Unpacking Editorial Agreements in Collaborative Video Production

Abstract

Video production is a collaborative process involving creative, artistic and technical elements that require a multitude of specialised skill sets. This open-ended work is often marked by uncertainty and interpretive flexibility in terms of what the product is and should be. At the same time, most current video production tools are designed for single users. There is a growing interest, both in industry and academia, to design features that support key collaborative processes in editing, such as commenting on videos. We add to current research by unpacking specific forms of collaboration, in particular the social mechanisms and strategies employed to reduce interpretive flexibility and uncertainty in achieving agreements between editors and other collaborators. The findings contribute to the emerging design interest by identifying general design paths for how to support collaboration in video editing through scaffolding, iconic referencing, and suggestive editing.

Full text

Unpacking Editorial Agreements in Collaborative Video
Production
Pavel Okopnyi
University of Bergen
Bergen, Norway
Oskar Juhlin
Stockholm University
Stockholm, Sweden
Frode Guribye
University of Bergen
Bergen, Norway
ABSTRACT
Video production is a collaborative process involving creative, artis-
tic and technical elements that require a multitude of specialised
skill sets. This open-ended work is often marked by uncertainty and
interpretive exibility in terms of what the product is and should
be. At the same time, most current video production tools are de-
signed for single users. There is a growing interest, both in industry
and academia, to design features that support key collaborative
processes in editing, such as commenting on videos. We add to
current research by unpacking specic forms of collaboration, in
particular the social mechanisms and strategies employed to reduce
interpretive exibility and uncertainty in achieving agreements
between editors and other collaborators. The ndings contribute to
the emerging design interest by identifying general design paths for
how to support collaboration in video editing through scaolding,
iconic referencing, and suggestive editing.
CCS CONCEPTS
•Human-centered computing →User studies
;
Collaborative
interaction
;
Collaborative content creation
;
•Social and pro-
fessional topics →Computer supported cooperative work.
KEYWORDS
Collaborative Video Production; Video Editing; Interpretive Flexi-
bility; Feedback; Video Review; Editorial Agreement
ACM Reference Format:
Pavel Okopnyi, Oskar Juhlin, and Frode Guribye. 2020. Unpacking Edito-
rial Agreements in Collaborative Video Production. In ACM International
Conference on Interactive Media Experiences (IMX ’20), June 17–19, 2020,
Cornella, Barcelona, Spain. ACM, New York, NY, USA, 10 pages. https:
//doi.org/10.1145/3391614.3393652
1 INTRODUCTION
In recent years, there has been an increased interest in investigating
video production as an interactive and social practice [
15
]. This
has been fuelled by emerging practices, technical developments,
and video trac becoming the prevalent data in quantitative terms
both on the xed and mobile Internet [13].
Permission to make digital or hard copies of all or part of this work for personal or
classroom use is granted without fee provided that copies are not made or distributed
for prot or commercial advantage and that copies bear this notice and the full citation
on the rst page. Copyrights for components of this work owned by others than the
author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or
republish, to post on servers or to redistribute to lists, requires prior specic permission
and/or a fee. Request permissions from permissions@acm.org.
IMX ’20, June 17–19, 2020, Cornella, Barcelona, Spain
©2020 Copyright held by the owner/author(s). Publication rights licensed to ACM.
ACM ISBN 978-1-4503-7976-2/20/06. . . $15.00
https://doi.org/10.1145/3391614.3393652
New ways of producing video content have emerged as both
high-speed communications networks and production tools have
become cheaper and more readily available for distribution. Video
production, once exclusive to big companies and TV stations, has
also become more aordable due to the availability of less expen-
sive digital video cameras and non-linear video editing software
[
16
]. The video production workow is still a complex process [
2
].
On-going collaborative mechanisms of live production have been
described in a series of studies [
30
,
31
]. In the production of non-live
video, the workow also includes post-production, that is, work
occurring after shooting. Such work typically involves multiple
participants who possess certain skills. It includes video editing,
special eects application and sound mastering [
9
], and it relies
on an infrastructure of hardware devices and software tools [
11
].
Ultimately, it requires that people, tools, and artefacts are somewhat
organised [
3
]. The media format per se is open-ended, allowing for
multiple decisions, for example on style, colour, tempo and story
[
4
]. There is also interpretive exibility [
6
,
14
,
34
] in terms of what
the product is at a given moment and what it should be in the end.
It follows that collaborative processes are often marked by uncer-
tainties about what decisions to make and how the work should be
pursued.
In this paper, we focus on negotiations between editors and
other collaborators during the editing part of video production.
The motivation for this study is three-fold. First, as in any type of
collaborative work [
37
], negotiations are a critical part of video
production because editing depends on several collaborators in the
production coming together in what we call reviewing. Reviewing
involves interactions, and usually, a nal agreement on the result.
When an editor reveals the content for commenting by other collab-
orators or customers this can be characterised as a social interface.
Second, the topic is of interest because it makes the collaboration in
video production, with its opportunities and challenges, empirically
available, in a workow that is mostly functionally separated. Third,
given the advances in technology in hardware, software and net-
working technology, based on single users, a mismatch between the
collaborative production process and the functionality of the edit-
ing software has been identied as a challenge and an opportunity
in research and industry [28].
However, collaborative video editing still happens mostly be-
tween co-present collaborators or through asynchronous communi-
cation via email and messaging services [
28
]. In this study, we take
a step back from design to unpack current editing practices. To bet-
ter support design research, we argue for a widened scope to look
at how they reach editorial agreements in the collaborative process.
Unpacking the ways in which these agreements are made allows
117

IMX ’20, June 17–19, 2020, Cornella, Barcelona, Spain Okopnyi et al.
an empirically grounded understanding of user needs, and conse-
quently improved design exercises on how to support collaboration
in video editing beyond making comments on a timeline.
Thus we focus on social practices in and around editing. Speci-
cally, we consider mechanisms and strategies employed by collabo-
rating parties to reduce interpretive exibility and uncertainty and
achieve editorial agreements in video post-production.
The focus on unpacking social practices is aligned with what
Kirk et al. [
18
] stated in 2007, that is, that there was a need for a user-
centred approach given the emergence of mobile video production.
They identied a need to understand social interactions between
participants and co-authors [
19
], to inuence design that met user
demands. Such studies included, for example, explorations of how
storytelling patterns help novice users create videos [
17
] and that
of live video production [30, 31].
We conducted 11 in-depth semi-structured interviews with 13
participants who had experience in video editing of dierent scales
and in dierent roles and who also worked together in projects.
Since such work is often distributed and the interaction between
collaborators is mostly spread out in time [
28
], making it hard to
carry out observations, we chose to conduct interviews.
