Using REW to do Room Correction

What do you want to play? Frequency sweeps? Do you have Tidal?
Lots of questions but if so there’s an audio setup cd that has sweeps and specific tones.

I’m sure there are others.

1 Like

Yes but the advanced measurement tools need to be synchronized with the sweep, it seems.

I do have measurement CDs in my library.

Not sure then, sorry.

@AndersVinberg I too use XTZ to measure and appreciate that it spits out serious room modes that can be affected by PEQ. In my room, it is two dips. I am wondering how you had XTZ compute “8 filters”? Can you provide more info.? Did you add them to Stimulus EQ manually?

Not sure. I think I did one run, it found two filters, copied those to StimulusEQ, ran it again which modifies the signal by those filters, and then it found more.
But I also experimented with different mike positions…

I may be misunderstanding the intent. If you just want to measure and adjust your PEQ settings, and if you have access to a Windows, Equalizer APO is awsome and free. If you can connect your DAC to the windows machine, you can configure APO to work with your DACs DS driver. Then point REW to your DAC as output device. You can set your filters in APO (convolution or PEQ), remeasure/adjust/repeat. You can also play sweeps through REW or play tones that track your cursor, to pinpoint issues. Another option is to play pink noise and do spatial averaging in RTA mode.

The reason I am suggesting is, for a EQing perspective, it does not matter that much that you use Roon or the exact same setup till it hits the amp, as long as your amp and speakers are the same. There is enough uncertainty introduced by the measurement process, subtle changes are irrelevant.

Yes, that’s what I do with XTZ.
But that assumes Roon and XTZ filters work the same.
To begin with, Roon has capabilities that XTZ lacks.

Thanks - that is how i did it too.

I have not experimented with the difference between one measurement location and three… Will need to do that.

I am having trouble easily toggling between DPS presets in ROON… Seems like that would be easy but the new present is not being applied straight away. I will keep trying but any advise? Am i missing a trick?

If i understand right, once you have set the Roon PEQ, you want to analyze the output signal.

In XTZ, to get Room Nodes discovered, XTZ uses its own signal sweep and measures the response producing location of room nodes and proper frequency and Q to counteract these nodes. Presumable it does this when all Roon eq is off.

If you apply XTZ’s recommend PEQ setting to counteract the nodes, and run XTZ’s signal sweep, shouldn’t you be be testing the signal with the PEQ applied and if it is, then wouldn’t that be reflected in the new measurement produced?

Yes, but it uses XTZ’s implementation of those filters.
I can’t test Roon’s implementation.
And more importantly, XTZ doesn’t tell me if it sounds good since it can’t play music.

I did that: created filters in XTZ, tested the measurement in XTZ, manually transferred them to Roon, and it sounded weak and flat. So what do I do now?

Did I mistype the filters? Is Roon’s implementation different? I can’t measure Roon.

Does it sound equally bad in XTZ? Can’t listen in XTZ.

I adjust the filters in Roon, using trial and error, until it sounds better. How flat is that? I can’t measure that. I can transfer the modified filters to XTZ and measure them, are those measurements valid?

And for every trial and error cycle, do I do all of that?

Not a reasonable activity for an adult.

Hi - i think there are two points here.

  1. Not being able to test Roon’s implemention. Here i believe that you can. You apply the suggested XTZ filter to ROONs PEQ and apply. When you do XTZ’s room measurement (no filters in XTZ applied) then the signal produced by XTZ and played out of the ROON DAC should reflect these filters.

Now i have a thought - in my system i send the XTZ signal via a preamp with a digital out to a DAC Roon ready PreAmp (PS Audio). That sends the signal to my amp. Perhaps your system does not work that way?

  1. XTZ quality. Am i at an early stage in testing the XTZ filters. It found two serious room modes. I corrected for them, but had to almost immediately leave for a business trip so have not a/b’ed the results sufficiently to have an opinion on the differences of the applied PEQ in Roon.

If you run white noise (or pink - i have to check - i can never remember) through Roon after you apply your PEZ filters, your XTZ RTA would give you an idea on how flat the signal produced is. Perhaps best to confirm this with XTZ (and choice of white or pink noise).

  1. When I run XTZ signal like that it will go to the DAC, but not through Roon and hence not through Roon DSP. Roon doesn’t have an input.

  2. Yes, I have played pink noise through Roon with Ron’s DSP (from a test CD I have in the library, as above I can’t feed XTZ pink noise through Roon) it does give me a feeling for how flat it is. That’s reasonably comforting. But (a) that’s low resolution measurement, third octave, so serious anomalies could hide within one band. And (b) XTZ doesn’t have an averaging mode, its slowest mode is 1 s, so the bars bounce around. I ran if four times and saved the curves, plus paused it, so now I have five measurement curves - they are often 5 dB apart.

I can muddle through, but it is unsatisfactory.

Anybody with better methods?

Here’s a way to test Roon filters using REW. It doesn’t need any special hardware.

REW works by playing a sweep (from low to high frequency). It has an option to use a timing signal (use Acoustic Timing Reference). Checking this option means that REW waits until it hears the timing signal before starting to measure.

So, here’s what I did:

  1. Record the test sweep that REW produces (I did this on a Mac using Soundflower and Audacity then saved the output as a FLAC file).
  2. Add the FLAC file to a folder which Roon watches. Roon can now play the file.
  3. Start REW and tell it to measure and to wait for timing reference.
  4. Play the FLAC file in Roon - REW starts measuring as soon as the timing reference is detected.

Image (below) shows 3 things measured in this way. Top line (red) is speaker response with no DSP applied in Roon. Second line (green) is with convolution filter (generated from REW) applied by Roon. Third line (magenta) is with Roon applying a 2KHz, 36dB high pass filter (just to check that things are working as they should).

The only thing you can’t do is to test a low pass filter - this filters out the timing signal that REW generates - but that’s not much of a disadvantage for me at least. What it does mean is that I can now see the actual impact of filters being applied in Roon.

I could write this up as a short tutorial (with screenshots) later this week if anyone thinks that would be useful.

12 Likes

That’s great. Wish XTZ had that.
REW is inexpensive but it doesn’t seem to support the mic that came with XTZ.
Hmm…

I used UMIK-1:

https://www.amazon.co.uk/miniDSP-UMIK-1-Measurement-Calibrated-Microphone/dp/B00N4Q25R8

1 Like

Also, according to this link, you can use the XTZ microphone with REW as long as you have the calibration file (which can, I guess, be downloaded from XTZ site).

http://www.hometheatershack.com/forums/rew-forum/33261-xtz-microphone-use-rew.html

1 Like

yes please! A tutorial would be great, I am interested in knowing how you created the sweep with the timing reference.

thanks!
Marco

1 Like

That’s terrific.
Thanks.

A question maybe someone can answer here: in manual PEQ your are only able to address frequencies between 20Hz - 22000Hz. I already asked the support team to change this because I also want to be able to EQ in the 12 - 20Hz zone. If I use Convolution will this also be limited to 20Hz and above or can I address lower frequencies?

A little background. Although theoretical beyond human hearing this 12 - 20Hz area does influence spatial clues, depth of the sound stage etc. And it is also sensitive to room nodes. If you have a node around let say 56Hz it also will be a problem at 14, 28, 112 and other multiples in the bass zone.

I’d imagine @brian would have the answer to that one.