Room Measurement Tutorial for Dummies Part 1

[Sal1950]

Ah, sometimes I love computers! Found the original images and I think I placed them all correctly in the OP.

Good show Captain !
Glad you didn't have to depend on my memory.

 

[Descartes]

Will we ever see an AVR using AI to properly calibrate the system! I am technologically incompetent when it comes to all these software. I just like to turn it on and get the best sound I can without having to spend hours tinkering with adjusting everything. I once paid a supposedly professional engineer to come and calibrate my system and it was a great disappointment!

 

[Sal1950]

Will we ever see an AVR using AI to properly calibrate the system! I am technologically incompetent when it comes to all these software. I just like to turn it on and get the best sound I can without having to spend hours tinkering with adjusting everything. I once paid a supposedly professional engineer to come and calibrate my system and it was a great disappointment!

I'm not tech incompetent but am well past the age where I find this kind of thing enjoyable.
I'm all for a fully auto DRC system that I don't have to fool with any more.

 

[Miniyouuuu]

AI will help, as Dirac helps with very easy indications. But the human factor is still there. Dirac proposes a target curve, are you sure that is your desirable target?
You need to measure the room in order to let AI understand the physical conditions of the room. Are you sure the micro is set in the correct place? Do you define multiple spots of measurement accurately? Maybe future 3D scanners could help, but AI will only take conclusions based on human-provided data.

 

[Keith_W]

Reading this thread makes me believe there is no hope for humanity. This is ASR, where we believe that measurements and a science-based approach can really deliver results. Yet here we are, hoping that AI comes to the rescue and can interpret measurements for us. Feh to all that, get off your bloody arses and go and LEARN! Learning is pleasurable, learning keeps your mind young, and knowledge about how to use REW makes you irresistible to women!

Can you spot the REW expert in this photo?

 

[Sokel]

and knowledge about how to use REW makes you irresistible to women!

It depends, mine looks down on me when she passes by, she says that red trace should be 255, 0, 0 or non at all
Humanity suffers from all shorts of pains

 

[mmares056]

I don't need no stinking AI to help me interpret my measurements. . I like to learn how to use all room acoustic measuring tools and the software and what the software does. For me personally, learning how to interpret measurements referencing the software documents and getting help from Forums like this one is very satisfying and I take pleasure in doing it.

 

[dimedrol]

I have finally bought a measurement mic and calibrated my Genelecs! Best thing ever is when you install a correction curve in APO and turn it on and off while playing music. I found dense, fast and not too bass-heavy metal to be the most revealing, something like Inflames - Scream. When I do a side by side comparison like that, without the correction the speakers now sound like they are playing in a tin bucket! My great and highly praised Genelecs, how dare you! And I've listened for them like it for a long time! Not all songs get a big quality boost, but some are literally fixed now. Very happy! Also, as far as finding optimal sub settings: phase, crossover frequency, volume - using the mic I have finally found those too! To my great surprise I was actually off by much, despite once spending hours marrying the sub with the monitors!
An update: I have by now measured both of my audio systems, the 2.1 monitors and the 2.0 hi-fi speakers. I gotta say, both of them had significant dips on the bass which in the 2.1 system was almost fully resolved by the subwoofer. I had to extend it all the way to 150Hz for the best linearity. The 2.0 system however is helpless with a huge 30db dip at 70Hz. Just got me thinking that a sub is rather a necessity in stereo unless you have a whole lot of room treatment/speaker/listening position variability, which I don't.

 

[Zinda]

There are few more things transformative in audio than learning how room acoustics works. You never look at any system the same again. Like moving from film photography to digital, measurements give you instant feedback on what your system is doing. Even if you choose to not make any changes to the system, performing measurements can advance your knowledge of audio. So I hope everyone of you reading this post will try to measure your room.

My promise to you in return is to give you the world's simplest and most foolproof way to perform room measurements. And do so in a manner that get gets you professional results with substantial ability to go as deep as you want.

