Room Measurement Tutorial for Dummies Part 1

[digitalfrost]

I could only measure sub and sat at full range then. Possible, but then I'd had to somehow mix the two measurements together? I've found a cheap mic pre that allows me output the signal via ADAT. I can then go from there into the Digiface and use the mic pre as master clock. That should solve this problem.

 

[Krunok]

Possible, but then I'd had to somehow mix the two measurements together?

If you want to display several measurements on the same graph in REW you should use "Overlays".

 

[Daverz]

These 2 IRs are practically perfectly aligned, you don't need them to be perfeclty one over the other.

Yeah, I was grasping for some reason that my DRC-FIR results were so mediocre when I remember getting pretty good results previously.

Turns out there was a bug in the Bash script I wrote to do all the boilerplate for sox, drc and bruefir, and my results were only correct for 48 kHz, the rate I was using to test the filters.

The corrected sound is fantastic now, with very focused imaging and a strong center image.

 

[Krunok]

Yeah, I was grasping for some reason that my DRC-FIR results were so mediocre when I remember getting pretty good results previously.

Turns out there was a bug in the Bash script I wrote to do all the boilerplate for sox, drc and bruefir, and my results were only correct for 48 kHz, the rate I was using to test the filters.

The corrected sound is fantastic now, with very focused imaging and a strong center image.

I'm glad to hear that!

Btw, I think it would be nice if you can take time and show how you did it in a separate thread. It was your work that inspired Pascal (developer of BRuteFIR plugin for Volumio) and me to add support for DRC-FIR to the plugin.

 

[Daverz]

I'm glad to hear that!

Btw, I think it would be nice if you can take time and show how you did it in a separate thread. It was your work that inspired Pascal (developer of BRuteFIR plugin for Volumio) and me to add support for DRC-FIR to the plugin.

I think that must have been somebody else. I didn't join the forum until March.

I'll update my notes for the process I follow.

 

[Krunok]

I think that must have been somebody else. I didn't join the forum until March.

I'll update my notes for the process I follow.

Do you own Vandersteens Quatro?

 

[Daverz]

Do you own Vandersteens Quatro?

Yeah, but I thought the Volumnio plugin predated my post in March.

 

[Krunok]

Yeah, but I thought the Volumnio plugin predated my post in March.

Then it was you. It does, but DRC-FIR support was added inspired by your posts.

 

[q3cpma]

Is there a fully free software way of doing all of this on POSIX? I know BruteFIR exists, but getting the data is the hard part: REW is proprietary (and made in Java, which is a heavy dependency). Then there's the mic driver compatibility...

If anyone has experience doing room correction on POSIX (Linux, *BSD), please tell me.

 

[Krunok]

Is there a fully free software way of doing all of this on POSIX? I know BruteFIR exists, but getting the data is the hard part: REW is proprietary (and made in Java, which is a heavy dependency). Then there's the mic driver compatibility...

If anyone has experience doing room correction on POSIX (Linux, *BSD), please tell me.

You can use Windows to make measurements in REW and create your filters, after that you copy filters to BruteFIR and forget about Windows and REW.

 

[Daverz]

Is there a fully free software way of doing all of this on POSIX? I know BruteFIR exists, but getting the data is the hard part: REW is proprietary (and made in Java, which is a heavy dependency). Then there's the mic driver compatibility...

If anyone has experience doing room correction on POSIX (Linux, *BSD), please tell me.

I think most USB 2.0 audio devices should just work in Linux. My Umik-1 works fine.

REW has a Linux installer. You'll need the openjdk package installed.

DRC-FIR comes with some utility programs for doing sweeps and generating an impulse response. I still prefer to do this with REW.

(I do my REW measurements on a Macbook, compute the filters on a Linux Mint desktop, and copy them over to my Ubuntu music server, and the transport is an RPi running Linux!)

 

[digitalfrost]

I ordered a Behringer ADA 8200. Bit of overkill with it's 8 channels, but it was cheapest microphone -> ADAT interface I could find. Sadly I couldn't get HOLMimpulse to work, which used to work before, but I could make measurements with REW.

I set the ADA to 48khz master clock mode and configured my RME card to use this as clock source. You can easily create a single WDM soundcard for each 8 channels in the MADIface settings, you just have to use the first two to in analog world to make the applications happy (L/R). With ASIO, you can do whatever you want anyway.

