Multi-Sub Optimizer Reference (page 13)

# Specifying Optimization Options

Before running an automated optimization, the optimization options should be set. On the main menu, choose Tools, then Optimization Options.... Select the configuration whose optimization options you wish to set. Unless you are specifying optimization options for a sub-only configuration, each configuration will have four property pages for setting optimization options. For sub-only configurations, there is no Target Curve property page.

### Method Property Page

The Method property page of the Optimization Options dialog is shown below.

On the Method property page, you have two choices of optimization method. The choice you make depends on whether or not you will be adding global EQ (such as Dirac) to the main listening position (MLP) later, or whether the EQ determined by MSO will be the final one. These are not hard and fast rules though. For instance, sometimes As flat as possible without additional global EQ might work better than Best match of MLP with other listening positions even if you plan to use additional global EQ later. Experience has shown that As flat as possible without additional global EQ can produce a smoother response at the MLP than Best match of MLP with other listening positions, thus making global equalization of the MLP easier.

#### As flat as possible without additional global EQ

If you don't anticipate using any global EQ after running MSO, this is the option to choose. Using this method, MSO tries to get the frequency response at each listening position as flat as possible without regard to the response matching of the MLP with the other listening positions. At first blush, using this method might appear inferior to Best match of MLP with other listening positions with regard to getting the smallest seat-to-seat variation in frequency response possible. However, data examined so far has shown that Best match of MLP with other listening positions produces rather small improvements in seat-to-seat response variation compared to As flat as possible without additional global EQ.

#### Best match of MLP with other listening positions

If your intent is to apply further global EQ after optimizing with MSO, you'll usually want to choose this option. When you choose Best match of MLP with other listening positions, MSO tries to get the response at the MLP to be as flat as possible, while matching the response at each non-MLP position as closely as possible to the response at the MLP. Response errors between the MLP and non-MLP positions in this mode that are independent of frequency count as simple level differences and do not count in the error calculation. In theory, when using this mode, if you apply further global EQ after optimizing with MSO to make the response at the MLP even flatter than MSO makes it, the flatness of the responses at non-MLP positions will be further improved as well.

#### Choose main listening position (MLP)

When you choose the Best match of MLP with other listening positions optimization method, you must also choose which listening position is the MLP. Choose the appropriate position from the combo box. If you have a measurement group that is named "MLP", it will be chosen automatically for you.

#### Reference level for MLP

If you are setting the optimization options for a sub-only configuration, you'll need to set a reference level, in dB SPL, for the MLP here. MSO will attempt to force the MLP's average level over frequency in the Frequency range to optimize (which you specify in the Criteria property page) to this value. See the Configurations section for a discussion of sub-only configurations. See below for a discussion of the Frequency range to optimize option.

### Criteria Property Page

The second property page is called Criteria. It allows for specifying the frequency range for optimization, the frequency range over which the reference value is computed, and the maximum time the optimizer is allowed to run. The Criteria property page is shown below.

#### For Ordinary Configurations With Subs and Main Speakers

For each measurement group, the RMS value of the deviation of the response of that group from a reference level or curve is computed over frequency. The frequency range over which this error computation is performed is specified in the Frequency range to optimize field. The reference level is not specified directly as a numeric value, but indirectly using a range of frequencies. This range is specified in the Frequency range to compute reference. Where applicable, the average value of the combined response of mains and subs at a given listening position is computed over the frequency range specified in the Frequency range to compute reference fields. This is the reference level

When the optimization method chosen is As flat as possible without additional global EQ, this reference level is computed separately for every listening position. The reference level may therefore shift from one listening position to another, and also during the process of the optimization as filter parameter values are changed. When the optimization method chosen is Best match of MLP with other listening positions, the reference level is only computed for the main listening position. In this case, the "reference" for the non-MLP positions is the response curve of the MLP itself. Errors between the curves of the MLP and non-MLP positions that are frequency-independent are subtracted out before computing the RMS value of the error for that position over frequency. Therefore, level shifts between the MLP and other positions do not count toward the total error.

In order to get the flattest possible response, the reference level needs to shift a little as the shape of the response curve changes, but not too much. If it shifts too much, an undesired baseline shift between subs and mains might occur. To prevent the reference level from shifting too much, the reference level frequency range needs to include some frequencies where the response is nominally flat and unaffected by filters you add to the subs.

#### For Sub-Only Configurations

When you optimize sub-only configurations, you must always specify an MLP in the Criteria property page. You must also specify a fixed reference level, which only applies to the MLP. When the As flat as possible without additional global EQ option is chosen, the fixed reference level is always used for the MLP, and for the other listening positions it is computed using the frequency range specified in Frequency range to compute reference fields. When the Best match of MLP with other listening positions option is chosen, the reference for the non-MLP positons is the MLP itself, so the Frequency range to compute reference is not needed, and the controls that allow you to specify it are disabled.

#### For All Configurations

The final option on the Criteria page is the time duration the optimizer is allowed to run. The optimizer is always searching for a better solution as long as it's running. The maximum allowed duration can be specified here, and may require a bit of experimentation to get right. If too long a duration is specified, the optimization is easily stopped via the Stop Optimization button on the toolbar.

### Group Weights Property Page

The third property page is called Group Weights. It is shown below.

For each measurement group, the RMS amplitude deviation from the reference in dB is computed over frequency. Then these RMS errors are again combined in an RMS fashion over all measurement groups to get a single number whose value would be 0 dB for perfectly flat response at all listening positions. This second stage of RMS error computations can be weighted, for example to prefer a slightly flatter response at the main listening position. This is done by giving the MLP a weight of 1.0 and the other listening positions a weight of less than 1.0, maybe something like 0.75 for all non-MLP positions. If you choose the Best match of MLP with other listening positions optimization method, it is suggested that you leave all these weights at their default value of 1.0.

### Target Curve Property Page

Finally, a house curve can be imported in the Target Curve page. This property page is shown below, and will not be shown for sub-only configurations.

A target curve is a text file with two columns. The first column is the frequency in Hz, and the second column is the curve value in dB. Currently, there is no target curve editor in MSO, so you must create the curve yourself with a text editor. A workaround exists as follows.

- Download the target curve example.
- While observing the example's graph, adjust the low-frequency boost, center frequency and Q of the shelving filter to get the curve you want.
- In the Filters node under Sub Channel 1, choose Save this Channel's Response as Target Curve.
- Specify the file name and folder in the Save As dialog box.
- A text file will be created with a curve that goes from 5 Hz to 400 Hz.

It is recommended that the Use logarithmic interpolation option be checked. In this mode, if there are two widely-spaced data points on the curve, the line connecting them will appear straight when the curve is plotted on a graph having a logarithmic frequency scale. For instance, Earl Geddes recommends a curve with a slope of between 3 and 6 dB per decade from 200 Hz down to 20 Hz. This can easily be accomplished using logarithmic interpolation and a text file having only two entries as follows, which shows a 6 dB per decade slope.

20.0 6.0

200.0 0.0

The curve's value will be constant at 6.0 dB below 20 Hz, and constant at 0.0 dB above 200 Hz.

For more complex curves, Use cubic spline interpolation will draw a smooth curve between the target curve points.

It is *not* recommended to establish a house curve via individual EQ of each sub. Rather, individual EQ should be used to establish a flat response, and EQ applied at a later stage via a shared filter DSP channel EQ with the filters of the individual channels locked. This will be discussed later.