’m quite fond of this little film, as it marks a obvious progression from my former piece Lux ex Tenebris. I hope you enjoyed it as well!