Content uploaded by Johannes Schöning
Author content
All content in this area was uploaded by Johannes Schöning on Nov 13, 2017
Content may be subject to copyright.
It’s All Around You: Exploring 360°Video
Viewing Experiences on Mobile Devices
Marc Van den Broeck
Nokia Bell Labs
Antwerp, Belgium
marc.van_den_broeck@
nokia-bell-labs.com
Fahim Kawsar
Nokia Bell Labs
Cambridge, UK
fahim.kawsar@nokia-bell-labs.com
Johannes Schöning
University of Bremen
Bremen, Germany
schoening@uni-bremen.de
ABSTRACT
360
°
videos are a new kind of medium that gives the viewers a
sense of real immersion as they glimpse the action from all angles
and directions. Naturally, professional and amateur lm-makers are
actively adopting this new medium for transformative storytelling.
Despite this phenomenal progress in 360
°
video creation, current un-
derstanding on users’ viewing experience of these videos is limited.
In this paper, we present the rst comparative study on the user
experience with 360
°
videos on mobile devices using dierent inter-
action techniques. We observed 18 participants’ interaction with six
360
°
videos with dierent viewport characteristics (static or mov-
ing) on a smartphone, a tablet and a head mounted display (HMD)
respectively and measured how they interact with the content. We
then conducted semi-structured interviews with the participants
in which they explained their interaction with and viewing ex-
perience of 360
°
videos across three devices. Our ndings show
that 360
°
videos with moving viewports elicit higher engagement
from the viewers, and oer superior viewing experience. However,
these videos are cognitively demanding and require constant user
attention. Our participants preferred the condition with dynamic
peephole interaction on a smartphone for watching 360
°
videos
due to the simplicity in exploration and familiarity with navigation
controls. Many participants reported that the HMD oers the most
immersive experience however it comes at the expense of higher
cognitive burden, motion sickness and physical discomfort.
CCS CONCEPTS
•Human-centered computing →Human computer interac-
tion (HCI);User studies;Interaction techniques;
KEYWORDS
360
°
video; Head Mounted Displays (HMDs); Mobile Devices; Video
Consumption; User Experience
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 prot 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 ACM
must be honored. Abstracting with credit is permitted. To copy otherwise, or republish,
to post on servers or to redistribute to lists, requires prior specic permission and/or a
fee. Request permissions from permissions@acm.org.
MM’17, October 23–27, 2017, Mountain View, CA, USA.
© 2017 Association for Computing Machinery.
ACM ISBN 978-1-4503-4906-2/17/10. . . $15.00
https://doi.org/10.1145/3123266.3123347
1 INTRODUCTION & MOTIVATION
360
°
video is a rapidly growing new kind of medium that allows
viewers to sense the action from all angles and directions. These
videos oer viewers a unique experience as they feel as if they are
actually in the action, as it unfolds before their eyes.
Naturally, lm-makers are actively adopting 360
°
videos to trans-
form storytelling and make viewers engaged with the content more
emotionally. Such captivating storytelling is naturally appealing
for the professional lm industry. However, due to the availabil-
ity and advancement of low-cost camera equipment, 360
°
videos
are now becoming a mainstream medium for engrossing and im-
mersive journalism [
1
], brand advertisement [
11
], and live sports
[
12
]. Recognising the potential, on-line platforms such as YouTube,
Facebook, Vimeo among others are now actively supporting 360
°
video formats. Furthermore, YouTube has set up an on-line studio
that oers various resources so that anyone can produce their omni
directional lms. While the creation process of a 360
°
video has
improved greatly in recent years, this technology has yet to nd the
appropriate, ergonomically usable and emotionally engaging
delivery mechanism that resonates with a mass audience.
It is natural to expect that a 360
°
video is most immersive when
experienced through a virtual reality device, e.g., a head-mounted
display (HMD). However, there has not been many systematic ex-
ploration to assess whether HMDs would oer rich experiences
with 360
°
videos and paves the way for a wider adoption. Recently,
multiple mobile platforms have enabled support for 360
°
videos. For
example, the YouTube application on both iOS and Android now
supports watching and exploring 360
°
videos though a dynamic
peep-hole exploration technique [
22
] or touch-based navigation.
Despite this progress, very few studies have reported a reection
on user experience with these devices while watching 360
°
videos
or more generally the delivery modality that evokes emotional en-
gagement with 360
°
content - whether a personal story or artistic
endeavours.
In this work, we examine two specic aspects of this larger
puzzle, namely i) How dierent viewport characteristics of a
360
°
video inuence viewers’ behaviour, and ii) which device
viewers prefer to watch a 360
°
video? Viewport composition
can be static or moving. In the former static viewport (SVP) case
the cameras are statically placed in the centre of the scene while
shooting a 360
°
video, where as in the later moving viewport (MVP)
case, the cameras move while recording the video. We hypothesise
that this composition dynamics will have a profound inuence on
how a viewer will explore and experience 360
°
video scenes and
the overall video. Exploration indicates the degree of user activity
of looking around in the 360
°
video to discover more content. To
this end, we quantitatively logged 18 users’ exploration patterns
on a smartphone, a tablet and an HMD respectively while watch-
ing six 360
°
videos produced with both static viewport (SVP) or
moving viewport (MVP)
1
We then interviewed the participants to
understand their subjective experience with these videos on dif-
ferent devices taking into account exploration, navigation, and
immersion factors.
