Spacial Average Measurements

[AudiocRaver]

AudiocRaver maybe there is now a method for identifying early reflections.

You have delved far more deeply into this topic than I have, kudos and thanks for all the great info and links, as well as for John'a and Matt's and others' guidance. So many great things to learn, so little time. What really excites me about topics like this is that they indicate future paths toward highly useful products and new possibilities. Keep it up!

 

[Matthew J Poes]

You have delved far more deeply into this topic than I have, kudos and thanks for all the great info and links, as well as for John'a and Matt's and others' guidance. So many great things to learn, so little time. What really excites me about topics like this is that they indicate future paths toward highly useful products and new possibilities. Keep it up!

I don't believe its possible to zero in on a reflection using any kind of averaged measurement, even a vector averaged measurement. The approach I've always used is to check the distance to the first reflection in the impulse after the noise occurring right around the direct signal. This is normally floor bounce, ceiling bounce, or wall bounce. Precise measurements can help you figure out which is the most likely. If you move the mic around, this slightly changes these distance, which slightly changes the position of the reflection in the impulse.

I know some experts use mic shading to do this too, I'm not totally sure how that is done. I've used a trick with a figure 8 patter mic, an Omni mic, a shadow, and two channel collection, but I don't think it would work in a small room like our theaters or living rooms. REW is also not 2 channel and most don't have a figure 8 mic.

In any case, I find that examining the impulse response is good enough to look at the reflections and figure out where they are. Once you pin point them you can see the effect of treatments by the reduction in the magnitude of the pulse at those timing points.

 

[60secondsassassin]

Thanks John, Thanks Matthew,
Following your advices and with the help of this page and this one, I was able to target my room mode peaks.
Bass are more cleaner not faded like with LS filter and not to boomy.
Really good, thanks

 

[Marzolino]

Dear all, it's my first post .
I'm enthusiast of REW and got with it a good equalisation in my listening room , but just for low frequencies up to about 300 Hertz (minimum phase zone, according with the theory explained in the REW manual).
Now I'm thinking about the mid and high frequencies equalisation, tryng the Moving Mic Measurement method (MMM) by Jan-Luc Ohl https://www.ohl.to/audio/downloads/MMM-moving-mic-measurement.pdf, already quoted in this forum.
What's the method for import in REW these MMM data and run the REW EQ?
I'm thinking the following steps:
1) load in a USB key a 30 sec pinknoise file;
2) play this file in the stereo system and make a Moving Microphone Measurement;
3) load in Audacity - I'm experienced with Audacity - the pinknoise file, make the FFT Spectrum and export it with .txt format;
4) do the same above with the measuremente MMM file;
5) import into REW the two .txt files and, in "All SPL" page, calculate the difference graph between them;
6) apply the REW EQ to this difference file.

Could it work? or have you some advice in order to equalize the mid-high frequencies with the MMM method?

Thank you in advance

 

[AudiocRaver]

Dear all, it's my first post .
I'm enthusiast of REW and got with it a good equalisation in my listening room , but just for low frequencies up to about 300 Hertz (minimum phase zone, according with the theory explained in the REW manual).
Now I'm thinking about the mid and high frequencies equalisation, tryng the Moving Mic Measurement method (MMM) by Jan-Luc Ohl https://www.ohl.to/audio/downloads/MMM-moving-mic-measurement.pdf, already quoted in this forum.
What's the method for import in REW these MMM data and run the REW EQ?
I'm thinking the following steps:
1) load in a USB key a 30 sec pinknoise file;
2) play this file in the stereo system and make a Moving Microphone Measurement;
3) load in Audacity - I'm experienced with Audacity - the pinknoise file, make the FFT Spectrum and export it with .txt format;
4) do the same above with the measuremente MMM file;
5) import into REW the two .txt files and, in "All SPL" page, calculate the difference graph between them;
6) apply the REW EQ to this difference file.

Could it work? or have you some advice in order to equalize the mid-high frequencies with the MMM method?

Thank you in advance

Interesting approach.

For frequency response only, the MMM tells yu a lot, and I like it as a way to get a snapshot of a speaker's performance.

Setps 1, 2 are OK, "meke the FFT spectrum" ? not sure about the Audacity's capability that might help you do that, or what the output is like. REW can play pink noise and allow a MMM measurement with the RTA function, save that to a measurement panel and do tthe rest, or can record from an external pink noise file.

At highest frequencies, the mic's orientation makes a difference of a couple dB.

 

[Marzolino]

Interesting approach.

For frequency response only, the MMM tells yu a lot, and I like it as a way to get a snapshot of a speaker's performance.

Setps 1, 2 are OK, "meke the FFT spectrum" ? not sure about the Audacity's capability that might help you do that, or what the output is like. REW can play pink noise and allow a MMM measurement with the RTA function, save that to a measurement panel and do tthe rest, or can record from an external pink noise file.

At highest frequencies, the mic's orientation makes a difference of a couple dB.

 

[Marzolino]

Hi, thank you for the reply and for your opinion about the MMM method!
About Audacity I've many times extract from a .wav file the equivalent .txt file. Simply go in Audacity to Menu File => Open File ; when the waveform is displayed go to Menu = > Analysis = > Show Spectrum , here select the block size (1024, 2048, 4096, the higher the blocksize, higher the resolution an more rows in the .txt file). In the files annexed (blocksize = 2048) you can see the same sequence in the Hz column, both in the Pinknoise data and in the Mic Measurement data, so they can be easy compared.
The difference between signal data and measurement data gives the advantage to make non considerable the non linearity in the signal file; it can also be done with Paste & Copy the data in a datasheet like Excel, instead of the 5) step.
You can see that the Hz sequence is'nt Log but linear, so the low freq data are few and the hi freq are crowed, but for hi-freq analisys that's good.
My doubt is the behaviour of REW in equalisation over a imported file .txt format .
REW works with a signal generated by itself and syncronous with a windowing system, but what's the analysis method for EQ calculation on an imported .txt file? Will it give reliable risults ?
What do you think about?
Anyway I think I'll make a test, asap when my relatives,and room noise, are out !
Bye