ROOM ACOUSTICS MEASUREMENT AND ASSESSMENT METHODS ASSESSMENT OF ROOM ACOUSTICS CAN BE DIVIDED INTO: OBJECTIVE MEASUREMENT METHODS SUBJECTIVE ASSESSMENTS OBJECTIVE METHODS 1. Signal to Noise Ratio 2. Loudness 3. Decay curves --Level.v.Time Reverberation Time T 30, T 20, EDT RT ratios Initial Time Delay Gap, ITDG 4. Energy Distribution in the Decay Lateral Energy Fraction, LF Clarity -- C 80, etc. 4. Room Resonance (modes) 6. Transfer Functions 1. Signal to Noise Background noise must be low enough so as not to affect the listener. Simple method A Weighted Level No information on failing frequencies if too high Better methods Rating Methods (NR, RC) Speech Interference Level, SIL NR (Noise Rating), NC (Noise Criterion), RC (Room Criterion) plot measured octave band sound pressure levels RECOMMENDED Offices NR25-55 Hospitals Wards NR30 Domestic Bedroom NR25 Lounge NR30 Concert Hall NR20 Recording Studios NR15 This would be NR46 Determined by the highest curve that is cut SIL (Speech Interference Level) Based on average level in the 500Hz, 1, 2, 4 khz octave bands. Determines the possibility of face to face communication so is a measure of privacy How far apart do I need to position desks in an open plan office? 1
2. LOUDNESS Loudness gives a measure of whether or not the audience will be able to hear the performance at the correct loudness level. For an orchestra the sound power of the instruments determines the maximum size of the auditorium For electronic music the level in all parts of the auditorium can be measured and adjusted to give an even and sufficient level. Loudness is given by the strength factor, G which can be determined from the reverberation time, RT. RT mid is the average of 500 Hz & 1000 Hz The best concert halls have a strength factor G between 4.0 and 5.5 db 3. DECAY CURVES Measured with interrupted sound source and a plot of the decay through time. RT and EDT calculated from the average of many decays. Compare with recommended RT for type of space and performance --- found from published data (next slide) Faults such as flutter echoes can be detected from the decay curve (and by listening) Recommended RT s WARMTH AND BRILLIANCE Warmth --- smoothness to the music. If a sound field is too warm, the hall can be undesirably "dark." Brilliance --- bright, clear, ringing sound. With too much brilliance, the sound is harsh, brittle, and metallic sounding. Balancing warmth and brilliance is achieved by balancing the ratio of low frequency Reverberation Time (RT) to high frequency RT. Thus, the Bass Ratio (BR) has been suggested as an objective measure of Warmth: A bass ratio between 1.1 and 1.25 is desirable in halls with a high RT, A bass ratio between 1.1 and 1.45 is recommended for any hall with an RT of 1.8 sec or less. http://www.concerthalls.unomaha.edu/discussion/warmbrilaural.htm INITIAL TIME DELAY GAP (ITDG) The ITDG is the time delay between the direct sound and the first reflection shown on the idealised reflectogram below. ITDG affects our perception of the spaciousness of the hall short gap intimacy long gap -- remoteness ACTUAL REFLECTOGRAM ITDG = 5 ms ITDG ITDG = 47ms 2
Measured ITDG for Various Halls (rated as good Concert Hall ITDG (ms) Amsterdam, Concertgebouw 21 Boston, Symphony Hall 15 Vienna, Gr. Musikvereinssaal 12 Basel, Stadt-Casino 16 Berlin, Konzerthaus 25 Cardiff, Wales, St. David s Hall 25 New York, Carnegie Hall 23 Tokyo, Hamarikyu Asahi 16 Zurich, Großer Tonhallesaal 14 4. ENERGY DISTRIBUTION in the Decay Note:- 25ms = 8.5m path difference (direct-reflected) CLARITY Clarity refers to how clear the sound quality is. Can you hear every separate note distinctly, or do the notes tend to blur into one another? Some blending is often desired for music, but for speech and opera, greater clarity leads to better speech intelligibility. C80, the Clarity Index, is the ratio of energy <80ms to >80ms used for music. (C50 for speech) C 80 = +5dB, C 80 = - 13 db Spatial Impression Reflections from side walls affects the spatial impression. Lateral reflections can be measured using directional figure-of-eight microphones. Lateral / total energy fraction, LF determined E LF = E Walls Total early LF affects auditory source width (ASW) Late LF (>80ms) affects listener envelopment (LE) Popular concert halls have early LF = 0.2 (mid frequencies) Recommendations for room acoustic criteria (Gade et al [2007]) 5. ROOM RESONANCES Measured using a swept sine signal and plotting the results or by white noise and FFT Evenly spaced modes are OK 3