In unpacking the ways in which editing is done, we are inspired
by the sociology of knowledge and the so-called empirical program
of relativism (EPOR) [
6
,
34
]. An important premise in this line of
inquiry is that the content does not hold the resolution for disagree-
ments. First, we display the interpretive exibility in media and
in social interaction during editing, such as in choosing colours,
aspect ratio, emotions and style. Second, we identify various social
closure mechanisms through which the exibility is reduced, such
as exclusion of participants from ther early stages of production,
introduction of mediators, and enforcement of contracts. These
ndings are then used to suggest new ways to design for collabo-
ration in editing suites. In addition to the mechanisms that have
already been investigated, we suggest that design of technology
should draw on three types of collaborative closure mechanisms –
scaolding, iconic referencing, and suggestive editing.
2 PREVIOUS RESEARCH
Video editing is the manipulation of video clips and other media, for
example, music and images, to arrange them into a motion picture.
It is typically carried out with non-linear video editing software
that allows random access to dierent parts of footage, audio and
more. Most of the editing work is conducted by individuals [
28
],
which is reected in software design. Editing suites, such as Adobe
Premiere and FinalCut, provide extensive means for a single user
to manipulate the video in several dimensions, including sound
and transitions; however, they include few supports to collaborate
through the systems.
In our study, we focus on collaboration in the post-production
phase of video production. Especially, we consider how production
teams negotiate editorial decisions. This entails, for example, clients
giving feedback to the editor and production team or feedback
between an editor and producer. In this section, we discuss relevant
studies of feedback in collaboration and existing video feedback
services, as well as studies that have focussed on designing feedback
features.
2.1 Feedback in Collaborative Creativity
There have been new suggestions about collaborative features in
editing suites that support current key cooperative processes, that
is, the sharing and handling of feedback in the editing process. Tech-
nical aspects of collaborative editing have been studied [
26
], and
new general collaborative functionalities have been investigated
(e.g. [
39
]). There have also been studies seeking to provide new
features that support the specic task of commenting on edits in
video production.
One of the forms of communication that helps achieve results is
feedback – an assessment of performance or a product by clients
who hire creators to produce something. Creators seek feedback,
because it helps them understand their clients’ needs [
1
]. As a part
of the creative process, feedback is interactive. Through it reviewers
and creators “co-construct problems space” allowing for changes
to be implemented [12].
The quality of feedback is crucial to its processing [
25
]. Its timing
and specicity largely aect the allocation of resources and per-
formance quality, which inuence overall performance as creators
“invest more resources on tasks for which higher quality ... feedback
is available” [25, p. 1076].
Creative workers employ dierent strategies to communicate
their work and receive feedback. In a face-to-face environment,
authors can have richer communication with reviewers due to non-
verbal cues, and the context of commentaries can be immediately
understood; they also have opportunities to discuss the video in de-
tail and brainstorm about new changes [
28
]. However, the complex
nature of media, such as music [
33
], visual design [
27
] or schematics
[
36
], can lead to inaccurate, incorrect and ambiguous information
being passed from one party to another. Receiving feedback can
sometimes be frustrating, and as suggested by Phalip et al. [
32
],
dealing with it can be even harder in a face-to-face environment. To
alleviate communicational challenges designers and engineers can
refrain from sharing works in progress with their clients, colleagues
and supervisors and create tailored presentations containing only
selected elements [27, 36].
In a distributed setting, it is challenging to make sure that all
participants are aware of the object and topic of the discussion [
21
].
Thus, remote feedback needs to deal not only with what Phalip et.
al [
32
] labelled as the ambiguity of content (“what is said”) but also
with the ambiguity of scope (“what is talked about”).
2.2 Existing Video Feedback Services
There is a strong interest in feedback communication services in
the video production industry which is evident in the existence of
many live web services, such as Frame.io
1
, Vidhub
2
, Commentbub-
ble
3
, Wipster
4
, Filestage
5
, LookAt.io
6
, ReviewStudio
7
, Screenlight
8
,
and others aiming to simplify the video production workow by
facilitating feedback. On their websites, these services emphasise
1https://frame.io/
2https://vidhub.co/
3https://commentbubble.com/
4https://wipster.io
5https://lestage.io/
6https://www.lookat.io/
7https://www.reviewstudio.com/
8https://screenlight.tv/
118

Unpacking Editorial Agreements in Collaborative Video Production IMX ’20, June 17–19, 2020, Cornella, Barcelona, Spain
that “[d]isorganized client feedback can slow” video production,
and they oer tools to share and comment on videos with automatic
time coding. These tools mainly focus on the support of indexical
referencing by introducing a new visual layer on top of the video.
In this layer, users can highlight certain areas of the frame, add
annotations, and provide structured feedback to video producers.
The commentaries are represented on the timeline as markers, in-
dicating that there is a comment associated with the certain point
in time or time frame.
These services are usually stand-alone web applications. Some,
such as Frame.io and Wipster, integrate with popular video editing
software, allowing video editors to view commentaries directly in
the software. Thus, those services normally require a video editor
to upload a video, share the link to it with a reviewer and wait
for the commentaries. The reviewer will then access the service
interface and leave comments on the video. The editor will see the
comments in the service interface or, if it is integrated with the
service, the video editor.
2.3 Designing for Feedback
There have been some attempts to research and design for reviewing
of video and visual media in dierent settings, from real-life video
production sites [
2
] to online platforms [
28
] and virtual reality [
24
].
There is a certain design challenge here, as reviewers can refer to
both very specic and general objects or topics in their comments
[5].
Inspired by scaolding theory, Luther et al. developed a review-
ing system for visual design prototypes, such as user interfaces and
posters; the system provided a structured interface giving reviewer
a set of pre-determined design critique statements to choose from
[
20
]. This approach allowed even novice users to produce feedback,
which would be considered almost as good as feedback by experts
[42].
O’Leary et al. developed a tool for designers allowing them to
take snapshots of dierent versions of their work and annotate them.
This helps designers keep track of and act on received feedback.
However, the lack of integration with the design tool (Adobe XD)
causes some friction in the workow. [27]
Pavel et al. [
28
] developed a prototype of a reviewing system that
allows a reviewer to video record comments that will be attached
to certain points on the timeline and share the video review with
the editor.
In an experiment with a tangible video editor that allows multiple
participants to edit the video together [
43
], collaborators tended to
communicate actively, in contrast, in the editing session with a tra-
ditional software editor, it was found that one participant dominant.
In another experiment [
40
], a tangible video editor aimed to benet
from the implicit group awareness [
8
] – a sense of co-presence, un-
derstanding of other participants’ work, and availability for direct
communication.
Bartindale et al. designed a system for actors and lm crew to
review video on set, annotate it, and provide context for their acting
work. However, in the eld trial, the system was mostly used by cast
for “simple browsing, comparing and playback of clips to reect on
performances” [
2
]. With that limited knowledge of current practices
in video editing and specic design suggestions, we argue that there
is a need for more user studies.
