Help with miniDSP 2 X 4

[RCHRDM]

I recently bought a miniDSP 2 X 4 and want to load equalization settings from REW into the DSP.

I installed the 2-way Crossover Advanced plugin software in the miniDSP.

This software has two sets of equalizers, one on the input and one on the output of both channels (left and right). Each equalizer has six controls or frequencies and a button to download from REW.

How to I do an analysis of a room and upload 12 different frequencies (six in the input stage and six in the output stage) to the miniDSP?

Richard

 

[John Mulcahy]

The EQ window allows various miniDSP equalisers to be selected and there is an option to save the filter settings to a file which the plugin can import. There is a miniDSP app note for the process. The equaliser option for the 2x4 allows 6 filters, if you wanted to use more than that you would need to generate 2 different sets of 6 filters.

 

[RCHRDM]

Thanks, John.

I am new to this room EQ stuff. Are six filters or frequencies normally sufficient to adjust a room's acoustics?

I guess I would have to somehow manually set up a second set of six filters based on the values created by REW?

I wonder why the miniDSP does the equalizer like that with two separate sets of filters?

Richard

 

[John Mulcahy]

There are usually crossover filters between inputs and outputs so the filters act on different signals or on different outputs depending on how the crossovers are being used. Six filters is plenty to deal with modal issues below 200 Hz or so, which is the range where filters are most appropriate.

 

[RCHRDM]

Thanks! I will follow the REW manual and give it a try.

Richard

 

[RCHRDM]

I think I figured out how to get a room analysis. However, I don't understand the the Equalizer data yet.

I have attached a plot from REW. The upper curve is the room data and the lower curve is the Predicted.

Why is the predicted more or less a mirror image of the room graph? Shouldn't it be a flatter version of the original room sample or curve? That is the objective of equalization, right?

I probably have some settings wrong.

Richard

 

[John Mulcahy]

Those are relatively high in the frequency range for EQ and so quite broad filters, they are bringing the overall level towards the target rather than addressing individual features. Best paying most attention to the range up to 200 Hz and making sure the "Allow narrow filters below 200 Hz" option is selected in the Filter Tasks pane.

 

[RCHRDM]

John

I am unclear about the "Predicted" line. I assumed that the predicted line is the curve that we would like to have in the room, once the appropriate corrections have been made. Why does it seem to be the same as the measured line?

Please explain the predicted line.

Richard

 

[John Mulcahy]

The predicted line is how the response would look after applying the filters. The desired response is labelled "Target".

 

[RCHRDM]

John

Here is another plot from REW. The predicted looks the same as the measured to me. I guess I don't get it yet.

Richard

 

[John Mulcahy]

The green shaded area is the effect of the filters. In the region where the filters have meaningful effect the predicted response has been brought down to the target at about 90 dB. At the centre frequency of filter 1 the difference is about 14 dB. Where the filters have little or no effect the predicted will be close to or the same as the target.

 

[RCHRDM]

Thanks, that helps a lot.

Questions:

1. It appears that the room has a steep "low" area between 40 and 50 Hz., but there is apparently no correction for that?

2. Likewise, there is no correction for the variations between 100 to 200 Hz?

3. I have attached the plot from the miniDSP screen with the filters applied. Does this look right?

Richard

 

[John Mulcahy]

1. It appears that the room has a steep "low" area between 40 and 50 Hz., but there is apparently no correction for that?

The whole response below about 55 Hz lies below the target, REW will not apply any corrections below the first point the response crosses the target to avoid the risk of driving speakers beyond their low frequency extension limits. Even if that were not the case, narrow dips are not boosted as they are very location dependent and attempting to address them would create response peaks at positions even a very short distance away from the measurement point.

2. Likewise, there is no correction for the variations between 100 to 200 Hz?

What frequency span did you set as the match range in the Filter Tasks pane?

3. I have attached the plot from the miniDSP screen with the filters applied. Does this look right?

Yes.