There are a number of videos and guides online on how to do this but in my opinion, they all use the wrong approach. Namely, they attempt to show you in one go how to do everything. Problem with that is a number of things can go wrong along the way that lead to failure and giving up. My approach may seem odd but please trust me. It is the best way to do this and is actually what I did myself to learn to use the tool. And comes from seeing so many people try to take the full bite and fail.

Room EQ Wizard
The tool of choice is a computer program called Room EQ Wizard or REW for short. Sadly there is no wizard in the program. Once you learn how to use the program, it is powerful and quick to use. Until that point though, the program can seem impenetrable even if you are computer savvy and know about measurements. The pitfalls are non-standard user interface, use of heavy technical jargon ("tukey window" anyone?) that even caused me to not know which end was up. It was only through significant experimentation and creating "controls" where I knew what the outcome had to be, did I manage to figure out how the program works. The path I am going to show you here will keep you from facing these difficulties.


0. First step is downloading the program. You used to have to register in a forum to get it but it now has its own web site. Downloading is hassle free, no spam, not garbage. http://www.roomeqwizard.com/

REW is written in "java" which means that it is pretty portable and runs on Mac and Windows. The downside is that it is written in Java which means the stupid thing (Java that is) will sit in the background and forever pester you to upgrade it, and opens your computer to security holes. I enable and disable it as I need it. I suggest for now leaving it enabled until you get through the program. A few days of use is not painful.

This is ALL you need for this part of tutorial as far as hardware and software! Yes, you heard me right. You don't need to buy a microphone. You won't even be hooking up your laptop to your audio system. All I want you to do is get through running the program and importantly, learn the steps to generate test tones and measure the response. The sound will be produced by your laptop speakers. And measurements performed using its built-in microphone.

No, this won't be the real measurements of your room since your laptop speakers suck and the microphone sucks nearly as much. The reason I want you to start this way is that the hardest part of this is using the program. Last thing we need is adding other bits to it that can also cause problems of their own. Most computers have working microphone and speakers so we are good to go there for testing. Once we are comfortable with this step we can move to adding hardware and making real measurements.

1. Install REW and run it. My version checks for new versions all the time at start-up. Cancel out of that if you see it and let the program start. You should see this screen:View attachment 483682

2. Ignore everything but the one button to the top left that I have highlighted in yellow called "Measure." Hit that button. REW will pop up an error message complaining about Calibration. Ignore that! And tell it to "continue anyway."
View attachment 483683

You should then see this pop up window:

View attachment 483685

Ignore everything but the two VU meters on the right. There are two numbers at the bottom and red indicator on the bar graphs. See if the numbers are dancing up and down by a bit. Whistle or shout into the laptop and see if they jump way high. If you have used your laptop for skype or anything elses that uses the mic, you should be good to go here.

If you don't see the VU meters changing, your computer microphone is not working. Stop here. Don't do anything else until you figure this out. For Windows users, go to the sound control panel, select the microphone and make sure it is not set to mute, or its level at zero. Many laptops have a function key that mutes and unmutes the microphone. Check to make sure you have not activated the muting this way.

3. Assuming you are seeing the levels changing with sound, next step is to hit the button on the left that says "Check Levels." You should hear a hissing sound like FM radio with no-station. Those of you who use AVRs, should be familiar with this test tone for calibrating your speaker levels. This only plays for a few seconds during which the VU meters should shoot way up.

If nothing plays, then you have a playback problem in your computer. Start your favorite media player and have it play some music on your laptop. If you have changed the default output to USB DACs, etc. you need to defeat those so that the music comes out of the laptop speakers. Once there, then REW should work just the same and the broadband test noise playing by hitting Check Levels again.

REW will then report on a level in green, yellow or red. This is the amount of headroom you have in dB SPL. REW likes to have about 10 dB SPL of headroom. Anything lower and REW complains with a red SPL value:

View attachment 483686

Use the text field called "Level (db FS)" and hit the down button on it and keep testing until the level shows green. You can see mine set at -25.0 in this snapshot.