3 METHODS AND SETTING
In this section, we briey describe the video production process, and
our approach to investigating feedback practices and the analysis
of gathered data.
3.1 The Video Production Process
Video production is a complex process that usually involves many
participants. An organisation or individual needing to create a video
hires video professionals, comprising a video production team that
is usually organised and led by a producer – a person responsible
for both client handling and production arrangement. Under a
producer’s supervision, various professionals (e.g., screenwriters,
directors, video editors, sound editors, colourists, etc.) collaborate to
create a video. Depending on the size of the project and its budget,
a video production team can include from several to dozens of
participants.
The video production process traditionally consists of three
stages, which are as follows: pre-production, which includes plan-
ning and organising the lming process; production, which is
footage lming; and post-production, in which the footage is edited.
Editors cut lm into a coherent video clip or movie, sound editors
manipulate the sounds and music of the video, colourists apply l-
ters and so on. In many cases, especially in small video productions,
a single person can acquire several roles; for example, the editor
may be also responsible for colour correction and sound mastering.
At some point, the client reviews the video and provides feedback to
the production team, typically in the form of time-coded comments.
In this work, we focus on communication between clients and
production teams, how this communication is organised and what
strategies production teams apply during this communication.
3.2 Data Gathering
We performed a series of in-depth semi-structured interviews. We
applied a purposive sampling strategy and recruited people with
experience in video editing at dierent scales, from amateur clips
to feature-length movies. In ve cases, interviewees gave us con-
tact information for colleagues and former clients who agreed to
participate in this study.
Dierent types of video production contexts involve dierent
kinds of collaboration. In live video production, for example, they
rely on large, tight and co-located teams [
9
,
30
,
31
] that interact in
various ways, such as through audio channels and camera move-
ments: these have been unpacked in observational studies. In this
research, we focus on post-production: this process can involve
large teams where individuals have dierent roles, but these dis-
tributed settings do not easily lend themselves to observations.
Thus, the choice of the interview method and the topic is moti-
vated by the focus on the collaboration that often occurs between
geographical places and asynchronously.
In the interviews, we also asked the informants about providing
examples of feedback they had received from clients, and we ob-
tained several such examples. In addition, one of the clients who
was interviewed provided examples of feedback to the production
119

IMX ’20, June 17–19, 2020, Cornella, Barcelona, Spain Okopnyi et al.
team. These textual records were then discussed in some of the
follow-up interviews (i.e. we asked how they perceived and acted
on the feedback).
We conducted 11 in-depth semi-structured interviews with 13
participants, including two interviews with pairs of collaborators
and nine one-on-one interviews. Each interview lasted from 25
minutes to 2 hours. In the Table 1 we list some details about each
interview, including the countries in which participants had their
experiences. Eight of the interviewees were video production pro-
fessionals who worked with video daily. Five were customers or
superiors of video producers/editors. Five had experience working
for TV stations in various capacities – as editors, journalists, camera
operators and so on. During the interviews, participants recalled
their recent projects and gave accounts of their collaborative video
editing experience. We recruited people who were involved in video
editing in a professional capacity and we were specically aiming
for people who had worked together. To recruit participants we
reached out to video professionals in our personal networks and
some of them put us in contact with former and present colleagues;
in one instance, we supplemented the data collection with inter-
views of two clients. Nine of the interviewees had collaborated or
were collaborating with each other as follows: I02 (editor and pro-
ducer pair) had worked together on one project; I03 (cameraman and
editor/editing director), I04 (producer), and I05 (editor) had worked
together in several dierent projects; and I07 (writer/producer), I8
(editor/camera operator), I10 and I11 (clients) had worked together
on a big educational video project. Their accounts of their shared
projects allowed us to analyse the collaborative process from their
dierent perspectives.
Eight interviews were supplemented with follow-up questions
conducted in person or via messaging services and social media
websites. In the follow-ups, we focused on specic cases of (mis-
)communication with co-authors and customers and strategies for
handling these cases.
3.3 Data Analysis
We used qualitative content analysis to analyse the interview tran-
scripts. This is a research technique for making valid inferences
from texts (or other meaningful matter) to the contexts of their use.
It is an approach for qualitative data reduction and sense-making
based on taking a volume of qualitative material and attempting to
identify core consistencies and meanings [
35
]. This enables us to
re-articulate the meanings of the empirical data set in light of the
context, allowing research questions and answers to arise in the
interaction between the known literature and corpus [19].
The data was coded inductively to identify relevant aspects of
the interaction between the collaborators, without the intention of
revealing underlying or discursive meanings. The process of coding
and categorising was done iteratively in collaborative sessions, as
well as individually by one of the researchers, who was the main
person responsible for the analysis. The discussions and analytical
work from the joint sessions laid the foundation for the coding
scheme. It was applied and constantly rened and improved by the
responsible coder, who revisited and coded the rest of the material
between the collaborative sessions.
4 ANALYSIS AND RESULTS
From the analysis of our interviews we identied nine relevant
themes for understanding reviewing in video editing, referring
to either the interpretive exibility of content or various ways
collaborators reach agreements.
Below, we present the most relevant categories in the coding
scheme, to address our questions concerning the interpretive ex-
ibility in editing, as well as the ways participants reach editorial
agreements.
4.1 Interpretive Flexibility in Editing
The comments on the video reect the possibilities of interpreting
the video-editing process in many ways. They relate to ambiguities
and uncertainties in both the process and the medium that add
to the challenges of reviewing and collaborating. In other words,
there is an interpretive exibility [
6
] when addressing dierent
video features. In the following, we discuss examples of such open-
endedness in dierent features of video production, that is, both
in ostensibly concrete elements of the video, such as audio and
colouring, and partly in overlapping and nebulous qualities, such as
montage, style and narrative. All these elements are up for negotia-
tion and discussion throughout the editing phase, notwithstanding
any pre-production agreements or plans.
4.1.1 Interpreting Video in Words. Discussion and textual com-
ments in and on video editing have a specic professional lan-
guage with idiosyncratic terms, jargon, vocabulary and assump-
tions. Nonetheless, it is hard to describe the audio-visual, time-based
medium. As one of the interviewees put it: It is very dicult to
put into words the full range of audio-visual eects experienced when
viewing video, and everyone’s perception works dierently (I04).9
In pre-production an idea for a video must somehow be presented
in the form of a textual task description, including the look and feel
of the video. This is a challenge according to one editor: A task
description is a type of art; because you need to describe what you need
in text and [do it] briey (I05). Interpretive exibility can prevail
throughout the production process, without participants reaching
any agreements. Collaboration in editing generates continuous
instances of uncertainty and vagueness between editors and clients,
for example, as stated by one editor: Often, clients ask you to do it
“awesome” but they cannot dene “awesome”. ... They try to describe on
paper what they want to see, but it’s hard. Our language is insucient.
We need something else, another instrument (I04).
