As the majority of the footage I need is dependent on weather, I was able to shoot tonight with both the 600D and the X20. After the previous unsuccessful shoot with the Fuji X20, I wanted to see if this was down to the particular conditions, or whether the lack of manual controls (including lack of manual focus) means that the clips are soft.
After reviewing the footage from today, I found that the Canon 600D was far more successful than the Fuji X20. Only the final close-up footage was successful to use in the final video, meaning that this camera is in a further niche. Although, as I had two tripods available, this meant that I could simultaneously film on both cameras, meaning that I was able to obtained the aforementioned successful footage from the Canon 600D.
As the Canon 600D is college property, in order to use it I need to book it out over several days, leaving a limited time in which I have to film. Because of the low-light capabilities of the camera, I decided to focus today on shooting the sunset and night-time footage. As each sunset is different, I was keen to capture as much footage as possible so it can be shortened down into a time lapse clip. After the sun had set, I continued filming, focusing on the moon which will be a reoccurring element within the chorus, with the intention to film the moon in various stages (at least 4) to portray the transition of time. Similarly to the sunset footage, these natural shots are entirely dependent on weather, and luckily tonight was a clear night, with a clear view of a waxing crescent moon.
As the moon is a distant object, the focal reach of the 18-55mm kit lens caused the moon to appear quite small in the frame. Because of this, I switched lenses to my Tamron 70-300mm lens, which will zoom in far enough to define the moon as well as fill the frame. The exposure of the camera was also very important, as it is a bright object, requiring a lower exposure in order to maintain this definition. This also demonstrated the need for planning; as there was a possibility that I would be shooting footage of the moon, I needed to have taken a zoom lens with me to achieve this.
During the practical pre-production shoots, I found that having only one tripod meant that I could only shoot one piece of footage on each camera at one time. As I found that the Fuji X20 really retained true colour in my test footage, I decided that it would be effective to have it filming the sunset at the same time as the Canon, resulting in two differing yet similar pieces of footage which I can review later. In order to achieve this, I booked out a tripod from College, resulting in two tripods, hence allowing two simultaneous clips to be captured. However, because of the Fuji X20's automatic autofocus, the bright elements in the night sky confused the camera, causing it to constantly shift focus, meaning that all 5 clips that were shot with this camera were unsuccessful. I will bear this in mind in future shoots, using the X20 only during the daytime.
Wednesday's green screen shoot, resulted in a 1:24:57 long clip which needed to be edited down into clips which I can use. The way in which the green studio was set up meant that it would have been an ineffective use of time if I created a new sequence for each take, meaning that the first review of footage would be fairly surface level, to contain where the aspects of performance were.
I created the table below which isolates each piece of footage, which I then determined to be usable or unusable, with a comment as to this decision.
After reviewing the footage, I trimmed the original clip for each of the times noted in the table above. This resulted in 25 clips which I can then further review and compare.
Further Review:
After condensing the original video into the 25 clips above, I re-watched these clips critically, isolating only the best areas of each video. This resulted in a further refinement which I can transfer directly into Premier Pro. As I have been very critical with this particular review, I considered my intentions outlined in the storyboard when selecting these edits, but expect to return to the 25 above to complete the video and ensure variety.
These notes were composed in the iOS app Paper by FiftyThree, which enabled me to keep each of these timestamps in a digital journal, remaining far more organised.
In order to remain organised for the session in the TV Studio, I have written myself using the multi-platform software Evernote, which I can access at any time through my phone or a computer. I hope to get as much usable footage from this session as possible for the final video, so that another session may not be needed, giving me more time to edit and refine the piece.
After the first double exposure test, I researched into online tutorials which could refine the double exposure effect. Once I had experimented myself into refining the double exposure technique (evidenced in the second double exposure test) I found this double exposure After Effects tutorial below on Youtube.
Described as 'medium difficulty' this tutorial used techniques which I have used in the past, as well as some which I haven't yet come across. Free footage was also available with this tutorial, and I decided to use the tree, dust, and lens flare footage to see if I could get a final result similar to that in the video. I used my own green screen footage of James. Documented below is the progression of my own work following the tutorial, compiled in the online software Canva.
Whilst the end result perhaps does not have as much definition on James's facial features, there are definitely certain aspects to this video which I will continue to refine:
• Using 'Add' to blend the background into the double exposure, creating a more harmonious composition.
• The use of a pan in the background footage, and then duplicate this in a new layer and fragment it, adding depth to the overlaid footage.
• Creating an edge blur by manipulating the blur and shrink/grow transformations within the screen matte to further blend the double exposure with the background.
Below is the results from an After Effects I attended. As I intend to use After Effects heavily, these sessions will further my knowledge of this software and aid the editing process.
This session covered animation and key frames, using shape and text layers to create various moving parts in a title sequence. Using the transform tool I manipulated: opacity, rotation, scale, anchor point, position, skew, and blur to create varying effects. In order to create smoother transitions between these effects I used Easy Ease on some of the key frames, which removes any jitter in the movement between each key frame.