If your microphone or system is not working, which you should have caught earlier, you will see something like this:View attachment 483687

This is basically "digital zero" with the program unable to capture anything. In my case, my mic was set to mute.

Here is an OK level:
View attachment 483688

Notice that all I had to do is lower the db SPL a couple of notches.

Alas, this exercise can be futile. The test tone REW uses here is different than the one it uses to make its real measurements. The reason to go through this is because the controls to adjust the levels is the same when we make the actual measurements. And this step aids in further troubleshooting if the basics are not working right.

5. Now that the basic checks are done, we are ready for our first measurement. How exciting! Before you can do that, you need to fill out the two text fields I had highlighted in the user interface. Those are the frequency range of the measurements. It wants to know the start and stop frequencies. Put in 20 for the first box and 20000 for the second. Now hit the "Start Measure" button.

At first you likely hear nothing even though the program seems to be doing something. That is because your laptop speakers can't reproduce 20 Hz. After a while, the frequency gets high enough that you start to hear a tone that is changing in pitch. It rapidly changes in frequency all the way up to 20 Khz. For the same reasons as the bass, and your hearing, likely you won't hear the last part of this sweep either. That's OK. We are just learning about the system.

During this measuring, REW puts up a "headroom" message and a changing SPL number much like when you were checking levels. Once again, if this number runs into values in red, your signal is too hot. Likewise blue means your levels are too low. Use the Levels control to change the levels or the speaker level in your PC if you can't boost it enough in REW.

REW seems happy with numbers in 10 to 20. I like to get it as close as I can to 10 so that the measurements I make are well above the room noise floor. For this testing anything in green is fine because we are just playing. But when the time comes to make real measurements, this becomes important.

Alas, adjusting levels in this mode can be tedious as every time REW is done with the measurement, it goes back to the main program interface with the new measurement captured:View attachment 483689

If your levels were wrong, ignore what you see and click "Measure" again and go back to the measurement window and adjust the levels. Keep doing this until you get satisfactory levels.

6. Congratulations. You just made your first room measurement! Or I should say the capabilities of your laptop speaker and mic. You should see two things here. One is your measurement in a much smaller window on the left. But also a much expanded version on the right as you see in mine. If you see the little graph on the left but not on the right, then use the scroll bar to the left of the large graph and go up and down and you should be able to find it.

The X (horizontal) axis is the frequency. The Y (vertical) axis is the playback "SPL Level." I put that in quotation mark because the absolute numbers there are completely wrong. We could have set our mic and speaker level to anything and change those numbers. Good news is that this does not matter!!! In acoustics measurements, we only care about the shape of things, i.e. relative values. We care that at say 50 Hz, we had a peak that was 10 db higher than at 1 Khz. We don't care what the actual measured value was at either point.

Attempting to get accurate SPL levels is the biggest mistake users of REW make! Do not fall in that trap. There is no useful information in there. Getting accurate levels is possible but requires extensive detective and configuration work with no net value. Get an SPL meter if you want to calibrate your system.

Back to our measurements, you can see the response of my laptop above with the TV playing in the background and the limit of 10 KHz. As expected my laptop speakers don't play below 300 Hz. And high frequency response was shot after about 7 or 8 Khz. Isn't this type of information revealing? Could you have guessed these limits from just looking at your laptop or playing music?

As you keep making measurements, REW stacks each one of them on the left. By clicking on any of them, you see that graph.

6. Now is the time to get comfortable with the program. Make another measurement but this time cup your hand over the microphone or speakers and see how the graph changes.

To see all the measurements at once to compare them, click on the button that says "ALL SPL." You see a list of checkboxes at the bottom. Click them on and off and graphs appear and disappear. The color of the graph matches the measurement lists on the left. Speaking of that, click on where the date is and type whatever you want. I put in the name of the measurement such as, "speaker covered by hand." There is also a white box for any comments you want to add.

To delete a measurement click on the red "x" on the measurement bar on the left. Go ahead, delete all of them and then make a new measurement. Get comfortable with the interface.

