How to disable smoothing with FDW?

[shaunp]

Hello,

I've just started measuring with FDW enabled to filter out reflections from my measurements and hoped I could ask for some help.

-I don't know what FDW values to use, so currently I'm just plugging in values from any guides I can find online. Is there a way to check or dial it in and get confidence? Offhand, I think I should check the impulse response and confirm reflection spikes are eliminated if the settings are good. Other than that, is there anything else I should do or other criteria I should check?

-I saw my measurements have 1/48 smoothing applied automatically after enabling FDW. I'd like to export the measurement without smoothing (to Rephase), but selecting 'Remove smoothing' still leaves 1/48 smoothing enabled. Is there a way to do this? My 'width in cycles' is set to 15. Other info is that 'Allow 96 PPO log spacing' is unchecked, but my measurement info shows 'Freq spacing is 96 PPO.

Thanks,
Shaun

BTW - I'm new and really enjoying REW

 

[John Mulcahy]

The FDW involves conversion to log spacing, hence the 96 points per octave spacing. That inherently requires some smoothing before the conversion, all log spaced measurements are first smoothed to half the log sample resolution. Effectively 1/48 octave is as unsmoothed as an FDW result can get.

 

[shaunp]

Thanks John. I'll try out some other windowing options and see if I can get more resolution. It's really helpful info.

Here's the background for my post and reason I'm looking for more measurement resolution:
--------
I'm measuring in REW, applying phase EQ corrections from Rephase and remeasuring in REW afterwards to check the results. I'm seeing some phase corrections having no effect, minimal effect, or even opposite effects after I apply them. So I hope more measurement resolution can help me understand the reason. I really don't know why yet. If you or anyone has any ideas, I'd love to hear it.

My problem and setup+procedure details are below.

Input path:Beyerdynamic MM-1 mic>Focusrite Clarett+ rec IF>REW
Output path:REW>CamillaDSP(to apply EQ adjustments)>Focusrite Clarett+ rec IF>Kali Audio IN8v2 spkrs
EQ adjustments path:REW>Export Measurement to Rephase>Adjust EQs>Generate wav>Input wav to CamillaDSP for EQ adjustments

Basically, my procedure is below and issue is at step 3a)

First pass:
1a) REW(FDW on+mic cal file+Rec IF cal file)>CamillaDSP(No EQ)>Rec IF>1 Speaker>Mic>Rec IF>REW(Export Measurement)

1b)Rephase(Import 1a Measurement)>Add Phase EQ(aiming for flat resp)>Generate wav(64 bits IEEE mono)

Second pass:
2a) REW(FDW on+mic cal file+Rec IF cal file)>CamillaDSP(wav from 1b)>Rec IF>1 Speaker>Mic>Rec IF>REW(Export Measurement)
NOTE: Measurement has many 1b phase corrections, but not flat yet, so will try Rephase again to adjust.

2b)Rephase(Import 2a Measurement+disable all previous phase EQs to avoid double EQ)>Add new Phase EQs(aiming for flat resp)>Reenable all phase EQs for wav>Generate wav(64 bits IEEE mono)

Third pass:
3a) REW(FDW on+mic cal file+Rec IF cal file)>CamillaDSP(wav from 1b)>Rec IF>1 Speaker>Mic>Rec IF>REW(Export Measurement)
NOTE: Measurement is similar to 2a (much flatter than original 1a measurement), but some parts better and other parts worse. 2b changes are not really taking effect, something is amiss!

Thanks,
Shaun