REW Beta Release Overlay chart new feature suggestion

[GunghoUK]

Hi all, please forgive my ignorance of REW but I am trying to evaluate the output response of guitar pickups, using REW, by connecting the generator to a coupling coil strapped to the pickup under test. The transmitter coil has nominal parameters of resistance = 162 Ohms and inductance = 800 microhenries and is connected to the headphone jack of my soundcard.

In order to purely compare the frequency responses not actual signal levels, in the overlay chart, it would be ideal if all the active traces could be normalised (vertically range converted 0 to 1) to see their relative differences more accurately.

Is this currently possible elsewhere or am I missing something already in REW?

Additionally it would be ideal if the output generator amplitude could be reduced (or increased) by 6dB per octave, for reactive (or capacitive) loads like a guitar pickup or speaker. Alternatively maybe a -6b/Oct filter could be added to the measured responses of all or selected graphs to emulate the generator reduction/oct function.

This would then give a true frequency response for the inductive transmitter load and the inductive nature of the pickups themselves. To that end the inductance & resistance of the generator load would need to be inputted beforehand hence the need to describe my transmitter load coil characteristics.

I hope my ramblings have made sense. Thanks for any feedback and thank you for the awesome REW.


EDIT : So sorry. Just realised I've posted in the wrong forum section. Please forgive me and move as appropriate. Thank you

 

[John Mulcahy]

You can align a set of responses by right clicking on the graph and using Align SPL.

You could apply a 10 Hz first order low pass filter to get -6dB/octave, and either save the filtered result as a measurement using the "Generate measurement from predicted" action in the EQ window filter tasks pane or just view the filtered results on the Overlays - Predicted SPL graph. Right clicking on that graph gives you an option to copy filters from one measurement to all the other selected measurements, so you would only need to manually set the filter on one measurement.

 

[John Mulcahy]

Note that the comments above refer to the V5.40 beta build.

 

[GunghoUK]

Thanks for the reply but I'm not sure you understand my measurement results. The output plot looks like a band pass filter with a very low Q, due to high coil resistance, and a variable centre resonant frequency. See attachment. I think what I need is to apply is a -6db/octave filter to the entire response curve but I cant see that option currently.

 

[John Mulcahy]

Your measurement vs your measurement with 1st order low pass (so roll-off of 6 dB/octave) at 10 Hz, aligned at 1 kHz:

 

[GunghoUK]

Yup that's the goal. Now to figure out where the settings are.

 

[John Mulcahy]

It's a 10 Hz first order low pass filter ("LP1") applied in the EQ window.

 

[GunghoUK]

Thanks. I'll give that a go

 

[GunghoUK]

Sorry one more question. Why did you choose a 1kHz alignment frequency. In my ignorance I would have thought the peak response frequency of approx. 3kHz might have been the obvious choice.

 

[John Mulcahy]

Just to pick a frequency.

 

[GunghoUK]

Perhaps the end tangental/linear point of the original curve?

 

[John Mulcahy]

I suspect that particular alignment is irrelevant since it is responses after filtering that will be aligned with each other and they should have more obvious choices.

 

[GunghoUK]

Yes, in hindsight, alignment seems irrelevant,. EQ application on measurement causes the signal to drop into the system noise floor at the highest frequencies so I'll need to apply the EQ post measurement which is more logical and was suggested in my original post. Thanks for your help now I know how to go about it in REW and just need to sort out the settings to achieve the SPL alignment without a simple 0 - 1 normalisation of the individual plot data. Thanks.