Windowing to minimize room effects

[Veleric]

Guys,
I'm new here so bear with me! I've made some DML speaker prototypes and started making measurements last night. It's my first time using REW. Well, maybe second but I didn't get very far the first time. Presently, I'm not interested in the room effects, since I have no idea yet which room (if any) these speakers might eventually end up in. I understand that windowing can help minimize the room effects, but I don't what settings to use. Are there recommended window settings to use for this purpose? Are there other suggestions besides windowing (say microphone location, etc) that minimize room effects in the measurement? I'm sure this has been asked a million times, so I apologize for that.
Thanks,
Eric

 

[John Mulcahy]

The general aim would be to place the window so it cuts off the response just before the first reflection. How soon that occurs depends on where the speaker is positioned relative the the floor, ceiling and walls, the further away it can be placed from reflecting surfaces the wider the window can be. The narrower the window, the worse the frequency resolution and the higher the frequency below which the response is not valid. For a 3 millisecond window, for example, the lowest frequency and the frequency resolution would be 333 Hz.

 

[Veleric]

So, if I'm in say a 20x20 room with 9 ft ceilings, I could set the window for 4 ms and remove all the reflections, but have minimum frequency (and resloution) of 250 Hz, or set the window for say 10 ms and get minimum freq (and resolution) at 100 Hz, but realize that I have not eliminated the first reflections from the floor and ceiling. Is that basically correct?

 

[John Mulcahy]

Yes. If the environment suits then outdoor testing with the speaker on a platform or with the speaker lying on the ground and the mic suspended above it can ease window constraints.

 

[Veleric]

John,
Defaults are 125 ms for left window and 500 for the right window. Is the right window the one you have been referring to, and hence the one I would adjust to minimize the room effects? Or does the left window also have some influence?
Thanks
Eric

 

[Veleric]

Another newbie question: I'm using a Dayton Audio imm-6 mic. It's made for an iphone but I have a trrs to trs adapter and that at least seems to be working. The mic has a calibration file, but I can't figure out how to apply it. When I go to the Cal Files tab, there's a place to add a soundcard cal file, but not a mic cal file. The section under mic cal files says to select an input device and input but there is nothing shown to select. What am I missing?
Thanks,
Eric

 

[John Mulcahy]

Yes, the right window is the one that matters when the window reference time is at or near the IR start, the default is for the reference time to be at the peak which is very close to the start for full range measurements.

Input device and input selections are made on the Soundcard preferences. When you have made a selection for them there will be an entry for that input combination in the mic cal files list, you can also specify the cal file in the dialog that is shown when the Cal Files button on the Measure dialog is pressed.

 

[akebrake]

Thanks John, when on the subject...

Say I’m analyzing an already captured measurement.
Shortening the right hand window, to gate out late reflections and increase time resolution. The corresponding low cut off goes up. 20ms = 50 Hz,
but the freq resolution shown is 8 Hz? Can y expand on that?

 

[John Mulcahy]

The frequency resolution figure is based on the total window width, left+right. The validity of that figure depends on the window positioning relative to the impulse start, in effect how much of the response actually lies within the window.

 

[akebrake]

Thanks!
I have to think a bit about it

 

[Veleric]

Input device and input selections are made on the Soundcard preferences. When you have made a selection for them there will be an entry for that input combination in the mic cal files list, you can also specify the cal file in the dialog that is shown when the Cal Files button on the Measure dialog is pressed.

Okay, that makes sense but I still don't know exactly what I'm doing here. For Output device options I have 3 choices:
1 Default Device
2 Primary Sound Driver
3 Speakers (Realtek High Definition Audio)
I've been using (1) Default Device and that seems to work, but I'm not clear on what the three choices really mean/represent. I think that what I want is simply so that I make the choice that sends the output through the HDMI cable connected between my PC my AVR that feeds the speakers. And that is what happens when I choose (1) but I'm still not certain if that's the true correct choice.

For input Device I have 4 options:
(1) Default Device
(2) Primary Sound Capture Driver
(3) Microphone (Realtek High Definition Audio)
(4) Stereo Mix (Realtek High Definition Audio)

Here, if I choose any of the first three, the system seems to be "hearing" the imm-6 mic, but only choice (3) provides an "Input" choice that requests a mic calibration file. So in this case I'm thinking (3) is the correct choice. Up until now I have been using (1) Default.. which did not allow an "Input" to be chosen and apparently hence did not allow the addition of a calibration file.