The diculties of interpreting content concern, among other
things, colours, style and rhythm. An editor paraphrased feedback
from a client as follows: From this frame to that frame, I don’t like
how this woman’s face looks; it’s too pale (I04).
The editors found it challenging to articulate the desired colour
tone and brightness level, as well as what a collaborator desires. In
another example, an editor had to handle a comment stating that
the blue is not blue enough (I03).
Furthermore, the “style” of a video can imply several features,
such as choice of colour scheme, frame rate and montage. The fol-
lowing example shows how establishing a working relationship
involves knowing how to interpret features like style and colour:
9Some quotations have been translated from Russian, including this one.
120

Unpacking Editorial Agreements in Collaborative Video Production IMX ’20, June 17–19, 2020, Cornella, Barcelona, Spain
N
Participants (number of,
gender, country)
Length (min-
utes)
Main Role(s) in Video Production
Experience in Video Pro-
duction
Video Type
01 1, male, USA 148 Writer, Producer, Editor, Camera Not Professional Documentary and Educational
02 2, male, Russia 112 Writer, Producer, Editor, Camera Not Professional Educational
03 2, male, Russia 69 Producer, Editor Professional Corporate and TV
04 1, male, Russia 58 Writer, Producer Professional Corporate
05 1, female, Russia 88 Editor Professional Corporate, TV, Full-length
06 1, male, USA/Ecuador 84 Writer, Producer, Editor Professional Full-length feature lms
07 1, female, Norway 71 Writer, Producer, Editor Professional Educational
08 1, male, Norway 67 Editor, Camera Professional Educational
09 1, male, Norway 25 Editor Professional Commercial
10 1, male, Norway 66 Writer, Reviewer Not Professional Educational
11 1, male, Norway 53 Writer, Reviewer Not Professional Educational
Table 1: Interviews Metadata
Those [directors] that we have known for a long time, we know their
styles. Frame rate, close-ups or wide shots. One loves to see the video
in sepia, another wants high contrast. ... With new directors, we must
work more, discussing their vision and how they would like their
videos. (I03) Interpretations of the rhythm of a video are also con-
sidered to bring uncertainty and vagueness. The montage – that is,
the sequencing and pacing of shots and scenes in a video – inuence
viewers’ experience of a video [
23
]. One director paraphrased a
comment he has given to an editor: “You know, it doesn’t feel right,
like, the rhythm does not feel right” (I06). In this comment, the lm
director is recalling his conversation with an editor who worked
on his movie. The director did not like the pacing in the open-
ing sequence of scenes of the movie, which lacked the emotional
communication the director was trying to capture.
4.1.2 Context-Related Uncertainties. There are ambivalences de-
riving from the context in which a video is produced, rather than
the edit per se. The context can either be the production process
or the viewing situation. The main challenge in the collaboration
is to agree on how to interpret the video given the actual state of
the production process. From editor’s point of view, this process is
not necessarily obvious to other stakeholders: They watch [the raw
video] and start screaming, “Why it is so cheap?”. ... After 15 minutes
of [arguing], I understood that it was the rst time in their life when
they had seen a frame in a cine prole directly from a camera without
colour correction (I04).
In this interview excerpt, the video producer is describing a case
in which his clients did not understand that high-quality video is
achieved in a long process of editing, including colour correction.
More experienced video editors or producers project certain quali-
ties onto the unpolished footage. They can eectively distinguish
between how it looks now and focus on the right aspects when
making an assessment in an early stage of post-production. This
may not be easy for those with less experience. Furthermore, the
reviewing needs to consider the stage of the production process: In
some cases, we can show a work half-done, in others – it is better not
to. If [clients] don’t belong to the media world, they won’t understand
that this is a draft and they will start giving negative feedback on
good work (I03).
This interviewee, who was an editor, explained that clients in-
experienced in video production tend to focus their feedback on
small details that will be addressed at a later stage. To avoid this
non-constructive feedback, they employ the strategy of excluding
clients from the editing process and do not show the draft video
until it is almost nished.
Another type of uncertainty is related to the need to understand
their potential viewers and their context: Not so long ago we did
a project where we were working a lot with sound and music. The
client said they didn’t hear anything. It turned out that they had
crappy speakers in their laptop (I05). In this case, the editor had
designed the audio expecting the clients to use headphones or a
high-quality sound system but the audio seemed poor because the
client was using the poor equipment. This example highlights that
some aspects of the review and assessment of the video will rest
on assumptions by participants in the collaborative process. Such
assumptions may not be shared by other participants, and this can
be a source of misunderstandings and disagreements.
4.1.3 Confusion when Co-Present. We identied ambiguity in in-
terpreting edits even when collaborators were physically co-located:
One time when I was submitting the video in person, it was at their
headquarters in [another city]. We also had the CEO and company
owner on Skype ... instead of constructive feedback everybody was
just shouting their opinions. I was contending with 10 people who did
not belong there (I04).
Although co-present and having the opportunity to engage in
indexical gesturing over hours of discussion, the meeting ended
without any agreement. The producer, who was presenting the
video, could not argue with 10 people. Confronted with many com-
ments, his arguments on costs and other practicalities fell short.
4.1.4 Interpretive Flexibility in Referencing. An important aspect
in reviewing is nding a way to describe where an issue is found in
the video, that is, making a reference. Below, we describe two such
approaches – reference to time and reference to objects – and their
associated ambiguities. Time-coded comments are appreciated by
editors. They send the video to the client for a round of review
with instructions on how to provide feedback with references to a
timeline which is visible in the video per se or in the software: Of
course, you have to tell them ... the feedback needs to be time-coded,
so you tell them: “Make notes when you watch the video, write down
the time code and the comment you have” (I07). Such feedback can
be communicated both in person and via email and messengers.
However, the interviewees also provided examples where such in-
structions did not work: Sometimes, you need to specify, as they
say, “At 56 seconds, the frame is not suitable”, and I have a cut [be-
tween frames] at 56 seconds and there is a need to specify (I05). The
121

IMX ’20, June 17–19, 2020, Cornella, Barcelona, Spain Okopnyi et al.
time code ("56 seconds") refers to several frames, since the video
is normally made up of 25 frames per second. In this example, the
editor struggled to understand what was meant by the comment,
since the time sequence contained dierent contents. Time codes,
or any other time references, refer to intervals, rather than frames,
resulted in ambiguity in this case.
Collaborators may refer to objects in the video to position the
reviewing comment. This is illustrated in the following, with an
editor citing a previous comment by a director: “You have a shot
with a girl walking in the frame”, but I have a lot of shots with a
girl walking (I03). A reference by content is confusing – even this
seemingly concrete and detailed description was insucient in this
case.
All the above accounts of situations reveal uncertainties and
disagreements even when there are apparently clear references to
structural elements in the video or specic scenes in the footage.
