getting a set of eq filter to match a flat frequency response to a given measurement

[guynaa]

hello, i made a measurement in REW.
my goal is: create a set of eq filter to turn a flat frequency response into the frequency responsive of the measurement.

when i try go into the eq page regularly, it tried to correct the measurement INTO a flat frequency response (the opposite of what i wanted).

is there a way to achieve that?

 

[Wayne A. Pflughaupt]

Don’t quite get what you’re doing. Are you taking a measurement of the speakers, applying EQ filters via your receiver or outboard equalizer, and then taking another measurement? Or something else?

Regards,
Wayne

 

[guynaa]

i have measured a filter i want to mimic with eq

 

[Wayne A. Pflughaupt]

Sorry, but I have no idea what that means.

Regards,
Wayne

 

[guynaa]

lets say a i have a measurement that looks like it has a peak at 300Hz with 1.5Q and 5db Gain.
if i was to go to the eq tab, and press "match response to target", i would get a suggested with a PK filter at 300Hz with 1.5Q and -5db Gain. (minus 5db)

the matching tries to "cancel out" the measurement back to the target.
i would like go get the opposite of that, match a flat response to that measurement
so i would get an eq filter of: 300Hz with 1.5Q and 5db Gain

 

[Wayne A. Pflughaupt]

So you want to induce (for instance) a 5 dB peak @ 300 Hz to an otherwise flat measurement?

Regards,
Wayne

 

[guynaa]

yes, assuming the original measurment has that 5dB peak.

all i want is a set of filters try to closely match the frequency response of the measurement
for example if the measurement has a big bass boost,
i want the eq to also have a big bass boost.

 

[Wayne A. Pflughaupt]

Got it. It’s pretty easy, really. Using your previous example:

lets say a i have a measurement that looks like it has a peak at 300Hz with 1.5Q and 5db Gain.

if i was to go to the eq tab, and press "match response to target", i would get a suggested with a PK filter at 300Hz with 1.5Q and -5db Gain. (minus 5db)

The “match response to target” function got you the filter Q and frequency. So, instead of cutting 5 dB, simply boost 5 dB, and you’ve “duplicated” the peak.

Regards,
Wayne

 

[guynaa]

thats true, the problem is that when i do that i sometimes get filter with very big negative gain (-30dB),
that would force me to use a filter with 30dB,

this happens because the program does not know i want to reverse the gain, it has protection so the gain wont be too big, but not too negative..
thats why i need to reverse the impulse response or something like that/

 

[Matthew J Poes]

thats true, the problem is that when i do that i sometimes get filter with very big negative gain (-30dB),
that would force me to use a filter with 30dB,

this happens because the program does not know i want to reverse the gain, it has protection so the gain wont be too big, but not too negative..
thats why i need to reverse the impulse response or something like that/

You can ignore the suggested amount of boost and simply use a level you are ok with.

What are you trying to achieve? Why are you trying to apply this particular target boost? Adding a lot of boost can cause amplifier and signal clipping and generally isn't a great idea.

 

[Wayne A. Pflughaupt]

Not sure what to tell you. Most people take measurements to eliminate undesirable peaks in response, not make them twice as bad.

Regards,
Wayne

 

[John Mulcahy]

Try loading the target response you want as a calibration file, that will effectively invert it so your flat measurement turns into an inverted version of the response. The filters will then, in trying to flatten it again, create something like the original target response.