SYN-AUD-CON Acoustic Test and Measurement Seminar The Early Sound Field: Properties, Perceptual Attributes, and Measurement Methods

1
SYN-AUD-CONAcoustic Test and Measurement
SeminarThe Early Sound Field Properties,
Perceptual Attributes, and Measurement
Methods NEIL THOMPSON SHADE President/Principa
l ConsultantAcoustical Design Collaborative,
LtdRuxton, MD Director, M.A. Acoustics
ProgramPeabody Institute of Johns Hopkins
UniversityBaltimore, MD  28 29 August 2002
2
Overview of Talk

• For the Early Sound Field I Will Discuss
• History
• Simple Definition
• Differences and Acoustical Attributes
• Psycho-Acoustic and Perceptual Factors
• Related Metrics Used for Evaluation
• Doing Measurements and Interpretation

3
Some History on Early Reflected Sound Field

• H. Matthews, London, 1826 book Observations on
  Sound Shewing the Causes of Its Indistinctness
  in Churches, Chapels, Halls of Justice c. with a
  System for Their Construction
• That part of the sound which proceeded in a
  direct line would arrive first and that
  reflection from an oblique wall, having to
  perform a longer journey, would arrive later.
  Echo does not politely wait until the speaker is
  done but the moment he beings and before he has
  finished a word, she mocks him as with ten
  thousand tongues.
• From Buildings for Music, Forsyth, M. (1985)
  Cambridge, MA MIT Press

4
Some History on Early Reflected Sound Field

• Joseph Henry studied audibility of reflections at
  Smithsonian Lecture Hall (around 1850)
• CONCLUSION Echoes not perceptible if listener
  within about 50 ft of reflecting surface.
• Helmut Haas studied audibility of secondary
  sources (19491951)
• CONCLUSION Level of secondary source can be up
  to 10 dB higher than primary source and not be
  perceived if within 30 to 35 ms of secondary
  source.
• Schultz and Beranek studied late-to-early sound
  index (19631965)
• CONCLUSION Sound arriving within 50 ms of
  direct sound is useful for determining sense of
  reverberance with actual program sources. Called
  it running reverberance.

5
Some History on Early Reflected Sound Field

• Riechardt studied speech clarity (C50) and
  musical clarity (C80) (19731975)
• CONCLUSION Speech clarity important for sound
  within 50 ms and music clarity important for
  sounds arriving within 80 ms.
• Harold Marshall studied lateral reflections
  (19671968)
• CONCLUSION Reflections within 50 to 80 ms after
  direct sound from side walls increase perceived
  envelopment and spatial impression.
• Michael Barron studied lateral reflections based
  on Marshalls work (19711972)
• CONCLUSION Delay of 20 to 80 ms has slight
  influence on spatial impression, but level
  differences between primary and secondary sources
  more important.

6
Importance of Early Sound Field

• Early Sound Field Provides the Room Sound
• Rooms with Same Reverberation Time Will Sound
  Different
• Due to Early Reflections
• Often Desirable to Have Strong Early Reflections
• Speech Improves Clarity and Level
• Music Helps With Intimacy, Definition,
  Spaciousness, and Musical Ensemble
• Strong Early Reflections Directed at Audience Can
  Alter the Late Reflected Sound Field Both in
  Duration (Time) and Strength (Level)
• Early Reflections Not Desirable in Studio Rooms
  
• Effects Sense of Localization and Timbre of Sound

7
Importance of Early Sound Field
Haas/Precedence Effect Data
8
Importance of Early Sound Field
Barron Spatial Perception Data
9
Importance of Early Sound Field
Toole/Olive Spatial Perception Data
10
Importance of Early Sound Field
Toole/Olive Audibility of Single Reflection Data
11
Description of Early Sound Field

• Early Sound Field Is Room and Position Dependent
• Based on Room Size, Geometry, and Location of
  Sound Reflecting Surfaces
• Approximate Guidelines for Early Sound Limits
• Large Rooms (Concert Halls, Large Churches,
  Arenas) 0 to 150250 ms
• Medium Rooms (Recital Halls, Most Churches, Large
  Lecture Rooms) 0 to 50100 ms
• Small Rooms (Studios, Screening Rooms,
  Classrooms) 0 to 1030 ms
• Idealized Room Response
• 1. Direct Sound (LD)
• 2. Initial Time Delay Gap
• 3. Early Reflections (LRE)
• 4. Build-up of Reverberant Field (LR)
• 5. Steady State Energy Balance (LR)
• 6. Decay of Reverberant Field (LR)
• 7. Ambient Noise (LAMB)

