Evaluation
Pressure plots
MAPP 3D provides two different methods to evaluate the performance of a system design:
The first shows sound pressure distribution on geometry for a narrow range of frequencies—up to one octave wide—in the Model View tab. These are often referred to as pressure plots.
The second is to add virtual microphones to the model in strategic locations to evaluate the broadband response and maximum acoustic output. See Measurement View.
Pressure plots displayed on geometry in the Model View tab enable visualization of coverage and loudspeaker-to-loudspeaker interactions on any geometry selected for prediction. A pressure plot can be up to one octave wide, a periscope view of the entire band spectrum. Below, the image on the left is a one-octave wide prediction, centered at 4 kHz. The data from the microphone location in the Model View (left, below) is represented in the Measurement View (right, below). When the prediction bandwidth is one octave, the data points highlighted in the Measurement View are averaged and the level is represented using colors in the Model View.
Model View Periscope View of Bandwidth (left) and Measurement View, One Octave, Centered at 4 kHz, Highlighted (right)
Generally, a one-octave prediction at 4 kHz is representative of high-frequency coverage. A one-octave prediction at 250 Hz is representative of mid-band frequencies. As frequency decreases to the subwoofer range, 1/3-octave predictions from 100 Hz down to the lowest frequencies the system reproduces are effective in analyzing coverage and loudspeaker-loudspeaker interactions.
Prediction planes are acoustically transparent in the model, except for geometry selected as Ground Plane. See Ground Plane.
The graphic representation of sound pressure in MAPP 3D has several parameters located on the main application window and in the FILE > PROJECT SETTINGS menu, both on the Appearance and SPL tabs. These are described below.
Model View pressure plots
To generate a pressure plot on geometry, the model must include:
At least one loudspeaker on a visible layer
At least one prediction plane (geometry selected for Prediction) on a visible layer
Processor channel must be assigned and un-muted with level above -∞ dB
Select the following for a prediction:
Generator level
Signal type
Width of prediction, up to one octave
Range of frequencies – no options when width is one octave
enter frequency
Click PREDICT (cmd-R) to generate a pressure plot on the prediction planes. Click CLEAR to remove pressure plot from prediction planes.
Model View Tab, Prediction Parameters
Signal generator
The generator level and signal type affect the SPL results.
The generator level is adjustable between -90 dB and +50 dB
The generator signals available are:
Pink Noise is generated by SIM, which has a 12.5 dB difference between peak and average levels (crest factor).
B-Noise is SIM pink noise filtered with a B-weighting curve. Use B-Noise to more accurately predict the maximum output of a system when speech is the input signal.
M-Noise is a signal that closely represents the peak-to-average ratio as a function of frequency expected of music, currently under consideration by an AES standards committee. Below 500 Hz Pink Noise and M-Noise are functionally the same. For more information, please see m-noise.org.
Weighting Curves – A, B, C, and Z
SPL tab parameters
Adjust the Project Settings parameters to change how pressure plots are displayed. Use the FILE > PROJECT SETTINGS menu (cmd-shift-P) and select the SPL tab to access these settings.
Project Settings – SPL Tab
SPL scaling
When SPL is selected, the average SPL (not peak) is represented by colors for the selected bandwidth. The default range and value settings can be changed by clicking the MANUAL buttons (above) and entering new values.
The Max Value is the loudest level plotted. If levels exceed this value, the area is plotted in white. Increase the Max Value until white is no longer displayed, see below.
The Min Range is the number of dB below the Max Value that will be plotted. If the prediction plane is plotted black, the SPL values are below the level range plotted. To plot as colors other than black, decrease the Max Value or increase the Min Range value, see below.
Model View – SPL, Change Only to Max Value Level: 85 dB (left), 100 dB (middle), and 115 dB (right)
When a loudspeaker is predicted with the Max Value = 75 dB, SPL values of 70 dB are plotted in red, (1) below. When the Max Value is set to 87 dB, 70 dB is plotted as yellow, (2) below. It’s the same data, just scaled and plotted differently.
For (2) below, when the Range = 42 dB, 42 dB of attenuation below the Max Value of 87 dB is plotted. In (3) below, when the Range = 30 dB, levels between 87 dB and 57 dB are plotted. 70 dB is now plotted as a light blue color. Again, it is the same data, just scaled and displayed differently.
