acoustic and perceptual features of eating places
play

Acoustic and perceptual features of eating-places . PerMagnus - PowerPoint PPT Presentation

Acoustic and perceptual features of eating-places . PerMagnus Lindborg Nanyang Technological University, Singapore (asst. prof.) KTH Royal Institute of Technology, Stockholm (doct. cand.) permagnus@ntu.edu.sg www.permagnus.net Monday, July 29,


  1. Acoustic and perceptual features of eating-places . PerMagnus Lindborg Nanyang Technological University, Singapore (asst. prof.) KTH Royal Institute of Technology, Stockholm (doct. cand.) permagnus@ntu.edu.sg www.permagnus.net Monday, July 29, 2013

  2. Lindborg: EAT study Soundscape perception How much does soundscape quality matter to the experience as a whole? Which sound sources are positive or negative? What are the acoustic differences between places by category e.g. fastfood/ restaurants, cheap/expensive, Western/Asian etc? Soundscapes are important parts of 'servicescapes' (Novak et al. 2010, North & Hargreaves 1996). Eating-places in Singapore EAT112: broad, exploratory EAT15: multi-modal ratings EAT40: (priciness) Monday, July 29, 2013

  3. Lindborg: EAT study Perceptual features Semantic differential (Osgood C. et al. 1957) circumplex models of affect (Russell J. 1980, Thayer) Swedish Soundscape Quality Protocol (e.g. Pleasantness- Eventfulness , PLEV ) (Axelsson Ö, Berglund G., & Nilsson M.E.) Music emotion research, e.g. Geneva Music Emotion Scales (GEMS, Zentner, Grandjean et al.) Music (emotion) perception, ecological approach (Friberg A. et al., Vestfäll D., Andean J.) Acoustic features feature unit Sound Pressure Level (SPL) sone Loudness N10 Leq A, C - different spectral shape, based on Noisiness zeroXrate rate hearing sensitivity curves at different intensity CminusA dB SPL levels (psychoacoustics ear-physiology) rolloff Hz Spectral shape need for richer & more nuanced measures!!! spectral spread Hz 2 N10m90 sone “notice-events” (De Counseel, B.) “events” event density Hz -1 simulators for acoustic design projects tempo BPM (Hellström B., Lundén, P.) Monday, July 29, 2013

  4. Lindborg: EAT study EAT112: Categorisations 112 eating-places. On-site SPL, audio, annotations, photography (Oct. 2012). Monday, July 29, 2013

  5. Lindborg: EAT study EAT112: Categorisations 112 eating-places. On-site SPL, audio, annotations, photography (Oct. 2012). 112 eating-places by categories Bar&Buffet Cafe Fastfood Hawker Restaurant Chinese MixFusion OtherAsian Bar&B Cafe Fastfood Hawker Resto sum Chinese 0 6 3 5 7 21 MixFusion 4 5 5 13 4 31 OtherAsian 2 3 2 4 12 23 Western 1 11 11 0 14 37 Western sum 7 25 21 22 37 112 Monday, July 29, 2013

  6. Lindborg: EAT study 112 eating-places by categories EAT112: Categorisations Bar&Buffet Cafe Fastfood Hawker Restaurant Chinese 112 eating-places. On-site SPL, audio, annotations, photography (Oct. 2012). MixFusion OtherAsian Western Leq by Food Style (broad) in 118 eating-places Priciness by Food Style in 118 eating-places 82 1.0 80 C-weighted 78 0.5 76 Priciness Leq 74 0.0 A-weighted 72 -0.5 70 68 Chinese MixFusion OtherAsian Western Chinese MixFusion OtherAsian Western Monday, July 29, 2013

  7. Lindborg: EAT study EAT112: Swedish Soundscape Quality Protocol (SSQP) SSQP 'qualia' for 112 eating-places Chinese Bar&Buffet Cafe eventful MixFusion 3 Fastfood OtherAsian Hawker e c Restaurant Western x i t c o i a t i n h 2 g c 1 Eventfulness annoying pleasant 0 -1 m o n m -2 o t l o a n c o u s uneventful -3 -3 -2 -1 0 1 2 3 Pleasantness Monday, July 29, 2013