Our results show that although MVP videos are cognitively
demanding, they oer richer and engaging viewing experience
than SVP videos. Dynamic peep-hole interaction oered in modern
smartphones is preferred over HMDs due to the familiarity and
simplicity in content exploration. Concerning immersion, HMDs
oer the best user experience; however, this immersive experience
comes at the expense of demanding cognitive resources and physi-
cal discomfort.
2 RELATED WORK
A number of past studies have reported user experience with mo-
bile video consumption exploring impact of content, environment,
device interaction, and modalities. Our work draws on these past
research and focus primarily on assessing 360
°
videos viewing ex-
periences. In the following, we position our work with respect to
these areas of related research.
In one of the earliest studies on mobile video consumption prac-
tices, O’Hara et al [
24
] reported the dierent ways mobile videos
are integrated into people’s everyday lives. This rich ethnographic
study identied motivations and values that drive mobile video
consumption and elicit engaging social interactions in dierent
socio-dynamic contexts. As omni-directional content started ap-
pearing, Benko et al. [
2
] explored multi-point interactions with
immersive omni directional visualisations in a dome. Bleumers [
3
]
and also Zoric et al. [
37
] took this research into the living room
and further studied the consumption of 360
°
videos on static TVs.
Similar to the work of Zoric et al [
37
], Pece [
26
] identied chal-
lenges that come with this new interactive content, e.g., balancing
active and passive viewing, enabling orientation in the panoramic
image space and supporting both individual exploration and social
conversations around it. These studies uncovered a set of concrete
usability challenges that arise while consuming omnidirectional
content in dierent mobility settings.
A number of recent research explored interaction experiences
with panoramic videos. Kasahara et al. [
18
] studied the viewing
experience of rst person omnidirectional video. Huber et al. [
16
]
further explored dierent user interfaces for mobile video brows-
ing and later explored eective interaction with 360°videos using
HMDs [
27
], and Jumisko-Pyykkö et al. [
17
] researched the user
experience when watching a mobile 3D video in a car. These work
highlighted the challenges that device form and interaction aor-
dance put on users’ viewing experience of omnidirectional videos.
Our work is shaped by these ndings, as we extrapolate on their
suggestions. However, we investigate the viewport dynamics and
device preferences instead of interaction dynamics.
Concerning content and immersion, Fonseca et al. recently re-
ported a study that shows HMD oers superior immersion and
can trigger a higher emotional response from captivating content
1.
than other handhelds [
10
]. In a similar endeavour, Ramalho et
al. described the eect of multi-sensory immersion on engaging
and emotional user experience with 360
°
videos [
30
]. These work
primarily investigated the emotional dynamics in content consum-
mations. In contrast, our work looked at viewport composition
dynamics and device preferences independent of the content and
its emotional appeal.
Besides research on content consumption experience when watch-
ing videos, there is a large body of research on interaction tech-
niques and modalities. Peephole interaction [
29
] with mobile de-
vices was heavily studied e.g., by Rohs et al. [
31
,
32
] and recently
repeated in dierent contexts [
8
]. Many research also looked at con-
trollers [
6
,
20
] or device types [
13
,
25
,
34
] on immersion in games
and other VR worlds. Already in 1999 Pierce et al. [
28
] published
guidelines for successful HMD-based experiences. Current HMD
manufacturers are pushing hard to close the gap between simula-
tion and reality as demonstrated by previous research [
21
], e.g., by
introducing new hardware (e.g., the Oculus Rift consumer version)
or algorithms (e.g., Mayo Clinic’s Galvanic Vestibular Stimulation
technology
2
). While this body of research is very relevant to the
work presented here, we intend to understand device dynamics
with respect to 360
°
videos. Perhaps, the closest research to ours
was reported by Boonsuk and his colleagues, where they evaluated
the optimal interface for panoramic 360
°
videos [
4
]. Their primary
focus was on uncovering the degree of spatial cognition that a 360
°
video demands. In contrast we look at the mobile user experience
with 360°videos.
3 USER STUDY
We conducted a user study to investigate the viewport as well as
the device dynamics of 360
°
videos in more detail. We explored
the interaction with and viewing experience of two dierent type
of 360
°
videos on three dierent devices with dierent interaction
modalities.
3.0.1 Participants. We recruited a group of 18 people from
our research facilities (9 males, 9 females, age range 20 – 54). For
recruiting, we used stratied sampling with snowball sampling
within each stratum. One participant could not complete the ex-
periment due to cyber-sickness while wearing the HMD. So we
recruited another participant of the same gender for completeness.