Model View Predictions, SPL Mode – Same Data Represented Differently Based on SPL Max Level and Min Range Settings
The SPL values are only for the range of frequencies selected. When narrower prediction bandwidths are selected, the SPL values decrease. The pressure plots below use the same parameters for the same loudspeaker, the only change is the bandwidth of the prediction.
Model View, SPL – Less SPL as Bandwidth Narrows
Resolution
Resolution determines the number of colors used to represent pressure. Select Full Resolution, or level deviation per color used. The number of colors used to plot SPL or Attenuation can be helpful in different ways. Full Resolution uses a large number of colors to represent pressure, which are impressive images to include with a proposal. Using the 3 dB/color or 6 dB/color resolution settings makes determining coverage areas more obvious.
When changing between resolutions, no re-calculation is performed, the application only re-renders the prediction data.
Model View – Resolution Settings, Full Resolution, 0.5 dB, 1 dB, 3 dB, and 6dB Per Color
Attenuation range
The Attenuation selection uses colors to represent attenuation, normalized to the point on all of the prediction planes that has the highest pressure level. The color at the top of the scale (0 dB) represents this point. All other points are equal or less in level, represented by colors indicating level of attenuation.
To modify the range of SPL represented, click MANUAL and enter a higher or lower range of attenuation to display. This is the level of attenuation from 0 dB that will be plotted.
Model View – Attenuation, Min Range 42 dB, 30 dB, and 18dB
Attenuation examples
When using Attenuation mode, the level is relative to the loudest source. It is not intuitive that a subwoofer predicted at one-octave, 4 kHz appears to be very loud (below). It is in fact 30-40 dB lower in level at 4 kHz than within its operating range. Because the subwoofer is the loudest source in the model at this range of frequencies, the 0 dB color is plotted nearest this loudspeaker.
Model View – Subwoofer Prediction, One-Octave, 4 kHz
If a small, full-range loudspeaker (MM-4XP) is added to the model, the 0 dB color is plotted nearest the full-range loudspeaker because it is much louder than the subwoofer at 4 kHz.
Model View – Subwoofer and Mid-High Loudspeaker Prediction, One-Octave, 4 kHz
Appearance tab parameters
Adjust the Project Settings parameters to change how pressure plots are displayed. Use the PROJECT SETTINGS menu (cmd-shift-P) and select the APPEARANCE tab to access these settings.
Project Settings – Appearance Tab, Mesh Density and Pressure Map Opacity
Mesh density
The Mesh Density selection determines the resolution of the SPL map, and affects both calculation time and how detailed the pressure plot will be. A higher density has more points on the listening planes, creating a highly detailed pressure plot that takes longer to calculate than lower mesh densities (see below).
Model View Prediction – Mesh Density Setting Very Low (left), Very High (right)
Pressure Map opacity
Pressure Map Opacity changes the opacity of the prediction data in the Model View tab, 1.0 is opaque, 0.1 is almost transparent. (see below).
Model View – Pressure Map Opacity 1.0 and 0.7
Data Near Loudspeaker
Each loudspeaker data set has a small data void around the loudspeaker, intentionally.
Model View – Data Not Available Within One Meter
Example uses
Example 1: The Model View image below is of a one-octave, 4 kHz, 3 dB/color prediction. The colors indicate level on-axis , -3 dB down , -6 dB down for the octave centered at 4 kHz. When in SPL mode, the SPL values are for the bandwidth of the prediction only, one octave in this case. Relative pressure levels are available in Attenuation mode, the same as in MAPP XT.
Model View – One Octave, 4 kHz, 3 dB/Color, SPL Prediction – On-Axis (1), -3 dB (2), -6 dB (3)
Example 2: To evaluate coverage of a loudspeaker model using a standard level of variance for designs, +/- 3 dB or 6 dB, one option is to select Attenuation, Min Range = 12 dB. The first two colors indicate where coverage is within the design limits. The center of the red area is relative 0 dB. Between the red and yellow color is 3 dB less level. Between the yellow and green color is another 3 dB less level.
Model View – Attenuation, Range = 12 dB, Colors Represent Relative Level
Below, three different loudspeaker models are evaluated with slight location and aiming changes. On the left, the coverage is too narrow, even if the loudspeaker were better aimed. In the middle, there is too much level variation from front to rear. On the right, almost the entire half of the seating area is within 6 dB for the octave centered at 4 kHz. Fills may be necessary for rear of seating. For model selection, placement, and aiming, this viewing strategy is efficient.
