[mc_lover]

Hi there.
I am not sure if Dynaudio I use every is working properly or not. it seems to me that its mid-range driver is stiff. It sounds unlike it used to be.

Is there a function in REW to measure just the frequency response of a speaker like those made in an echoic room? If i am correct, a REW regular frequency sweep measures interaction between speaker and room together.

Any suggestion what parameter to set is very much welcome.
Thank you

 

[jtalden]

You are correct that the room will have influence, but this will not be a problem to see if a MR driver is "working properly".

The TW/MR drivers are best placed 1m or more from any boundary. If placed on a stand away from the walls, then you measure down someplace well below 1kHz with good accuracy. This should be fine to tell if a MR is functioning normally. The closer the boundary and large objects the higher the frequency the impact to the response will be.

> Set the mic about 12" away from the TW driver on axis.
> Set the sweep measurement sweep from 200-20kHz and measure so the level is about 70dB.

The response will be reasonably representative, but will not be the same a good anechoic measurements. The reflections of the boundaries will have some impact in the midrange. Also the baffle step may not be properly represented. These limitation are not important to see if a MR is working properly.

If you want to remove the impact of the boundaries look at the resulting impulse and:
> Place the REW left window just before the initial rise of the impulse.
> Place the REW right window just before the first refection peak. It will occur someplace between 2-4 ms after the initial large impulse peak if the speaker is placed as suggested. The time is dependent on how close the nearest boundary is.

This will still does eliminate all sources of errors as compared to an anechoic response. but it will be reasonably close.

A search for "near field loudspeaker measurement" should turn up several more detail descriptions.

 

[Wayne A. Pflughaupt]

Doesn’t REW have a gated function that can approximate an anechoic measurement?

Regards,
Wayne

 

[jtalden]

Doesn’t REW have a gated function that can approximate an anechoic measurement?

Yes, this is just a terminology difference. A gated measurement is the same as setting the left and right windows appropriately.

There is still the minor differences that I mentioned as compared to a truly anechoic measurement however. These are not important to the OP's objective however.