Such uncertainties are clearly prevalent in the collaborative process.
In the next section, we unpack the ways in which such interpretive
exibility is reduced and agreements are achieved.
4.2 Ways of Reaching Editorial Agreement
Here, we analyse the ways in which participants achieve edito-
rial agreement. These include agreements reached by reference to
written statements and other documents, closure by rhetorics of
practicalities and organisational closure mechanisms.
4.2.1 Agreements Through Wrien Statements. To reduce exibility
in the interpretation of a video, editors rely on written communica-
tion: I prefer written communication, not phone calls. Because if you
discuss everything by phone, then it starts: “Oh, I asked this, but you
did not hear me” (I05).
In personal meetings, editors employ note-taking and even audio
recording: If a problem emerges in a face-to-face meeting, that means
they are too lazy to take notes, but I usually make my own notes.
Sometimes, I record the meeting with a voice recorder, and I have a
good memory. (I04)
The editors assume authority by referring to their written notes
in case of a disagreement with a client.
4.2.2 Closure by Documentation. In pre-production, there are many
artefacts – ways of documenting and formalising – available for
planning and specifying what the video should be. They can also
be used to specify content, style and narrative. This limits the pro-
duction process and its scope. Documentation can be of various
types, for example, contracts, storyboards and screenplays.
A contract establishes the legal and monetary relationship be-
tween parties. It is also used to delimit interpretive exibility. With-
out specied limitations, the client may want to redene the task
during the production: For example, we have a task to shoot this, with
this camera movement. The client [on site] asks the camera operator,
“Can you also shoot that thing?” (I03).
The interviewee describes a recurring situation in which a client,
who was present during the lming process, tried to extend the
task by asking a camera operator to lm additional scenes. These
scenes were not included in the plan or contract. In cases like
this, a mechanism the production team can use is to point to the
documentation to avoid unnecessary costs: Usually, you have to
decline. You tell them, “Look, we have a storyboard or a screenplay,
we did everything according to it” (I04). This example shows how
documentation is brought into the collaborative process to reduce
the interpretive exibility of what the end product should be; we
term this closure by documentation.
Another related mechanism is limiting the number of editing
iterations: In the contract, we usually state, for example, that we will
do three iterations of changes. So, the client agrees that it is enough.
... I have a maximum of one or two iterations because I work with
experienced clients (I04). The process risks being prolonged if it is
not delimited in the contract. In the following, the editor described
her experience of re-editing a video, which was originally shot by
another production team but became embroiled in an extended
editing process: There is a team that works on a video, they do
multiple edits, and they don’t have a limit for edits in the contract. It’s
a scandal (I05). The exible and open-ended nature of the process
may even be used by parties for reasons related to external context:
Once, we had a situation in which the customer came up with new
ideas for edits just to prolong the project because they did not have
the money to pay us on time (I03).
While the documentation and artefacts aim to specify and delimit
the scope of the product and process, there will inevitably be certain
aspects of the process and inherent uncertainties that will need to
be resolved during production.
4.2.3 Rhetorics of Practicalities. The participants sometimes evoked
local practical constraints to reduce interpretive exibility. In this
case, one editor created a specic, local “deadline” that she would
suggest to delimit collaborators: I specify this every time. I say, "Do
you need it by eleven? Because after twelve I will be busy with another
project". And they say, "Yes, by eleven", or "We will watch it after
eleven and send changes during the day" (I05).
This mechanism draws on local circumstances to reduce inter-
pretive exibility. We see it as a rhetorical device since there are
always such circumstances, although they are not always evoked
as they were in this case.
4.2.4 Iconic Referencing. Another way to reduce interpretive exi-
bility is pointing to some sort of template or example. Clients use
such references, for example, videos which they like or nd suitable:
Intelligent clients provide you with references, other videos, and tell
you what they like – colour scheme, dynamics, cuts and so on (I04).
These examples contribute to reducing the open-endedness of the
editing process and managing expectations. They also oer produc-
ers an opportunity to reuse existing elements instead of designing
them, from discussions with a client, including new fonts, eects,
colours, lters and so on.
4.2.5 Organisational Closure Mechanisms. Another way to limit
uncertainty and interpretive exibility is by applying what we
term organisational closure mechanisms. They work by arranging
collaborative activity in ways that decrease the options. Below,
we present three examples covering how video workers seek to
limit the number of participants involved in review sessions, how
the production team prefers to have one representative from the
stakeholders who knows the history of the project and how the
team appoints a single negotiator that deals with client feedback.
122

Unpacking Editorial Agreements in Collaborative Video Production IMX ’20, June 17–19, 2020, Cornella, Barcelona, Spain
One editor spoke about an experience with multiple people from
the client side who were involved in the discussion of the video.
They lacked an understanding of what changes were possible and
how much edits would cost, for example, re-mastering sound would
be very costly, and re-shooting a scene from another camera angle
would be eectively impossible. The editor commented, You have a
project group and you agree on the script, we shoot it and send them
the movie. But then there is a person who did not participate in the
initial discussion, and they say, “Let’s do it this way” ... However, they
don’t know how much it will cost to implement some changes (I08). A
strategy used by video producers to avoid such situations is limiting
the communication to a single person from the client side who also
has a mandate to make decisions: If ... you were negotiating with a
company CEO that he will be the one accepting the video when it’s
done ... there is no sense in talking to managers (I03). The editor
explained that when a video is ordered on behalf of a company, its
representative, typically a person from marketing, may not want to
take responsibility for the video and could be hesitant both to accept
the work and suggest edits. Thus, the production team prefers to
communicate with an executive, such as the CEO, who does not
answer to a supervisor and can be called the main stakeholder in
the process.
A similar strategy can be applied to the production team, includ-
ing operators and editors. To restrict the conversation and focus
the production, a video producer may serve as a single negotiator
who is responsible for handling communication with a client: All
the changes come through the producer, who passes the changes [list
of suggestions] to the performers. ... Or, everybody starts to express
themselves and they cling with tongues. I think, this does not happen
in big serious projects, they must have a hierarchy (I04).
The producer acts as a negotiator between the production team
and client. This way, they aim to arrange and limit the client’s
opportunity to participate in the project.
In sum, all these mechanisms, including the reduction in the
number of participants, preference for an appointed reviewer, and
appointing negotiators, aim to arrange the collaboration in a certain
way to reduce feedback.
5 DISCUSSION AND DESIGN IMPLICATIONS
We have studied how a part of the collaboration in video editing
– reviewing edits – is conducted. We unpacked how participants
come to agreements, given the uncertainty and ambiguities. The
latter are often handled during iterations of reviews and re-edits,
and often include written communication. Although the study is
motivated by eorts to increase collaboration remotely, it reveals
challenges linked to interpretative exibility in editing, both dur-
ing co-located edits with the possibility for indexical gestures, for
example, when partners are co-present in a physical meeting –
and when conducted remotely and asynchronously. The challenges
identied in both situations indicates that replicating local indexical
gesturing remotely [
28
], is not the only possible design path. Here,
we broaden the design space, emphasising the variation of closure
mechanisms. We then discuss how they tend to favour the editors
over other collaborators and suggest design directions that would
re-balance their positions.