Model View – Attenuation, Range = 12 dB, Evaluate Loudspeaker Location, Orientation, and Model Selection for Coverage
Measurement view
MAPP 3D provides two different methods to evaluate the performance of a system design:
The first is by showing sound pressure distribution over listening planes for a narrow range of frequencies—up to one octave wide—in the Model View tab. See Pressure plots.
The second is to add virtual microphones to the model in strategic locations to evaluate the broadband response and maximum acoustic output. This option is described below.
Analysis using only one of these methods is only partial analysis. Although it is true that most design work can be effectively accomplished using only Measurement View data, Model View predictions can inform microphone placement and help identify instances of coverage beyond the audience area(s).
To display prediction data in the Measurement View tab, at least one loudspeaker and one microphone must be added to the model.
Add microphones
Microphones are added to display broadband information and SPL values at the microphone locations.
INSERT > MICROPHONE or right-click in model, select INSERT MICROPHONE.
Use a Z-axis value representative of an average ear height.
Once inserted, microphones can be moved, copied, arrayed, mirrored, etc., in the same ways the geometry and loudspeakers are modified.
Insert Microphone Window
Model View – Microphone with Loudspeaker and Geometry
Microphone placement
There are many ways to approach system design and optimization; some provide more repeatable and homogeneous coverage results than others. Microphone placement is driven by design and optimization methodology. Below are some typical microphone placement methods.
Line Arrays: Generally, microphones are placed incrementally on-axis of a line array, at least at the beginning, middle, and end of coverage. For deeper coverage areas, additional microphones are placed incrementally.
Point Source Arrays: Generally, microphones are placed on-axis to the first element, at the -6 dB point1 of the first element, and on-axis of the adjacent array element.
Fills: Generally, on-axis to the larger system, at the -6 dB point1 of the larger system, and on-axis to the fill loudspeaker.
Note
1Within the range of frequencies that are directional, usually reproduced by a horn, generally 2 kHz to 20 kHz.
Frequency and IFFT response (SIM)
The Measurement View tab is based on the architecture and nomenclature of Meyer Sound’s SIM dual-channel FFT measurement platform, which is used in our anachoic chamber to collect data for MAPP 3D.
Frequency and IFFT response (SIM) tab overview
Measurement View Tab – Frequency & IFFT Tab
Plotter View Selection for each of the four panes, see options below this list
Zoom tool for vertical scale
Measurement Selection for processor, processor channel, microphone, smoothing level, IFFT window time, and processor trace toggle
Result Amplitude selected for this plotter
Vertical Scale dB/division selector
Room + Processor Amplitude selected for this plotter
Horizontal Zoom for each plotter
Expand/Collapse Plotter to occupy entire window
Result Phase response selected for this plotter
IFFT impulse response selected for this plotter
Propagation Delay amount of delay added to generator signal to synchronize with microphone
Air Attenuation current settings
Store Live Trace and name it for later recall—stores both the Headroom and Frequency & IFFT traces
Visual Options select to zoom and pan plotter view, link to zoom/pan like plotters
Trace Level Offset entered in dB
Propagation Delay for associated trace
Microphone name selected for live/stored trace
Processor Output Channel active for live/stored trace
Trace Recall drop-down to select stored traces
Show/Hide and Color selection for traces
Measurement View – Frequency & IFFT, Pane View Selection Options
Frequency & IFFT response (SIM) signal patching
A transfer function is the difference between two signals. In order to compare them, the signals compared must be synchronized. Delay is added to the generator signal and to the processor signal to synchronize them with the microphone signal. This delay is added in MAPP 3D by clicking the Propagation Delay AUTO button. The delay time is displayed next to the AUTO button. A delay time can also be manually entered here.
Measurement View, Frequency & IFFT Response Tab – Auto Propagation Delay
Three transfer functions are available: Processor, Speaker+Room, and Result, which are depicted below and used in MAPP 3D.
Note
In MAPP 3D, “Room” is the same as “Speaker + Room” used below.
SIM Signal Patching and Transfer Function Names
For devices (processors and loudspeakers) that have the same performance at low and high levels, the transfer functions do not change for broadband, high density (not sparse) signals like pink noise, B-Noise, and M-Noise, which are selectable as the generator signal in MAPP 3D. The loudspeaker data sets in MAPP 3D are stored at the onset of limiting, the boundary of linear operation, which we use as the definition of maximum acoustic output.
