ICT Resource Microphone Placement

Transcription

1 ICT Resource AV Design Calculators, Tools and Resources Microphone Placement Page 1 Deakin esolutions Information Technology Services Division Audio Visual and Networks Unit ICT Volume 2.7 AV Design Calculators, Tools and Resources ICT Resource Microphone Placement Document History Ver. Primary Author(s) Description of Version Date Completed Neil Clarke Initial drafting 26/05/ Joy Gin Peer review, published to web site 02/06/ Neil Clarke Drafting continues 08/07/ Neil Clarke Drafting continues Overview section added 09/08/ Neil Clarke Drafting continues 22/09/ Neil Clarke Drafting continues MIC20 added 10/02/ a Neil Clarke Drafting continues MIC20 adjusted 23/03/2013 This version contains preliminary guidance subject to refinement in the light of experience. In particular maximum microphone distances may be over or under estimates in particular acoustic environments. No microphone system can be expected to produce satisfactory results if the acoustic environment does not comply with the requirements set out at ICT Volume 2.3 Videoconferencing Standards section Copyright Deakin University Deakin ssolutions 23/03/2013

2 ICT Resource AV Design Calculators, Tools and Resources Microphone Placement Page 2 Contents 1. Overview Scope of document... 6 Structure of document: Design objective Range of cases Worked examples: Large tables Wide tables (preferred) Large Wide tables including immersive telepresence style Large wide arc tables Large wide V tables Large wide trapezoidal tables Large wide U tables Large wide rectangular tables Small Wide tables including D@YD style Small kidney tables Small wide oval tables Small wide arc tables Small half-cirle tables Small wide V tables Small wide diamond tables Small wide trapezoid tables Small wide rectangular tables Square and Circular tables Large Square and Circular tables Large circuler table Large square table Small Square and Circular tables Small circular table Small square table Long tables (non-preferred) Long Trapezoidal tables Small long trapezoid tables Large long trapezoid tables Deakin ssolutions 23/03/2013

3 ICT Resource AV Design Calculators, Tools and Resources Microphone Placement Page Long U tables Small long U tables Large long U tables Wider Large long U tables Long Rectangular tables (non-preferred) Small long rectangular tables Large long rectangular tables Wider Large long rectangular tables Appendix A Definitions and Theory Design rule General distance limits Definition of terms and symbols Theory Cardioid microphone Sensitivity vs angle Sensitivity vs distance Overall sensitivity The design rule microphone sensitivity variation Use of front-of-house loudspeakers only Minimum distance limits Maximum distance limits Appendix B Related considerations Tabletop as reflector to increase gain Tabletop as noise conductor/amplifier Installation of Audio Technica ES947 and ES947W microphones Height above table Lean Direction a person is facing Appendix C Omnidirectional microphones Appendix D Additional sensitivity contour plots Appendix E Further information Deakin ssolutions 23/03/2013

4 ICT Resource AV Design Calculators, Tools and Resources Microphone Placement Page 4 List of Figures and Tables Figure 1: Range of aspect-ratios, small rectangular tables (single microphone)... 7 Table 1: Seats per microphone large wide tables... 8 Figure 2: Large wide arc table, case of R 0.4r sensitivity field for each microphone... 9 Figure 3: Large wide V table sensitivity field for the microphone at the apex... 9 Figure 4: Wide Arc table example of 3 microphones at R 0.4r Figure 5: Wide V table example of 3 microphones Figure 7(a): Large U table Figure 7(b): Large U (or rectangular) table, with island table for microphones Figure 9: Large wide rectangular table corner (a) symmetrical, (b) asymmetrical Figure 9: Large wide rectangular table straight edges (c) off-axis, (d) 45, (e) on-axis Wide rectangular table aspect ratio range Figure 10: Small kidney table and corresponding small oval table (microphone at Rmin > 0.5W) Figure 11: Small wide oval tables at aspect ratios (a) L/W = 2/3 and (b) L/W = 1/ Small half-circle table dimensions Figure 13: Small wide diamond table at the maximum aspect ratio of W=2.3L and F=0.025W Figure 14: Small wide trapezoidal table at the maximum aspect ratios of W/L=5/3 and V/L=1 and required microphone placement at F=0.05V [sic] Figure 15: Small wide rectangular table at the maximum aspect ratio of W/L=7/4 and required microphone placement at F=0.03W Figure 16: Large square table for up to 18 seats (6+6+6) Figure 17: Small circular table Figure 18 Small square table, microphone placed at preferred position F=0.05W Long trapezoidal table dimensions Long trapezoidal table aspect ratio range Figure 19: Small long trapezoidal table, at absolute max aspect ratio L/W = 3/2 and required microphone position F=0.1V Long U table aspect ratio range Figure 20: Small long (a) U and (b) rectangular tables, preferred max aspect ratio 3/2 and preferred position F=0.1W Long rectangular table dimensions Long rectangular table aspect ratio range Figure 21: Long rectangular table with 2 microphones at preferred maximum aspect ratio 3/ Figure 22: Long rectangular table with 3 microphones at preferred maximum aspect ratio 9/ Figure 23: Alternative layout for wider large long rectangular table Table A.1: General distance limits Figure A.1: Table dimensions and definitions Figure A.2: Cardioid microphone sensitivity vs angle polar plot (microphone facing up the page hence upsidedown heart shape) Figure A.3: Inverse square law sensitivity reduces with distance Deakin ssolutions 23/03/2013

5 ICT Resource AV Design Calculators, Tools and Resources Microphone Placement Page 5 Figure A.4: Sensitivity field plot shows contour lines of equal sensitivity Table A.2: Sound volume variations Figure A.5: Upper limit for range of single cardioid microphone C in meeting room with good acoustics Omnidirectional MIC20 microphones Figure 24: Alternative layout options for large wide trapezoidal tables Figure 25: Wide rectangle up to (a) 12 seats (2+8+2); (b) 18 seats (3+12+3) Figure 26: Wide rectangle up to 20 seats (6+8+6) or 24 seats (7+10+7) Figure 27: Long rectangular table for up to 20 seats (7+6+7) Figure 28: Small Kidney table showing fine detail, and oval table overlay kidney sensitivity variation within:. 42 Figure 29: Small wide oval table at absolute minimum aspect ratio L/W = 9/20, sensitivity variation ±3 db around the entire usable perimeter Deakin ssolutions 23/03/2013

6 ICT Resource AV Design Calculators, Tools and Resources Microphone Placement Page 6 1. Overview 1.1 Scope of document This document specifies the quantity, positioning and orientation of all table-top microphones at Deakin University, in particular as used for videoconferencing. Ceiling microphone arrays for use with flexible (reconfigurable) furniture are not covered in this document. Structure of document: 1. Overview (this section) 2. Wide tables (preferred) 3. Unity aspect-ratio tables (i.e. square and circular) 4. Long tables (non-preferred) 5. Appendix A: Definitions and Theory 6. Appendix B: Related considerations 7. Appendix C: Omnidirectional microphones not recommended 8. Appendix D: Additional examples 9. Appendix E: Further information 1.2 Design objective The objective is to achieve uniform sensitivity around the usable perimeter of the table, so that people can sit anywhere around this perimeter and be captured at approximately equal volume. This is not achieved through close-micing, which is not suitable for the application for several reasons: it creates hot spots and cold spots: sensitivity is dependent on exact sitting position of participants, with the result in practice that some people are too loud and others are too quiet; the microphones end up in amongst the participants work areas: the microphones become too sensitive to paper-shuffling etc noises; thus: the further away the microphones are, the less the variation in sensitivity with position, and the less the sensitivity to paper-shuffling noises. Maximum distances are set by the room acoustics: environmental noise and reverberation. Microphone sensitivity drops with the inverse square of distance (figure A.3), thus distances must not be so great as to drop into the noise floor. Room acoustics must comply with the requirements set out at ICT Volume 2.3 Videoconferencing Standards section All microphones in use at Deakin are cardioid directional microphones due to their superior performance, and whose sensitivity pattern is depicted throughout this document. Note that the sensitivity contour diagrams are independent of scale, and thus can be scaled as appropriate, e.g. 1 unit could be 1 seat or 1 m or half the width of the table in other words the same diagram can be scaled to the actual table/circumstance at hand, with the provisos that: Maximum scaling: maximum scaling is limited by noise floor, as noted above; Minimum scaling: diagrams start to become inaccurate below 1 unit = 0.6 m (refer section 6.3), however such small scaling would result in microphones too close to participants anyway, as noted above re close-micing. Please refer to Appendix A for all definitions of terms and symbols; formal design rules and underpinning theory, in particular: microphones must be at least 0.6 m from usable perimeter (table edge); aim for less than ±2 db variation around the primary usable perimeter; worst case ±3 db variation around the entire usable perimeter; microphones must face away from presentation surfaces this ensures that (a) microphones inherently face towards talkers when they are facing towards the presentation surfaces; and (b) microphones facing towards FoH place unnecessary stress on the echo-cancellation system; microphones must be fixed in their correct orientations and positions. Front (refer figure A.1): Rear Deakin ssolutions 23/03/2013

7 ICT Resource AV Design Calculators, Tools and Resources Microphone Placement Page Range of cases For the typical case of rectangular tables here is the most extreme range of cases for which a single microphone delivers acceptably uniform sensitivity around the usable perimeter: Figure 1: Range of aspect-ratios, small rectangular tables (single microphone) (a) Widest: (b) Square: (c) Longest: W max = 1.75L L = W L max = 1.5W F = 0.03W (3% of W) F = 0.05W (5% of W) F = 0.1W (10% of W) Beyond this maximum width (L too small): centre (rear) seat is too hot; side seats are too quiet Beyond this maximum width: must use more than 1 microphone (refer figure 25) Beyond this maximum length (L too long): rear (end) seats too quiet Beyond this maximum length: must use more than 1 microphone (refer figures 21 and 22) Worked examples: (a) Widest: W = 1600 (1.6 m) L min = 4/7 W = 920 (0.92 m) F = 0.03W = 50 (5 cm) R = L F = 870 (0.87 m) (b) Square: W = 1600 (1.6 m) L = 1600 (1.6 m) F = 0.05W = 80 (8 cm) R = L F = 1520 (1.52 m) (c) Longest: W = 1600 (1.6 m) L max = 1.5W = 2400 (2.4 m) F = 0.1W = 160 (16 cm) R = L F = 2240 (2.24 m) If the front perimeter is often used (has people sitting at it in non-videoconference meetings) then F may be increased to a maximum of 300 mm (if the microphone lies between seating positions) or an absolute maximum of 450 mm if the microphone is directly aligned with a seating position, however an absolute minimum of R min >0.5W (or more generally R>S as per section 5.4.7) must always be preserved: (e.g. in above examples R must always be greater than 0.8 m of particular relevance for case (a)) Large tables If the absolute scaling is simply too big for a single microphone or (single row of microphones) to satisfactorily cover, having regard to background noise and/or reverberation (Appendix A section and the 0.8 D c limit at figure A.5) then additional microphones are required in both directions (e.g. figures 16, 27). These breakpoints (as to when the table becomes too big) are dependent on the individual room acoustics, but preliminary guides for satisfactory and excellent acoustic conditions are given in the charts against each table type covered below. Deakin ssolutions 23/03/2013

8 ICT Resource AV Design Calculators, Tools and Resources Microphone Placement Page 8 2. Wide tables (preferred) Figures 2, 3 and 9 (in particular 9(a), (d) and (e)) are the primary workhouse elements for combining together in all large multi-microphone tables that are wider than can be accommodated by a single row of microphones along the longitudinal (front-rear) axis of the table. Resulting microphone arrays are generally variations on the illustrations at figures 16, 24, 25, 26 and Large Wide tables including immersive telepresence style Large (i.e. multi-microphone) 1 wide tables are designed by joining together corners (e.g. figures 9(a) and (b)) and straights (e.g. figures 9(c), (d), (e)) to provide coverage of the full seating area. Figure A.4 (Appendix A) shows that there is surprisingly great flexibility of suitable microphone positions, orientations and angles for corners and straights. Figure 3 shows a further example. Figure 17 shows the extreme similarity of the sensitivity contour and a circular arc. The maximum number of seats that can be satisfactorily covered by each microphone is given in Table 1 below, but is typically around 4. Microphones should be as far away from the participants as possible, subject to the limits set out in Table A.1 in Appendix A, and the depth of the table itself, and specific limits set out for specific cases below (in figures 2 to 9). Microphones must be angled so that Front-of-House (FoH) speakers lie within the allowed grey shaded regions as shown in figures 2 to 9, i.e.: Nearer FoH speakers should be >125 off-axis from the microphone (plan view); Furthest FoH speakers should be >125 off-axis from the microphone (plan view); i.e. microphones must not face towards the front of house. Table 1: Seats per microphone large wide tables Edge shape Minimum Preferred Corner, symmetrical figure 9(a) 2 (P 600) Corner, asymmetrical figure 9(b) 2 (S 600) Straight, off-axis figure 9(c) 2 (S 600) Straight, 45 angle figure 9(d) 2 (P 600) Straight, on-axis figure 9(e) 2 (R 600) Wide arc figure 2 2 (R 600) Maximum satisfactory acoustic conditions * Maximum ideal acoustic conditions ** (P 900) 4 (R 900) * Acoustic characteristics conforming to ICT 2.3 Videoconferencing Standards section (refer figure A.4): ** Only under the most ideal acoustic characteristics (reverberation and background noise) as set out at ICT 2.3 Videoconferencing Standards section Small (i.e. single microphone) tables are covered in section 2.2 for wide tables and also later in this document for other table shapes (aspect ratios). Deakin ssolutions 23/03/2013

9 ICT Resource AV Design Calculators, Tools and Resources Microphone Placement Page Large wide arc tables Positioning of microphones in wide arc tables starts with figure 17 where it is seen that essentially perfect uniformity of sensitivity is achieved when microphones are placed at R=0.8W (where W is the diameter of the arc) [or more conventionally R=1.6r where r is the radius of the arc] and that very little variation results for significant ranges of positioning, e.g. R=(1.6±0.2)r produces a sensitivity variation of less than 1 db. This illustrates that there is very good alignment between the shape of the contour plot and the shape of the table. However for very shallow arcs, literal application of the above formulae would result in microphone positioning much too far from the table. In which case the wide arc is more like a series of straight line segments (compare figures 2 and 9(e)). Wide arc tables are covered by a number of microphones placed around the arc. The exact number of microphones depends on the ratio of the depth of the table (and hence R how far back from the edge the microphones can be placed) and the radius r of the arc, as well as the overall dimensions of the arc. Figure 2 depicts the case of R 0.4r. Figure 2: Large wide arc table, case of R 0.4r sensitivity field for each microphone Figure 3: Large wide V table sensitivity field for the microphone at the apex A worked example for 3 microphones is shown below at figure 4. A worked example for 3 microphones is shown below at figure Large wide V tables Coverage of large wide V tables is constructed by a combination of figure 3 at the central apex and figure 9(e) in the wings. The exact number of microphones depends on the ratio of the depth of the table (and hence R how far back from the edge the microphones can be placed) and the overall width W of the table. A worked example for 3 microphones is shown below at figure 5. Deakin ssolutions 23/03/2013

10 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page 10 Figure 4: Wide Arc table example of 3 microphones at R 0.4r Delivers sensitivity variation of only ±1 db around the entire usable perimeter and can easily accommodate 12 seats around the arc.

11 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page 11 Figure 5: Wide V table example of 3 microphones

12 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Large wide trapezoidal tables Large wide trapezoidal tables will generally be constructed: for shallow side angles: as for rectangular tables (e.g. as depicted in figures 26 and 27) using elements from figure 3 above and 9(d) below; or for broad side angles: using the same elements as for a wide V table, i.e. figure 3 at the corners and figure 9(e) along the straights. Less commonly, a 3 rd alternative for mid-range side angles where the rear end of the table can be comfortably covered by a single microphone is depicted in figure 24 in Appendix D below.

13 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Large wide U tables Large wide U tables are constructed from elements depicted in figure 2 above and 9(d) below, as illustrated in figure 7 for the cases of (a) 3 microphones around the arc, and (b) with island table: Figure 7(a): Large U table

14 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page 14 Where the tables themselves are narrow (700 mm or less), with the result that microphones would be too close to participants, an island table can be used for mounting the microphones. While this is less preferred due to loss of gain (refer appendix A section 5.1), this option has the advantage of providing greater isolation of participant noises that would otherwise be transmitted through the table (finger tapping, paper shuffling, etc.). Figure 7(b): Large U (or rectangular) table, with island table for microphones Above figure shows case of 3 microphones. For wider tables, additional microphone/s can be placed around the arc, as illustrated in figure 7(a). And for longer tables additional pairs of microphones can be placed down the length of a (longer) island table. The same island table strategy can be also adopted with very similar results for large rectangular layouts of narrow tables, using layouts illustrated in figure 7(b) and 16.

15 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Large wide rectangular tables The components to build a large rectangular table (wide, long or square) are given in figure 9. Figure 9: Large wide rectangular table corner (a) symmetrical, (b) asymmetrical (a) (b) Figure 9: Large wide rectangular table straight edges (c) off-axis, (d) 45, (e) on-axis (c) (d) (e) Elements (a), (d) and (e) are the most commonly used. Worked examples are given in Appendix B (figures 26 and 27). An example of a large square table is given in figure 16. A further example, employing element (b) is given in figure 23, which may be useful for tables of specific aspect ratio and size. Element (d) is very useful in a wide variety of situations to ensure that FoH speakers remain in the grey-zone, while also ensuring that the considerations of section 5.5 (in Appendix A) are catered for. A further use of element (d), or in extreme cases element (c), is in place of element (e) for the case where the table is too narrow (e.g. V or trapezoidal tables) to place the microphone far enough away from the participants so as to keep the sensitivity variation within the required limits. Wide rectangular table aspect ratio range Microphones Angle Minimum Preferred maximum Absolute maximum 1 0 W = 1 L (1/1) W = 1.7 L (5/3) W = 1.75 L (7/4) 2 45 W = 1.6 L W = 4 L W = 4.1 L 3 45, 0, 45 W = 3 L W = 5.6 L W = 5.6 L 4 * 45 W = 5.6 L W = 8 L W = 8.2 L * Greater than 4 microphones is rarely required, because 4 microphones can cater for up to 16 seats in width.

16 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Small Wide tables including style Small kidney tables Small kidney tables, suitable for up to 5 seats around the primary usable perimeter (or 6 in ideal acoustic conditions) are frequently used in collaborative learning applications, with one or more such tables in the learning space. A particular requirement in such (informal, uncontrolled) collaborative learning environments is that the microphone sensitivity is retained within the specified tolerances as best as possible throughout the unusable front perimeter, in case students sit throughout that area, with their backs to the presentation surface. This is achieved as shown in figure 10(a), by placing the microphone as close as possible to the inner edge of the centre of the kidney. The microphone must be fixed in place, the cable completely concealed, with no part of the microphone or assembly overhanging the table edge. The placing of the microphone as depicted in figure 10(a) achieves a satisfactory 4 db (±2 db) variation (i.e. from +2 to +6 db) around the Primary Perimeter and 6 db (±3 db) around the Entire Perimeter as required. This can be seen more accurately in greater detail in the larger version of the diagram included in Appendix D. This gives rise to the requirement that the minimum aspect ratio for a small kidney table is such that R/W=0.5 where R=L, i.e. the length as measured along the minor axis through the middle of the kidney (i.e. perpendicular to the major width axis). Figure 10(b) shows an identically dimensioned oval table, and also illustrates that the same microphone position produces the same results around the Primary Perimeter, but is not optimum for Entire Perimeter. Optimum microphone placement for oval tables is F<0.05W as illustrated in Figure 11 below. Figure 10: Small kidney table and corresponding small oval table (microphone at Rmin > 0.5W) A more detailed version of these diagrams is provided in Appendix D.

17 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Small wide oval tables Suitable for up to 5 seats, or 6 under ideal acoustic conditions. Figure 11: Small wide oval tables at aspect ratios (a) L/W = 2/3 and (b) L/W = 1/2 Absolute maximum aspect ratio of L/W = 20/9 and required microphone position are shown in Appendix D Small wide arc tables Suitable for at least 6 seats, refer sample case at figure 2 above Small half-cirle tables Suitable for up to 5 seats, or 6 under ideal acoustic conditions. Small half-circle table dimensions Dimension from table edge at closest point Minimum Preferred maximum Absolute maximum table width (diameter) W= 1200 W= 2100 W= 2400 distance to microphone at 90 off-axis S = 0.5W = 600 S = 0.5W = 1050 S = 0.5W = 1200 distance to microphone from front of table F = 0 * F = 50 * F = 50 * All microphone distances are measured to the active centre of the microphone pickup. * But not so that any part of the microphone, cable or connector overhangs the edge of the table Small wide V tables Suitable for up to at least 6 seats, refer illustration at figure 3 above.

18 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Small wide diamond tables Figure 13: Small wide diamond table at the maximum aspect ratio of W=2.3L and F=0.025W Small wide trapezoid tables Figure 14: Small wide trapezoidal table at the maximum aspect ratios of W/L=5/3 and V/L=1 and required microphone placement at F=0.05V [sic]

19 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Small wide rectangular tables Figure 15: Small wide rectangular table at the maximum aspect ratio of W/L=7/4 and required microphone placement at F=0.03W

20 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Square and Circular tables 3.1 Large Square and Circular tables Large circuler table Large circular table is implausible (too much unusable table-top area). Please refer instead to realistic related shapes: Wide arc (section above) Small semi-circular (section above) Small circular (section below) Large square table Large square table is a specific case of large rectangular table (refer section above), and are constructed using elements (b), (d) and (e) from figure 9 as illustrated below and in Appendix D (figure 26, 27). Figure 16: Large square table for up to 18 seats (6+6+6)

21 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Small Square and Circular tables Small circular table Suitable for up to 6 seats, or 7 under ideal acoustic conditions; plus an optional 2 in the secondary seating area. The secondary seating area is the area shaded in purple below, where people will typically take extra chairs from the front of the table. Figure 17(b) depicts essentially perfect sensitivity uniformity (<±0.1 db) around the entire primary usable perimeter for microphone placed at F=0.2W, while figure 17(a) shows optimum adjustment (F=0.1W) to accommodate additional chairs in the purple shaded secondary seating area, in which case the sensitivity variation is still excellent at <1 db (<±0.5 db) around the primary usable perimeter. Figure 17: Small circular table (a) optimized for primary and secondary seating: microphone placed at F=0.1W (b) optimized for primary seating only: microphone placed at F=0.2W Figure 17(b) also depicts the optimum placement of a single microphone at the head (rear) end of a long U table (refer section 4.2 below), i.e. ideal position R=0.8W ; figure 17(a) further illustrates that this ideal placement can vary to R=0.8W ± 0.1W with little impact (sensitivity variation <1 db around primary semicircle) Small square table Figure 18 Small square table, microphone placed at preferred position F=0.05W

22 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Long tables (non-preferred) Wide tables are strongly preferred over long tables due to the resulting superior sight lines. 4.1 Long Trapezoidal tables Trapezoidal tables are preferred over rectangular tables due to the resulting improved sight lines. Long trapezoidal table dimensions Dimension from table edge at closest point Minimum Preferred maximum Absolute maximum table width, front W= 1200 W= 2100 W= 2400 table width, rear V ~ 0.5 W V = W (rectangular) distance to microphone at 90 off-axis S = 600 S = 1050 S = microphone: distance to microphone from front of table F = 0.05 V F = 0.1 V F = 300 distance to microphone on-axis R = 0.9 W R = W V R = W V All microphone distances are measured to the active centre of the microphone pickup. Long trapezoidal table aspect ratio range Microphones Minimum Preferred maximum Absolute maximum 1 L = 1 W (1/1) L = 1.3 W L = 1.5 W (3/2) 2 L = 1.35 W L = 2.5 W (5/2) L = 2.6 W 3 L = 2.55 W (>5/2) L = 3.6 W L = 3.7 W 4 * L = 3.65 W L = 4.6 W L = 4.7 W * A perceived need for greater than 4 microphones usually indicates that the allowable maximum 2H..6H viewing distance range would be exceeded Small long trapezoid tables Figure 19: Small long trapezoidal table, at absolute max aspect ratio L/W = 3/2 and required microphone position F=0.1V

23 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Large long trapezoid tables Sensitivity fields are very similar to the large long rectangular tables depicted in section below, but with head-end pattern similar to figure 19 above. With 2 microphones: Suitable for (aspect ratio up to 5/2) the front microphone is placed at F=0.05W from the front of the table; the rear microphone is placed at approximately the middle of the table (R=0.5L) but subject to the limits set out in the table above and such that the coverage handover boundary (the V in the field plot) is between two seating positions, not aligned with it. With 3 microphones: Suitable for (aspect ratio 3.6) the front microphone is placed at F=0.05W from the front of the table; the rear microphone is placed at approximately R=L/3 but subject to the limits set out in the table above and the middle microphone is placed at approximately the midpoint between the other two microphones (i.e. at approximately F=L/3) but such that the coverage handover boundaries between all microphones (the V in the field plot) is between seating positions, not aligned with them. 4.2 Long U tables Long U table aspect ratio range Microphones Minimum Preferred maximum Absolute maximum 1 L = 1 W (1/1) L = 1.5 W (3/2) L = 1.8 W (9/5) 2 L = 1.7 W (>5/3) L = 3.0 W (3/1) L = 3.6 W (18/5) 3 L = 3.3 W (>10/3) L = 4.5 W (9/2) L = 5.3 W (53/10) 4 * L = 5.0 W (>5/1) L = 6.0 W (6/1) L = 7.0 W (7/1) * A perceived need for 4 microphones usually indicates that the allowable maximum 2H..6H viewing distance range would be exceeded. Figure 17(b) (section above) depicts the optimum placement of a single microphone at the head (rear) end of a long U table, i.e. ideal position R=0.8W (~0 db sensitivity variation around the entire head-end semicircle); figure 17(a) further illustrates that this ideal placement can vary to R=0.8W ± 0.1W with little impact (sensitivity variation <1 db around head-end semicircle); and figure 20(a) (section below) illustrates that the preferred maximum is R=1.4W (or R=1.6W under ideal acoustic conditions).

24 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Small long U tables Figure 20: Small long (a) U and (b) rectangular tables, preferred max aspect ratio 3/2 and preferred position F=0.1W Large long U tables Sensitivity fields are very similar to the large long rectangular tables depicted in section below, but with head-end pattern as per figures 17 or 20(a) above. With 2 microphones: Suitable for (aspect ratio up to 3/1) the front microphone is placed at F=0.05W from the front of the table; the rear microphone is placed at least R=0.8W from the rear of the table, or at approximately R=L/2 for longer tables, but such that the coverage handover boundary (the V in the field plot) is between two seating positions, not aligned with it. With 3 microphones: Suitable for (aspect ratio 9/2) the front microphone is placed at F=0.05W from the front of the table; the rear microphone is placed at least R=0.8W from the rear of the table, or at approximately R=L/3 for longer tables; the middle microphone is placed at approximately the midpoint between the other two microphones (i.e. at approximately F=L/3) but such that the coverage handover boundaries between all microphones (the V in the field plots below) is between seating positions, not aligned with them.

25 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Wider Large long U tables Where a U table exceeds the width that can be covered by a single microphone, as per the limits set out in the tables in section 4.2 above, the microphone array would typically be constructed as for any other large U mostly using elements from figure 2 and 9(d), as illustrated in figure Long Rectangular tables (non-preferred) Long rectangular table dimensions Dimension from table edge at closest point Minimum Preferred maximum Absolute maximum table width W= 1200 W= 2100 W= 2400 distance to microphone at 90 off-axis S = 0.5 W = 600 S = 0.5 W = 1050 S = 0.5 W = 1200 distance to microphone from front of table F = 0.05 W F = 0.1 W F = 300 * distance to microphone on-axis R = 0.8 W R = 1.4 W R = 1.6 W All microphone distances are measured to the active centre of the microphone pickup. * If the room is frequently used for non-videoconferencing purposes: the front microphone may be placed at up to F=300 mm to keep it out of the way of local users, or an absolute maximum of F=450 mm if there are an odd number (1 or 3) of seating positions along the front of the table. Long rectangular table aspect ratio range Microphones Minimum Preferred maximum Absolute maximum 1 L = 1 W (1/1) L = 1.5 W (3/2) L = 1.7 W (5/3) 2 L = 1.7 W (>5/3) L = 3.0 W (3/1) L = 3.3 W (10/3) 3 L = 3.3 W (>10/3) L = 4.5 W (9/2) L = 5.0 W (5/1) 4 * L = 5.0 W (>5/1) L = 6.0 W (6/1) L = 6.7 W (20/3) * A perceived need for 4 microphones usually indicates that the allowable maximum 2H..6H viewing distance range would be exceeded Small long rectangular tables Suitable for up to 8 seats (3+2+3) with 1 microphone; or up to 11 seats (4+3+4) in ideal acoustic environments. Refer figure 20(b) above. Microphone is placed at F=0.1W from the front of the table.

26 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Large long rectangular tables with 2 microphones: Suitable for up to 8 seats long x 3 seats wide (total 19 seats) with 2 microphones; or up to 10x4=24 seats in ideal acoustic environments. Refer figure 21 below. Microphone placement: the front microphone is placed at F 1 =0.05W from the front of the table; the rear microphone is placed at approximately the middle of the table: R 2 =(L-F 1 )/2 but such that the coverage handover boundary (the V in the field plots below) is between two seating positions, not aligned with it. Figure 21: Long rectangular table with 2 microphones at preferred maximum aspect ratio 3/1 with 3 microphones: Suitable for up to 12 seats long x 3 seats wide (total 27 seats) with 3 microphones; or up to 15x4=34 seats in ideal acoustic environments (except that the 2H..6H viewing distance limit range will usually restrict numbers of seats to lower limits than this). Refer figure 22 below. Microphone placement: the front microphone is placed at F 1 =0.05W from the front of the table; the rear microphone is placed at approximately R 3 =(L-F 1 )/3 and the middle microphone is placed at approximately the midpoint between the other two microphones (i.e. at 2 approximately F 2 =L/3) but such that the coverage handover boundaries between all microphones (the V s in the field plots below) is between seating positions, not aligned with them. Figure 22: Long rectangular table with 3 microphones at preferred maximum aspect ratio 9/2 2 or more precisely: F 2 = F 1 + (L-F 1)/3 when F 1 has been offset due to local usage of the front of the table.

27 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Wider Large long rectangular tables Where a rectangular table exceeds the width that can be covered by a single microphone, as per the limits set out in the tables in section 4.3 above, the microphone array would typically be constructed as for any other large rectangle mostly using elements (a) and (d) from figure 9, as illustrated in figure 27 for up to 6 seats wide, or figure 26 for even larger tables. Alternatively, for some specific table aspect-ratios and sizes, element 9(b) might be useful as set out in figure 23 below. Figure 23: Alternative layout for wider large long rectangular table

28 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Appendix A Definitions and Theory 5.1 Design rule Table-top microphone arrays shall be designed so that the sensitivity variation range: around the entire usable perimeter of the table must be no more than 6 db (i.e. maximum ±3 db around its mid value); and along the primary usable perimeter of the table should preferably be no more than 4 db (i.e. preferably ±2 db around the mid value); and in the secondary seating area must also be within the 6 db range; and around the front perimeter of the table should preferably also be within the 6 db range (so that if the front perimeter is used, e.g. for audio conferencing, that people in these seating positions are also reasonably audible but not excessively loud); and these sensitivity values to apply at normal sitting height of 0.5 m above the table surface. All microphones shall be cardioid microphones and only of the types specified at ICT Volume 2.3 Videoconferencing Standards section 6.8. All table-top microphones shall be fixed in their correct positions and orientations, so that end-users cannot re-position them. 5.2 General distance limits The following microphone placement distance limits shall apply: Table A.1: General distance limits Dimension from table edge at closest point Minimum Preferred maximum Absolute maximum distance to microphone at 90 off-axis S = 600 S = 1050 S = 1200 * distance to microphone on-axis R = 600 R = 2100 R = 2400 * distance to microphone on-axis R = 600 R = 2.8 S R = 3.2 S * Individual table designs (above) may specify different, usually tighter, limits than those given in table A.1 above. * Preferred maxima must not be exceeded except under the most ideal acoustic characteristics (reverberation and background noise) as set out at ICT Volume 2.3 Videoconferencing Standards section Distance from talker to microphone shall always be less than 0.8 D c where D c is as defined in section below. The proportion by which a design sits below this 0.8 D c limit (i.e. the headroom ) shall be used as a measure of the predicted performance quality of the design. For example a design that sits within 0.6 D c has a headroom of 25% or (via 3 the inverse square law) 2.5 db. 3 calculated using: -10log 10((0.6/0.8) 2 ) = 2.5 db refer section and figure A.3.

29 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page 29 Figure A.1: Table dimensions and definitions L F R W + V S = Primary Usable Perimeter = Secondary Seating Area = Front Perimeter (unusable) inset: oval table: Clipart (table, chairs, displays) courtesy Cisco 5.3 Definition of terms and symbols Front = towards the presentation surface (towards Front-of-House) Rear = away from the presentation surface Wide = table aspect ratio such that W > L, where W measured parallel to presentation surface Long = table aspect ratio such that L > W, where L measured perpendicular to presentation surface Small = table requires only 1 microphone, according to the design rules set out in this document Large = table requires >1 microphone, according to the design rules set out in this document W = width of the table at its widest point, measured parallel to presentation surface (for the case of trapezoidal tables, this means at the front of the table) V = (for trapezoidal and other relevant cases:) the width at the rear of the table L = length of the table, measured perpendicular to presentation surface F = distance from microphone to front table edge, measured perpendicular to table edge P = distance from microphone to usable table edge, measured perpendicular to table edge R = distance from microphone to table edge, measured on-axis S = distance from microphone to table edge, measured 90 off-axis r = radius of arc or circle (for case of full circle, r = W/2 ) seat = seat width, usually 0.6 m (occasionally larger)

30 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page 30 Entire usable perimeter = entire table perimeter, excluding the front perimeter (i.e. the sum of the green and purple sections in figure A.1 above). Front perimeter = any part of the table perimeter that is at an angle greater than 90 to the presentation surface, i.e. anywhere the seat would be facing even to a small extent away from the presentation surface (i.e. the red sections in figure A.1 above). Primary usable perimeter = where properly seated persons mouths may be centred a person cannot generally sit centred less than 0.5 seats (0.3 m) from a table edge; at corners an additional 0.25 to 0.5 seats must also be allowed on one or other or both sides of the corner to allow for leg and seat clearance in total a distance of 0.5 to 1.0 seats (0.3 m to 0.6 m) from each corner is therefore excluded from the primary usable perimeter (i.e. the green sections in figure A.1 above). Secondary seating area = that part of the entire usable perimeter that is not part of the primary usable perimeter plus any additional area defined by specific individual table designs (above) (i.e. the purple sections in figure A.1 above). Notes: For D@YD style class-rooms where student tables may be slightly angled with respect to the presentation surface: V, W and L are measured relative to the major axis of the table itself, not relative to the presentation surface or room. Distances to microphone (F, P, R and S) are always measured to the active centre point of the microphone s pickup, not the microphone s external housing. Distances are shown in mm, e.g means 1200 mm. All diagrams are shown with Front-of-House to the left of the page, as per figure A.1 above. 5.4 Theory Cardioid microphone Cardioid (directional) microphones are always used because they greatly cut down pickup of background noise and reverberation (local echo). When correctly oriented (i.e. facing away from the far-end audio which emanates from front-of-house speakers [1] ) cardioid microphones also greatly cut pickup of far-end audio, reducing the stress on and improving the performance of the videoconference system s echo cancellation system. Thus: cardioid microphones effectively reduce background noise and both near-end and far-end echo Sensitivity vs angle The sensitivity of cardioid directional microphones reduces with angle off-axis as depicted in figure A.2. (The name cardioid derives from its very roughly heart-shaped sensitivity pattern.) Figure A.2: Cardioid microphone sensitivity vs angle polar plot (microphone facing up the page hence upside-down heart shape) 300deg (~-3 db) 270deg (~-6 db) 240deg (~-12 db) 330deg (~-1 db) 210deg (~-24 db) 0deg (0 db) deg 30deg (~-1 db) 150deg (~-24 db) 60deg (~-3 db) 90deg (~-6 db) 120deg (~-12 db) [1] Deakin Standards require that in all videoconference applications far-end audio comes via front-of-house (program audio) speakers only.

31 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Sensitivity vs distance Sound level reduces with the square of the distance between source and microphone, hence the inverse square law. Figure A.3: Inverse square law sensitivity reduces with distance That is, power level and hence microphone sensitivity reduces by 6 db for each doubling of distance, as depicted in figure A Overall sensitivity Combining the above two relationships (figure A.2: sensitivity vs angle and figure A.3: sensitivity vs distance) we get the overall sensitivity vs angle and distance, depicted as contour curves of equal sensitivity in figure A.4. Figure A.4: Sensitivity field plot shows contour lines of equal sensitivity If we travel obliquely towards the microphone (e.g. along any of the dashed lines in figure A.4) there is first-order cancellation of the two effects: the increasing sensitivity with reducing distance is approximately neutralized by the reducing sensitivity with increasing angle off-axis. With the result that there is only a nett ±1 db variation in sensitivity along the entire length of any of those dashed line trajectories. The blue region depicts a region in which the sensitivity does not vary by more than ±3 db, allowing a wide range of table shapes and sizes, and distances and angles between microphone and audience. [Note that the curves in figure A.4 are similar to (derived from) but different than the shape of the polar plot in figure A.2. Figure A.4 is an X-Y field plot, not a polar plot.]

32 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page 32 How to read the sensitivity field plots (above and below throughout this document): Length is depicted in arbitrary units: 1 unit could be 1 seat (e.g. 0.6 m); or 1 m or any other convenient unit related to the size of the table, for example: many of the plots depict tables of width 2 units, so for a table that is 1.6 m wide this would mean that 1 distance unit = 0.8 m. The microphone is facing to the right of the diagram and has its active centre at the intersection of the longitudinal (on-axis) and lateral axes. Plan view: The microphone, field (and table not shown above) are viewed from above. Contour lines show paths of equal sensitivity. Thick lines are 3 db apart. Thin lines are 1 every 1 db. Black lines are less sensitive (too quiet) than blue lines. Red lines are more sensitive (too loud) than blue lines. The 0 db reference level (green contour line) is arbitrarily defined as the sensitivity at 1 distance unit at 90 off-axis. However this absolute level is unimportant. What is important is that the entire usable perimeter of the table remains within any 6 db range. For example: the blue shaded region in the above diagram shows a 6 db region ±3 db around the arbitrary 0 db reference level. Any other 6 db range would be equally acceptable, e.g. from 6 to 0 db or from 0 to +6 db The design rule microphone sensitivity variation The smallest change in volume that the human ear can detect is generally considered to be around 1.5 db. Therefore the preferred design rule limit of ±2 db is only fractionally above the limit of human hearing and the worst-case limit of ±3 db is also very acceptable. To put this range into further perspective, below are some additional comparative figures: Table A.2: Sound volume variations Difference Example 1~1.5 db Limit of human ear to detect volume change 6 db Someone talking to you from 2 seats away as compared with from 4 seats away 7~8 db Front-of-house loudspeaker volume at distance 2H as compared with at 6H from front 12 db Someone talking to you from immediately adjacent seat as compared with from 4 seats away 18 db Someone talking to you from immediately adjacent seat as compared with from 8 seats away Use of front-of-house loudspeakers only The information in table A.2 also validates the requirement in ICT Volume 2.3 Videoconferencing Standards section 6.7 that far-end audio be played through front-of-house speakers only in all meeting rooms and not through ceiling speakers. If locally people can be comfortably heard from one end of the table to the other, where the volume variation can be as high as 18 db, then front-of-house audio can be even more comfortably heard, because its volume variation will be as little as 7~8 db from front (2H) to rear (6H) of the audience listening area Minimum distance limits The minimum distance between microphone and table edge (throughout the primary usable perimeter) under any circumstances is set at 0.6 m for the following reasons: To provide adequate workspace between the participant and microphone, for the general convenience of the user, and more importantly so that paper-shuffling, ipads and other user activities do not generate excessive background noise into the microphone. To reduce the variation in sensitivity around the perimeter between hot-spots (in front of microphones) and cold-spots (between microphones) and hence decouple the microphone design from exact seating positions, which can rarely be accurately controlled in practice. To reduce the sensitivity to forward/backward lean. Another uncontrollable environmental variable is the extent to which an individual person will lean forward or backwards in their chair. A person sitting correctly/normally and participating actively/normally in the discussion will normally sit with their mouth almost exactly above the table edge. Reducing the dependency on this variability (of different people leaning forwards and backwards in their seats) is achieved by placing the microphones further from the table edge.

33 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page 33 Also: The bigger the physical scaling (i.e. the bigger the distance from microphone to table edge), the bigger in absolute terms is the blue region in figure A.4, and hence a range of compounding benefits: o o o o the more flexibility of table sizes and shapes that fit within the blue envelope; the more a particular table doesn t get too near the edges of the blue envelope; the fewer the number of microphones needed in a particular case; and hence the less often a C90 or (more importantly) a ClearOne is required (which is important for both cost and performance reasons). In summary: It can be seen from figure A.4 that the contour lines spread out further away from the microphone. This means that the dependence (in db/m) of sensitivity on exact position progressively reduces at greater distances. Conversely the contour lines get very close together close to the microphone, increasing the dependence of sensitivity on the exact position of the participants with respect to the microphone. Greater distances from the microphone produce more consistent results. Other considerations: Minimum distance on-axis vs off-axis: Due to the 6 db variation in sensitivity (figure A.2) between on-axis and 90 off-axis, the distance between microphone and table edge (R) on-axis should ideally be around twice the distance between microphone and table edge (S) at 90 off-axis, i.e. R 2S, however different ratios and ranges apply for different specific table designs as set out in the body of this document the extremes being, for cases where seating is opposite both R and S is: R>S and R<3S. Microphone visibility: To further reduce the extent of paper-shuffling (etc.) noise pickup, it is important that tabletop microphones are obvious to untrained end-users (i.e. no invisible recessed microphones) Maximum distance limits The sensitivity contour diagrams throughout this document are (correctly and deliberately) dimensionless, i.e. they apply equally to tables of any scale, e.g. a table of width 1 m or a table of width 1 km. However other considerations define practical upper limits: While direct sound (the wanted signal) drops off accurately with square of distance (figure A.3), total power drops off less rapidly due to unwanted reflections from walls, ceilings and other hard surfaces in the room. These unwanted reflections muddy the intelligibility of speech, by: reverberation (local echo); and attenuation of higher frequencies (comb filter effect). Similarly, environmental background noise limits the audibility of speech. These reflections and background noise are roughly independent of distance between talker and microphone, thus as distance to microphone increases, the reflections and noise progressively drown out the intelligible speech. Practical distance limits are very dependent on individual room acoustics. Ideal and suitable acoustic environments are specified in ICT Volume 2.3 Videoconferencing Standards section Indicative corresponding extreme and preferred maximum distance scales are set out above against each type of table. Upper limit The upper range limit for placement of cardioid microphones is 0.8 D c where D c ( Distance critical ) is defined as the distance from the talker where the direct sound power from the talker equals the total power from all reflected waves. For meeting rooms (including seminar rooms and small-to-medium classrooms) with good acoustics, D c can be approximated as the perimeter of the room, with the exception of the corners. The extent to which the corners lie outside 0.8 D c depends on the acoustic properties of the surfaces and (to a generally lesser extent) the position of the talker within the room. For such rooms, the 0.8 D c limit should not need to be (and should not be) approached. The proportion by which a design sits below the 0.8 D c limit (i.e. the headroom ) is used as a measure of the predicted performance quality of the design. For example a design that sits within 0.6 D c has a headroom of 25% or (via

34 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page 34 the inverse square law) 2.5 db. This means there would be an additional 2.5 db of headroom of direct signal power over the reverberant power, additional to what would be provided by the 0.8 D c design rule upper limit. The design rule itself provides a further headroom of around 6 db. For rooms with poorer acoustics, and for lecture theatres and larger classrooms and function spaces, D c detaches from the walls, resulting in the range limit becoming smaller islands around each microphone. Just as the blue region in figure A.4 indicates the allowable range (e.g. for a secondary audience sitting away from the primary seating area), so too does the pink region in figure A.5 below indicate the disallowable region for a secondary audience. Figure A.5: Upper limit for range of single cardioid microphone C in meeting room with good acoustics D c C maximum range for cardioid microphone at C based on max 0.8 D c to talker T 0.8 D c from talker at T room perimeter

35 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Appendix B Related considerations 6.1 Tabletop as reflector to increase gain The tabletop can act as an effective reflector increasing gain by up to 6 db (less for higher frequencies). Therefore table layouts with large gaps between microphone and participants are to be avoided because the size of any gap proportionately reduces the gain benefit. If architectural considerations prefer a gap, these should be infilled with glass or other suitable material. 6.2 Tabletop as noise conductor/amplifier Conversely, the tabletop also acts as a conductor of finger/hand/pen tapping noises, paper shuffling, etc. Even worse, tabletops can act as an amplifier of these noises, in much the same way as the sounding board of a stringed musical instrument (e.g. acoustic guitar). Therefore it is most important that the table is not made of ply-wood or other thin material and must not be hollow. To prevent noise conduction: Table microphones must be isolated from the table by suitable rubber or foam mounts. Care must be taken to ensure that any locking, locating or securing screws/nuts/etc. do not introduce a conduction path an isolating layer must be preserved between any such mounting assembly and the microphone or table. The tabletop must not be made of any type of ply-wood or other thin material. The tabletop must not be hollow. The tabletop should be made of thick solid wood so as to act most like a solid ground-plane and not like a sounding board. Regardless of material used, the tabletop must be solid and greater than 30 mm thick, preferably 50 mm. Any laminates must be securely bonded to the solid core, consistently across the entire surface-area. Placing the microphones on a separate island table, e.g. as illustrated in figure 7(b), with a small gap to the meeting table is an effective strategy to eliminate all noise conduction. Embedded (e.g. Lavalier) type microphones must not be used. Omnidirectional microphones must not be used. Microphones must be easily recognisable by end-users. Microphones must not be placed in a user s workspace, i.e. microphones must be at least 600 mm from table edge around the entire usable perimeter Installation of Audio Technica ES947 and ES947W microphones To minimize sound transmission, ES947 microphones must always be fitted with the mechanical isolators provided with the microphone, above and below the table as shown. This means that a 23.5 mm diameter hole must be drilled all the way through the tabletop to prevent the table coming into contact with the microphone body at any point. The microphone must be fixed in its correct orientation by removable adhesive (i.e. uncoloured silicone glue) applied to all surfaces of the upper (above table) isolators, so as to prevent subsequent accidental rotation (e.g. due to fiddling by end-users). Excess glue must be carefully, cleanly and promptly removed so that no glue is visible; glue must not be allowed to enter the microphone grille. When tightening the nut, care must be taken to: not alter the correct alignment of the microphone the small dot in the rim of the housing indicates the front of the microphone; not overtighten or crush the acoustic isolators.

36 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Height above table All sensitivity field contour plots in this document have been computed adjusted for the 3 rd dimension, i.e. the height of the talker s mouth above the tabletop, using the following data: height of mouth above table 500 mm; plot distance unit: 1 unit = 900 mm. This scaling has been chosen so that the field plots are accurate to within about 1 db for all plot scalings from 1 unit = down to 1 unit = 600 mm. The effect of considering the 3 rd dimension is to slightly soften (reduce) the dependence of sensitivity on (plan-view) distance. 6.4 Lean One uncontrollable aspect of participants behaviour is the extent to which they lean over the table or lean back in their seats. This document assumes that the nominal average seating position is where the mouth is vertically above the table edge. The consideration at section 5.3 above slightly reduces the dependence on lean, but the most important way of reducing this dependence is to place the microphones further away from the participants. Hence this document requires a minimum microphone setback from table edge of 600 mm under all circumstances, however greater separation is preferable to reduce dependence on lean. 6.5 Direction a person is facing The more significant impact of participants behaviour on audibility is the direction a talker is facing when speaking, i.e. towards or away from the microphone. Generally, people face towards the people they are addressing (conversely it is a common everyday experience to not be clearly heard if facing away from your audience). Hence, if microphones face generally away from the display surfaces (i.e. placed between front-ofhouse and the participants and facing roughly towards the rear of the room) then this automatically ensures that talkers are facing roughly towards the microphones when they face to address the far-end audience. Hence this document in all configurations specifies microphones placed towards the front of the room and facing towards the rear of the room, and never facing towards the display surfaces. This automatic alignment of microphone with direction people are facing when talking is particularly effective for the preferred wide ( telepresence -style) table configurations recommended by this document. This strategy is also reasonably effective for long table configurations, however in those configurations there is a greater tendency for people to be facing away from their far-end audience, placing greater onus on participants to consciously face towards the far-end when talking.

37 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Appendix C Omnidirectional microphones Omnidirectional microphones are not recommended. The Cisco MIC20 microphone is the only exception. The MIC20 is only suitable for smaller rooms and the maximum number of MIC20 in a system is two. Omnidirectional microphones shall be placed as follows: in the centre for square and circular tables along the major axis of symmetry for oval tables along the major axis of symmetry for long rectangular tables near the front edge for wide rectangular tables Dimensional limits and microphone placement for specific cases are as follows: Omnidirectional MIC20 microphones Mics Table shape Aspect ratio (max) Position Mic 1 Position Mic 2 1 Square W = L S = W/2 R = L/2 n/a 1 Circle W = L S = W/2 R = L/2 n/a 1 Wide Rectangle W = 3.1 L S = W/2 F = 50 n/a 1 Wide Oval W = 1.75 L S = W/2 R = L/2 n/a 1 Long Oval L = 1.75 W S = W/2 R = L/2 n/a 1 Long Rectangle L = 1.67 W S = W/2 R = L/2 n/a 2 Wide Rectangle W = 5.5 L S = W/4 F = 50 S = 3W/4 F = 50 2 Widest Rectangle W = 6 L S = W/2 1.4L F = 50 S = W/ L F = 50 2 Wide Oval W = 3 L S = W/4 R = L/2 S = 3W/4 R = L/2 2 Long Oval L = 3 W S = W/2 F = L/4 S = W/2 R = L/4 2 Long Rectangle L = 2.8 W S = W/2 F = L/4 S = W/2 R = L/4 2 Longest Rectangle L = 3.1 W S = W/2 F = L/2 0.7W S = W/2 R = L/2 0.7W e.g. for Long Rectangles: if L > 1.67 W then two microphones are required: Long Rectangle: L > 1.67 W. and L < 2.8 W W/ L 0.5 L 0.25 L 0.5 L 0.7 W 1.4 W 0.5 L 0.7 W Longest Rectangle: L > 2.8 W. and L < 3.1 W W/2 L

38 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page Appendix D Additional sensitivity contour plots Figure 24: Alternative layout options for large wide trapezoidal tables

39 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page 39 Figure 25: Wide rectangle up to (a) 12 seats (2+8+2); (b) 18 seats (3+12+3)

40 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page 40 Figure 26: Wide rectangle up to 20 seats (6+8+6) or 24 seats (7+10+7)

41 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page 41 Figure 27: Long rectangular table for up to 20 seats (7+6+7)

42 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page 42 Figure 28: Small Kidney table showing fine detail, and oval table overlay kidney sensitivity variation within: 4 db (±2 db) throughout the Primary Usable Perimeter; 6 db (±3 db) throughout the Entire Perimeter, including all of the unusable Front Perimeter caters for collaborative learning applications where students may cluster around all sides of the table.

43 ICT Volume AV Design Calculators, Tools and Resources Microphone Placement Page 43 Figure 29: Small wide oval table at absolute minimum aspect ratio L/W = 9/20, sensitivity variation ±3 db around the entire usable perimeter