Anchor Point & Rotation: The yellow star has been manipulated using both the anchor point and rotation transformation tools, with the anchor point the location which the star rotates around. By selecting the start and ending point with two key frames, this caused the star to rotate once around the anchor point.
Scale: Using 3 key frames for the large circle, I used the scale tool to grow and shrink the shape. Setting the shape from 0% (invisible), grow to 140% (larger than normal size), then down to 100%, causing the shape to grow to larger than normal and shrink back to normal size.
Transparency: On the smaller circle, I used two key frames at the beginning and the end of the clip to make the shape appear from nothing. In order to achieve this, the value of the first key frame was 0%, and the second key frame's value was set at 100%.
Position: Used on the "Title Text" text layer, here the position transformation moved the text layer from out of the frame to the centre of the frame, using two key frames and motion blur to smooth this transition. This transition can be created through placing the object in the start position before moving it to it's final position, and the software blocks in this movement automatically.
One thing I found with the previous multiple exposure was the final result appeared quite flat and stencil-like, with just the sheer outline masking the time lapse clip. In this test I wanted to experiment with creating depth within the original green screen footage.
Despite being only 4 seconds long, this clip took around 3 hours to edit, which I believe will decrease as I become more confident using After Effects. Because of this, I am keen to refine the technique and various effects which create the final result to hopefully lessen the time spent creating each double exposure in the final video.
In order to emphasise the impression of 3D on this 2D footage, I created a vignette which will draw the focus to the centre of he frame, where the action is happening. Also, the vignette will lesson the appearance of the double exposure 'floating' in the sequence.
In the previous experiment, Displacement Maps were used to alter the time lapse footage in an attempt to add depth and interest to the background footage. Here I used a Displacement Map on the green screen footage layer, so the face will displace the background footage, and add definition to James's features. Whilst this does create depth, this looks almost Solarised, which results in a loss of realism in this sequence. The Displacement Map was effective in adding depth to this piece, I will be experimenting with adjusting shadows and highlights to create a more natural aesthetic.
After my previous unsuccessful attempts of getting a time lapse of the night sky, I decided to try to capture star trails instead. The image above is composed of 17 images each with 30 second exposures with an ISO of 600. A time consuming but rewarding process, this star trail captures just a part of the sky, but I am pleased with the effect I achieved, especially for my first attempt. As I do not own a remote trigger I had to continually wait and press the shutter, which was laborious. The images were taken using a Fuji x20 and have had no colour correction.
In order to combine all of the photos I had taken, I uploaded them into Photoshop, and changed each Layer Blending Mode to 'Lighten' which allowed all of the images to be visible in the end result.
On Saturday, upon seeing the storm approaching, I captured these time lapses of the clouds rolling in, demonstrating the growing intensity of the storm. After taking shots on both the iPhone 6 and the Fuji x20, I reviewed each of these clips and was pleasantly surprised by the quality of the Fuji x20. The first three shots of the video were filmed with the Fuji x20, and the final shot with the iPhone 6. Due to the fading light, noise is especially present in the final shot taken with the iPhone, with the higher ISO creating this in an attempt to capture as much light as possible. As I could manually control the ISO on the Fuji x20, the quality of these shots in low light are a vast improvement on the iPhone, making them suitable for night time shooting.
Setting up my tripod for some sunset time lapse footage I had planned, I noticed some dark stormy clouds in the distance. These clouds developed into a large storm, however I only had access to my Fujifilm x20, which I had never filmed on before. Not wanting to miss this opportunity, I set up my tripod and camera towards the storm and captured these shots of lightning.
I'm very pleased with the final results, which I slowed down in post-production using Premiere Pro. The vivid purple and pink colours are striking against the dark sky, and the overall framing of each shot is more than usable within double exposure.
I edited these clips together to demonstrate the particular clips which are usable. I had to crop each of the original files as they were around 10 minutes long, this is due to the lightning being fairly random, and I could only predict the general direction the bolts are coming from. This quickly approaching storm encouraged me to think on my feet in order to not miss the opportunity, developing my problem solving skills which will be useful in future shoots.
In preparation for the storyboard, I created sketches when I had an idea of how to illustrate certain lyrics on screen. I first used pen and paper to sketch out and plan some shots I intended to use.
I then switched to the iOS app Paper by FiftyThree, which allowed me to sketch on the go, with a wealth of colours which would not have been available otherwise. I found this a really useful software which allowed me to revisit sketches, so if I had a quick idea I could jot it down, and refer and refine it later.
As the track I have chosen has a male vocalist, I was keen to represent this through my casting choices. I decided to ask James who has featured in the pre-production tasks to be the actor in my video as I have already developed experiments with him, so I have a more solid background when filming the actual video.
James and I also share a similar timetable, which is helpful as the TV studio (where the majority of footage will take place) is only available 9-5 Monday-Fridays, which enables me to organise sessions around our free periods.