The views available in the Frequency Response & IFFT panes are (defaults in bold):
Result Amplitude: the amplitude transfer function between the processor input and the microphone
Result Phase: the phase transfer function between the processor input and the microphone
Room Amplitude: the amplitude transfer function between the processor output and the microphone
Room Phase: the phase transfer function between the processor output and the microphone
Processor Amplitude: the amplitude transfer function between the input and output of the processor
Processor Phase: the phase transfer function between the input and output of the processor
Room + Processor Amplitude: the Room and Processor amplitude transfer functions displayed in the same plot
Room + Processor Phase: the Room and Processor phase transfer functions displayed in the same plot
IFFT: the Inverse Fast Fourier Transform, or the impulse response. Only available as the difference between source and the microphone (Result), which includes the processor, loudspeaker, and air propagation time.
Frequency and IFFT data usage
Observe the Result Amplitude trace as equalization and level adjustments are made if there is a target system response curve. The Result curve represents the spectral balance difference between the processor input and the loudspeaker output. If the Result trace is equal amplitude (flat), the system response will represent the input without changing the spectral balance. The proper spectral balance of a system is variable, usually based on a specification or a user’s expectation. It should be determined what the desired spectral balance of system is before committing to a design. Ensure the design has enough available headroom to reproduce the spectral content at the desired or specified level without exceeding available headroom.
Observe the Room + Processor Amplitude with the processor inverted for more precise equalization. Adjust equalization filters to match the inverted Processor trace to the Room trace; the Result will be equal amplitude (flat). It is not recommended to make equalization choices from only one microphone location for most applications. One exception would be a small control or listening room where there is only one primary listener.
Use the IFFT to synchronize arrivals of different sources:
measure, synchronize, and store the trace of the source that arrives latest in time
recall this trace (lower-left of window under Trace Name)
mute the first source, unmute the second source
click the AUTO button
The difference between the stored Propagation Delay (listed in the stored trace under Result) and the current Propagation Delay is the delay time to enter in the processor channel for the earlier arriving source to synchronize the two sources.
Signal generator
The generator level and signal type affects the SPL results.
The generator level is adjustable between -90 dB and +50 dB.
The generator signals available are:
Pink Noise is generated by SIM, which has a 12.5 dB difference between peak and average levels (crest factor).
B-Noise is SIM pink noise filtered with a B-weighting curve. Use B-Noise to more accurately predict the maximum output of a system when speech is the input signal.
M-Noise is a signal that closely represents the peak and average signals present in music. Currently under consideration by an AES standards committee. For more information, please see M-Noise.org.
Weighting Curves – A, B, C, and Z
View options
Click-drag in any pane to display horizontal and vertical axis values.
Select ENABLE WINDOW CENTERING to vertically center amplitude plots.
Select Enable Window Centering to Plot Traces at Vertical Center
The LINK ALL option links the Zoom and Pan settings for all like traces. All amplitude plots will have their Zoom/Pan settings linked. All phase plots will have their Zoom/Pan settings linked.
Zoom enables the use of the mouse wheel to zoom in/out of any pane.
Pan enables click-drag panning within a pane to view the entire plot when zoomed in.
Trace display
Up to five traces can be displayed in the data plots at the same time: the live trace and four previously stored traces. Each trace will be identified by the selected Output Channel, Microphone, and Propagation Delay. Enter offset values in dB for any trace.
Trace management
TOOLS > TRACE MANAGEMENT
Traces can be deleted by opening the Trace Management window from the Tools drop-down menu. Select the trace(s) to delete, click DELETE. When traces are stored, both the Frequency & IFFT and the Headroom data is stored using one name.
Headroom tab overview
Measurement View Tab – Headroom Tab
Vertical Zoom of dBSPL axis
Microphone Selection for live trace
Headroom for Individual Loudspeakers remaining(+) or exceeded(-), relative to maximum linear output
Trace Toggles show/hide different plotter data, see below
Store Live Trace stores both the Headroom and Frequency & IFFT traces
Visual Options select to zoom and pan plotter view
SPL Values maximum SPL, dependent on generator signal type and level
Horizontal Zoom of plotter pane
Microphone name selected for live/stored trace
Trace Recall Selection drop-down to select stored traces
Show/Hide corresponding trace
Vertical Scale dB/division selector
Trace toggles
Make selections to show/hide available traces.