All participants owned a mobile device (a smartphone or a tablet
or both), but none of them had experiences with actively watching
360°videos.
3.0.2 Apparatus. We used a 5.1 inch Samsung Galaxy S6 as
the smartphone on which the 360
°
videos were navigated using
a dynamic peephole navigation [
22
], a 10.1 inch Samsung Galaxy
Tab 2 as the tablet on which the 360°videos were navigated using
touch as an input modality, and Samsung Gear VR as the HMD on
which the content can be explored using full body input [35].
All study participants watched two 360
°
videos one with a static
viewport (SVP) and one with a moving viewport (MVP) on each of
the three dierent mobile devices. This resulted in six 360
°
videos
for each participant. In the smartphone and HMD conditions we
2http://www.vmocion.com
Type Name Narration
SVP Waldo[9] Guided exploration of a scene to look out for a person named Waldo [3:53 min].
SVP Refugee[33] Short documentary about the refugee crisis in Europe [3:16 min].
SVP Horror[19] Japanese horror movie [3.02 min].
MVP Dive[7] Dive with sharks near the sunken wreckage of the Ray of Hope [4:10 min].
MVP Mountain[14] Last steps to the Mont Blanc summit [3:17 min].
MVP Car[23] Convertible drive on the Pacic Coast Highway [2.16 min].
Table 1: List of static viewport 360°videos (SVP) and moving viewport 360°videos (MVP) used in the user study.
logged the 2D orientation data from the built-in gyroscope of Sam-
sung Galaxy S6, and on the tablet, we logged the touch (and scroll)
input on the screen using a custom logger application. We use
the orientation data to study the degree of 2D exploration in the
smartphone and HMD condition. For the tablet, we analysed the
scroll movements with respect to the screen resolution to acquire
participant’s degree of exploration on the 2D plane.
3.0.3 Video Selection & Characteristics. The six 360
°
videos
(professional as well as semi-professional videos) used in the study
were selected from the 360
°
videos YouTube channel
3
. All 360
°
videos were ranked within the top 50 trending 360
°
videos in this
channel in March 2016.
First, we went through all videos and grouped them by their
main viewport characteristics. Independent from the content shown
in the videos about 40% of the 360
°
videos mainly used one or
multiple static viewports (SVP) (i.e. the cameras were placed on a
stand in the centre of the scene to record the 360
°
video ), whereas
the rest mainly used moving viewports (MVP) (i.e. the cameras
were moved while recording the video). Most of the videos also
combined scenes with SVP followed by scenes with MVP and vice
versa. Therefore we rst selected those 360
°
videos from the list
that mainly used one VP throughout the whole video and were
about 3–4 minutes long. Videos that were changing the viewport
dynamics between SVP and MVP multiple times were removed as
possible study video candidates. We excluded extreme MVP videos,
which rapidly alter the VP in 3D dimensions (e.g., roller coaster
rides or ghter jet ights) to avoid cyber sickness of our participants.
Naturally, the content of MVP 360
°
videos is dierent to the content
of SVP 360
°
videos; as such we carefully assessed the content of the
videos and selected only those videos with comparable content but
having dierent viewport characteristics. Table 1 lists all the videos
and their links that we have used in the study after following the
selection procedure mentioned above.
3.0.4 Design. Our experiment followed a 3x2 factorial within-
subject design with the dierent devices (tablet, mobile, HMD) and
the viewport characteristics (SVP, MVP) as the factors. The three
videos for each viewport were randomly assigned to the conditions.
The order of device condition was counterbalanced and presented in
blocks. For example, both videos (SVP, MVP or vice versa) with the
HMD happened in one block without allowing the user to switch
to another method. With 18 subjects the test application on the
devices recorded 108 trials.
3https://goo.gl/NLNr7H
3.0.5 Procedure. Initially, participants were given a written
task description and were asked about their demographics and de-
vice experiences. Next, the participants tested each device condition
while seated in a revolving chair in front of a table. We used a 360
°
video with an SVP in this training phase [36].
After each device condition, the participants were requested to
complete a SUS questionnaire
4
. After completing all conditions, we
conducted semi-structured interviews with the participants and had
them rate the quality of the 360
°
videos as well as their preferred
device. To stimulate discussion of possible reasons for selecting a
particular type of device, we presented the participants with three
possible factors: exploration, navigation, and immersion. Later, we
analysed data by coding the selection reasons against each device.
Each interview was audio-recorded for later analysis. The total
time each participant took was about 50 minutes.
4 RESULTS
In this section, we discuss the study results from two main perspec-
tives. First, we discuss participants’ viewing experience with 360
°
videos with dierent viewport characteristics. Then, we discuss
which devices were preferred for watching 360°videos.