  8. Lindborg: EAT study SSQP 'qualia' for 112 eating-places EAT112: ANOVA Chinese Bar&Buffet Cafe eventful 3 MixFusion Fastfood OtherAsian Hawker c e Western Restaurant i x t o c i a t i h n 2 c g 1 Eventfulness annoying pleasant 0 -1 ANOVA with Pleasantness, Eventfulness, Occupancy, Overall m o n m -2 o t l o a n c o u s Sonic, and Overall Visual as response variables to three factors: -3 uneventful Situation (Indoors or Outdoors), Food Style (4 different), and -3 -2 -1 0 1 2 3 Design Style (5 different) revealed four significant effects: Pleasantness Rated Overall Sonic and Visual Environments Overall Visual Quality and Design Style compared with Occupancy (F=5.54, p=0.000***, eta 2 =0.18). 5.0 Overall Sonic Quality and Design Style 4.5 (F=2.90, p=0.026*, eta 2 =0.10). 4.0 rating 3.5 Occupancy and Design Style (F=2.66, p=0.037*, eta 2 =0.094). 3.0 2.5 Occupancy and Situation 2.0 (F=5.56, p=0.020*, eta 2 =0.051). … no significant effect on SSQP ‘qualia’ 0.96 0.94 occupancy 0.92 0.90 Bar&Buffet Cafe Fastfood Hawker Restaurant Monday, July 29, 2013

  9. Lindborg: EAT study Rated Overall Sonic and Visual Environments compared with Occupancy Result: Quality ratings correlate negatively with Occupancy 5.0 4.5 4.0 rating 3.5 3.0 Correlation between Occupancy and 2.5 2.0 ratings of Sonic and Visual Environment Overall Quality 0.96 1.02 0.94 occupancy 0.92 0.90 1.00 Bar&Buffet Cafe Fastfood Hawker Restaurant 0.98 0.96 occupancy 0.94 0.92 0.90 0.88 1 2 3 4 5 mean ratings (Likert scale) Monday, July 29, 2013

  10. Lindborg: EAT study EAT112: Multivariate regression MVR of response features onto independent variables (ß coefficients) r 2 adj. Priciness SoundMass VariaFocus Pleasantness 0.025 -0.028 -0.092 0.216* Eventfulness 0.020 0.098 -0.060 -0.219* Occupancy 0.073 0.26** -0.127 -0.225* (SSQP Overall Sonic / Visual Quality not used in MVR bcs heteroskedasticity = non-normality of distributions) Monday, July 29, 2013

  11. Lindborg: EAT study EAT15: Multimodal descriptions Studio ratings (N=31) of 15 places from EAT112. Comparing descriptions of different modality: acoustic, semantic (scales and free-form), visual (position and colour). MVR including personality features of raters [3]. Monday, July 29, 2013

  12. Lindborg: EAT study EAT15: Acoustic features & SSQP ‘qualia’ Biplot of Qualia in circumplex model PCA biplot of acoustic features (EAT15) (95% confidence ellipses) (rotated 4.95 rad, mirrored around x-axis) eventful 3 RafMar e c x 1.0 i t X_Zero.crossingrate.bx c o i a t i n h c g Wishbo 2 6 SelClb 0.5 PC2' � VariaFocus (31.5%) 12 14 13 1 BeachB CoffB StbxE 9 Eventfulness 2 unpleasant pleasant 5 11 1 Vanner 0.0 N10.bx 3 0 CryJad 4 Fuller spectspread.noRMS.bx Jacks TWG 8 7 StbxW -1 LiuSan -0.5 Tobys pulseclar.bx 15 m -2 o n regularity.bx m o t l a o -1.0 n c 10 rhythmpeak.bx o u s -3 SelRes uneventful -3 -2 -1 0 1 2 3 -1.0 -0.5 0.0 0.5 1.0 PC1' � SoundMass (35.7%) Pleasantness Monday, July 29, 2013