Show Peak Spectrum
Show RMS Spectrum
SPL values
The SPL values are dependent on signal generator level and signal type selected.
Average SPL (A-weighted) – broadband pink noise or B-Noise value
Average SPL (C-weighted) – broadband pink noise or B-Noise value
Average SPL (Z-no weighting) – broadband pink noise, B-Noise or M-Noise value
Linear Peak SPL – broadband pink noise, B-Noise or M-Noise value
Max Linear Peak SPL – ultimate broadband peak SPL that can be achieved if all elements of the system are phase-matched and time-aligned to the microphone position. This is a static value, unchanged by Generator level adjustments or processor output gain adjustments. While this is not a realistic operational level, it is helpful when determining whether a loudspeaker system is capable of achieving a specified SPL value.
Headroom data
The Headroom tab is used to determine the maximum SPL by 1/3 octave band and broadband output. This measurement is a single ended measurement, not a transfer function. Selecting different generator levels and signals will affect the SPL results. The available headroom for each loudspeaker of the system is listed (3 above).
The data in the plotter changes color depending on which signal generator choice is made:
Pink – Pink Noise
Blue – B-Noise
Yellow – M-Noise
Trace display
Up to five traces can be displayed in the data plots at the same time: the live trace and four previously stored traces. Each trace will be identified by the selected Output Channel, Microphone, and Propagation Delay. Enter offset values in dB for any trace.
Trace management
TOOLS > TRACE MANAGEMENT
Traces can be deleted by opening the Trace Management window from the Tools drop-down menu. Select the trace(s) to delete, click DELETE. When traces are stored, both the Frequency & IFFT and the Headroom data is stored using one name.
MAPP 3D SPL vs. product datasheet SPL
The data presented in MAPP 3D matches real-world measurements and our datasheets. However, comparisons need to be completed in a manner that is representative of the loudspeaker to microphone distance and the acoustic environment. Meyer Sound Laboratories typically lists Linear Peak SPL on the product datasheets for broadband models as:
measured in free-field at 4 m, referred to 1 m. Loudspeaker SPL compression measured with M-noise at the onset of limiting, 2-hour duration, and 50-degree C ambient temperature is < 2 dB.
For subwoofers, the product data sheet typically lists Linear Peak SPL as:
measured in half-space at 4 m referred to 1 m. Loudspeaker SPL compression measured with M-noise at the onset of limiting, 2-hour duration, and 50-degree C ambient temperature is < 2 dB.
MAPP 3D, Lab 1.0 HR Project – Loudspeaker Face at 0 m ,0 m, 0 m, Microphone at 4 m, 0 m, 0 m
When loudspeaker data sets are acquired in our anachoic chamber, the distance between the loudspeaker and microphone is 4 meters. But, the industry typically specifies maximum SPL at a 1 meter distance. To convert a free-field, 4 meter test result to a 1 meter test result, the inverse square law is applied* and +12 dB is added to the result of the 4 meter test in free-field. This SPL value is listed on the loudspeaker data sheet making comparison to other 1 meter specifications simple.
For comparison of half-space, 1 meter measurements, where both the microphone and loudspeaker are on the ground, sound is radiated in one-half the area of the free-field measurement. Another +6 dB SPL is added because the radiation area is halved. Add +18 dB SPL to the result of a 4 meter, free-field measurement to compare it to a 1 meter, ground plane measurement.
To replicate this example in MAPP 3D:
open the Template Lab 1.0 HR (FILE > OPEN TEMPLATE)
move the microphone to 4 m, 0 m, 0 m (x, y, z coordinates)
place the loudspeaker face at 0 m, 0 m, 0 m (x, y, z coordinates)
predict using any of the noise sources
free-field: add +12 dB to the predicted Max Linear Peak SPL value in MAPP 3D
half-space: add +18 dB to the predicted Max Linear Peak SPL value in MAPP 3D
This result matches the Linear Peak SPL listed on the product datasheet.
* When the distance from a sound source is doubled, -6 dB SPL is lost. When the distance is halved, +6 dB SPL is gained. In this instance, 4 meters halved is 2 meters, and +6 dB SPL is gained. When 2 meters is halved, another +6 dB SPL is gained. In free-space, the SPL difference between a 4 meter measurement and a 1 meter measurement is +12 dB SPL.