4.1 Understanding Viewport Dynamics
First, we assessed how dierent viewport compositions inuence
viewers experience independent from the content shown in the
videos. In particular, we were interested in answering whether dif-
ferent viewports demand dierent exploration patterns. We noticed
that MVP videos, no matter what content was shown, consistently
require more exploration from viewers than SVP videos for all three
devices. On average, all participant’s exploration rate with MVP
videos was 17.3
°
per 30 seconds, on the contrary for SVP it was
14.6
°
per 30 seconds. We run a paired samples t-test to evaluate the
dierence in the exploration rate on viewports across all partici-
pants and found signicant dierence in the exploration rate for
SVP and MVP (t(53)=−12.933 and p< .01).
As an illustration of this observation, we randomly picked two
participants’ trace for two videos with varying viewports across
three devices from our sample and plotted their exploration be-
haviour in Figure 1.
This observation is counterintuitive as the natural expectation
is that when the viewport is static viewers will explore more to
4
The System Usability Scale (SUS) is a technology agnostic survey that are used to
assess the usability of a variety of products or services. It is composed of ten statements
contributing to a single score ranging from 0 to 100 [5].
Figure 1: Two randomly selected participants’ exploration
behaviour with SVP and MVP videos in the mobile, tablet
and HMD conditions. Here, exploration indicates the degree
of user activity of looking around in the 360°video to dis-
cover more content. We notice that SVP videos typically
have an initial exploration phase, in contrast MVP videos
are explored throughout.
experience the action from all angles and directions. However, our
results suggest the opposite, i.e.; viewers explore the viewports as
they keep changing instead of focusing on the moving action frame.
Qualitative feedback from the participants shed some light on
this phenomenon. A majority (14) of the participants mentioned
that static viewport becomes monotonous after several seconds
(5-10s) of initial exploration while moving viewport constantly
encourages them to explore the action frames from dierent angles.
In the interview one participant commented:
“SVP are boring. You look around, and then you
are done.. . . " (P5)
Similar comments were received from other participants (P1, P8,
P17).
Also, an additional promise of 360
°
videos is that the viewing ex-
perience is dierent for dierent individuals. Therefore, we looked
at the exploration behaviour at individual scale. We refer to Figure
1 again here, in which a pair of randomly picked participants watch-
ing the same 360
°
video in the same device
5
. We observe that for
all the cases, the same video was explored dierently by dierent
participants. For example, while watching the
Horror
video on
mobile the two participants’ exploration rate was 13
°
and 19
°
per 30
5For readability we present the rst 135 seconds of each video.
seconds respectively; for
Dive
on HMD these rates were 21
°
, and 31
°
per 30 seconds respectively. Similar results were observed for other
pairs. Indeed, we found that across all participants, their degrees of
exploration were not correlated (Pearson’s
r=−
0
.
28
,
0
.
14
,−
0
.
12
for mobile, tablet and HMD respectively, and
p> .
01 for the same
movie).
Another interesting observation was that SVP videos elicit explo-
ration in short bursts followed by a short period of no exploration.
This is particularly the case at the beginning 10 to 30 seconds of the
videos during which viewers initially explore the viewport (e.g., as
in
Horror
or
Waldo
) before settling on to the centre action frame
with occasional exploration.
Although MVP videos are more appealing for exploration, it
comes at the expense of demanding user attention, which often
causes confusion and even frustration among the viewers. This
is because as the viewport changes, viewers want to experience
the 360
°
video from all angles, and in the process sometimes they
miss interesting actions, or they become confused. In particular,
while watching a video on an HMD that oers a high degree of
immersion, such viewport changes makes viewers confused. One
specic comment was:
“You are navigating in the scene and all of a sud-
den the entire scene changes. That is confusing.. . ."
(P1)
Another remark was:
“ I looked left, then I looked right. When I looked
left again , the scene was dierent .. . " (P17)
One interesting situation that we encountered multiple times
in our study was that the sudden transition of viewports without
adequate transition cues in a video caught participants by surprise
while using a HMD (e.g., as in
Car in 0:31 min
). Due to the
high immersion factor, if a scene transition happens unexpectedly
it causes physical distress to viewers. As such the lmmakers need
to provide eective and artistic cues either visual or auditory, to
avoid such situations in a 360°video .
Videos are primarily about the lmmaker conveying a message,
with artistic composition of action frames to construct the most
compelling story. However, given the freedom of exploration that
the viewers possess in 360
°
videos we concur that directed sto-
rytelling would be dicult without implicit or explicit guidance.
This is especially essential for MVP videos as viewers frequently
explore the viewports, and in the process, they might miss the most
interesting frames in the video. Multiple remarks (11) from our
participants emphasised this fact. One participant mentioned:
“ In a regular video, the viewport is picked by a pro
lmmaker. Now I might miss out on something... . "
(P11)
Although these concerns with MVP based 360
°
videos, our par-
ticipants rated MVP videos higher than SVP videos, as shown in
Figure 2(a,b).
