Comprison of soundscpe on the ground floor of tue-houses in Hnoi nd open urn spce in Bordeux Tun Anh Nguyen GRECAU Bordeux, Ecole Ntionle Supérieure d Architecture et de Pysge de Bordeux, Tlence, Frnce ABSTRACT: Ech urn spce hs its own soundscpe: specific sound level distriution ccording to its urn form. Urn morphology nd soundscpes re the topics of mny reserch fields. The definition of the reltionship etween urn morphology nd its soundscpe gives the clerest pprecition of urn forms through their soundscpes. This pper lso pproches the dominnt urn morphology fetures nd specific soundscpes on the studied res of Hnoi nd Bordeux. Besides, comprison etween estern nd western urnism soundscpes is mde. Keywords: Soundscpe, urn morphology, urn form, open spce, tue house. 1. INTRODUCTION The reserch dels with Hnoi s old town nd Sint- Michel qurter in Bordeux. Hnoi s old town strted s gthering of rtisns on trditionl socio-economic sis resulting in multi-purpose spce used for living, trding nd crfting. The sis structurl elements re the tue-houses, which hve nrrow front (from 3 to 4 meters) nd long depth (from 20 to more thn 50 meters). Becuse of the fct tht those houses host t the sme time goods production nd sles, they need to hve ig ground floor spces. Tht is why most of the houses hve their min ground floor doors open. Trding goes on from the morning till ten o clock in the evening during the week, resulting in noisy environment throughout the dy. The sound levels chnge clerly from indoor spces to outdoors ones [1]. Sint-Michel qurter in Bordeux hs lso n ncient history, which hs given it its own chrcter nd specific fetures. The vehicle circultion wys set its contour, it is linked to the rest of the city through secondry rods nd it is well connected to the surrounding qurters. The circultion xis converges into Sint-Michel squre [2]. U-shped streets nd nrrow ded-ends divide the re into locks of smll dimensions resulting in low numer of dwellings per lock. The smll distnce etween consecutive streets lso results in mny junctions nd empty spces. We cn lso find in the studied zone, plce of vitl importnce for the qurter: the Meynrd squre, which is ig open spce. The permnent soundscpe is different in the open spces nd in the closed ones. 2. EQUIPMENT, RESEARCH METHOD AND REPRESENTATION DATA 2.1 Equipment The equipment used ws the following: - Dt recorders (Tscm DA-P1, Sony TCD-D1) - DAT tpes - Sennheiser microphones (MKE 2002 nd MKE gold) set s dummy hed clled SEB (Binurl record system). - Symphonie Crd. - Clirtor. To trnsfer the recorded udio signls from the DAT mchine to the computer, we used the 01dB softwre dbf 32. Both the recordings done in Hnoi nd in Bordeux were crried out using this equipment. (Fig 1) Figure 1: Sound recording nd dt nlysis equipment. 2.2 Reserch method The method proposed here is sed on new concept clled the soundwlk. This kind of mesurement consists on wlking long pre-defined route on the re under study while recording the sounds tking plce throughout the wlk. The wlker hs to crry two microphones, positioned next to their ers so tht inurl recording is done. In order to chieve the est results we performed the most pproprite kind of soundwlk on ech site: simple, continuous or prllel. In ddition to this method, we implemented fixed sound recording method. This
kind of mesurement consists on stying on specific spot for certin time (not shorter thn 5 minutes) while recording the sounds tking plce there y mens of the inurl recording technique explined efore. Pictures were lso tken while doing the sound recordings. (Fig 2) Figure 2 : Soundwlker nd sound recording on fixed point. 2.3 Dt representtion After trnsferring the recorded signls to computer, it is possile to otin grphic representtion of the verge sound level over frequency nds (thirds of octve) for given time period. A 3D view of tht would e clssic representtion. However, 2D representtion, clled the coustic imge, enles n esier dt interprettion y showing the time on the horizontl xis, the frequency on the verticl one nd rnge of colours representing the sound levels. This colour scheme is explined in key, which shows the colours corresponding to the different sound level nds (10dB nds). [3] (Fig 3) Figure 4 : Hnoi s old town. Nowdys, the tue-houses hve ecome multipurpose spces where people live nd goods re produced nd sold. Most of the production nd trding ctivities tke plce y the houses min ground floor doors nd tht is the reson why those doors re usully open. Besides, production ctivities require lrge spces nd opening the min doors, helps mximize tht fctor. [1] (Fig 5) Figure 5: Ground floors in the streets. 