  13. Lindborg: EAT study EAT15: Free word association feelings: perceived evoked sound Monday, July 29, 2013

  14. Lindborg: EAT study EAT15: Free words (rated in VA quadrants, N=3) compared with SSQP ‘qualia’ Biplot of Qualia in circumplex model (95% confidence ellipses) eventful e c x 1.0 i t c o i a t i n h c g 6 0.5 12 14 13 9 Eventfulness 2 unpleasant pleasant 5 11 1 0.0 3 4 8 7 -0.5 15 m o n m o t l a o -1.0 n c 10 o u s uneventful -1.0 -0.5 0.0 0.5 1.0 Pleasantness Monday, July 29, 2013

  15. Lindborg: EAT study EAT15: Visual sorting task compared with SSQP ‘qualia’ Interface 'white square', mean absolute blob position (50% conf.ellipses) 0.15 6 12 14 2 0.10 7 15 3 4 13 1 5 11 0.05 9 0.00 8 10 -0.05 -0.10 -0.05 0.00 0.05 0.10 Monday, July 29, 2013

  16. Lindborg: EAT study EAT15: Visual sorting task compared with SSQP ‘qualia’ Biplot of Qualia in circumplex model Interface 'white square', mean absolute blob position (95% confidence ellipses) (50% conf.ellipses) eventful 0.15 e c x 1.0 i t c o i a t i n h 6 c g 12 14 2 6 0.10 0.5 12 7 14 15 3 13 4 13 1 9 Eventfulness 2 5 unpleasant pleasant 5 11 11 1 0.05 0.0 3 4 8 7 -0.5 9 0.00 8 15 m 10 o n m o t l a o -1.0 n c 10 o u s -0.05 uneventful -0.10 -0.05 0.00 0.05 0.10 -1.0 -0.5 0.0 0.5 1.0 Pleasantness Monday, July 29, 2013

  17. Lindborg: EAT study EAT15: Multimodal description are largely congruent Congruency as MDS similarity (Kendall’s tau ) between descriptions. (LAB=colour in CIE Lab space) SMVFa PLEVz LABz PLEVz 0.299 * (28) LABz 0.166 (34) 0.505 *** (34) POSz 0.218 (28) 0.586 *** (28) 0.552 *** (34) PLEVz = SSQP Pleasantness-Eventfulness LAB = colour in CIE Lab space (3 dimensions) POS = vertical-horisontal visual sorting task SMVF = acoustic features (SoundMass-VariaFocus) -z = values are z-transformed for each participant Monday, July 29, 2013

  18. Lindborg: EAT study EAT15 9 
“don’t
like” [Starbucks] music
(11) 6 
“don’t
like” background
(3) eventful noise
(7) birds
(4) kitchen
(3) good
(2) e c x i t [Pasir
Ris
Beach
Bar] c o construcCon
(4) loud
(2) i a t i n h c g loud
(4) pop
(2) song
(2) 6 9
“like” 12 14 music
(5) 13 [Coffee
Bean] 9 birds
(5) 2 unpleasant pleasant 5 11 1 5
“like” 3 4 Juice/coffee/boiler
machine
(3) 8 7 5
“don’t
like” 15
“don’t
like” blending/coffee
machine
(4) crickets
(10) 15 kitchen
noise
(2) m o n m traffic
(4) o t l vacuum
cleaner
(3) a o n c 10 o u air
(2) s 15
“like” insect
(2) [Robertson
Quay] uneventful outdoor
(2) 10
“don’t
like” 10
“like” peaceful/calm
(3) fan
(11) voices
(2) sound
(4) noise
(5) fan
(4) birds/chirping
(2) talking
(4) [Seletar
Country
Club] machine
(3) Monday, July 29, 2013

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend