Inﬂuence of Texture Fidelity on Spatial Perception in Virtual Reality

Andrei Iuliu Lucaci, Morten Trier Bach, Poul Anker Jensen and Claus Brøndgaard Madsen

Computer Graphics Group, Department of Architecture and Media Technology, Aalborg University, Aalborg, Denmark

Keywords:

Virtual Reality, Spatial Perception, Verbal Assessment, Architectural Visualization.

Abstract:

In this paper, we investigate the inﬂuence of texture ﬁdelity on spatial perception in a standalone virtual reality

application. To investigate this, we implemented a detailed virtual representation of an actual physical environ-

ment, namely a small one-bedroom apartment. The virtual apartment representation was tested in two different

visual styles: high ﬁdelity realistic textures, and a ”paper model” texture. Some test subjects experienced the

virtual models using the actual physical apartment as transitional environment, other subjects experienced the

model at an unrelated physical location. The environments were evaluated with 20 participants aged 20 to 61

and results indicated a systematic overestimation of distances in virtual reality for all conditions. The results

showed that a higher texture ﬁdelity had a positive inﬂuence on precision but no signiﬁcant inﬂuence on accu-

racy. It was also showed that transitional environments negatively inﬂuenced precision, but had no signiﬁcant

inﬂuence on accuracy. Self assessments of presence from the experiment supported previous claims about a

correlation between the level of detail in an environment and presence, but not a correlation between presence

and distance perception.

1 INTRODUCTION

Architectural visualization, in its essence, is a pro-

cess that transports a concept from a two dimensional

perspective to a three dimensional image or a physi-

cal mockup. Due to the advancement of computers,

there are also a plethora of software solutions that can

facilitate creating a computer generated architectural

model. Recently, the idea of combining Virtual Re-

ality (VR) and game engines brings a new level of

creativity and control to achieve high-realism scenes

(Team, 2018).

Standalone VR headsets increases the potential of

VR by affording greater ﬂexibility than computer-

tethered devices, and have more features than a

browser or a mobile phone. In recent years, advance-

ments in Head-mounted Display (HMD) technology

has made the idea of owning a VR headset without

the need of a VR Ready computer more palpable, as

is indicated by the popularity of the Oculus Quest 2

(Lang, 2021).

While HMDs show promise, they still have some

issues. One such issue is that spatial perception in VR

has repeatedly been shown to be inaccurate, which

brings into question the use of VR as a medium for

architectural visualisation (Loyola, 2018).

https://orcid.org/0000-0003-0762-3713

On the other hand, studies researching spatial per-

ception in VR suggests that a high ﬁdelity, low la-

tency, immersive environment can decrease the in-

accuracies in user’s distance perception (Interrante

et al., 2006) (Phillips et al., 2009a) (Phillips et al.,

2009b).

With this in mind, the aim of this study was to

“investigate the inﬂuence of texture ﬁdelity on spatial

perception in standalone virtual reality for architec-

tural visualization.”

2 SPATIAL PERCEPTION IN

VIRTUAL REALITY

Spatial perception in Virtual Reality (VR) has been

extensively studied and the consensus appears to be

that spatial perception in VR is compressed compared

to spatial perception in the real world (Interrante et al.,

2006) (Loyola, 2018). I.e., in VR people generally

feel that the space is smaller than its real physical

counterpart. If such spatial mis-perception is indeed

the case, the concept of using VR as a tool in archi-

tectural design and validation processes is obviously

compromised, as it can lead to misrepresentation by

architects and clients alike (Loyola, 2018).

244

Lucaci, A., Bach, M., Jensen, P. and Madsen, C.

Inﬂuence of Texture Fidelity on Spatial Perception in Virtual Reality.

DOI: 10.5220/0010890100003124

In Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 1: GRAPP, pages

244-251

ISBN: 978-989-758-555-5; ISSN: 2184-4321

Figure 1: Spatial perception was assessed with verbal assessment in virtual environments with different texture ﬁdelities

modelled after a real life room (left image).

A series of studies by Interrante et al. investigates

the inﬂuence of immersion and ﬁdelity on spatial per-

ception in VEs (Interrante et al., 2006) (Phillips et al.,

2009a). Their main ﬁndings were that distance per-

ception appears to not be signiﬁcantly compressed

in high ﬁdelity, low latency, immersive environments

and even less if the VE represents an exact replica of

the room in which the user is located in, in the real

world (Interrante et al., 2006) (Phillips et al., 2009a).