Diving
and
Car
received the highest ratings (both
are MVP videos), followed by
Refugee
(SVP),
Mountain
(MVP),
Horror
(SVP) and
Waldo
(SVP). On an average 62% of the partici-
pants rated MVP videos as very good or good as opposed to 32%
rating SVP videos as very good or good. A Chi-square test also con-
rmed that participants’ rating towards 360
°
videos signicantly
Figure 2: (a) Participants’ subjective rating of dierent 360°
videos and (b) comparison of SVP and MVP based videos
diered by viewport dynamics (
χ2(
4
,N=
108
)=
16
.
088 and
p< .01).
Looking further into this through subjective data, we found those
participants preferred MVP videos primarily due to high immersion,
excitement and emotionally engaging experience. On the contrary,
our participants (17) constantly remarked that SVP videos are only
good for documentaries, that are slow paced, explorative and does
not demand high attention (e.g., Refugee). One comment was:
“ For documentaries, I see a plus for static view-
ports and for all the rest you miss the fun... . " (P18)
4.2 Understanding Device Dynamics
In this section, we shift our focus on devices, and our objective is
to answer - which device participants prefer to watch a 360
°
video
and why. 55% of the participants selected a smartphone as their
preferred device to watch a 360
°
video followed by a tablet (28%)
and an HMD (17%). While their explicit ratings clearly showed a
preference towards a smartphone, the SUS scores as illustrated in
Figure 3(a) for the three devices were less conclusive - smartphone
with
µ
: 75
.
83
,σ
: 9
.
63, tablet with
µ
: 75
.
28
,σ
: 12
.
03 and HMD
with
µ
: 66
.
53
,σ
: 13
.
15. Hence, we conducted one-way repeated
measure ANOVA on the SUS means of the groups and observed
a statistically signicant dierence (
F(
2
,
34
)=
3
.
585
,p< .
05)
across the groups.
To examine this result further, we looked at three factors of
these devices - exploration, navigation and immersion. We con-
ducted a two-way contingency table analysis to see if these fac-
tors inuenced participants’ choices about using a specic de-
vice for watching 360
°
videos, and found a signicant association
(
χ2(
4
,N=
54
)=
25
.
123 and
p< .
001). Figure 3 (b) shows how
these factors inuenced participants’ device preference.
Indeed, our interviews revealed that participants felt most com-
fortable with the mobile device due to the simplicity of the peephole
exploration and familiarity with navigation technique (stop, fast
Figure 3: (a) SUS scores of mobile, tablet and HMD and (b) dif-
ferent factors inuencing participants’ preference towards
a device
forward, etc.) in a mobile. However, multiple comments (7) were
raised regarding the ergonomic diculties faced with the mobile,
as participants mentioned that they wouldn’t be keeping their arm
raised for a long time. Tablets were primarily preferred due to the
familiarity and its simple navigation controls.
However, it was mentioned (11) that since one has to click and
drag around the screen to see dierent views, 360
°
videos are not
immersive in a tablet, put extra burden and as a result the charm
wears o quickly. HMD by far was mentioned as the most immersive
device for watching 360
°
videos. In fact, 89% of the participants
mentioned that HMD gives them a completely dierent sensation,
and they feel as if they’re actually in the scene as it unfolds before
their eyes. One particular comment was:
“You are isolated from external elements. I no
longer thought about the room I’m in . . . " (P1)
Also, with HMD users can explore the viewport more naturally.
Multiple participants (4) mentioned that the exploration comes
naturally to them in a HMD. We explicitly looked at participants’
exploration characteristics over three devices. On average a partici-
pant exploration rate with HMD was 28
°
per 30 seconds, followed by
mobile (21
°
per 30 seconds) and tablet (9
°
per 30 seconds). one-way
repeated measure ANOVA test conrmed that that participants’
degree of exploration 360
°
videos signicantly diered by device
types (F(2,34)=4.745,p< .01)
This result clearly suggests that a viewer will explore a 360
°
video more freely on a HMD. However, these natural exploration
and immersion come at a high cost of poor ergonomics. All of our
participants mentioned that the HMD was too heavy to wear, and
they felt physically stressed. As we mentioned, one of participant
quit the study after 30 seconds of wearing the HMD as she felt
sick caused by motion, pixelated screen, and weight of the HMD.
Furthermore, as mentioned earlier, sudden viewport changes with-
out eective cues also contributes to cyber-sickness. Albeit these
constraints, several of our participants (9) mentioned that if the
weight of the HMD and screen resolution are improved, they would
prefer HMD for watching 360°videos.
It was clear from our study that tablet (and as a generalisation a
device with static position) is not immersive enough to emotionally
engage the audiences with 360
°
videos. While, the mobility oered
by mobile and HMD brings the engagement higher, it also demands
dedicated physical setting, i.e., the very nature of the way a viewer
explore the viewports of a 360
°
video would require rotatable seating
furniture. This aspect was highlighted by multiple participants (11).
