Room Measurement Tutorial for Dummies Part 2

[Furyosa]

Hello! Is there an updated version of this REW guide? I'm on REW and on step 2 of part 1 I'm already lost, my REW doesn't look anything like that. After hitting "click anyway" I don't get the page with the start button and sound meters. I get a different page with a state button and after it runs a sweep, the data is nowhere to be found.

 

[Sal1950]

REW Help Index

 

[posvibes]

This is not a criticism, this is not a request, this is not a reminder.

Just a query.

Is there or will there be a Part 3 to this topic?

 

[Macatouille]

Bumping this with the same question: would love a part 3. Like that the first 2 parts are written so plainly where so many others are filled with jargon.

 

[ClaudeJ1]

Bumping this with the same question: would love a part 3. Like that the first 2 parts are written so plainly where so many others are filled with jargon.

Eigenton ratio. Room treatments. Sub bass FIRST, then the rest.

 

[dorirod]

Is there a page that shows the do's and don'ts on applying EQ? I know some of this may depend on specific conditions, preference, etc. Looking a robust list of things to catch most egregious mistakes a dummy like me might make like:
-don't EQ up below this frequency
-can EQ up within this frequency range if speaker allows for it (good directivity, distortion, etc.)
-which sort of peaks/valleys is worth doing EQ on (especially if you're limited to say 5 or 10 only)
-don't bother with EQ above this frequency
-max (or min rather) Q value to use (e.g. don't bother if some auto EQ offers Q of 24 or something)

Thank you!

 

[levimax]

Is there a page that shows the do's and don'ts on applying EQ? I know some of this may depend on specific conditions, preference, etc. Looking a robust list of things to catch most egregious mistakes a dummy like me might make like:
-don't EQ up below this frequency
-can EQ up within this frequency range if speaker allows for it (good directivity, distortion, etc.)
-which sort of peaks/valleys is worth doing EQ on (especially if you're limited to say 5 or 10 only)
-don't bother with EQ above this frequency
-max (or min rather) Q value to use (e.g. don't bother if some auto EQ offers Q of 24 or something)

Thank you!

To be safe and good place to start:
First do what you can with speaker placement, listening position, rugs, and furniture placement.
Don't EQ above Schroeder frequency of room, if you't calculate it use 300 Hz
Smooth measured response using Var Smoothing then EQ peaks down and don't EQ dips. REW EQ automatically calculates filters and works well. Just set up accordingly.

This should help a lot and do no harm. Probably all you need in most cases.

 

[kolestonin]

Smooth measured response using Var Smoothing then EQ peaks down and don't EQ dips. REW EQ automatically calculates filters and works well. Just set up accordingly.

Could you please elaborate a bit on that? You mean there is no need to EQ the dips as REW will do it automatically? And then why it does not automatically EQ the peaks as well?
or maybe that the dips do not have such a big impact so there is no need to bother? But then why?(given it cost you nothing)

I am not disputing you. I am just asking as I have some minimum experience and 2-3 unsuccessful attempts with REW and I am gathering information so to give it another try

 

[levimax]

Could you please elaborate a bit on that? You mean there is no need to EQ the dips as REW will do it automatically? And then why it does not automatically EQ the peaks as well?
or maybe that the dips do not have such a big impact so there is no need to bother? But then why?(given it cost you nothing)

I am not disputing you. I am just asking as I have some minimum experience and 2-3 unsuccessful attempts with REW and I am gathering information so to give it another try

There are always exceptions but generally trying to "boost a dip" is difficult because they are usually caused by "destructive interference" and adding more power doesn't help. In addition dips, especially narrow ones, tend to not be particularly audible. So a good place to start is to not try to fill dips.

Peaks on the other hand respond well to EQ and are also very audible.

With this in mind I set up REW to not try to boost dips and it will always cut down the peaks. The level also comes into play and I usually use "calculate target level from response" see snip below.

Also when setting the "lower level" be realistic about you speakers' response and set this high enough so REW doesn't create a boost where your speakers are rolling off. In the snip I have it set to 15 Hz but I have full range speakers with a room mode at ~20 Hz so I am actually trying to smooth off a peak not boost anything below 20 Hz.

Hope this helps.

 

[Pieter1267]

REW is very complicated but if I was to write a short guide.

