FDW to remove the room reflections

[tjcinnamon]

I want to measure a speaker from 2m away from 500Hz to 5000Hz. What’s the best FDW setting I can use to negate reflections?

Also, should I measure full range and set FDW or should I just measure a bit before 500Hz and a bit after 5000hz?

 

[John Mulcahy]

Generally better making full range measurements where possible.

For a good result down to 500 Hz you would typically need the right window to be around 3 - 4 ms. You can then add FDW somewhere between 1/3 and 1/12 octave, but results will depend heavily on how far the speaker and mic are from room surfaces, the further the better. You can see the difference between the windowed and unwindowed impulse responses on the Impulse graph.

 

[tjcinnamon]

Generally better making full range measurements where possible.

For a good result down to 500 Hz you would typically need the right window to be around 3 - 4 ms. You can then add FDW somewhere between 1/3 and 1/12 octave, but results will depend heavily on how far the speaker and mic are from room surfaces, the further the better. You can see the difference between the windowed and unwindowed impulse responses on the Impulse graph.

Thank you!

Unfortunately the speaker and mic are about 1m away from room surfaces. Does that narrow down a value between 1/3 and 1/12?

 

[John Mulcahy]

No. To remove reflections requires a right side window shorter than the time to the reflection. That in turn raises the minimum valid frequency and reduces the frequency resolution of the result. You might get an acceptable result from playing with the FDW setting but that depends on the specific details of the environment. If you need an accurate result from 2m best moving the equipment to a larger space and searching for "loudspeaker ground plane measurements".

 

[tjcinnamon]

No. To remove reflections requires a right side window shorter than the time to the reflection. That in turn raises the minimum valid frequency and reduces the frequency resolution of the result. You might get an acceptable result from playing with the FDW setting but that depends on the specific details of the environment. If you need an accurate result from 2m best moving the equipment to a larger space and searching for "loudspeaker ground plane measurements".

That makes sense. Can windowing be used for nearfield and would that solve my problem.

I’m trying to closely align some drivers as best I can for phase cancellation. The better I can match them while “in air” the better the crosstalk cancelation will be

 

[John Mulcahy]

Can windowing be used for nearfield

Yes.

would that solve my problem

Not entirely. If you want the result to have low frequencies you need a sufficiently long window. If the window is longer than the time for reflections to arrive they will still be in the result, but the direct sound would be much louder than the reflections for a nearfield measurement, reducing their effect.