One particular comment was:
“We need new furniture. I can’t see myself using
this in my couch. I need a couch that can turn.. . . "
(P7)
Similar comments were received from other participants. How-
ever, this personalised physical setting also has interesting social
consequences. As one participant commented:
“With a box on your head, you miss the social as-
pect of watching a movie together, eating a snack,
etc.. . . (P18)
Videos - whether they are movies, sports, ads or otherwise -
play a critical role as a social trigger to bring families, friends, and
colleagues together by oering them a shared experience. With
the advent of 360
°
videos and consequent demand for personalised
setups, we see an interesting dynamics emerging here. With per-
sonalised rotatable seating arrangements, immersive content, and
improved consumption device viewers will experience 360
°
videos
uniquely. However, this engaging experience comes at the expanse
of time spent on the social interactions while enjoying a video.
5 DISCUSSION & CONCLUSIONS
The study results gave us an illuminating picture of participants’
experience of watching dierent 360
°
videos with various mobile
devices. We briey highlight a set of implications here that emerged
from our study.
360°videos need guidance:
Our participants preferred MVP
360
°
videos over SVP videos. In addition, MVP 360
°
videos
triggered more interaction with the content by the partic-
ipant. Therefore, we recommend using MVP, if the lm-
makers prefer to have the users interacting with the content.
Even so, we observed and initial 10 to 30 seconds exploration
phase in the SVP 360
°
videos , we recommend to guide fur-
ther interaction explicitly (e.g., as in
Waldo
) or implicitly
(e.g., as in Horror) after the initial exploration phase.
Scene transitions need careful composition:
MVP 360
°
videos are cognitively more demanding. Producers should
carefully guide users attention to avoid abrupt VP changes
(e.g., as in
Car
in 0:31 min.) This is particularly true for de-
vices that oer a high degree of immersion (e.g., an HMD).
In these cases careful VP changes are one important factor to
reduce and avoid cyber-sickness. Similar as for the SVP 360
°
videos we recommend guiding users with visual or auditory
cues to avoid bad VP changes. We have observed that videos
with VP changing supported with one or multiple cues (e.g.,
in
Mountain
with the path and in
Car
with the road) are
in general cognitively less challenging than free and fast
VP changes in 3D space (e.g., as in roller coaster rides etc.).
Therefore, we would also recommend to cut dierent ver-
sions of one 360
°
video with especially trimmed VP changes
for the use on an HMD or mobile with dynamic peephole
interaction.
360°videos need new entertainment constellation:
Current HMDs are not ready yet to allow a smooth video
viewing experience. Multiple technical challenges (e.g., screen
resolution, form factor, weight distribution, etc.) need to be
addressed to provide a better viewing experience. Even so,
the 360
°
videos used in the study were just about three min-
utes long, the participants complained about the clumsiness
of the HMD. Nevertheless, the HMD provides by far the
highest degree of immersion. Overall, the dynamic peephole
interaction technique on the smartphone provided the best
viewing experience. Some participants also recommend to
use a bigger device (e.g., the tablet) for the interaction, if the
device is not heavy. In both cases (HMD and smartphone)
participants also argued that both interactions would re-
quire rotatable new seating furniture in their living rooms.
Past research has also uncovered similar insights highlight-
ing swivel chairs oer superior viewing experience [
15
].
Indeed, we are already observing such new entertainment
constellation emerging commercially for 360
°
videos, e.g.,
VR Cinema6.
The results and observation presented in this paper should be
taken in the light of the study conditions and the devices and thus
may not necessarily be generalisable to all situations across all
360
°
videos. For example, our experimental settings did not include
combination of SVP and MVP videos. We have reported both view-
port compositions have advantages and disadvantages. In future
we want to study the eect of combined viewport on users’ view-
ing experience. Another limitation of our work is the content bias.
Given, the selected videos are chosen from YouTube without any
control on the story and emotional appeal, we have not quantied
the content inuence. However, we consider this is a fantastic route
to follow in the future with respect to 360
°
videos and in particular
assess the relationship of emotional attributes and users’ viewing
experience. Another potential area of exploration is the impact of
users’ posture and dierent environment constellations. In future
work we also want to study the auditory component that naturally
adds to the overall 360°video viewing experience.
To conclude, our results suggest that 360
°
videos with moving
viewports trigger more exploration and oer superior viewing ex-
perience across all devices. However, these videos demand higher
user attention, and can often cause motion sickness if viewport
transitions are not composed carefully. Especially when watched
on HMDs. The most interesting nding of our study is the fact that
our participants preferred peephole interaction on a smartphone
for consuming 360°videos.
While the general wisdom is that the HMD is the most suitable
device for 360
°
videos, our study suggests that HMD is not ready yet
to resonate with a mass audience due to higher cognitive burden,
motion sickness and physical discomfort even so HMDs clearly
provide the highest degree of immersion. Taken together these and
the rest of our ndings oer subtle guidelines towards immersive
and dynamic visual experience with 360
°
videos on mobile devices.