Click on the button on top right that says "Limits." Change the left and right number and that would change the range of frequencies seen. If you want to see just the bass, set the low to 20 Hz and high to 200 Hz. That won't show much now given the limitation of the laptop but is essential when the time comes. Top and bottom set on the left adjust the high and low of the display. Again, ignore the second set on the right.

Go to File menu and click Save. What REW saves is the entire set of measurements shown in the left vertical bar, not single ones!

This is all I want you to do in this lesson. The rest is quite a bit easier. Have trouble? Ask questions here. Let's get everyone to finish this part before we move on.

Part 2: http://www.audiosciencereview.com/f...om-measurement-tutorial-for-dummies-part-2.5/

I get better results using a phone app than REW and thats if I can even get REW to stay running without errors. I can run a complete test using my phone before my noisy computer even is ready to start a program. Running my computer while testing is defeating the purpose of the test since it's the loudest noise in the room.

The problem is finding an app for Android that has all elements needed to run a complete test without needing external sources to run it.

I have no issues using another phone as a source to generate test tones but it's just easier to have that in the app.

The best one I've found is called Spectrum RTA by a Russian Dev named Leonid. He's a nice guy and will answer emails. He has pulled his app from play store but has several revisions found on many other app sources like Aptoide. I found it's best to download all revisions since some won't work on some devices and some versions of Android. The newest might not be best for newest OS but installing them all will allow you to find the best one.

His app has almost everything you could want to do any testing. 4 different noises , 2 sweeps, frequency generator, control over how it looks and all kinds of settings that are useful. A pretest noise cancelling to reduce the noise floor , peak hold, recording, scalable smoothing and screen sizes and rotation. Change to music notes, choice of mics and setting mic correction, you can use external mics and send source tones to use Bluetooth or wired or use a Daytona mic that has the extra wired output along with the aux input or headphone jack.

Theres much more but you get the idea. So easy to use and results are useable in the real world.
I have moved on past running tests now. The amount of correction is so slight that it can be corrected with 3 frequencies reduced 1 to 3db each. Cut at 80, 200 and a boost at 1.25k. using modified martin Logan Prodigy's powered by a Sunfire 300 and a QSC GX5 on woofers.

I do another test where I let an app run and record info while I play music and at the end of the album I'll check the plot left from the peak hold. It should be close to a flat line that tips down as it goes higher in frequency. Same affect as pink noise or white noise. I'll run 2 phones and see how closely the plot looks to compare the mics also run different apps to see which ones are more accurate. Most apps will produce way different plots and can be discounted as worth using. Only 3 or so apps show similar plots out of 30 or more I've tested and it's way worse when I tested computer programs. Out of 30 I found none that could reproduce the same results between them. More than half wouldn't run properly it didn't allow the use of an external sound card with a powered test mic running to USB and computer analog inputs are just awful and can't be used. Last time I tried using my computer as a source for any sound I could hear my mouse movements through my speakers and that was over the fan noise! I gave up on a computer for anything but photo shopping and to watch a movie through my tvs speakers. HDMI sound is fine but analog is noisy and awful, my amps have no digital input.

 

[Blew]

Last time I tried using my computer as a source for any sound I could hear my mouse movements through my speakers and that was over the fan noise! I gave up on a computer for anything but photo shopping and to watch a movie through my tvs speakers. HDMI sound is fine but analog is noisy and awful, my amps have no digital input.

Perhaps you need a new laptop

 

[MaxwellsEq]

Last time I tried using my computer as a source for any sound I could hear my mouse movements through my speakers

You probably have an earth loop.

 

[murl]

I remember being in a “high end” audio store several years ago right after purchasing a lifetime membership to Roon when it was $500 and discussing my source being a computer with the owner. It may as well been blasphemy to him. He tried to sell me on some source,
CDs ripped to it with rubber feet and virgins generating the power inside magically isolated from the noise of the world connected to a chord Hugo M scaler and the DAC. I don’t understand as much about this hobby as many of you but try to take a scientific approach to this like I do my job. Have never understood the people advocating for a really expensive power cable connected to the outlet in your house that may run 30 cents a foot.