TLDR – I made an experimental documentary and I’m about 60% happy with it.


This film is a simple documentary/interview style piece which focuses on the relationship between image and sound. It centres on the ideas of validity of information, audience assumptions, and the notion of confessions.

It will feature 4-5 people sitting (individually) in a studio and looking directly into the camera, without speaking. As the camera lingers on each person (intercutting if/when necessary), different voiceovers will read the confessions of each person on screen. I am keeping the visuals as simple as possible to draw the attention of the audience to the sound.

Neither the voice reading the confession, nor the actual confession being heard, will necessarily match the person on screen, so audiences will not be sure at any point whose confession they are hearing. One confession, may also be a false story, added by me.

I am doing this because I am interested in the documentary genre, but specifically the type of documentary that offers up information to the audience in the form of a message, such as An Inconvenient Truth (2006), Bowling For Columbine (2002), or Blackfish (2013). With the documentary genre enjoying an increase in popularity since the turn of the century [1], my thoughts began to turn to the blurring lines between information and entertainment. While these films are inherently factual, they are often created with an agenda and with the intent of being a commercial success, so I wondered how impartial documentaries tended to be, if audiences can trust the information which they get from films, and if it even matters.

Furthermore, in today’s environment of 24 hour news saturation and ‘fake news’, the information that is presented to viewers as ‘fact’ will have a certain amount of inherent bias from those who present it. Frank Esser states that the average length of soundbites on American news (in this case, a segment of uninterrupted speech by a political candidate broadcast on television news) has shrunk from forty two seconds in 1968 to eight seconds in 2004 [2]. With such a short duration of speech available to audiences, it is left to journalists to select which quotes are used, and in what context, meaning the original meaning or spirit of the quote can be manipulated for almost any given purpose. With 57% of Americans getting their news from television [3], the stories they are watching have been chosen and shaped by a producer in order to obtain ratings, so how reliable and ‘honest’ can it be considered? The same is true in documentary filmmaking – nothing can ever be captured and presented to an audience without bias, so how much does a filmmaker’s decision making process influence the final outcome of the film or the information that the audiences take from it?

[1] (10/03/17)

[2] (02/04/17)

[3] (02/04/17)


This film requires 4-5 ‘actors’ to do the following three things:

  1. Write down a confession (in private). I need each participant to write down something which they have never told anyone before but that they would like to. It needn’t involve any specific names or places, and should be no longer than 4-5 lines.
    • I am thinking along the lines of a short story which doubles as a personal confession, which could be relatable to others, e.g. – “When I was 14 I was in Chemistry class in secondary school and I broke a Bunsen burner. I panicked and swapped it with the boy next to me when he wasn’t looking. When the teacher found out he had broken the equipment he was given detention after school. I never told him the truth but I still feel awful when I think about it.”
  2. Appear on camera. They will be sitting on a stool and staring at the camera for around 60 seconds. There will be no acting
  3. Record their voice in the sound studio (off camera) reading someone else’s confession.


Filming will take place in a studio setting on campus. I intend to keep things very simple – a white backdrop and a single chair in frame. The shots will be medium close ups of each actor sitting on the chair so as to focus on their upper body and face during the confession. I can’t decide if I want to use a slow tracking in shot or not, so what I think I might do is film in a higher resolution than needed, maybe 2k, to allow me to zoom in during post production to get a similar feel if I want to. A slow track/zoom into the actor’s face may increase the intensity of the statements being heard. I also may have a second camera on set to capture smaller body movement such as if the actor shuffles their feet, rubs their hands together, or bites their lip. These small nuances could increase the tension on screen during the confessions. A dolly zoom might be interesting as it’s often used in films to portray a figurative change of perception, so that theme could work well here.

I also intend to finish the film in black and white for two reasons – firstly, to minimise the visual difference between people on screen (by removing colour as a distraction and lowering the contrast between shades, the intention is to make everyone look similar and create an ‘everyman’ vibe), and secondly to highlight the metaphorical concept of the world being neither black, nor white, but shades of grey.

Stylistic influence – A short film by Tatia Pileva called First Kiss (2014) [4]. It’s a short documentary capturing the moment when strangers are asked to kiss for the first time. At present it has over 119 million views on Youtube. It also won a Golden Lion at Cannes Lion 2014, and a Clio award.
The audio recordings will take place on the same day, but in the sound studio on campus. I will have individuals go into the sound booth, pull out two of the written confessions from a hat, and read them both into the microphone. From a technical point of view this will give me the clearest sound possible for the edit. I will have the actors pick and read two confessions for two reasons. Firstly, if they pick their own confession, they can still read it (along with the other one) and neither I nor the audience will know the truth. Secondly, from a technical point of view, it will allow me more flexibility in the edit suite if each person offers more than one reading. Certain things may sound better when read by different people, so it allows me to make certain stylistic choices in post-production.

There may be music played in the background, though the track hasn’t been chosen yet. That, in itself, can be a vital component of the film, either supporting or contradicting the intended theme or meaning of the piece. In fact, Lara Cory (2013) mentioned in her blog that quite often the audio of a piece has the power to outweigh the image, being a far more efficient emotive indicator [5]. By adding music, I would be making an editorial decision to try and enhance or adjust the tone or emotion of the film, which in itself is a manipulation of the audience reaction. Should documentaries that wish to remain impartial include music?

The film will be 2-3 minutes in length, or however long it takes to read the confessions.

[4] (10/03/17)

[5] Music Without Words (16/03/17)


The purpose of the film is three-fold:

Firstly, to highlight the current trend of ‘fake news’, audience assumptions, and the validity of information sources. Last year (2016) was the year that ‘fake news’ really came to the forefront of popular culture. From Brexit to the Trump campaign, our news and social media outlets were flooded with so many stories that it became almost impossible to tell the truth from the lie. A recent study from Stanford University showed that 80% of American middle schoolers believed that the ‘sponsored content’ on their social media feeds was actual news [6]. These children are the voters of tomorrow, and they are painfully misinformed already, even at such an early age. As filmmakers, especially in an age where anyone can upload a video to Youtube and get millions of hits, it is important to understand the power that our work can have over public influence.

Secondly, to question the relationship between sound and image – if there is no obvious link between the image we see and the sound we hear, do audiences assume a connection, and how reliable is the information they then take from the scene? Reliability and trust are important concepts to consider when making a documentary film, so understanding the framework of the genre, and how information is derived from both sound and images is key to delivering an intended message. The difference between implication and inference can be vital for filmmakers, so knowing how audiences read a scene can be just as crucial as knowing how a scene is put together. Michael Chion (as quoted by David T. Johnson) states that sounds are generally only processed in relation to visualised spaces, in a concept called synchresis, and he notes the remarkable tolerance that audiences have for any sounds to seem as though they belong to the image [on screen]. “For a single body and a single face on the screen, thanks to synchresis, there are dozens of allowable voices.”[7] It is this theory that I will explore.

Thirdly, it questions the concept of a confession, and whether or not it counts if someone else is making the confession on your behalf, or if no one knows that you have actually made and owned the confession? While this is perhaps the most specific (and least important) question of the three, is does lend itself to the broader topic of information ownership and reliable sources. Second hand information like this, when not presented directly by the source, leaves itself vulnerable to misunderstanding and misinterpretation. If you can’t trust the source, can you trust the information?

[6] Evaluating Information: The Cornerstone of Civic Online Reasoning (16/03/17)

[7] Documenting the Documentary (16/03/17)


The production of this film really took places over three distinct stages: firstly, there was the collection of the confessions. Almost all of the confessions are real, from real people. One of them is my own, one of them is a lie. It was quite difficult getting people to send me confessions – there was probably a reason they had never told anyone, so the prospect of putting it on film was probably quite unnerving. I spoke to a few friends and got a couple of them to email me something, and then got a couple more from my Twitter followers after a rather pleading tweet one day. This was essentially ‘the writing’ stage of the film, and for one that consists of only about ten sentences, it was incredibly hard. Never underestimate the power of good writing.

The second stage was the audio recordings in the sound studio. Following on from the earlier problems of actually getting the confessions, no one either wanted to, or was able to, attend and record the script. In the end I doubled up an earlier session I had booked with Laure for An Exquisite Agony and had people fill in. We all read all of the confessions, that way I gave myself the most freedom when it came to the edit. I had several versions of each piece, read by several people. The actual recording was fun, although it took a while to get fully relaxed an in the zone, but within the hour we were flying. I think sound studio work is something I really enjoy. There’s something about the creativity of the art, but the controlled environment of the studio that seems to really blend well for me.

The final stage was the filming.  Having not enticed anyone to stand in front of a camera for 30 seconds, I was left to do it myself, just so I had something complete. In the end, I didn’t really care too much about what the film looked like, because I knew with only one person on screen it wouldn’t hold as much impact. I like to consider this a work in progress still. With that in mind, I set my camera on a tripod, stood in front of a white sheet that I hung from the wall, played the audio I had saved, and stared at the camera for two minutes. The idea was to try and act a little when each confession was read, but I found it difficult to direct myself acting, especially when I couldn’t see anything. A monitor would have helped a lot. The whole process took about 15 minutes and the less said about it the better.



I have a final version of this film, but it’s not one I’m happy with, or even particularly like. I had to make do with no actors on camera, so I thought I would just film myself so I could have some images to go with the audio I cut together. It’s not great, so for a very limited time, here it is online:


For this film I was trying to create a simple piece which would highlight and question the relationship between image and sound. It would ostensibly be a documentary film but I wanted to the audience to consider the authenticity of what was being presented to them.

I think this sort of thing is important today because so much of so many people’s news and information comes from videos and social media outlets, that often what is presented is fact is not, and this misleading information can have a profound effect on society as a whole (see Brexit and Trump).

When I was recording the monologues for this piece, I underestimated how different each person’s reading of the confessions would be. I had expected each recording to be relatively similar in tone, inflection, rhythm, and style, with not much more than the voice or the accent changing. They were obviously different in a huge way because I wasn’t using professional voice actors, and I wasn’t used to trying to direct people to change their voice. I also found it more difficult than expected to find people willing to write down a confession, and appear on screen, however briefly. Basically, I underestimated the difficulty of the whole process – at least those parts that relied on other people.

Because of that, I don’t think this film reflects my initial idea in terms of style, level of professionalism, or even content. What I do still think works is the concept of questioning the authority of any given information. I think it would have worked much better with a variety of actors on screen and the use of professional (or at least, trained) voice actors for the monologues, because with just one person on screen but a selection of different voiceovers, it begins to feel different. It feels less like a documentary and more like a piece of fiction, maybe with each voice representing a voice in the character’s head, rather than different individual people.

What I would like to do now is, first of all, remake the film to the original plan – with multiple people on screen, each reading a different person’s confession. I think it would drive my point home a little harder and a little clearer. I’d like it to gain some traction online because, with another general election coming up shortly, social media is about to become flooded with opinions and lies, and I think it’s important for people to be aware of how easy it is to create deliberately misleading information. The whole point of the idea is to make people question their assumptions about reliability and realism in the media, and while this particular film won’t achieve that, I still believe in the principal behind it.


  • Have a strong script written early,
  • Cast the film correctly,
  • Don’t direct yourself.