Help! «No filters to export. The selected filter set has no filters that have an effect, nothing to export»

[anzoni]

Hello,

I'm just getting started with this great program and I've already made several measurements of my equipment, for now to discover how REW works.

Unfortunately I got into a loop from which I cannot get out when I try to save and export as .WAV the configurations of the filters generated by the EQ module in order to introduce them into Foobar2000; I'm doing something wrong when saving these settings.

I have tried:

Attempt 1
- Open measure (full L)
- Smoothed 1/6
- Open EQ window
- Calculate target level from response
- Match response to target
- Predicted curve and filter numbers appear
- Open EQ filters panel with the generated filter data
- Save filters by pressing the diskette icon in the EQ filters panel – a REW filter set file is generated and saved (filtres generats1...)

- "All SPL" screen - "all files" open "filters generated1..."filtres generats1 not found"

Attempt 2

- Open measure (full L)
- Smoothed 1/6
- Open EQ window
- Calculate target level from response
- Match response to target
- Predicted curve and filter numbers appear
- Open EQ filters panel with the generated filter data
- Save filters by pressing the diskette icon in the EQ filters panel – a REW filter set file is generated and saved (filtres generats2...)
- Close the EQ filters panel
- Filter Task – Generate measurement from predicted

- All SPL window appears new equalized measurement curve with filters (EQ full L2)
- Save new curve (EQ full L2) from the All SPL window

- Close EQ window
- Close full L measurement on All SPL screen
- File – Export – Export filters impulse response as WAV – fill in the filters impulse response export panel: “No filters to export. The selected filter set has no filters that have an effect, nothing to export»

Attempt 3
- Open measure (full L)
- Smoothed 1/6
- Open EQ window
- Calculate target level from response
- Match response to target
- Predicted curve and filter numbers appear
- Open EQ filters panel with the generated filter data
- Save filters by pressing the diskette icon in the EQ filters panel – a REW filter set file is generated and saved (filtres generats3...)
- Close the EQ filters panel
- Filter task – Generate measurement from filters

- All SPL window appears new curve (Filters full L3) more or less symmetrical to the original measurement, but more smoothed and at a lower SPL level (+- 0 SPL)

- All SPL window save this new curve - file REW measurement Filter full L3 is created- Close the All SPL window original measurement (Full L) and leave the new curve activated

- File – Export – Export filters impulse response as WAV – fill in the filters impulse response export panel: "No filters to export. The selected filter set has no filters that have an effect, nothing to export»

The result is still the same: I am not able to properly save the filter settings and then export them as a .WAV file



Vers. REW 5.20.13

Tested in Windows 7 64-bit i Windows 10 64-bit

Java ver. 8

I need help to take this step, explained in a very basic way to be able to overcome it.

Thank you so much.

 

[John Mulcahy]

The filters are part of your original "full L" measurement. Select that measurement and then use File -> Export filters impulse response as WAV.

 

[anzoni]

Thank you very much Joan for the quick reply. We will try as you say.

 

[anzoni]

Hello,

New measurement session to compare original and filtered curves. Now I was able to export the WAV filters, thanks!
But maybe there is something I'm still not doing right because the measurements, after filtering, seem very similar to the original measurement, and not very similar to the "predicted" curve. Is this possible?

thank you so much


Full L vs. predicted L

Full R vs. predicted R

Full L vs. Full L filtered

Full R vs. Full R filtered

Target settings from L & R

 

[John Mulcahy]

Your L and R measurements are RMS averages of 4 individual measurements (presumably 4 different positions). Your filtered measurements are averages of 2 measurements, it isn't clear what went into those averages.

If you measure a single position, apply some filters, then measure again from the same position the predicted curve should be a very close match for the measurement with the filters.

If you use an average of a set of measurements to produce some filters the prediction is likely to match an average of filtered results measured at the same positions. If you do not use the same measurement process for the unfiltered and filtered results you will not see results that match the prediction.

Any individual position will not match the prediction for the average, since no individual position will be the same as the average. You could, however, define the same filters for the individual measurements and the individual predicted results should be a good match for filtered measurements made at those individual positions.

 

[anzoni]

Hello,

I took the 4 measurements of L and R in exactly the same place, without changing the location of the microphone, with the intention of compensating for the background noise of the city, the four of each channel are practically calculated, as well as the resulting average.


What I don't understand is how the single measurement I made once the filters were installed results in an average of two measurements, as you showed me.

At the end of the week I will try to generate the filters on a single reading, as you recommend.

Thank you very much for your help and patience

 

[anzoni]

Many thanks for the help; finally of two clean measures two clean filtered measures!