6 ACKNOWLEDGEMENTS
This research was supported by SeRGIo, an icon project realised
in collaboration with IMEC, with project support from VLAIO
(Flanders Innovation & Entrepreneurship) and by the Volkswagen
foundation through a Lichtenbergprofessorship.
6https://thevrcinema.com
REFERENCES
[1]
BBC. 2016. BBC 360 Degree Channel. (2016). https://goo.gl/cseiFj Last accessed:
March 31, 2016.
[2]
Hrvoje Benko and Andrew D. Wilson. 2010. Multi-point Interactions with Immer-
sive Omnidirectional Visualizations in a Dome. In ACM International Conference
on Interactive Tabletops and Surfaces (ITS ’10). ACM, New York, NY, USA, 19–28.
[3]
Lizzy Bleumers, Wendy Van den Broeck, Bram Lievens, and Jo Pierson. 2012.
Seeing the Bigger Picture: A User Perspective on 360 degree TV. In Proceedings
of the 10th European Conference on Interactive Tv and Video (EuroiTV ’12). ACM,
New York, NY, USA, 115–124.
[4]
Wutthigrai Boonsuk, Stephen Gilbert, and Jonathan Kelly. 2012. The Impact of
Three Interfaces for 360-degree Video on Spatial Cognition. In Proceedings of the
SIGCHI Conference on Human Factors in Computing Systems (CHI ’12). ACM, New
York, NY, USA, 2579–2588. https://doi.org/10.1145/2207676.2208647
[5]
John et al. Brooke. 1996. SUS-A quick and dirty usability scale. Usability evaluation
in industry 189, 194 (1996).
[6]
Paul Cairns, Jing Li, Wendy Wang, and A Imran Nordin. 2014. The inuence of
controllers on immersion in mobile games. In Proceedings of the SIGCHI Conference
on Human Factors in Computing Systems. ACM.
[7]
Discovery Channel. 2016. MythBusters: Shark Shipwreck. (2016). https://goo.gl/
lGJvyh Last accessed: March 31, 2016.
[8]
Ashley Colley, Wouter Van Vlaenderen, Johannes Schöning, and Jonna Häkkilä.
2016. Changing the camera-to-screen angle to improve AR browser usage. In
Proceedings of the 18th International Conference on Human-Computer Interaction
with Mobile Devices and Services. ACM, 442–452.
[9]
Corridor Digital. 2016. Where’s Waldo 360
°
. (2016). https://goo.gl/T8gPQP Last
accessed: March 31, 2016.
[10]
Diana Fonseca and Martin Kraus. 2016. A Comparison of Head-mounted and
Hand-held Displays for 360&Deg; Videos with Focus on Attitude and Behavior
Change. In Proceedings of the 20th International Academic Mindtrek Conference
(AcademicMindtrek ’16). ACM, New York, NY, USA, 287–296. https://doi.org/10.
1145/2994310.2994334
[11]
JR Futrell. 2016. See the future of video by looking behind you: Introducing 360
degree video ads. (2016). http://goo.gl/n8wISz Last accessed: March 31, 2016.
[12]
John Gaudiosi. 2016. This company streams live sports events to virtual reality.
Here is how. (2016). http://goo.gl/Q9lq79 Last accessed: March 31, 2016.
[13]
Vivian Genaro Motti and Kelly Caine. 2014. Understanding the wearability of
head-mounted devices from a human-centered perspective. In Proceedings of the
2014 ACM International Symposium on Wearable Computers. ACM.
[14]
Google. 2016. Mont Blanc Street View - 360
°
video of the Mont Blanc Summit.
(2016). https://goo.gl/sfdPSI Last accessed: March 31, 2016.
[15]
Jan Gugenheimer, Dennis Wolf, Gabriel Haas, Sebastian Krebs, and Enrico Rukzio.
2016. SwiVRChair: A Motorized Swivel Chair to Nudge Users’ Orientation for
360 Degree Storytelling in Virtual Reality. In Proceedings of the 2016
CHI Conference on Human Factors in Computing Systems
(CHI ’16). ACM, New York, NY, USA, 1996–2000. https://doi.org/10.
1145/2858036.2858040
[16]
Jochen Huber, Jürgen Steimle, and Max Mühlhäuser. 2010. Toward more e-
cient user interfaces for mobile video browsing: an in-depth exploration of the
design space. In Proceedings of the international conference on
Multimedia. ACM, 341–350.
[17]
Satu Jumisko-Pyykkö, Mandy Weitzel, and Dominik Strohmeier. 2008. Designing
for user experience: what to expect from mobile 3D TV and video?. In Pro-
ceedings of the 1st international conference on Designing
interactive user experiences for TV and video. ACM, 183–192.
[18]
Shunichi Kasahara, Shohei Nagai, and Jun Rekimoto. 2015. First Person Omni-
directional Video: System Design and Implications for Immersive Experience.
In Proceedings of the ACM International Conference on In-
teractive Experiences for TV and Online Video (TVX ’15).