5.1 Reaching Editorial Agreements
Negotiations are critical in collaborative work [
37
]. The question
of how collaborators reach editing agreements resonates with a
long-standing research theme in the area of computer supported
collaborative work (CSCW). This focusses on how work ts with
formal descriptions of workows and work processes, given that
the latter are inherently incompletely described (e.g., [
7
,
10
,
37
,
38]). We unpack the mechanisms employed to reduce interpretive
exibility in video editing. For example, participants organised face-
to-face communication and gesturing. In distributive asynchronous
collaboration, they relied on documents, such as contracts and story
boards, as well as organisational mechanisms, to suppress attempts
to reinterpret the results. Moreover, they organised feedback-giving
process around their working cycles and limited the inuence of
customers by excluding them from the editing process.
5.2 The Politics of Editing
Suchman [
38
] argued that descriptions and formalisation of roles,
responsibilities and intents in workows are important for the poli-
tics of workplaces, in that they favour one perspective – commonly
the managerial perspective – over another. In our analysis, we see
organisational asymmetries generated through editors’ struggle to
control the collaborative process, for example, by what we refer to
as “organisational closure mechanisms”. Such strategies reduce un-
certainty and interpretive exibility while decreasing the possibility
for clients to take an active part in production.
5.3 New Design Paths
In academic research and industry, there is currently an interest
in increasing the digital opportunities for collaboration in video
production. As mentioned above, many existing collaborative tools
rely on largely indexical referencing, for example, the provision of
feedback in the form of time-coded commentaries. We present three
additional design paths for distributed and remote collaborations
in editing. They relate to the identied closure mechanisms and,
at the same time, aim to support clients and other non-editors in
participating more actively in the editing process. We term these
design paths scaolding,iconic referencing and suggestive editing.
The goal of the study and purpose of identifying the design paths
is to show where new design explorations can be done and iden-
tify these as possible design spaces. The design paths are meant
to inspire new designs and give a general direction for such explo-
rations. The generation of relevant and concrete design suggestions
requires expertise not only in video production, such as exhibited
by our interviewees, but also in systems design and software de-
velopment. We think that such design explorations would require
a research-through-design study involving designers, developers,
and users to explore these paths and populate them with concrete
design proposals that could be iteratively implemented and evalu-
ated. In this way, we contribute a user study that shows the need to
move beyond the research on concrete design attempts, which have
already been tried out in other design studies and commercially
available applications.
5.3.1 Scaolding the Collaboration. Below, we discuss the advan-
tages and disadvantages of “scaolding” mechanisms, by which we
123

IMX ’20, June 17–19, 2020, Cornella, Barcelona, Spain Okopnyi et al.
mean a pre-set structure that supports participant collaboration
[
22
,
41
]. Our approach is like that reported by Luther et al. [
20
,
42
].
In our case, such mechanisms should be integrated within a collab-
orative editing suite to mitigate participants’ level of commitment
and knowledge of video production. The proposed mechanisms
include organisational scaolding,procedural scaolding, and docu-
mentational scaolding.
Organisational Scaolding: Such structure makes visible the over-
all functional separation of the project including in the editing. The
scaold establishes this structure, in this specic detailed work.
Dierent participants are given certain roles in the review process,
which depend on their general responsibilities. For example, a per-
son from marketing might review the video in terms of how their
brand is represented in the video. Another representative might be
assigned the task of fact checking the video. Yet another collabora-
tor might be instructed to check whether the video’s narrative ts
the initial intentions, including look and feel as conveyed through
the video’s montage, pace and emotion.
These mechanisms can be implemented through assigning roles
to dierent stakeholders in a video reviewing software and provid-
ing further scaolding through instructions that correspond to the
dierent roles and responsibilities.
Procedural Scaolding: As noted in the analysis, experienced
video editors assume certain qualities in an otherwise unnished
video edit. For example, they will disregard poor sound quality
assuming that it will be improved after so-called mastering. A pro-
cedural scaolding mechanism would help inexperienced reviewers
to focus on such elements of the current version, where they could
give comments that t in with the state of production. The example
in the analysis, where the reviewer thought the video looked cheap,
could be mitigated if, for example, the suite provided information
on upcoming colour corrections. A procedural scaold providing
instructions in line with the current production stage, as in the
example above, could guide the review process. It could also make
suggestions for what to look for in each of the iterations in the video
editing, considering how the workow and production pipeline are
organised and what is usually done in each stage.
Documentational Scaolding: As mentioned above, contracts,
story-boards and textual documentation have a key role in video
production. They help to articulate intents and the vision for the
nal product. Linking these documents to relevant parts of the
video suite strengthen them as structural support. For example if a
contract states that the video should contain a certain element, such
as an interview with a certain person, a link can be provided to the
segment of the video to show that this has been included and where
to nd it in the video. Moreover, textual documentation coul be
made into a checklist to be included in the software. It could point
out which of the elements are ready for review and what elements
remain to be included. Such documentational scaolding mech-
anisms already exist in many o-the-shelf project management
tools where there is support for tracking the progress of dier-
ent tasks. This could also be useful in a video production context
to operationalise the documentation into tasks, make transparent
what contractual obligations have been fullled and ensure that
the feedback is not concerned with unnished tasks.
Challenges in Scaolding. There is a downside to the additions
described above. Adding digital scaolding features to the dierent
parts of the video production process will involve formalising tasks
and providing detailed descriptions of intents, roles and respon-
sibilities. Such descriptions would still be inherently incomplete
and thus, run the risk of creating new problems. Such solutions
would also add extra work for the collaborators. Therefore, a chal-
lenge in the design of such scaolding features would be striking a
proper balance between adding support and over-formalising the
collaborative process.
5.3.2 Iconic Referencing. Video workers often employ indexical
referencing in their communication. For example, they refer to a
location by a time code or gesturing towards the screen with a
nger or via an overlay in collaborative video reviewing software.
Our ndings indicate that, while it is undoubtedly useful, this is
not always sucient. We also note that the use of examples, which
reect collaborators’ ideas for the video, is appreciated to reach
agreements on what to do. We term this practice “iconic referenc-
ing”, alluding to semiotic theory [
29
]. We suggest that it should be
investigated as a design path in collaborative video production tools.
For example, the editing suite could allow a reviewer to suggest
a concrete colour scheme. Such examples would then add to the
type of general statements identied as problematic by editors, for
example, when trying to work out the meaning of “blue is not blue
enough.” A more complex example would be an implementation of a
reference library with examples of video and audio clips. Examples
in this library could be focussed on, and grouped by, dierent video
features such as graphics and colours. The client would browse
the library and search for good examples that communicate, for
example, their understanding of “awesome” or suggested pacing.
5.3.3 Suggestive Editing. In our analysis we saw how editors worked
to exclude clients from inuencing the editing process to reach
agreements. For example, they did not send half-done work, as
reported by [
27
,
36
], and we showed that they asked reviewers to
provide feedback as time-coded commentaries. Co-located editing
sessions with clients, as observed in [
28
], allowed participants to
try alternative cuts and choose between several options in short
iterations.
We propose including reviewers in the editing process, allowing
them to suggest edits to the video remotely. This can be done in the
form of lean video editing software that is stripped of most features
except those that would be appropriate for other participants.
For example, after a video editor cut the footage and rearranged
the shots into a movie, they would submit this raw draft to the
client. The client would open the draft video in the special soft-
ware, which would allow them to perform only selected actions –
previewing the video, rearranging the shots or removing one or
several shots in the timeline, and attaching commentaries on the
rationale behind the changes. The changes would be recorded as
suggested and then conveyed to the video editor, who would decide
to accept/decline the changes, or re-cut the video. After the cut had
been done, the video editor (or another person) would reach the
next iteration of editing, which would involve colour correction.
Upon receiving the new draft, the client would open it again in
the special software, but this time they would only be able to alter
the colour scheme (white balance, hue, brightness, contrast, etc.)
by applying predened colour lters. This could go on through
124

Unpacking Editorial Agreements in Collaborative Video Production IMX ’20, June 17–19, 2020, Cornella, Barcelona, Spain
other stages of video post-production, such as sound mastering,
application of graphics, subtitling and voice-over recording.
In this way, the client would have an opportunity to participate
in the video editing process and, possibly, alleviate the need to
search for proper wording to describe the desired video.
6 CONCLUSION
The interpretive exibility and inherent uncertainty of the video
production process imply many questions for further research and
many design possibilities. In this paper, we took a perspective
that was inspired by the sociology of knowledge and the so-called
EPOR in our study of the video production process. The intent was
to unpack current collaborative editing practices. We identied
several ways that participants in collaborative video editing use
mechanisms and strategies to reach editorial agreements. Both
professional video makers and their clients encounter trouble when
communicating about videos as it is highly demanding to nd the
expressive ability to describe ideas and handle the interpretive
exibility of a video fruitfully.
We suggested design paths that consider the interpretive exibil-
ity of video and existing closure mechanisms. These paths should
be viewed as possible general directions at a conceptual level. We
suggest that further elaborations should be investigated as concrete
implementations, where the proper balance between various re-
quirements can be unpacked. The proposed paths aim to inspire
design that can support collaboration in video editing and con-
tribute to the development of collaborative features in non-linear
video editing software.
ACKNOWLEDGMENTS
The study is a part of the project “Better Video workows via
Real-Time Collaboration and AI-Techniques in TV and New Me-
dia”, funded by the Research Council of Norway under Grant No.:
269790. We would like to express our gratitude to Vizrt’s Even
Normann, Valentina Caruso, Roger Saetereng, Brage Breivik and
Ronan Huggard.
REFERENCES
[1]
Susan J Ashford, Katleen De Stobbeleir, and Mrudula Nujella. 2016. To seek
or not to seek: Is that the only question? Recent developments in feedback-
seeking literature. Annual Review of Organizational Psychology and Organizational
Behavior 3 (2016), 213–239.
[2]
Tom Bartindale, Guy Schoeld, Clara Crivellaro, and Peter Wright. 2016. Trylm:
Situated support for interactive media productions. In Proceedings of the 19th
ACM Conference on Computer-Supported Cooperative Work & Social Computing.
ACM, 1412–1422.
[3]
Susanne Bødker and Clemens Nylandsted Klokmose. 2011. The human–artifact
model: An activity theoretical approach to artifact ecologies. Human–Computer
Interaction 26, 4 (2011), 315–371.
[4]
Juan Casares, A Chris Long, Brad A Myers, Rishi Bhatnagar, Scott M Stevens,
Laura Dabbish, Dan Yocum, and Albert Corbett. 2002. Simplifying video editing
using metadata. In Proceedings of the 4th conference on Designing interactive
systems: processes, practices, methods, and techniques. 157–166.
[5]
Soon Hau Chua, Toni-Jan Keith Palma Monserrat, Dongwook Yoon, Juho Kim,
and Shengdong Zhao. 2017. Korero: Facilitating complex referencing of visual
materials in asynchronous discussion interface. Proceedings of the ACM on
Human-Computer Interaction 1, CSCW (2017), 34.
[6] Harry M Collins. 1981. Stages in the empirical programme of relativism.
[7]
Paul Dourish. 2001. Process descriptions as organisational accounting devices:
the dual use of workow technologies. In Proceedings of the 2001 International
ACM SIGGROUP Conference on Supporting Group Work. ACM, 52–60.
[8]
Paul Dourish and Victoria Bellotti. 1992. Awareness and coordination in shared
workspaces.. In CSCW, Vol. 92. 107–114.
[9]
Arvid Engström, Oskar Juhlin, Mark Perry, and Mathias Broth. 2010. Temporal
hybridity: footage with instant replay in real time. In Proceedings of the SIGCHI
Conference on Human Factors in Computing Systems. ACM, 1495–1504.
[10]
Elihu M Gerson and Susan Leigh Star. 1986. Analyzing due process in the
workplace. ACM Transactions on Information Systems (TOIS) 4, 3 (1986), 257–270.
[11]
Frode Guribye and Lars Nyre. 2017. The Changing Ecology of Tools for Live
News Reporting. Journalism Practice 11, 10 (2017), 1216–1230.
[12]
Spencer H Harrison and Elizabeth D Rouse. 2015. An inductive study of feedback
interactions over the course of creative projects. Academy of Management Journal
58, 2 (2015), 375–404.
[13]
Cisco Visual Networking Index. 2017. Global mobile data trac forecast update,
2016–2021 white paper. Cisco: San Jose, CA, USA (2017).
[14]
Iris Jennes and Wendy Van den Broeck. 2017. The Social Construction of Tar-
geted Television Advertising: The Importance of Social Arrangements in the
Development of Targeted Television Advertising in Flanders. In Proceedings of the
2017 ACM International Conference on Interactive Experiences for TV and Online
Video. ACM, 41–50.
[15]
Oskar Juhlin, Arvid Engström, and Elin Önnevall. 2014. Long tail TV revisited:
from ordinary camera phone use to pro-am video production. In Proceedings of
the SIGCHI Conference on Human Factors in Computing Systems. ACM, 1325–1334.
[16]
Oskar Juhlin, Goranka Zoric, Arvid Engström, and Erika Reponen. 2014. Video
interaction: a research agenda.