3.2 Soundscpe fetures Hnoi s old town dtes from the seventeenth century. It linked trditionl crft villges to direct trde sis. It emerged upon historic se nd nowdys it hs ecome ig shopping re for the city. It hs every dominnt feture of trditionl commercil qurter. The high popultion nd circultion densities generte noisy soundscpe throughout the dy. The sounds ssocited to the production ctivities lso contriute to mke the soundscpe noisy. [4] (Fig 6) Figure 3: Acoustic imge nd its key 3. THE CASE OF HANOI S OLD TOWN 3.1 Urn morphology fetures. As shown elow, Hnoi s old town is prticulr re of the city. It hs specific urn morphology with nrrow streets nd smll tue houses. Those fctors result in high numer of dwellings per lock. Figure 6 : High popultion nd circultion densities in the streets. 3.3 Sound recording results. The sound recordings were done on the ground floor of tue house in the lock under study. Two different recordings were done, one of them inside the
house nd the other one outside. Both of them were done in fixed position during 5 minutes. The recording point chosen for the inside recording is in the middle of the ground floor. The one chosen for the outside recording is on the sidewlk outside the house. (Fig 7, 8) frequency re rnging from 31.5Hz to 5Khz. The sound perception depends directly on the outside sound sources. The sound level is the sme on oth ers. Tht shows tht the trffic on the two-wy rod is continuous. The highest sound level (80-90dB) is due to the mopeds nd crs horns. (Fig 9) Figure 7 : Pln nd section of the house where the mesurements were tken (101, Thuoc Bc St) Outside recording point Inside recording point Figure 8 : Outside nd inside recording points. We cn clerly see tht the distnce etween the two recording points is short. The min sound source is the trffic noise in the street. Figure 10 : Acoustic imges otined from the inside recordings. The inside recording coustic imge shows very different sound level spectrl distriution in comprison to the outside recording one. This is specilly noticele over the zone where we cn clerly pprecite the tue houses ttenution effect over certin frequency nds (filtering effect). Both points coustic imges show high sound levels (70-80 db) distriuted over the sme frequencies. The comprison etween the coustic imges revels the effect of the diffuse field due to the ground floors urn form. (Fig 10) 4. BORDEAUX SAINT-MICHEL QUARTER CASE STUDY 4.1 Urn morphology fetures. Sint-Michel qurter is Bordeux city centre specific re. Dominnt fetures of ncient urn morphology re concentrted there. Sint-Michel qurter Figure 9: Acoustic imges otined from the outside recordings. This coustic imge shows high noise level in the street. The noise min level is 60-70dB. We cn lso notice tht this noise spreds over very lrge Figure 11 : Bordeux centre. A typicl zone of Sint-Michel qurter ws chosen for the study. The sme commercil ctivities nd trding fshions s in Hnoi s old town cn e found there. The studied re is divided into smll locks y
nrrow streets nd ded-ends. Those smll locks re divided into tenements. The long nd nrrow locks re uilt one fter the other. The empty spces s the courtyrds nd the smll squre lend in with the urn structure. Meynrd nd Duurg squres re ig open spces next to the studied re. [5] (Fig 11) Cours Victor-Hugo Qui des Slinières Figure 14 : Soundwlk Trffic s sound source Soundwlk pth Rue des Fures The soundwlk strts t the eginning of Rue des Fures. (t its junction with Cours Victor Hugo), goes cross Meynrd squre ( zone), then cross Duurg squre ( d zone) nd finlly heds towrds Qui des Slinières. The min sound source is the trffic, which is fcing the left er. Sint-Michel squre Figure 12: Big empty, open spces. Duurg squre 4.2 Soundscpe fetures: Three importnt fctors re to e highlighted round the re under study: - The high trffic density long Cours Victor Hugo. - A complex trnsport junction etween Bir- Hkeim squre nd Qui des Slinières. - The commercil nd leisure ctivities in Rue des Fures nd in Sint-Michel squre. The soundscpe in Sint-Michel qurter shows the noise generted y trffic nd commercil ctivities like in Hnoi