In their initial paper on the subject, they also suggest

that the user’s sense of presence might have an inﬂu-

ence on their estimates (Interrante et al., 2006).

In a more recent study by Interrante et al. they in-

vestigated distance estimation in a non-photorealistic

VE against a photorealistic VE (Phillips et al., 2009b).

The ﬁndings of their study supports their previous

ﬁndings as the participants were signiﬁcantly better at

assessing distances in the photorealistic environment

than in the non-photorealistic environment.

In their paper it is also discussed that a possi-

ble explanation for their ﬁndings could be that the

unrealistic nature of their representation of a non-

photorealistic environment possibly interfered with

the participants’ presence in the environment, as it de-

viates a lot from the real world (Phillips et al., 2009b).

Hornsey and Hibbard further support the ﬁnd-

ing that distance estimation is biased towards un-

derestimation, (Hornsey and Hibbard, 2021). They

also found that the more pictorial cues are added to

a scene, the better the distance estimation will be

(Hornsey and Hibbard, 2021). Finally, they found

that texture gradients have a positive inﬂuence on dis-

tance estimation, especially at very short distances

(Hornsey and Hibbard, 2021).

Based on the above, we found it interesting to in-

vestigate whether the texture ﬁdelity alone has an in-

ﬂuence on distance assessment even if applied to the

same geometry. In order to measure the participants

perceived distances a few methods for this were re-

viewed.

2.1 Measurement Methods for

Perceived Distance

To evaluate the users’ spatial perception of our VE, a

measurement method was needed. Choosing a mea-

surement method for perceived distance can be difﬁ-

cult as perception is very abstract. Table 1 shows the

prominence of various different approaches in previ-

ous studies on spatial perception in VR.

Table 1: Measurement methods for perceived distance from

previous studies including whether they can measure ego-

centric or exo-centric distances.

Method Ego or

Exo

References

Verbal

Assess-

ment

Both (Loyola, 2018) (Peer and

Ponto, 2017) (Ng et al.,

2016) (Armbr

uster et al.,

2008) (Gagnon et al., 2020)

(Klein et al., 2009) (Kelly

et al., 2017)

Blind

Walking

Ego (Interrante et al., 2006) (Peer

and Ponto, 2017) (Kelly

et al., 2018) (Li et al., 2015)

(Kelly et al., 2017) (Klein

et al., 2009) (Phillips et al.,

2009b)

Resizing

Objects

Both (Kelly et al., 2018) (Kelly

et al., 2017) (Jensen et al.,

2020)

Time

Imagined

Walking

Ego (Peer and Ponto, 2017)

(Klein et al., 2009)

Resizing

Environ-

ment

Both (Jensen et al., 2020)

Blind

Throw-

ing

Ego (Peer and Ponto, 2017)

Having explored the methods shown in Table 1,

we chose to use verbal assessment as our measure

method. Even though some studies have suggested

that blind walking is more reliable (Ng et al., 2016),

Inﬂuence of Texture Fidelity on Spatial Perception in Virtual Reality

245

evidence found by Klein et. al suggests that this is

only the case in larger outdoor areas and that the

methods have similar reliability in indoor scenarios

such as ours(Klein et al., 2009).

3 IMPLEMENTATION OF THE

VIRTUAL ENVIRONMENT

In a mobile Virtual Reality (VR) context, Unity’s

built-in Universal Rendering Pipeline (URP) appears

to be the optimal render pipeline. Since the Oculus

Quest platform can be considered a mobile platform,

the ease-of-use and mobile centered optimizations of

URP make it a solid foundation to achieve the level of

quality desired, considering the computational limita-

tions of the platform.

3.1 Virtual Reality Implementation

VR Implementation for the application was done us-

ing Unity’s XR Interaction Toolkit, which affords

adding an XR Rig to the environment. The XR Rig

is automatically connected to the Head-Mounted Dis-

play (HMD) being used by a user and the camera

follows the HMD position, while controller objects

follow the controllers being used. The Field-of-View

(FOV) set for the rendered camera is 98 degrees Ver-

tical, and 128 degrees Horizontal.

3.2 Modelling the Virtual Environment

Recreating a real indoor environment in VR required

us to create 3D models of the environment and furni-

ture. The foundation of the environment for our ex-

periment was created with a method similar to what

was previously explored by Ozacar et al. (Ozacar

et al., 2017), starting with a ﬂoor plan made in Sweet

Home 3D, which was exported to Unity as a 3D

model with walls, ﬂoors and ceilings.

3D models representing furniture recreated from

the real world were placed in the environment in ac-

cordance with scale and distance translated from the

metric system to Unity’s measuring system. These

models were created in either Blender or Maya and

textured with Substance Painter. For details like text,

logos or stickers some manipulation of the textures

were done in Adobe Photoshop. The models were

created with a relatively low amount of polygons to

limit the memory usage of the application as we are

working with standalone headsets, as was also high-

lighted by Unity as an important factor in optimizing

performance of applications (Technologies, 2020). To

accurately represent the chosen apartment, pictures

of the furniture were taken and used as reference, as

shown in Figure 2.

Figure 2: A sample of the modelled furniture, with the left

image being the real life version, in the respective order a

cupboard, a desk, an ofﬁce chair and a mirror.

3.3 Virtual Environment Illumination

Lighting in the environment was achieved by combin-

ing a mixed directional light source (baked indirect

lighting and real-time shadows), a baked area light

source, and global illumination (GI). The directional

light was chosen to simulate direct sunlight with a

downwards facing angle entering through the win-

dow, and the area light was used to simulate the effect

of light from the sky. The sky area light was placed

just outside the window, facing into the room. In com-

bination these two types of light sources constitute a

typical, and efﬁcient, approach to emulating exterior

daylight affecting an indoor scenario, (Birn, 2014).

The reﬂectivity of a material in a virtual world has

a signiﬁcant effect on the aesthetics of the object that

it is applied to (Manson and Sloan, 2016). To increase

the materials ability to react to light in a realistic way,

Physically Based Rendering (PBR) was utilized. To

simplify the computation of reﬂections in the envi-

ronment, a reﬂection probe was used.

The reﬂection probe had a resolution of 256 pix-

els, and was used to ensure that light reﬂections (with

colours corresponding to objects they bounce off)

were present in the environment (see Figure 3).

Additionally, shadows were computed as shadow

maps, and ambient occlusion was applied to all ob-

jects. Finally a global volume for post processing ef-

fects was applied to the environment, allowing the in-

corporation of white balancing, gamma and gain al-

tering and tone mapping. The parameters for these

were set as shown in Table 2.

Table 2: Parameters used for Global Volume Post Process-

ing Effects.

Effect Parameter

Tonemapping ACES

Gamma 1.32

Gain 1.29

GRAPP 2022 - 17th International Conference on Computer Graphics Theory and Applications

246

Figure 3: A reﬂection probe from the environment. Top part

is the probe itself, with the bottom part being the cubemap.

4 EXPERIMENTAL DESIGN

The independent variable in the experiment was the

texture ﬁdelity of the environment with the two con-

ditions being: convincing textures and paper textures.

We used a between group design with two groups of

participants, each assigned to one of the conditions.

These conditions can be seen in Figure 4.

In the experiment we measured the participants’

spatial perception of the room, by asking them about

certain distances and getting them to verbally assess

them. The measurement method was chosen based on

the small size of the room as described in Section 2.1,

while other viable methods were presented in Table 1.

Figure 4: Side-by-side comparison of the two conditions,

with the left side being convincing textures and the right

side being the paper textures.

The primary interest of the study was to investi-

gate if convincing textures have an inﬂuence on the

user’s spatial perception of a Virtual Environment

(VE). To investigate the inﬂuence we came up with

the following set, “A”, of hypotheses.

Hypothesis A1: Participants experiencing the

scenario with convincing textures should assess

distance with higher accuracy and precision than

the ones experiencing the paper textures.

Hypothesis A2: Participants experiencing the

scenario with convincing textures should assess

distance with lower accuracy and precision than

the ones experiencing the paper textures.

Null Hypothesis A0: There is no signiﬁcant dif-

ference in the accuracy or precision of assessing

distance, between participants experiencing con-

vincing textures and paper textures.

Previous research done by (Steinicke et al., 2009)

(Interrante et al., 2006) (Phillips et al., 2009a), have

explored the effects of ”transitional environments”.

This phenomenon implies that depth perception is

positively affected by entering a VE from the phys-

ical environment it represents. Thus, in the design

of our experiment, half of the participants entered the

VE from the modelled apartment, with the other half

entering the VE from an unrelated location. The sub-

set, “B”, of hypotheses is as follows:

Hypothesis B1: Participants entering the VR

scenario from the real-life equivalent of the vir-

tual environment should assess distance with

higher accuracy and precision than the ones who

enter from a unrelated location.

Hypothesis B2: Participants entering the VR

scenario from the real-life equivalent of the vir-

tual environment should assess distance with

lower accuracy and precision than the ones who

enter from an unrelated location.

Null Hypothesis B0: There is no signiﬁcant dif-

ference in the accuracy or precision of assessing

distance, between participants entering from the

real-life equivalent of the virtual room and an un-

related location.

4.1 Participants

Participants for the experiment were contacted and re-

cruited through social media or email. For each of the

sub-experiments 10 participants were recruited - giv-

ing the experiment a total of 20 participants. The age

of the participants ranged from 20 to 61 - all with nor-

mal (20/20) or corrected vision.

Inﬂuence of Texture Fidelity on Spatial Perception in Virtual Reality

247

4.2 Apparatus

As we focused on standalone VR - the Oculus Quest

2 Head-Mounted Display (HMD) was used in the ex-

periment, as it has a higher resolution than the original

Quest (Technologies, 2021).

The inter-pupilary distance (IPD) of the Quest

2 can be physically set as 58, 63 or 68mm, which

was set to ﬁt each individual participant, as a wrong

IPD can result in the view being blurry (Corporation,

2020). The Quest 2 further supports a 89 degree hor-

izontal Field-of-View (FOV), 90 degree vertical FOV

and 127 degree diagonal FOV (Brown, 2021).

Figure 5: The teleportation spots are shown with a circle,

color-coded to match the distances assessed from them. The

ego-centric distances are shown with dotted lines, while the

exo-centric distances are shown with full lines.

4.3 Procedure

Firstly, the facilitators informed the participants about

the course of the experiment and ensured that they

were familiar with the HMD and the buttons they

would have to use on the controllers during the ex-

periment. In this introduction the physical IPD on

the HMD was also set to the best ﬁtting state for the

participants. Once the participants were ready they

would enter the environment matching the respective

condition they were assigned to.

The participants would be asked to walk around

and explore the room for 1 minute, as it has been

proven to improve spatial perception in VEs (Kelly

et al., 2018), before beginning the distance assess-

ment tasks.

After exploring the room, the participants would

be asked to press the trigger to teleport to the ﬁrst

spot in the room, which they had to assess distance

from. After teleporting, they were told to hold the

Oculus button until their position and orientation was

reset. This ensured that all participants were assess-

ing distances from the same location in the VE. After

assessing two distances in a spot, they would be asked

to teleport to the next spot following the same proce-

dure.

The spots and the distances are illustrated in a top-

down view of the environment in Figure 5. In total the

participants had to assess six distances ranging from

105cm to 440cm within the VE. While each partici-

pant was assessing distances, the facilitator would log

the assessments manually.

After ﬁnishing the assessments in Virtual Reality

(VR), the participants’ sense of presence in the en-

vironment was measured with the Igroup Presence

Questionnaire (IPQ) (igroup.org, 2016).

The questionnaire was included in the experiment

to see if the conditions had any inﬂuence on presence,

as it was suggested by Interente et al. that the ﬁdelity

of the environment, presence and depth perception

is correlated (Interrante et al., 2006) (Phillips et al.,

2009a).

Following the questionnaire, a follow-up unstruc-

tured feedback session was initiated using the follow-

ing inquiry:

Did you think any of the objects looked less or

more believable than others?

The purpose was to elicit the participants’

thoughts on the environment and the perceived dis-

tances.

5 RESULTS

The presentation of the results from the experiment,

described above in Section 4, is split into three cate-

gories: Distance Assessment, Presence and Feedback

on the Environment. To validate that the results of the

experiment were minimally inﬂuenced by the perfor-

mance of the application on the Head-Mounted Dis-

play (HMD), the framerate was logged. This showed

that the condition with convincing textures had a me-

dian framerate of 48fps, which is marginally less than

the paper condition which had a median framerate of

52fps.

5.1 Distance Assessments

In order to visualize and analyze the distance assess-

ments from the experiment, the amount of error in the

participants’ assessments for all distances were ex-

pressed in percentage making them comparable - sim-

ilarly to what was done in prior research by Loyola et.

al (Loyola, 2018). For our analysis of signiﬁcant dif-

ference between the groups, we chose to analyze the

absolute values to focus purely on their difference in

accuracy. Accuracy being deﬁned as: ”how much the

assessments deviate from the actual distances”. In the

plots, however, we will use the relative errors to show

GRAPP 2022 - 17th International Conference on Computer Graphics Theory and Applications

248

the precision of the assessments. Precision is here de-

ﬁned as: “how similarly participants within a certain

group assessed the distances”.

5.1.1 Distance Assessment Compared Between

the Conditions

The experiment showed no signiﬁcant difference (p-

value = 0.83 > 0.05) in the absolute error deviations

from the participants’ distance assessment between

the two conditions, convincing textures and paper tex-

tures. There is therefore no evidence that the condi-

tions inﬂuenced accuracy in distance assessment.

Figure 6: Visualization of the precision within each condi-

tion showed with relative error deviation in percentage.

The standard deviation of the relative error in per-

centage for the condition with convincing textures

was 23.17 percent, whereas standard deviation for

condition with paper textures was 58.98 percent. As

the standard deviation was lower for the condition

with convincing textures, compared to the condition

with paper textures, it shows that the condition with

convincing textures has increased precision of assess-

ment within the sample.

The plot in Figure 6 shows a comparison of the

relative error deviation in percentage between the

conditions.

5.1.2 Distance Assessment Compared between

Apartment and Unrelated Location

The experiment also did not show signiﬁcant differ-

ence (p-value = 0.25 > 0.05) in the absolute error

deviations from the participants’ distance assessment

between the locations. In our experiment there is

therefore also no evidence that being in the real life

equivalent of the Virtual Environment (VE) during the

experience has an inﬂuence on the accuracy of dis-

tance assessments.

The standard deviation of the relative error in per-

centage for the participants who experienced the ex-

periment in the real life equivalent of the VE was

Figure 7: Visualization of the precision within each location

showed with relative error deviation in percentage.

55.32 percent, whereas the standard deviation for the

ones who tried it in the unrelated location was 31.18

percent. As the standard deviation was lower for the

ones experiencing the experiment in the unrelated lo-

cation, compared to the ones experiencing it in the

real life equivalent of the VE, it shows that experienc-

ing it from the unrelated location has increased preci-

sion of assessment within the sample.

The plot in Figure 7 shows a comparison of the

relative error deviation in percentage between the

conditions.

5.2 Presence

The results from the presence questionnaires were av-

eraged and made into visual representations as pres-

ence proﬁles. No signiﬁcant difference in presence

was found between the participants who tried the ex-

perience in the real life equivalent of the VE (i.e apart-

ment) or in the unrelated location.

The means of the sub-scales of presence between

the conditions (i.e. convincing textures and paper tex-

tures) as shown in the presence proﬁle in Figure 8.

Figure 8: The presence proﬁle showing how the means of

each condition between the conditions are distributed for

each presence sub-scale.

The self-assessments of presence showed statis-

tically signiﬁcantly higher scores in the sub-scales

Spatial Presence (SP) and Involvement (INV) within

Inﬂuence of Texture Fidelity on Spatial Perception in Virtual Reality

249

the convincing textures condition compared to the pa-

per textures. Although for the general “sense of be-

ing there” (G) and sub-scale Experienced Realism

(REAL) there was no signiﬁcant difference between

the conditions.

5.3 Feedback on the Environments

Visuals

Seven of the participants stated that the outside of

the window looked slightly more “blurry” than the

rest. One participant described it as being immersion

breaking as it was ﬂat compared to the rest of the envi-

ronment and two others stated that the window frame

was jagged as well.

The mirror was also mentioned by multiple peo-

ple. Four participants stated that it was in accordance

with real life and one participant even stated that it

gave a him a better affordance of depth view. Al-

though, three participants stated that the mirror was

a bit jagged compared to the environment, while two

participants noted that it was weird to not see them-

selves in the mirror, and a single person stated that he

did not even recognise it as a mirror at ﬁrst.

Another interesting ﬁnding was that ﬁve of the

participants, who tried the experience from within the

real life equivalent of the VE, stated that the room

seemed smaller within the VE compared to real life.

This is particular interesting as most participant over-

estimated the distances. One participant also stated

that the VE was not as lit as the real life counterpart.

6 DISCUSSION

Our ﬁndings suggest, that the condition with convinc-

ing textures results in a higher precision than the pa-

per texture condition, however, no signiﬁcant differ-

ence was observed in terms of accuracy between these

conditions. This contradicts research by Interrante et

al. (Interrante et al., 2006) (Phillips et al., 2009a)

(Phillips et al., 2009b). A difference between the tex-

ture conditions that might have had an inﬂuence on

this is the illuminance levels. In the version with pa-

per textures, the white material being reﬂected by the

reﬂection probe illuminated the room more than the

colours from the convincing textures did. So even

though the lighting setup and parameters were the

same, the environment with the paper textures was

more illuminated. This can be seen in Figure 4.

Furthermore, the precision between locations was

found to be higher when experiencing the Virtual En-

vironment (VE) from the unrelated location than the

apartment, with accuracy being unaffected. This is in

contrast to previous research done by Steinicke et al.

(Steinicke et al., 2009).

Another interesting ﬁnding was that participants

in general seemed to overestimate the distances,

which contradicts previous research that has shown

tendencies of general underestimations in Virtual Re-

ality (VR), as presented in Section 2. It is also in-

teresting to note that ﬁve of the participants entering

the experience from the real life equivalent of the VE,

stated that the room seemed smaller in the VE, but

still overestimated the distances in their assessments.

In terms of the users’ sense of presence, the con-

vincing textures scored signiﬁcantly higher in spatial

presence and involvement. This correlates to prior

research which suggests that a higher visual ﬁdelity

of an environment positively inﬂuences presence (see

Section 2). However, the correlation does not, as

previous research have suggested, signiﬁcantly corre-

late with the distance assessments as described above.

There was no signiﬁcant change in any of the pres-

ence sub scales between the entering locations, which

contradicts previous research into the effect of transi-

tional environments on presence.

A possible threat to both validity and reliability,

as has been eluded to earlier in Section 2.1, is the

measurement method of verbal assessment in metrics,

which in a few earlier studies been suggested to be un-

reliable. It is likely that some of our outliers in the

experiment is due to the verbal assessment method

being unreliable for some individuals, since verbal

assessments in metric is not something that is com-

monly used in everyday life for most people. A few

participants also stated that they for some of their dis-

tance assessments, such as with the window frame,

used knowledge of familiar sizes of window frames,

instead of na

ıvely evaluating the distance in the VE.

Another threat to the reliability of the test is that the

experiment only had 20 participants. This makes it

impossible to properly generalize our ﬁndings.

However, since the majority of our ﬁndings con-

tradict previous research, we explored some of the

differences in methods between previous studies and

ours, and concluded that the main difference was that

we used a relatively small room compared to research

in this ﬁeld. We hypothesize that these size differ-

ences in the VEs may have been the main cause be-

hind the contradictions. We suggest an investigation

of this through a large-scale validation study with a

number of VEs of various sizes rather than just one,

and a larger sample of participants. This could be

done using the method we applied in this study, or,

more interestingly, a variety of the methods presented

in Table 1.

GRAPP 2022 - 17th International Conference on Computer Graphics Theory and Applications

250

7 CONCLUSION

The aim of the study was to investigate the inﬂuence

of texture ﬁdelity on spatial perception of a Virtual

Environment (VE). An experiment was conducted

with participants assessing distance in two different

VEs: one environment having convincing textures

and one having paper textures, both with the same ge-

ometry. The experiment was split into two sub exper-

iments investigating whether the location from where

you enter the VE has an inﬂuence on the assessments,

with one of the locations being the real life equivalent

of the VE.

The ﬁndings suggested that the environment with

convincing textures had higher precision than the one

with paper textures. However, there was no signiﬁ-

cant difference between the accuracy of assessments

in neither the texture ﬁdelities nor the entering loca-

tions. The entering locations also did not have any

inﬂuence in accuracy, while the precision of assess-

ments were higher for participants entering from an

unrelated location than the real life equivalent of the

VE. Both of these ﬁndings contradict previous re-

search into transitional environments.

A more interesting ﬁnding was the consistent

overestimations in distance assessments which con-

tradicts previous studies, which we hypothesize could

be caused by the evaluated VE being smaller in size

than those used in previous research. We suggest

that this is researched further by evaluating distance

perception in VEs of various sizes, using either the

method used in this paper, or a variety of those pre-

sented in Table 1.