[17]
Joy Kim, Mira Dontcheva, Wilmot Li, Michael S Bernstein, and Daniela Steinsapir.
2015. Motif: Supporting novice creativity through expert patterns. In Proceedings
of the 33rd Annual ACM Conference on Human Factors in Computing Systems.
ACM, 1211–1220.
[18]
David Kirk, Abigail Sellen, Richard Harper, and Ken Wood. 2007. Understanding
videowork. In Proceedings of the SIGCHI conference on Human factors in computing
systems. ACM, 61–70.
[19]
Klaus Krippendor. 2018. Content analysis: An introduction to its methodology.
Sage publications.
[20]
Kurt Luther, Jari-Lee Tolentino, Wei Wu, Amy Pavel, Brian P Bailey, Maneesh
Agrawala, Björn Hartmann, and Steven P Dow. 2015. Structuring, aggregating,
and evaluating crowdsourced design critique. In Proceedings of the 18th ACM
Conference on Computer Supported Cooperative Work & Social Computing. ACM,
473–485.
[21]
Jennifer Marlow, Scott Carter, Nathaniel Good, and Jung-Wei Chen. 2016. Be-
yond talking heads: multimedia artifact creation, use, and sharing in distributed
meetings. In Proceedings of the 19th ACM Conference on Computer-Supported
Cooperative Work & Social Computing. 1703–1715.
[22]
Neil Mercer. 1994. Neo-Vygotskian theory and classroom education. Language,
literacy and learning in educational practice (1994), 92–110.
[23]
Dean Mobbs, Nikolaus Weiskopf, Hakwan C Lau, Eric Featherstone, Ray J Dolan,
and Chris D Frith. 2006. The Kuleshov Eect: the inuence of contextual framing
on emotional attributions. Social cognitive and aective neuroscience 1, 2 (2006),
95–106.
[24]
Cuong Nguyen, Stephen DiVerdi,Aaron Hertzmann, and Feng Liu. 2017. CollaVR:
collaborative in-headset review for VR video. In Proceedings of the 30th Annual
ACM Symposium on User Interface Software and Technology. ACM, 267–277.
[25]
Gregory B Northcraft, Aaron M Schmidt, and Susan J Ashford. 2011. Feedback
and the rationing of time and eort among competing tasks. Journal of Applied
Psychology 96, 5 (2011), 1076.
[26]
Boris Novikov and Oleg Proskurnin. 2003. Towards collaborative video authoring.
In East European Conference on Advances in Databases and Information Systems.
Springer, 370–384.
[27]
Jasper O’Leary, Holger Winnemöller, Wilmot Li, Mira Dontcheva, and Morgan
Dixon. 2018. Charrette: Supporting In-Person Discussions around Iterations
in User Interface Design. In Proceedings of the 2018 CHI Conference on Human
Factors in Computing Systems. 1–11.
[28]
Amy Pavel, Dan B Goldman, Björn Hartmann, and Maneesh Agrawala. 2016.
VidCrit: video-based asynchronous video review. In Proceedings of the 29th Annual
Symposium on User Interface Software and Technology. ACM, 517–528.
[29]
Charles Sanders Peirce. 1902. Logic as semiotic: The theory of signs. Philosophical
writings of Peirce (1902), 100.
[30]
Mark Perry, Mathias Broth, Arvid Engström, and Oskar Juhlin. 2019. Visual
Narrative and Temporal Relevance: Segueing Instant Replay into Live Broadcast
TV. Symbolic interaction 42, 1 (2019), 98–126.
[31]
Mark Perry, Oskar Juhlin, Mattias Esbjörnsson, and Arvid Engström. 2009. Lean
collaboration through video gestures: co-ordinating the production of live tele-
vised sport. In Proceedings of the SIGCHI Conference on Human Factors in Com-
puting Systems. ACM, 2279–2288.
[32]
Julien Phalip, Ernest A Edmonds, and David Jean. 2009. Supporting remote
creative collaboration in lm scoring. In Proceedings of the seventh ACM conference
on Creativity and cognition. ACM, 211–220.
[33]
Julien Phalip, Matthew Morphett, and Ernest Edmonds. 2007. Alleviating commu-
nication challenges in lm scoring: an interaction design approach. In Proceedings
of the 19th Australasian conference on Computer-Human Interaction: Entertaining
User Interfaces. ACM, 9–16.
125

IMX ’20, June 17–19, 2020, Cornella, Barcelona, Spain Okopnyi et al.
[34]
Trevor J Pinch and Wiebe E Bijker. 1984. The social construction of facts and
artefacts: Or how the sociology of science and the sociology of technology might
benet each other. Social studies of science 14, 3 (1984), 399–441.
[35] Michael Quinn Patton. 2002. Qualitative research and evaluation methods.
[36]
Daniela Retelny and Pamela Hinds. 2016. Embedding Intentions in Draw-
ings: How Architects Craft and Curate Drawings to Achieve Their Goals. In
Proceedings of the 19th ACM Conference on Computer-Supported Cooperative
Work & Social Computing (San Francisco, California, USA) (CSCW ’16). As-
sociation for Computing Machinery, New York, NY, USA, 1310–1322. https:
//doi.org/10.1145/2818048.2819932
[37]
Kjeld Schmidt and Liam Bannon. 1992. Taking CSCW seriously. Computer
Supported Cooperative Work (CSCW) 1, 1-2 (1992), 7–40.
[38]
Lucy Suchman. 1993. Do categories have politics? The language/action perspec-
tive reconsidered. In Proceedings of the Third European Conference on Computer-
Supported Cooperative Work 13–17 September 1993, Milan, Italy ECSCW’93.
Springer, 1–14.
[39]
Naimdjon Takhirov and Fabien Duchateau. 2011. A cloud-based and social
authoring tool for video.. In ACM Symposium on Document Engineering. 65–68.
[40]
Lucia Terrenghi, Torsten Fritsche, and Andreas Butz. 2008. Designing environ-
ments for collaborative video editing. (2008).
[41]
David Wood, Jerome S Bruner, and Gail Ross. 1976. The role of tutoring in
problem solving. Journal of child psychology and psychiatry 17, 2 (1976), 89–100.
[42]
Alvin Yuan, Kurt Luther, Markus Krause, Sophie Isabel Vennix, Steven P Dow,
and Bjorn Hartmann. 2016. Almost an expert: The eects of rubrics and expertise
on perceived value of crowdsourced design critiques. In Proceedings of the 19th
ACM Conference on Computer-Supported Cooperative Work & Social Computing.
ACM, 1005–1017.
[43]
Jamie Zigelbaum, Michael S Horn, Orit Shaer, and Robert JK Jacob. 2007. The tan-
gible video editor: collaborative video editing with active tokens. In Proceedings
of the 1st international conference on Tangible and embedded interaction. ACM,
43–46.
126