Basic HTML Version
NORDIC LIGHT & COLOUR
111
There were similar but separate questionnaires for each group
of models. Each form consisted of two parts.
The first part contained one question from the
PERCIFAL
(Perceptive Spatial Analysis of Colour and Light) question-
naire (Arnkil et al. 2011), (Matusiak et al. 2011), (Fridell Anter,
Haggstrom et al. 2012). The intention was to get spontaneous
answers to the question:
Do you experience the room to be dark
or bright?
The participant made a mark on a 7-step scale from
very dark to very bright. The number of marks for each step
and for each room was calculated, see Figure 15 and 19.
The second part included four more questions about lighting
in the scale models and needed more conscientious answers.
In answering the questions from this part of the questionnaire,
observers had to arrange rooms into descending order. The
first question from this part was:
Which room has the highest
light level (the brightest room)
? Arrangement had to be from the
brightest to the darkest. All the answers (for each particular
group of scale models) were calculated as a percentage of
the total number of participants. Results are represented in
Figures 16 and 20.
The second question of the second part was:
Which room has
the more comfortable lighting?
Answers were arranged from the
most comfortable to the least comfortable. Results are presen-
ted in Figure17 and 21.
The third question:
Your personal preferences among these
rooms (in lighting).Why?
Results represented in Figure 18 and
22.
These two questions (about comfortable lighting and personal
preferences in lighting among the rooms) were needed to verify
reliability of the results by help of comparison of the answers.
The last question in the questionnaire was:
How much do you
think colour affects your perception of light level?
The subject had
to mark the level of the influence of colour/colour compositions
on the light level perception on the proposed scale.
High dynamic range imaging (luminance maps)
HDR images of all the interiors of the models were created.
Afterward, 13 sets of low-dynamic images for each room were
made with a Canon EOS300D digital camera (see Figure 11).
The camera was mounted on a tripod and situated in the plane
of subject`s eyes to simulate the viewing position of the obser-
vers. The following camera settings were used: White balance
– Daylight, Auto-Bracketing – off, Sensitivity – 100 ISO, Auto
Focus – off, Aperture – fixed, f/4. Exposure variations
were achieved by varying the shutter speed in manual exposure
mode with step 1 EV.
In addition to the photos, references to physical measurements
have been made with a calibrated hand held Luminance meter.
The readings were used for further calibration of the HDR
images. All the low dynamic range images were processed
and combined into HDR images with the help of Photosphere
software and were calibrated according to the measured lumi-
nance readings.
Statistical analysis
For the statistical analysis of the survey results, mode values
were calculated as most representative in the experiment with
ordinal data. Percentages represented on the graphs were cal-
culated relative to the total number of participants. As a main
tool for the statistical analysis the Friedman test was used.
This is a nonparametric statistical procedure for comparing
more than two samples that are related. The parametric equi-
valent to the Friedman test is the repeated measures analysis
of variance (ANOVA) (Corder & Foreman 2011).
RESULTS
Measurements
According to the manual measurements and further calcula-
tions, the difference in luminance values, measured in a central
point of the painted wooden pieces under the Artificial Sky, is
3% (see Figure 12). Therefore, the difference in the reflectance
of these samples is also 3% (Figure 13). Based on these data
we can conclude that the colours of the walls in the scale mo-
dels were equiluminant.
Illuminance values have a higher difference – 5.3% (Figure 14).
However, these measurements are not comparable to lumi-
nance readings due to different measuring conditions.
Perception of the light level in one-coloured models
Part 1. The subjects were asked to evaluate the light level in
models using a scale from dark to bright. It was not allowed to
compare the rooms. The answer to the question had to be done
immediately and spontaneously.
Results showed that the Yellow room was perceived as brigh-
test, the Grey room as darkest, and the rest of the rooms were
placed in medium positions, Figure 15.
Part 2. In the second part of the questionnaire subjects com-
pared appearances of observed rooms and made a deliberate
ranking of the scale models.