ACM, New York, NY, USA, 33–42. https://doi.org/10.1145/2745197.
2745202
[19]
kayt2525. 2016. THETA 360 – horror movie. (2016). https://goo.gl/
WqNkOy Last accessed: March 31, 2016.
[20]
Sei-Young Kim, Joong Ho Lee, and Ji Hyung Park. 2014. The eects of visual
displacement on simulator sickness in video see-through head-mounted displays.
In Proceedings of the 2014 ACM International Symposium
on Wearable Computers. ACM, 79–82.
[21]
Taro Maeda, Hideyuki Ando, Tomohiro Amemiya, N Nagaya, Maki Sugimoto,
and Masahiko Inami. 2005. Shaking the world: galvanic vestibular stimulation
as a novel sensation interface. In ACM SIGGRAPH 2005 Emerging
technologies. ACM, 17.
[22]
Sumit Mehra, Peter Werkhoven, and Marcel Worring. 2006. Navigating on
Handheld Displays: Dynamic Versus Static Peephole Navigation. ACM Trans.
Comput.-Hum. Interact. 13, 4 (Dec. 2006), 448–457.
[23]
Mercedes. 2016. 360
°
video drive in the SL along the Californian coastline –
Mercedes-Benz original. (2016). https://goo.gl/QgDiqi Last accessed:
March 31, 2016.
[24]
Kenton O’Hara, April Slayden Mitchell, and Alex Vorbau. 2007. Consuming video
on mobile devices. In Proceedings of the SIGCHI conference on
Human factors in computing systems. ACM, 857–866.
[25]
Randy Pausch, M Anne Shackelford, and Dennis Prott. 1993. A user study
comparing head-mounted and stationary displays. In Virtual Reality, 1993.
Proceedings., IEEE 1993 Symposium on Research Frontiers
in. IEEE.
[26]
Fabrizio Pece, James Tompkin, Hanspeter Pster, Jan Kautz, and Christian
Theobalt. 2014. Device eect on panoramic video+ context tasks. In Proceed-
ings of the 11th European Conference on Visual Media
Production. ACM, 14.
[27]
Benjamin Petry and Jochen Huber. 2015. Towards eective interaction with om-
nidirectional videos using immersive virtual reality headsets. In Proceedings
of the 6th Augmented Human International Conference.
ACM, 217–218.
[28]
Jerey S Pierce, Randy Pausch, Christopher B Sturgill, and Kevin D Christiansen.
1999. Designing a successful HMD-based experience. Presence 8, 4 (1999),
469–473.
[29]
Roman Rädle, Hans-Christian Jetter, Jens Müller, and Harald Reiterer. 2014. Bigger
is not always better: display size, performance, and task load during peephole map
navigation. In Proceedings of the 32nd annual ACM conference
on Human factors in computing systems. ACM, 4127–4136.
[30]
João Ramalho and Teresa Chambel. 2013. Immersive 360
°
Mobile Video with
an Emotional Perspective. In Proceedings of the 2013 ACM Inter-
national Workshop on Immersive Media Experiences (Im-
mersiveMe ’13). ACM, New York, NY, USA, 35–40. https://doi.org/
10.1145/2512142.2512144
[31]
Michael Rohs, Robert Schleicher, Johannes Schöning, Georg Essl, Anja Naumann,
and Antonio Krüger. 2009. Impact of item density on the utility of visual context
in magic lens interactions. Personal and Ubiquitous Computing 13,
8 (2009), 633–646.
[32]
Michael Rohs, Johannes Schöning, Martin Raubal, Georg Essl, and Antonio Krüger.
2007. Map navigation with mobile devices: virtual versus physical movement with
and without visual context. In Proceedings of the 9th international
conference on Multimodal interfaces. ACM, 146–153.
[33]
RYOT. 2016. The Crossing: A 360
°
Look Into the Journey of Refugees to Greece.
(2016). https://goo.gl/zJp7UP Last accessed: March 31, 2016.
[34]
Beatriz Sousa Santos, Paulo Dias, Angela Pimentel, Jan-Willem Baggerman, Carlos
Ferreira, Samuel Silva, and Joaquim Madeira. 2009. Head-mounted display versus
desktop for 3D navigation in virtual reality: a user study. Multimedia Tools
and Applications 41, 1 (2009), 161–181.
[35]
Mel Slater and Martin Usoh. 1994. Body centred interaction in immersive virtual
environments. Articial life and virtual reality 1 (1994), 125–148.
[36]
BBC Earth Unplugged. 2016. 360
°
video Red Kite Bird Feeding Frenzy. (2016).
https://goo.gl/mCJMBV Last accessed: March 31, 2016.
[37]
Goranka Zoric, Louise Barkhuus, Arvid Engström, and Elin Önnevall. 2013.
Panoramic video: design challenges and implications for content interaction.
In Proceedings of the 11th european conference on Interac-
tive TV and video. ACM, 153–162.