s old town. The trding fshions in Sint-Michel re outdoors mrkets, greengrocers nd off-licences. The commercil ctivity tkes plce every morning nd it is hevier during the weekend. [5] Figure 15: Acoustic imges otined from the soundwlk recording. The coustic imge shows how the equivlent sound level spectrl distriution is not uniform throughout the soundwlk. The zone Meynrd squre is clerly, n open spce where the sound field is well defined. The equivlent sound level on the left er is higher thn on the right one ecuse the left er is fcing the sound source (trffic) wheres the right one is fcing the open spce nd it does not get the reflected sounds. Also, the only sound sources in Meynrd Sq. re the humn voices. The equivlent sound level of 50-60dB is minly cused y the wind. (Fig 15) The soundscpe fetures on the fixed recording points will e now nlysed. Recording point Figure 13 : Rue des Fures nd Cours Victor-Hugo. 4.3 Sound recording results. In order to study the distinctive fctors of the open spce of qurter, two sound recordings were performed: soundwlk nd recording in fixed position. Both recordings were done in Rue des Fures. Figure 16: Fixed recording points Trffic s sound source These sound recordings represent specific points in Rue des Fures. Recording point 1 is locted on closed spce (reflected field) nd recording points 2 nd 3 re locted on n open spce (free field). We re not compring the results etween different
recording points ut nlysing the soundscpe fetures round them, especilly in the open spce (recording point two). (Fig 16) - The equivlent sound level spectrl distriution is different on ech sound field. In the U shped streets in Hnoi, the sound field is clerly diffuse due to the urn form of the ground floors. However, the equivlent sound level spectrl distriution is uniform in the street nd not inside the houses. In Sint- Michel qurter in Bordeux, the sound distriution is uniform in the U shped streets (diffuse sound field) ut it is not in the open spce. 5. CONCLUSION Figure 17: Acoustic imge otined from the recording done on point one. We cn see from the coustic imge otined from the recording done on point 1, tht the spectrum is quite different, due to trffic density reduction. The equivlent sound level distriution is equivlent for oth ers. However, the equivlent sound level in zone is different ecuse there is some noise coming from Cours Victor Hugo. The highest sound levels (80-90dB) re eqully distriuted over frequency on oth ers though ( zone). (Fig 17) c d Figure 18 : Acoustic imge otined from the recording performed on point 2. We cn see this coustic imge how the equivlent sound level spectrl distriution is not the sme on oth ers. There re noticele differences etween the left nd right er coustic imges ecuse the microphones were exposed to different sound sources in the open spce ( c zone). The highest sound levels (80-90dB) re eqully distriuted over frequency on oth coustic imges though (zone d). (Fig 18) We cn notice how the equivlent sound level over the low frequencies is the sme on recording points 1 nd 2. Tht mens tht the sound source is quite powerful. Some generl oservtions out the specific soundscpe nd urn morphology fetures in Hnoi s old town nd Sint-Michel qurter in Bordeux re the following: - The trffic noise influence on the soundscpe is quite noticele. - The equivlent noise level is the sme on oth studied res. c d Our conclusion is tht oth sites (Hnoi s old town nd Sint-Michel qurter in Bordeux) soundscpe nlysis enles min urn morphology fetures definition. This nlysis helps us to hve n exct oservtion of urn morphology through soundscpes. ACKNOWLEDGEMENT The uthor wnts to thnk Ivn Ricoy, sound engineer, for his help during his trining course t GRECAU. REFERENCES [1] Les Chiers de l Iprus, Hnoi-Le cycle des métmorphoses, Edition Recherche/Iprus, 2001. [2] Etude pédgogique, Anlyse des plces et progrmme d méngement, École d Architecture et de Pysge de Bordeux, 1996-1997. [3] C. Semidor, The soundscpe s component of urn comfort : the cse of Plce Pul Doumer, 22 th PLEA, Beirut, Linon, 13-16 Novemer 2005. [4] T.A. Nguyen, Morphologie urine trditionnelle et minces sonores dns l'ncien qurtier de Hnoi, Symposium BMB, Bordeux, 5-7 Avril 2005. [5] T.A. Nguyen, C. Semidor, Reltionship etween urn ctivities nd soundcpe: commercil res in Bordeux nd Hnoi, 23 th PLEA, Genev, Switzerlnd, 6-8 Septemer 2006.