scieee Science in your language
[en] (orig)

Supplementary material

Author: Vargas, Angélica; Quintero, Jesús M.; Bermeo, Leonardo
Publisher: Zenodo
DOI: 10.5281/zenodo.16296756
Source: https://zenodo.org/records/16296756/files/Supplementary_Material.pdf
Supplemen a y Ma e ial
Sep embe 2025
1 PBRT example
The Physically Based Rende ing Toolki (PBRT) is a ende ing sys em ha implemen s he ma hema -
ical p inciples o mode n pho o ealis ic image syn hesis. I applies ligh anspo algo i hms g ounded
in physics, p o iding a ep oducible pla o m o e alua ing ma e ials, illumina ion, and came a models.
To illus a e he s uc u e o PBRT scene desc ip ions, we p esen a ep esen a i e example. The con-
igu a ion includes a pe spec i e came a, wo ligh sou ces (an in ini e a ea ligh and a blackbody-based
dis an ligh ), h ee sphe es wi h dis inc su ace p ope ies (di use, conduc ing, and dielec ic), and a
ex u ed g ound plane.
Lis ing 1 p esen s he comple e scene desc ip ion in PBRT syn ax, and Table 1 summa izes he
unc ion o each code block.
Lis ing 1: PBRT scene con igu a ion.
1LookA 080# eye
20 .5 0 # look a poin
3001 # up ec o
4Came a "pe spec i e"
5" loa o " [25]
6Sample "hal on"
7"in ege pixelsamples" [2048]
8In eg a o " olpa h"
9Film " gb"
10 "s ing ilename" "sphe e_sample.png"
11 "in ege x esolu ion" [700]
12 "in ege y esolu ion" [400]
13 Colo Space "s gb"
14 Wo ldBegin
15 Ligh Sou ce "in ini e"
16 " gb L" [0.4 0.45 0.5]
17 Ligh Sou ce "dis an " " loa scale" [1.5]
18 "poin 3 om" [-30 40 100]
19 "blackbody L" [3000]
20 # --- GEOMETRY AND MATERIALS ---
21 # Sphe e A: Di use Re lec ance (RGB)
22 A ibu eBegin
23 T ansla e -2 0.5 0
24 Ma e ial "di use"
25 " gb e lec ance" [0.8 0.2 0.2] # Red high e lec ance
26 Shape "sphe e"
27 " loa adius" [0.5]
28 A ibu eEnd
29 # Sphe e B: Aniso opic Roughness (Conduc o )
30 A ibu eBegin
31 T ansla e 0 0.5 0
32 Ma e ial "conduc o "
33 " gb e a" [0.2 0.2 1.4]
34 " gb k" [3.9 2.4 2.5]
35 " loa u oughness" [0.1]
36 " loa oughness" [0.5]
37 "bool emap oughness" [ ue]
38 Shape "sphe e"
39 " loa adius" [0.5]
40 A ibu eEnd
41 # Sphe e C: Dielec ic (Index o Re ac ion)
1
42 A ibu eBegin
43 T ansla e 2 0.5 0
44 Ma e ial "dielec ic"
45 " loa e a" [1.5] # Index o Re ac ion
46 Shape "sphe e"
47 " loa adius" [0.5]
48 A ibu eEnd
49 # G ound plane wi h checke boa d ex u e
50 A ibu eBegin
51 Tex u e "checks" "spec um" "checke boa d"
52 " loa uscale" [16]
53 " loa scale" [16]
54 " gb ex1" [0.1 0.1 0.1]
55 " gb ex2" [0.8 0.8 0.8]
56 Ma e ial "di use"
57 " ex u e e lec ance" "checks"
58 T ansla e 0 0 -1
59 Shape "bilinea mesh"
60 "poin 3 P" [ -20 -20 0 20 -20 0 -20 20 0 20 20 0 ]
61 "poin 2 u " [0010 01 11]
62 A ibu eEnd
Table 1: Explana ion o he PBRT scene desc ip ion in Lis ing 1.
Lines Explana ion
1–3 Came a se up using LookA : de ines he eye posi ion (0,8,0), he look-a poin
(0,0.5,0), and he up ec o (0,0,1).
4–5 Pe spec i e came a wi h a 25◦ ield o iew.
6–7 Hal on sample wi h 2048 pixel samples.
8 Volume ic pa h- acing in eg a o .
9–12 Film block: RGB ilm, ou pu ile sphe e sample.png, esolu ion 700 ×400 pixels.
13 Colo space: sRGB.
14 Wo ldBegin: s a o scene geome y and ligh ing.
15–16 In ini e ligh sou ce wi h g ayish-blue spec um [0.4,0.45,0.5].
17–19 Dis an ligh sou ce a (–30,40,100), in ensi y scaled by 1.5, spec um de ined by a
3000 K blackbody.
22–28 Sphe e A: di use ed Lambe ian ma e ial, ansla ed o (–2,0.5,0), adius 0.5.
30–40 Sphe e B: aniso opic conduc o . Ma e ial de ined by e ac i e indices η=
[0.2,0.2,1.4] and ex inc ion coe icien s k= [3.9,2.4,2.5]. Roughness is aniso opic
wi h u= 0.1, = 0.5. Sphe e a (0,0.5,0), adius 0.5.
42–48 Sphe e C: dielec ic sphe e ( anspa en ). Index o e ac ion η= 1.5, ansla ed o
(2,0.5,0), adius 0.5.
50–62 G ound plane: bilinea mesh wi h checke boa d ex u e. The ex u e al e na es
be ween da k g ay [0.1,0.1,0.1] and ligh g ay [0.8,0.8,0.8], epea ed 16×16 ac oss
he plane.
2
2 Op imized Colou Space T ans o ma ion Ma ices
Algo i hm 1 was applied o compu e he op imized ans o ma ions om CIE XYZ o RGB o he ou
colou spaces implemen ed in PBRT. Fo each space, we p esen he op imized ans o ma ion ma ix
T∗
XY Z→RGB , as de ined in Eq. (3) o he main documen (see also Algo i hm 1), oge he wi h he
pe cep ual simila i y me ic M(IR, IS(µ)) de ined in Eq. (9). The igu es compa e he spec al e e ence
ende ing wi h i s co esponding op imized syn he ic coun e pa .
sRGB
The op imized ans o ma ion is
T∗
XY Z→sRGB =