REFERENCES

Armbr

uster, C., Wolter, M., and Kuhlen, T. W. (2008).

Depth perception in virtual reality: Distance estima-

tions in peri- and extrapersonal space.

Birn, J. (2014). Digital Lighting and Rendering. New Rid-

ers, San Francisco, California, 3rd edition.

Brown, R. (2021). Oculus quest 2: Full speciﬁcation.

Corporation, H. (2020). Adjusting the ipd on the headset.

Gagnon, H. C., Buck, L., Smith, T. N., Narasimham, G.,

Stefanucci, J., Creem-Regehr, S. H., and Bobby, B.

(2020). Far distance estimation in mixed reality.

Hornsey, R. L. and Hibbard, P. B. (2021). Contributions of

pictorial and binocular cues to the perception of dis-

tance in virtual reality. Virtual reality : the journal of

the Virtual Reality Society, 1.

igroup.org (2016). igroup presence questionnaire (ipq)

overview.

Interrante, V., Ries, B., and Anderson, L. (2006). Distance

perception in immersive virtual environments, revis-

ited. In Proceedings of the IEEE Conference on Vir-

tual Reality, VR ’06, page 3–10, USA. IEEE Com-

puter Society.

Jensen, T., Kasprzak, F., Szekely, H.-G., Nikolov, I.,

Høngaard, J., and Madsen, C. (2020). Preliminary

study on the inﬂuence of visual cues, transitional en-

vironments and tactile augmentation on the perception

of scale in vr. In HCI International 2020 – Late Break-

ing Posters, Communications in Computer and Infor-

mation Science, pages 156–164. Springer.

Kelly, J. W., Cherep, L. A., Klesel, B., Siegel, Z. D., and

George, S. (2018). Comparison of two methods for

improving distance perception in virtual reality.

Kelly, J. W., Cherep, L. A., and Siegel, Z. D. (2017). Per-

ceived space in the htc vive.

Klein, E., Swan, J. E., Schmidt, G. S., Livingston, M. A.,

and Staadt, O. G. (2009). Measurement protocols for

medium-ﬁeld distance perception in large-screen im-

mersive displays.

Lang, B. (2021). Quest 2 continues to grow by leaps and

bounds among vr headsets on steam.

Li, B., Zhang, R., Nordman, A., and Kuhl, S. A. (2015).

The effects of miniﬁcation and display ﬁeld of view

on distance judgments in real and hmd-based environ-

ments.

Loyola, M. (2018). The inﬂuence of the availability of vi-

sual cues on the accurate perception of spatial dimen-

sions in architectural virtual environments. Virtual Re-

ality, 22:235–243.

Manson, J. and Sloan, P.-P. (2016). Fast ﬁltering of reﬂec-

tion probes. Computer Graphics Forum, 35(4):119–

127.

Ng, A. K. T., Chan, L. K. Y., and Lau, H. Y. K. (2016).

Depth perception in virtual environment: The effects

of immersive system and freedom of movement.

Ozacar, K., Ortakei, Y., Kahraman, I., Durgut, R., and

Karas, I. R. (2017). A low-cost and lightweight 3d

interactive real estate-purposed indoor virtual reality

application. ISPRS Annals of the Photogrammetry,

Remote Sensing and Spatial Information Sciences, IV-

4/W4:307–310.

Peer, A. and Ponto, K. (2017). Evaluating perceived dis-

tance measures in room-scale spaces using consumer-

grade head mounted displays.

Phillips, L., Ries, B., Interrante, V., Kaeding, M., and An-

derson, L. (2009a). Distance perception in npr immer-

sive virtual environments, revisited.

Phillips, L., Ries, B., Interrante, V., Kaeding, M., and An-

derson, L. (2009b). Distance perception in npr immer-

sive virtual environments, revisited.

Steinicke, F., Bruder, G., Hinrichs, K., Lappe, M., Ries, B.,

and Interrante, V. (2009). Transitional environments

enhance distance perception in immersive virtual re-

ality systems.

Team, D. (2018). What is architectural visualization?

Technologies, F. (2021). Oculus compare headsets. https:

//www.oculus.com/compare/.

Technologies, U. (2020). Optimizing graphics performance.

https://docs.unity3d.com/2017.4/Documentation/

Manual/OptimizingGraphicsPerformance.html.

Inﬂuence of Texture Fidelity on Spatial Perception in Virtual Reality

251