12
Description of Early Sound Field
13
Psycho-Acoustic Aspects
Early Sound and Perception Factors

• Perceptual Attribute
• Reverberance
• Speech ClarityMusic Clarity
• Spaciousness
• Envelopment
• Intimacy
• Stage Support

• Physical Parameter
• Early Decay Time (EDT)
• C50C80
• Interaural Cross CorrelationCoefficient
  (IACCEARLY)Lateral Fraction (LFEARLY)
• Interaural Cross CorrelationCoefficient
  (IACCLATE)Lateral Fraction (LFLATE)
• Initial Time Delay Gap (ITDG)
• ST1

14
Definition of Metricsto Assess Early Sound Field

• Early Decay Time (EDT) Units sec
• Measures sound decay from 0 to -10 dB down and
  extrapolates decay to full 60 dB decay limit.

15
Definition of Metricsto Assess Early Sound Field

• Clarity C50 and C80 Units dB
• Measures energy ratio of early time (0 to 50 ms
  or 0 to 80 ms) to late time (50 ms to infinity
  or 80 ms to infinity).

16
Definition of Metricsto Assess Early Sound Field

• Inter-Aural Cross Correlation Coefficient (IACC)
• Measures difference in sounds arriving at left
  and right ears.
• IACC Early (IACCE3) Units none
• Measures sound over 0 to 80 ms time period at
  500, 1,000, and 2,000 Hz bands and averaged.
• Evaluates spaciousness (apparent source width).

17
Definition of Metricsto Assess Early Sound Field

• IACC Late (IACCL3) Units none
• Measures sound over 80 to 750 ms time period at
  500, 1,000, and 2,000 Hz bands and averaged.
• Evaluates sense of envelopment.

18
Definition of Metricsto Assess Early Sound Field

• Lateral Fraction (LF)
• Measures energy ratio of sound arriving from side
  walls to total sound.
• Lateral Fraction Early (LFE4) Units none
• Measures sound over 5 to 80 ms time period at
  125, 250, 500, and 1,000 Hz bands and averaged.
• Evaluates sense of spaciousness.

19
Definition of Metricsto Assess Early Sound Field

• Lateral Fraction Late (LFL4) Units none
• Measures sound over 80 to infinity ms at 125,
  250, 500, and 1,000 Hz bands and averaged.
• Evaluates sense of envelopment.

20
Definition of Metricsto Assess Early Sound Field

• Initial Time Delay Gap (ITDG) Units ms
• Time difference between arrival of direct sound
  and first significant reflection.
• Evaluates sense of intimacy.

21
Definition of Metricsto Assess Early Sound Field

• Stage Support (ST1) Units dB
• Measures energy ratio of sound arriving from 0 to
  10 ms to sound arriving from 20 to 100 ms.
• Evaluates reflections that provide useful support
  to musicians to hear each other.

22
Measurements and Data Interpretation

• General
• Recommended Equipment
• Omni-directional sound source
• Omni-directional microphone (pressure or random
  incidence)
• Generate impulse response or Energy Time Curve
  (ETC)
• Pink noise, MLS signal, or swept sine signal
• Data collection procedures in ISO 3382,
  Measurements of Reverberation Time of Rooms with
  Reference to Other Acoustical Parameters

23
Measurements and Data Interpretation

• Measurement Methods
• Measurement in octave bands 125 to 4,000 Hz
• Minimum distance mic-to-source given by
• DMIN 2Ö (V/cT)
• V room volume, m3
• C 343 m/s
• T mid-frequency reverberation time, sec

24
Measurements and Data Interpretation

• Measurement Methods
• Excitation time of signal must exceed room
  reverberation time.
• Minimum number of source positions 3
• Stage Left
• Stage Center
• Stage Right

25
Measurements and Data Interpretation

• Measurement Methods
• Minimum number of receive (mic) positions based
  on room size (number of seats)

Minimum Number of Mic Positions 6 8 10
Number of Seats 500 1,000 2,000
Total number of measurements need to multiply
the number of source positions (3) by the minimum
number of mic positions.
26
Measurements and Data Interpretation

• EDT
• Measure in octave bands 125 to 4,000 Hz.
• Average all measurement values in each octave
  band.
• Characterize EDT at 500 and 1,000 Hz average or
  250 to 2,000 Hz average values.