6. Transfer Functions There are various methods:- Speech Transmission Index (STI) Uses complex sound over a 7 octave frequency range to simulate speech Rapid Speech Transmission Index (RASTI) Simplified method uses only 2 octave bands Impulse methods Dirac Delta (Measures the RTF) This can be used to predict STI / RASTI & most other acoustical parameters SPEECH TRANSMISSION INDEX (STI) Developed in the early 1970 s, the STI is a machine measure of intelligibility. A signal (80-12kHz) modulated by low frequency sine waves and designed to approximate long-term spectral characteristics of speech is generated. When transmitted through the air the signal becomes corrupted by both background noise and reverberation reducing the effective modulation of the original signal. The STI is a measure of the amount of information in the original signal that is transmitted to the listener. If STI = 0 (completely unintelligible) If STI = 1 (perfect intelligibility). Methodology Instrumentation The Modulation Transfer Function (MTF) is measured Results If the MTF is measured over a wide range of frequencies it is possible to identify any acoustical problems A constant MTF at all frequencies indicates the cause as background noise A constantly decreasing MTF with frequency indicates excessive reverberation An MTF which first decreases and then increases with frequency indicates an echo. SUBJECTIVE METHODS 1. Musicians Criteria 2. Listeners Assessments A. Speech B. Music If measured in only two octave bands 500Hz and 2kHz (RASTI) Only the index is available 4
1. Musicians Criteria Room should respond to their instruments (should have reflections back from the listening area to the stage) Musicians should be able to hear each other (reflecting surfaces around the performing area) 2. LISTENERS -- SPEECH PERCEIVED ARTICULATION INDEX, AI Traditionally the method used to assess speech intelligibility is for different speakers to read out a series of phonetically balanced words or phrases. Listeners in different locations write down what they hear the words to be. Normally if a word is misunderstood it is because the consonants are not heard properly. The number of words understood out of the total is expressed as a percentage as the phonetically balanced (PB) word score or perceived %AI Method Typical Word List Listeners write down what they think they heard and then check this against the original. 95% accuracy --- best possible 80% accuracy --- audience will understand every sentence 75% accuracy --- listener has to concentrate 65% accuracy --- intelligibility is poor Note that AI can also be predicted from measurements of the room transfer function (which can be predicted using modeling) are bad bar bask box cane cleanse clove crash creed death deed dike dish end feast fern folk ford fraud fuss grove heap hid hive hunt is mange no nook not pan pants pest pile plush rag rat ride rise rub slip smile strife such then there toe use wheat Typical Phrase List Oak is strong and also gives shade. Cats and dogs each hate the other. The pipe began to rust while new. Open the crate but don't break the glass. Add the sum to the product of these three. Thieves who rob friends deserve jail. The ripe taste of cheese improves with age. Act on these orders with great speed. The hog crawled under the high fence. Move the vat over the hot fire COMPARISON OF PHONETICALLY BALANCED WORD SCORE AND STI Measured speech transmission index, STI % 5
2. LISTENERS MUSIC Links EVALUATION GUIDE TO BE COMPLETED BY EXPERIENCED LISTENER CLEAR SOUND...BLURRED SOUND (varies from clear or blurred distinct to or muddy) LIVE REVERBERANCE..DEAD REVERBERANCE (liveness or persistence of sounds) mid-frequency WARM BASS..COLD BASS (relative liveness reverberance at bass of bass or longer duration of compared with mid and treble frequencies) INTIMATE SOUND REMOTE SOUND (auditory impression orchestra) of apparent closeness of SATISFACTORY LOUDNESS UNSATISFACTORYLOUDNESS (indicate early or direct sound (symbol reverberant sound (symbol R) on scale D) and RICH DIFFUSION POOR DIFFUSION (envelopment of sound surrounds the listener from which many directions) GOOD BALANCE POOR BALANCE (observe between choir, and amongst sections musicians and soloists or of the orchestra) SATISFACTORY BACKGROUND NOISE UNSATISFACTORY BACKGROUND NOISE (from HVAC equipment, or intruding from ancillary spaces or outdoors) ECHOES No Yes Direction. (long delayed reflections that are clearly heard) OVERALL IMPRESSION. Paper giving the assessment of 58 concert halls http://www.leoberanek.com/pages/eightyeighthalls.pdf Concert Hall Acoustics http://www.concerthalls.unomaha.edu/discussion.htm History of acoustical parameters http://www.arauacustica.com/files/publicaciones_relac ionados/pdf_esp_70.pdf 6