Long story short: Wow. I had so many phase issues by using two USB soundcards. Even when the bass was okay, in the upper registers stuff was all over the place. This is not a subtle difference. Totally worth it to me. In hindsight, I should've gotten the Fireface UCX instead of the Digiface. But I wanted to run external DACs, because of the reviews here. Speaking of which - @amrim I would be very happy if you could review the RME Fireface UCX.

You can basically do everything with this thing. It's capable of multichannel playback and you can do measurements. Maybe a bit overkill for people only interested in stereo, but I would like to know how it compares to the latest ADI DACs, which we know are excellent.

 

[q3cpma]

I think most USB 2.0 audio devices should just work in Linux. My Umik-1 works fine.

REW has a Linux installer. You'll need the openjdk package installed.

DRC-FIR comes with some utility programs for doing sweeps and generating an impulse response. I still prefer to do this with REW.

(I do my REW measurements on a Macbook, compute the filters on a Linux Mint desktop, and copy them over to my Ubuntu music server, and the transport is an RPi running Linux!)

I see, so DRC-FIR it is, it seems.

 

[Daverz]

I see, so DRC-FIR it is, it seems.

You might also try DRC Designer, but this is also Java.

http://www.alanjordan.org/DRCDesigner/DrcDesignerHelp.html

If the issue is not being able to access the audio device in REW, the sweep could be recorded with, for example, Audacity and then imported into REW.

The way the author of DRC-FIR recommends recording the sweeps is to use a duplex audio device (that is, the device can record and play back digital files at the same time using the same clock) for recording and feed the sweep output back into the second input channel as a reference signal. You'd probably want something like a Focusrite with phantom power for the microphone. I'd recommend Audacity for making the actual recording.

I prefer using a USB microphone and using a timing reference chirp in REW to time align the measurement. REW can also estimate the impulse response offset, which is more robust than just looking for the absolute maximum.

 

[q3cpma]

Thanks for the details. Since I'm on Gentoo without any Java crap already installed, I'm still gonna choose DRC-FIR. I'll probably use arecord or ffmpeg to record the audio with my Swissonic UA-2x2 and a Behringer ECM8000 I'll buy for that occasion.
Just need to find a brutefir Alsa guide, now.

 

[Roasty]

I'll be honest; I had not heard of rew until I came into this forum. I've been going through the rew website and reading the rew thread here as well.

Can someone quickly confirm my understanding (or lack thereof) that to properly apply rew, there must be a way to input a correction EQ somewhere in the music chain? ie if my music chain does not have a PC or some other hardware capable of applying eq, I cannot utilise rew properly?

 

[NTK]

Can someone quickly confirm my understanding (or lack thereof) that to properly apply rew, there must be a way to input a correction EQ somewhere in the music chain? ie if my music chain does not have a PC or some other hardware capable of applying eq, I cannot utilise rew properly?

That would be mostly correct. The only other way I can think of at the moment is to apply the EQ offline. For example, you can use an audio editing program such as Audacity to apply equalization curve to the source file, save the altered version, and playback the altered version. However, I wouldn't think this is viable in the long term.

 

[g29]

I'll be honest; I had not heard of rew until I came into this forum. I've been going through the rew website and reading the rew thread here as well.

Can someone quickly confirm my understanding (or lack thereof) that to properly apply rew, there must be a way to input a correction EQ somewhere in the music chain? ie if my music chain does not have a PC or some other hardware capable of applying eq, I cannot utilise rew properly?

@Roasty , You can measure your system and see what issues you may have. Some of those issues maybe addressed by means other than correction filters (speaker placement, seating placement, room treatment [absorption, diffusion], dampening resonating materials, sub integration, etc.) and verify the remedies by taking after measurements. These changes should be done before DSP because they can minimize amount of DSP required. So, you can benefit from using REW without having DSP functionality, but that would be the next step.

 

[Roasty]

Thanks very much for chiming in, guys.
Looks like I can still explore REW, albeit in a limited way. Interesting aspect of audio which is new to me. Cheers!

 

[Julf]

Can someone quickly confirm my understanding (or lack thereof) that to properly apply rew, there must be a way to input a correction EQ somewhere in the music chain? ie if my music chain does not have a PC or some other hardware capable of applying eq, I cannot utilise rew properly?

Even without EQ, REW is still very useful in figuring out what is going on with your speakers and room.