Shaded Positional dependent range of values
27
Measurements and Data Interpretation

• Clarity
• Measure in octave bands typically limited to 500,
  1,000, and 2,000 Hz.
• Limit of direct plus early sound set to 50 ms
  for speech clarity and 80 ms for music clarity.
• Values averaged to obtain a single value for 3
  frequencies and all measurement positions.

Shaded Positional dependent range of values
28
Measurements and Data Interpretation

• IACC
• IACC mics at right and left ears in dummy head to
  evaluate differences in sound levels arriving at
  right and left ears.
• IACC Early (IACCE3)
• Measure in 500, 1,000, and 2,000 Hz octave
  bands over 0 to 80 ms time period.
• Values averaged to obtain a single value for 3
  frequencies and all measurement positions.

Shaded Positional dependent range of values
29
Measurements and Data Interpretation

• IACC Late (IACCL3)
• Measure in 500, 1,000, and 2,000 Hz octave bands
  over 80 to 750 ms time period.
• Values averaged to obtain a single value for 3
  frequencies and all measurement positions.

Shaded Positional dependent range of values
30
Measurements and Data Interpretation

• Lateral Fraction (LF)
• LF energy ratio of output of figure-of-8 mic with
  lobes pointed towards side walls (CH 1) to
  omni-directional mic (CH 2).
• Lateral Fraction Early (LFE4)
• Measure in 125, 250, 500, and 1,000 Hz bands
  over 5 to 80 ms time period.
• Values averaged to obtain a single value for 4
  frequencies and all measurement positions.

Shaded Positional dependent range of values
31
Measurements and Data Interpretation

• Lateral Fraction Late (LFL4)
• Measure in 125, 250, 500, and 1,000 Hz bands over
  80 ms to infinity time period.
• Values averaged to obtain a single value for 4
  frequencies and all measurement positions.
• ITDG
• Measure in center of seating area at 500 or 1,000
  Hz.
• Time difference between direct sound and first
  reflection.

32
Measurements and Data Interpretation

• ST1
• Measure with music stands, chairs, and percussion
  on stage at 250, 500, 1,000, and 2,000 Hz.
• Values averaged to obtain a single value for 4
  frequencies and all measurement positions.

33
Summary and Conclusions

• Studies about early sound field nearly 150 years
  old.
• Studies continue to evolve relating perception to
  measurement parameters.
• Early sound field is a major factor in room
  sound.
• May be more important than reverberation time.
• Early sound field measurements different than
  standard acoustical measurements.
• Special equipment or software needed to properly
  evaluate.
• Measurement data very positional dependent
  requires numerous measurement locations to
  quantify.

34
Summary and Conclusions

• Important early sound metrics include
• Early Decay Time (EDT)
• Speech Clarity (C50)
• Music Clarity (C80)
• Interaural Cross Correlation Coefficient
  (IACCEARLY)
• Interaural Cross Correlation Coefficient
  (IACCLATE)
• Lateral Fraction (LFEARLY)
• Lateral Fraction (LFLATE)
• Initial Time Delay Gap (ITDG)
• Stage Support (ST1)

35
Summary and Conclusions

• Appropriate values of early sound metrics
  dependent on type of music (organ, orchestral,
  chamber, soloist, pop/jazz) or speech programs
  (talk, drama, opera).
• Typical ranges
• C50 Opera gt 0 dB Talks and Drama gt 2 dB
• C80 Organ lt -4 dB Symphony -2 to 2
  dB Chamber 0 to 2 dB Pop/Rock gt 2 dB
• EDT Organ gt 2.5 sec Symphony 1.8 to 2.2
  sec Chamber 1.3 to 1.6 sec Opera 1.4 to 1.6
  sec Pop/Rock lt 0.8 sec

36
Summary and Conclusions

• Typical ranges
• IACCEARLY lt 0.50
• IACCLATE lt 0.50
• LFEARLY lt 0.40
• LFLATE lt 0.40
• Initial Time Delay Gap (ITDG) lt 25 ms
• Stage Support (ST1) -13 to -11 dB