0.025763 −0.012221 −0.003964
−0.009648 0.018673 0.000414
0.000575 −0.002107 0.010917

.
The esul ing simila i y me ic is
M(IR, IS(µ)) = 0.77.
Re e ence sRGB: IR(SPD) Syn he ic op imized sRGB: IS(µ)
Rec.2020
The op imized ans o ma ion is
T∗
XY Z→Rec.2020 =

0.015744 −0.003262 −0.002324
−0.007121 0.017266 0.000168
0.000208 −0.000505 0.011132

.
The esul ing simila i y me ic is
M(IR, IS(µ)) = 2.06.
Re e ence Rec.2020: IR(SPD) Syn he ic op imized Rec.2020: IS(µ)
3
DCI-P3
The op imized ans o ma ion is
T∗
XY Z→DCI-P3 =

0.022362 −0.008353 −0.003612
−0.008696 0.018478 0.000248
0.000387 −0.000822 0.010329

.
The esul ing simila i y me ic is
M(IR, IS(µ)) = 1.57.
Re e ence DCI-P3: IR(SPD) Syn he ic op imized DCI-P3: IS(µ)
ACES2065-1
The op imized ans o ma ion is
T∗
XY Z→ACES2065-1 =

0.010284 0.000000 −0.000001
−0.005263 0.014574 0.001043
0.000000 0.000000 0.010831

.
The esul ing simila i y me ic is
M(IR, IS(µ)) = 2.09.
Re e ence ACES2065-1: IR(SPD) Syn he ic op imized ACES2065-1: IS(µ)
4
3 Expe imen al Resul s
To alida e he p oposed me hodology, a psychome ic es was conduc ed in which pa icipan s compa ed
eal and i ual scenes ende ed using he PBRT sys em. The e alua ion ocused on h ee main aspec s:
ma e ial appea ance, ealism o illumina ion, and o e all ideli y o he ende ed scene ela i e o i s
eal-wo ld coun e pa . Pa icipan s answe ed a se ies o s uc u ed ques ions using a i e-poin Like
scale, whe e 1 co esponded o he lowes ag eemen and 5 o he highes . The ull se o e alua ion
ques ions is p esen ed in Table 2.
To analyze he esponses, he pe cen age o selec ions o each o he i e esponse op ions in e e y
ques ion was assigned a weigh W( om 1 o 5). These weigh ed alues we e hen used o compu e
a weigh ed sco e (W S) pe op ion. Fo example, conside op ion 5 o ques ion Q2in Table 2. I was
selec ed by 39.1% o pa icipan s, esul ing in a weigh ed sco e o 0.391×5 = 1.96. The inal sco e o each
ques ion is ob ained by summing he weigh ed con ibu ions o all esponse op ions, wi h a maximum
possible alue o 5. Fo he same ques ion Q2, he calcula ion yields 0 + 0.09 + 0.78 + 1.22 + 1.96 = 4.04.
Table 2: Pos - es su ey on Visual Realism and P esence: de ailed esponses wi h weigh ed sco es (W S) and
inal sco es.
Ques ion Answe Op ions (Vo e%, W S) Final
Sco e
Q1: How p esen did you eel
in he i ual scene?
(1) No p esen : 0%, W S = 0
(2) Sligh ly p esen : 0%, W S = 0
(3) Mode a ely p esen : 0%, W S = 0
(4) Qui e p esen : 48%, W S = 1.91
(5) To ally p esen : 52%, W S = 2.61
4.52
Q2: How would you a e he
ealism o he scene?
(1) Un ealis ic: 0%, W S = 0
(2) Somewha ealis ic: 4.3%, W S = 0.09
(3) A e age ealism: 26.1%, W S = 0.78
(4) Mos ly ealis ic: 30.4%, W S = 1.22
(5) Ex emely ealis ic: 39.1%, W S = 1.96
4.04
Q3: How well did you pe cei e
dep h in he scene?
(1) No dep h: 0%, W S = 0
(2) Ve y li le dep h: 0%, W S = 0
(3) Mode a e: 4.3%, W S = 0.13
(4) Qui e a bi : 39.1%, W S = 1.56
(5) Highly pe cep ible: 56.5%, W S = 2.83
4.52
Q4: How com o able did you
eel?
(1) Ve y uncom o able: 0%, W S = 0
(2) Uncom o able: 0%, W S = 0
(3) Neu al: 0%, W S = 0
(4) Com o able: 34.8%, W S = 1.39
(5) Ve y com o able: 65.2%, W S = 3.26
4.65
Q5: How clea and sha p was
he scene?
(1) Ve y blu y: 0%, W S = 0
(2) Blu y: 0%, W S = 0
(3) Neu al: 17.4%, W S = 0.52
(4) Clea : 39.1%, W S = 1.56
(5) Ve y clea : 43.5%, W S = 2.18
4.26
5