1. Please bother to read the REW guide. Its complicated but the start and set up isnt;
2. Ditto: help section at REW is good setting up stuff, like how to make a sweep;
3. Calibrate, always start with calibrating the input. Open the SPL meter and the generator;
4. Always, ALWAYS set your SPL and other screens right. The button ''limits' is your friend. Y axis in 5 db steps. Figure out if you want low freq or full range and set preferences;
5. Waterfall and spectogram say a lot about how your room is;
6. Avoid fixing nulls, even with autoEQ- better to lower the target response a tad;
7. Busy making and crossover? Then your null could be because of phase and time problems;
8. Learn a LOT about phase and time alignment on youtube, for instance if you want to marry subwoofers with main speakers;
9. MINIDSP has a lot of helpful pages;
10. MANY helpfull folks here that can help you with identified problems. Make screenshots with a clear set of the 'limits'', preferably if youve used and if so, wich smoothing;


That and get a UMIK1 or UMIK2 microphone. Its very plug and play and you;ll use it for years and years.

 

[JW20147]

Hey everyone! I've been following this thread for a while, and today I just got my rented UMIK-1 to measure my KALI LP-UNF. I'm not sure how good the results are, but I used 1/6 smoothing, and it seems like there are a lot of bends?
The green one is my right channel, and the red one is the left channel. I noticed that they are not aligned in some places. Is this difference normal?Thank you all for your discussion and answers. Have a nice day, everyone.

 

[gorman]

Hey everyone! I've been following this thread for a while, and today I just got my rented UMIK-1 to measure my KALI LP-UNF. I'm not sure how good the results are, but I used 1/6 smoothing, and it seems like there are a lot of bends?
The green one is my right channel, and the red one is the left channel. I noticed that they are not aligned in some places. Is this difference normal?Thank you all for your discussion and answers. Have a nice day, everyone.

I am absolutely not an expert and hopefully somebody more knowledgeable than me will soon chime in.

From what I can tell, your frequency response is rather normal. You don't mention equalization or room correction of any sort, nor any acoustic room treatment. As such, an uneven bass response is simply your room acting on sound, as it's normal to do. With one or, preferably, multiple subwoofers you could even it out, at least partially.
The bass deficiency in the 150-250 Hz area might impact negatively the quality of what you hear, I guess.

But, again, no expert here.

 

[nathan]

Hey everyone! I've been following this thread for a while, and today I just got my rented UMIK-1 to measure my KALI LP-UNF. I'm not sure how good the results are, but I used 1/6 smoothing, and it seems like there are a lot of bends?
The green one is my right channel, and the red one is the left channel. I noticed that they are not aligned in some places. Is this difference normal?Thank you all for your discussion and answers. Have a nice day, everyone.

That's actually quite good and normal. Remember, everything below 200 and often 300 hz is dominated by the room modes and not the speaker linearity. And even above that level, measuring a speaker in a room will introduce wiggles that aren't really there.

If you want to get a better sense of the speaker itself, take it outside, set it high on a platform (as high as you can reasonably set it), please the UMIK1 about 1 meter in front of the speaker, and run the test tones through it. Not perfect, and there is a whole DIFFERENT thread here about doing pseudo-anechoic measurements without special tools, if you want to do an even better job of measuring the speaker. But doing it outside is better than indoors.

You may also want to adjust the Y axis to cover just 50 to 80 spl so that you can see the variation there more accurately.....and I would toggle PSYCH smoothing on and off, which is more like what your ears hear than 1/6 smoothing.

 

[JW20147]

That's actually quite good and normal. Remember, everything below 200 and often 300 hz is dominated by the room modes and not the speaker linearity. And even above that level, measuring a speaker in a room will introduce wiggles that aren't really there.

If you want to get a better sense of the speaker itself, take it outside, set it high on a platform (as high as you can reasonably set it), please the UMIK1 about 1 meter in front of the speaker, and run the test tones through it. Not perfect, and there is a whole DIFFERENT thread here about doing pseudo-anechoic measurements without special tools, if you want to do an even better job of measuring the speaker. But doing it outside is better than indoors.

You may also want to adjust the Y axis to cover just 50 to 80 spl so that you can see the variation there more accurately.....and I would toggle PSYCH smoothing on and off, which is more like what your ears hear than 1/6 smoothing.

Hey! Thanks for your advice. I adjusted the image display and reposted it.What confuses me a bit is that after 200, it feels like the two frequency responses hardly overlap at all? In fact, it might indeed have something to do with my room: the left speaker has a wall behind it and is close to a wooden bookshelf on the left, while the right speaker only has a wall behind it.
However, I’m still concerned that this might be due to the placement of my speakers, such as the two speakers not being perfectly symmetrical, with the left or right side being one or two centimeters closer. Or perhaps the angles of inclination are different, such as 24° for the left and 25° for the right?
My approximate listening distance is 80 cm between the two speakers, with my ears 45 cm from each speaker (of course, I know this is terrible, but I'm currently a student living with my parents, so it's difficult to make significant layout changes to the room).Should I continue experimenting with speaker placement or just enjoy the music until I move out?
Once again, thank you all for your advice; your suggestions are very important to me.