I referenced existing double exposure videos in creating a simple outfit which will act as the foundation for the overlaid image. A grey t-shirt and black jeans combination in recent years has been seen as an outfit which could suit casual casual occasions, as well as offering layering oppertunities, and is hence popular amongst the target audience. In turn, this fulfils the viewers expectations as to what a younger person would wear, removing any distraction which an outfit can create.
Below is the example of the risk regarding trailing cables connecting the camera to the main power supply. Cables are contained lowering the danger of becoming tangled in them.
As I intend to use time lapse within my final music video, I wanted to develop and refine the skills I developed from the original set of tests. After experimenting with double exposure in after effects, I composed each sky shot with nothing in the foreground, preventing any element obscuring the outline of the figure which it is overlaid onto. Whilst the tests using the iPhone 6 proved successful, I'm keen to improve the quality of the time lapses, especially if they're going to be manipulated in post production. Also, the iPhone doesn't give me much control over settings such as ISO, limiting the controls to focus and exposure.
I first intended to use a Canon 600D, and set up the shot for a sunset. Considering I have been filming for an hour at a time with the iPhone, I didn't realise the continuous shooting would be such a drain on the camera's battery. After 16 minutes the battery went completely flat, leaving me with perhaps 10 seconds of footage once edited. This is clearly very limiting, especially if I wanted to get multiple time lapse shots over the course of a day. Despite the battery issues, I am pleased with the quality of the footage I did achieve, and would like to use this camera more, preferably with a charger handy!
As I intended to shoot across the weekend with the 600D and did not have access to a charger, I used my contingency plan which was to use my own 1100D. I have shot with this camera before, and the video results have not been brilliant, again with little control over the exposure and ISO of each sequence. My original intention for using the 600D was to get some low-light shots of the night sky, as well as creating some out of focus shots resulting in Bokeh which would look quite abstract after the double exposure process. After attempting this with the 1100D, the results are very noisy and at times over exposed, creating footage which I would dismiss for the final video. I did manage to shoot a single star going across the sky, but this again was not very impressive or particularly usable due to the lack of interest within the frame.
Finally, I revisited the iPhone 6. After gaining footage of a sunset, I found the results to once again be noisy in the low-light, meaning that I would not use this camera either for the final video. However, the next day the sky was very characterful, and I managed to get some footage of a storm approaching. I was actually very impressed by this footage, leading me to the conclusion that the iPhone's strengths lie in well-lit scenes.
Inspired by the possibilities of multiple exposure, I used footage from the green screen and time lapse tests to create a double exposure clip. This was created using the skills which I had previously developed, but I do intend to search for tutorials which will further the impression of 3D within this 2D sequence.
As a group we visited the Green Screen studio and had a taster into green screen technology. Above is a video I edited of this session, with each of the elements practised written in the bottom right corner. As this was only a short session we did not have much time to fully refine the perspective or lighting of each experiment.
Perspective is extremely important when using green screen. It is very easy for the people to appear either too large or small in the frame, which we discovered and resolved by adjusting the camera. The background is also vital in this, for example, if an aerial photograph was used, we would have appeared to be sitting on top of the buildings, which would not have been a convincing composition. For this reason, when selecting each image (both the tree and the clouds) I made sure that the ground was visible in each, with the clouds being a more fun example, as if we were sitting on clouds. Lighting is also very important in creating a convincing image, which we did not have enough time to really refine.
By using Chroma Key to key out the shade of green from the screen, turning it transparent, this also meant that everything with that shade of green would also become transparent. As you can see above, I handed out various items in this green colour, including swatches of fabric, a bag, and a phone case, which all appear invisible. This could be interesting to integrate with masking the image, such as a headless person as demonstrated by Alex above.
After changing the background, we experimented with using multiple video channels, utilising three of the possible 4 (main camera, background, second camera) evident from the extra rectangle in the top image. This gives the possibility for the repetition and overlaying of footage, as well as allowing split screen and a second camera all in real time.
Effects and Transitions can also be added within the green screen technology, here is an example of a vignette placed around the main camera, causing Lewis and I to fade around the edges, creating a more seamless edge.
The aforementioned utilisation of multiple video channels is evident here, with the same main camera channel repeated behind the original footage. The green screen has already been keyed out, leaving everything that green colour transparent. As the background and green screen footage are on separate channels, this meant that the two green screen layers can be overlaid with no change to the background. I have illustrated this below:
Own Tests:
As James and I intend to use green screen for our final music videos, we privately booked the TV studio so we could learn more about the software involved and gain some footage which we can alter. We spent around 2 hours in the studio, and I've edited this down in the clip above to some sequences which I could use when experimenting with double exposure. Within the video, the use of multiple video channels is evidenced, as is the manipulation of these. Also, the variation of cinematography (including close up and long shot) as well as the both of us in the frame, offering a range of footage which I can develop upon at this stage.
Also in this session we looked at: varying backgrounds, the limitations of the available video channels, exporting the recording, changing lighting, how lighting effects the keying of the background, and adding motion clips. All of which could prove useful in future sessions.