Table 3: Pos - es su ey on Ligh ing and Spa ial Pe cep ion: de ailed esponses wi h weigh ed sco es (W S)
and inal sco es.
Ques ion Answe Op ions (Vo e%, W S) Final
Sco e
Q6: How open o limi ed was
he scene?
(1) Ve y con ined: 0%, W S = 0
(2) Somewha limi ed: 26%, W S = 0.52
(3) Neu al: 9%, W S = 0.26
(4) Fai ly open: 61%, W S = 2.40
(5) Comple ely open: 4%, W S = 0.22
3.43
Q7: Pe cep ion o colou em-
pe a u e change?
(1) No change: 0%, W S = 0
(2) Too wa m: 13%, W S = 0.26
(3) Too cool: 4%, W S = 0.13
(4) Sligh change: 0%, W S = 0
(5) App op ia e: 83%, W S = 4.13
4.52
Q8: How com o able was he
scene?
(1) No a all: 0%, W S = 0
(2) Sligh ly: 0%, W S = 0
(3) Somewha : 13%, W S = 0.39
(4) Qui e: 65%, W S = 2.61
(5) Ex emely: 22%, W S = 1.09
4.08
Q9: How did you pe cei e he
ligh ing?
(1) Ve y gla ing: 0%, W S = 0
(2) Gla ing: 0%, W S = 0
(3) Mode a e: 4%, W S = 0.13
(4) So : 70%, W S = 2.78
(5) Ve y so : 26%, W S = 1.31
4.22
Table 4: Pos - es su ey on Ligh ing and Spa ial Pe cep ion: de ailed esponses wi h weigh ed sco es (W S)
and inal sco es.
Ques ion Answe Op ions (Vo e%, W S) Final
Sco e
Q10: B igh ness o he scene? (1) Too b igh : 0%, W S = 0
(2) Too da k: 0%, W S = 0
(3) Sligh ly in ense: 0%, W S = 0
(4) Balanced: 74%, W S = 2.96
(5) So : 26%, W S = 1.31
4.26
Q11: Pleasan ness o expe i-
ence?
(1) Ve y unpleasan : 0%, W S = 0
(2) Unpleasan : 0%, W S = 0
(3) Neu al: 0%, W S = 0
(4) Pleasan : 30%, W S = 1.22
(5) Ve y pleasan : 70%, W S = 3.48
4.70
Q12: How ale o sleepy we e
you?
(1) Ve y sleepy: 0%, W S = 0
(2) Somewha sleepy: 13%, W S = 0.26
(3) Neu al: 9%, W S = 0.26
(4) Qui e ale : 26%, W S = 1.04
(5) Ve y ale : 52%, W S = 2.61
4.18
Q13: How ee o es ic ed
did you eel?
(1) Comple ely es ic ed: 0%, W S = 0
(2) Some es ic ions: 22%, W S = 0.43
(3) Neu al: 13%, W S = 0.39
(4) Fai ly ee: 61%, W S = 2.44
(5) Comple ely ee: 4%, W S = 0.22
3.48
Q14: How noisy o quie was
i ?
(1) Ve y noisy: 0%, W S = 0
(2) Noisy: 0%, W S = 0
(3) Neu al: 4%, W S = 0.13
(4) Quie : 30%, W S = 1.22
(5) Ve y quie : 65%, W S = 3.26
4.61
Q15: How elaxed o ne ous
did you eel?
(1) Ve y ne ous: 0%, W S = 0
(2) Sligh ly: 4%, W S = 0.09
(3) Neu al: 26%, W S = 0.78
(4) Somewha elaxed: 22%, W S = 0.87
(5) Ve y elaxed: 48%, W S = 2.39
4.13
Q16: O e all sa is ac ion? (1) Ve y dissa is ied: 0%, W S = 0
(2) Dissa is ied: 0%, W S = 0
(3) Neu al: 0%, W S = 0
(4) Sa is ied: 22%, W S = 0.87
(5) Ve y sa is ied: 78%, W S = 3.92
4.78
6