 

[nathan]

Totally normal in a room for the response of each speaker to measure a little differently.

Should I continue experimenting with speaker placement or just enjoy the music until I move out?

Speakers in a room will always vary a bit from one another. Enjoy the music and don't worry about the graph (which is pretty normal).

 

[mmares056]

Hey! Thanks for your advice. I adjusted the image display and reposted it.What confuses me a bit is that after 200, it feels like the two frequency responses hardly overlap at all? In fact, it might indeed have something to do with my room: the left speaker has a wall behind it and is close to a wooden bookshelf on the left, while the right speaker only has a wall behind it.
However, I’m still concerned that this might be due to the placement of my speakers, such as the two speakers not being perfectly symmetrical, with the left or right side being one or two centimeters closer. Or perhaps the angles of inclination are different, such as 24° for the left and 25° for the right?
My approximate listening distance is 80 cm between the two speakers, with my ears 45 cm from each speaker (of course, I know this is terrible, but I'm currently a student living with my parents, so it's difficult to make significant layout changes to the room).Should I continue experimenting with speaker placement or just enjoy the music until I move out?
Once again, thank you all for your advice; your suggestions are very important to me.

Your FR looks normal. In-room frequency response is influenced by room acoustic characteristics. As an example the attached show the frequency response for my FL/FR.

 

[Sal1950]

Hey everyone! I've been following this thread for a while, and today I just got my rented UMIK-1 to measure my KALI LP-UNF. I'm not sure how good the results are, but I used 1/6 smoothing, and it seems like there are a lot of bends?
The green one is my right channel, and the red one is the left channel. I noticed that they are not aligned in some places. Is this difference normal?Thank you all for your discussion and answers. Have a nice day, everyone.

As has been said, that's quite normal and not really too bad at all with only the below 250hz wide swings glaring.
A most repairable ill using a good digital room correction system.

 

[Sal1950]

I adjusted the image display and reposted it.What confuses me a bit is that after 200, it feels like the two frequency responses hardly overlap at all? In fact, it might indeed have something to do with my room: the left speaker has a wall behind it and is close to a wooden bookshelf on the left, while the right speaker only has a wall behind it. However, I’m still concerned that this might be due to the placement of my speakers, such as the two speakers not being perfectly symmetrical, with the left or right side being one or two centimeters closer. Or perhaps the angles of inclination are different, such as 24° for the left and 25° for the right?

Such is life in the world of Hi Fi, room interactions from speaker reflections are a way of life.
Some improvements can be gained using various room treatments, more if you can afford a dedicated room rebuild.
In the main you can never expect the two speakers to measure exactly alike in a room.
Do your homework to learn what can be done but don't fall off a cliff over it.

 

[JW20147]

Such is life in the world of Hi Fi, room interactions from speaker reflections are a way of life.
Some improvements can be gained using various room treatments, more if you can afford a dedicated room rebuild.
In the main you can never expect the two speakers to measure exactly alike in a room.
Do your homework to learn what can be done but don't fall off a cliff over it.

Your FR looks normal. In-room frequency response is influenced by room acoustic characteristics. As an example the attached show the frequency response for my FL/FR. View attachment 456635

Hey, thanks! Your opinions really helped me a lot.
Today, I didn't dwell on whether the two curves were aligned. I tried REW's EQ, but unfortunately, the effect wasn't very noticeable.I think I may have been too conservative when setting the EQ in REW, which is why there wasn't much difference(the pinkish one is after EQ, and the reddish one is the original).
Thanks again for your advice.

 

[Sal1950]

Hey, thanks! Your opinions really helped me a lot.
Today, I didn't dwell on whether the two curves were aligned. I tried REW's EQ, but unfortunately, the effect wasn't very noticeable.I think I may have been too conservative when setting the EQ in REW, which is why there wasn't much difference(the pinkish one is after EQ, and the reddish one is the original).
Thanks again for your advice.

REW is an excellent measurement tool! It can also offer good correction results but the learning curve for FR modification is very steep and not something that can be mastered in a few days.
I would highly suggest finding a way to introduce Audyssey or Dirac room correction into the playback chain to smooth the below 250hz curve.
The most basic approach is to use one of these semi-automated programs for the curve correction and then follow using REW to measure the
results.