What are the correct/ideal choices for Output and input devices?

 

[Veleric]

More questions about setting up the mic:
On my PC, when I attach the imm-6 mic, a "Realtek HD Audio Manager" box pops up that gives me some options. Do any of these have any impact on the REW measurements?
(1) I can set a mic level. This I've been turning up all the way to 100%. Is that okay or should it be set lower?
(2) I can select "Noise Suppression" and/or "Acoustic Echo Cancellation". I've been leaving these off. Right choice?
(3) I can choose a "Default Format" where the options are 44.1, 48, 96 or 192 KHz. (all 16 bit). Which is correct choice? Does it matter?

 

[Veleric]

And...
When I load the mic calibration file, REW shows only about 20 lines of data, (covering about 20 to 40 Hz) but the actual file has about 100 or so lines of data (up to 20 khz). Is REW just showing a "preview" of the file (and it's really seeing all the lines of calibration data)? Or is my calibration file too long or some other issue preventing the whole thing from being seen by REW?
Thanks,
Eric

 

[John Mulcahy]

For input Device I have 4 options:
(1) Default Device
(2) Primary Sound Capture Driver
(3) Microphone (Realtek High Definition Audio)
(4) Stereo Mix (Realtek High Definition Audio)

Here, if I choose any of the first three, the system seems to be "hearing" the imm-6 mic, but only choice (3) provides an "Input" choice that requests a mic calibration file. So in this case I'm thinking (3) is the correct choice.

It is.

On my PC, when I attach the imm-6 mic, a "Realtek HD Audio Manager" box pops up that gives me some options. Do any of these have any impact on the REW measurements?
(1) I can set a mic level. This I've been turning up all the way to 100%. Is that okay or should it be set lower?

At 100% it is probably applying some gain to the input, but as long as it is not clipping it should be OK.

(2) I can select "Noise Suppression" and/or "Acoustic Echo Cancellation". I've been leaving these off. Right choice?

Yes.

(3) I can choose a "Default Format" where the options are 44.1, 48, 96 or 192 KHz. (all 16 bit). Which is correct choice? Does it matter?

48k is fine.

When I load the mic calibration file, REW shows only about 20 lines of data, (covering about 20 to 40 Hz) but the actual file has about 100 or so lines of data (up to 20 khz). Is REW just showing a "preview" of the file (and it's really seeing all the lines of calibration data)? Or is my calibration file too long or some other issue preventing the whole thing from being seen by REW?

It is just a preview. You can view the calibration file response by selecting that trace on the SPL & Phase graph.

 

[Veleric]

John,
Thanks for your replies. Regarding my first question in post 11 regarding my choices for output selection, any thoughts?
Eric

 

[John Mulcahy]

Would usually want to select the physical device being used, but your HDMI output doesn't seem to appear in the list. Odd, usually anything that appears as a playback device in the Windows audio settings would appear in the output devices list. That probably leaves you needing to stick with Default device for output.

 

[Veleric]

So my next question is about my mic and calibrating the sound card.
As I mentioned, I'm using a Dayton Audio iMM-6 Calibrated Measurement Microphone, plugged in to the mic/headphone jack of my PC. Since the mic is designed for use with an iphone, it has a 3.5 mm TRRS connector. I use a TRRS to TRS adapter between the mic and the PC's 3.5 mm TRS mic/headphone jack.
So this isn't exactly (as far as I can tell) one of the three options described in the REW Help document.
With this mic, do I need to calibrate the sound card? Any other issues I should be concerned about?
Thanks,
Eric

 

[John Mulcahy]

With this mic, do I need to calibrate the sound card? Any other issues I should be concerned about?

You aren't using the soundcard for output so calibrating it doesn't really make sense. No issues I'm aware of with using an iMM-6.

 

[Veleric]

You aren't using the soundcard for output so calibrating it doesn't really make sense.

John,
Thanks again, especially for suffering my dumb questions, which continue::

If my output choice was (3) Speakers (Realtek High Def Audio), that would be using the Soundcard, correct? And hence would make sense to calibrate, right?

But if my output choice is (1) Default Device, is it certain that the Default Device is not also the Soundcard?

Eric

 

[John Mulcahy]

You said you were using an HDMI connection for your output, so no analog signal from the soundcard to calibrate with there.