• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Room measurements with KEF R3

EEE272

Active Member
Joined
Jan 4, 2022
Messages
128
Likes
61
I showed in a graph earlier that the tilt is optimal. Roughly -5dB around 10-20KHz
This might be a bit of a stupid remark but the optimal result should depend on listening distance and room treatment.
In the standard room, Toole describes it at 0.4 per octave up to around 6K-10K where you typically get a fall off that is stronger.
The above graph does show -9dB on 300 to 10K that is a bit too much.
In this case, I cannot imagine that the direct sound is still flat. I know Dirac sets different targets etc. but none of these optimal graphs assume that you change the direct sound to something very non-flat. On the contrary, Toole even warns about it. Of course, ultimately, what sounds right is right ;)
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,425
Likes
7,941
Location
Brussels, Belgium
This might be a bit of a stupid remark but the optimal result should depend on listening distance and room treatment.
Not really, above the transition frequency the tilt correlates very strongly to the speaker tuning, the early reflections plot to be specific.

your statement is only relevant in the extremes, when you're sitting in the reverberate field (so far away from the speakers that it's equally loud everywhere around the room and direct sound is almost non-existent) or in an anechoic chamber (only direct sound, no reflections).

In the standard room, Toole describes it at 0.4 per octave up to around 6K-10K where you typically get a fall off that is stronger.

That's a weird statement, since the room should not really influence the tilt. it's a result of the directivity of the speaker increasing as frequencies get higher.

In this case, I cannot imagine that the direct sound is still flat. I know Dirac sets different targets etc. but none of these optimal graphs assume that you change the direct sound to something very non-flat. On the contrary, Toole even warns about it. Of course, ultimately, what sounds right is right ;)

I was actually referencing the response without EQ in my earlier post, which shows a very healthy tilt in my opinion.

index.php
 
OP
M

Marcin

Member
Joined
Jan 12, 2022
Messages
49
Likes
20
Location
Poland
@abdo123 What I did in the recent change is I moved the listening position closer the the speakers so that it complies with the 38% rule more or less. Without any EQ it showed that 37 Hz peak got lower but 64 Hz got bigger. Then I started to test EQ.
At first 37 Hz -6db: 37 Hz peak got smaller but 64 Hz got even bigger than before.
Then I disabled 37 Hz from APO Eq and enabled 64 Hz -6db. This change didn't do anything to 37 Hz but lowered 64 Hz so they are equal now.

When listening to music now I don't experience so much booming sound that was before at the left side of the room. Bass is more crispy and I can actually hear more texture in the low frequencies.
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,425
Likes
7,941
Location
Brussels, Belgium
@abdo123 What I did in the recent change is I moved the listening position closer the the speakers so that it complies with the 38% rule more or less. Without any EQ it showed that 37 Hz peak got lower but 64 Hz got bigger. Then I started to test EQ.
At first 37 Hz -6db: 37 Hz peak got smaller but 64 Hz got even bigger than before.
Then I disabled 37 Hz from APO Eq and enabled 64 Hz -6db. This change didn't do anything to 37 Hz but lowered 64 Hz so they are equal now.

When listening to music now I don't experience so much booming sound that was before at the left side of the room. Bass is more crispy and I can actually hear more texture in the low frequencies.
Good to hear. Remember that distance is kind of key. The closer you sit to a speaker the less of the room you hear and more of the original material / speaker.

The speaker's location seems pretty good now, but if you can get the listening position even closer to the speakers you will notice dramatic improvements in clarity very quickly.

Of course all of these things need to be balanced in a way that it makes sense for your life style and layout of the room.
 
OP
M

Marcin

Member
Joined
Jan 12, 2022
Messages
49
Likes
20
Location
Poland
Now is the closest I can get so that the room is still functional. The distance between me and the speakers actually didn't change much, it's still around 2.6 m, but the distance between me and the back wall increased to 190 cm which is almost 38% of the whole length measuring from back wall.
I could only move the speakers about 10 cm more to the front wall but I don't know if it will do any improvement.
 
OP
M

Marcin

Member
Joined
Jan 12, 2022
Messages
49
Likes
20
Location
Poland
Going back to what @EEE272 wrote - should I be worried that something is wrong with my speakers or amplifier?
 

Jdunk54nl

Addicted to Fun and Learning
Joined
Aug 5, 2020
Messages
968
Likes
1,048
Location
Arizona
Did you put in your microphone house curve? How did you point the microphone? Up at the ceiling or straight at the speaker? If up at the ceiling, did you use a 90 degree cal file? If straight at the speaker did you use the 0 degree cal file?

I am guessing you pointed it at the ceiling with a 0 degree cal file. In my experience, a lot of cheaper microphones will show some weird roll off around the 15khz region, more if you use a 0 degree cal file but point it at the ceiling.
 
OP
M

Marcin

Member
Joined
Jan 12, 2022
Messages
49
Likes
20
Location
Poland
Mic pointed at the speakers. Regular calibration file, not the one with 90 in the name.
 

TurtlePaul

Addicted to Fun and Learning
Joined
May 19, 2020
Messages
631
Likes
1,027
Location
New York
Looking at the first post, the 90 hz dip corresponds to the back wall bounce. By adjusting your speaker positioning you have moved that up to the point where you have nulls just above 100 hz. Try pulling the speakers out so the speaker cones are 1.35 meters from the front wall. That should make the room mode at ~64-65 hz and the bass null the same frequency and should sound phenomenal. Of course, that is far to pull them out into the room.
 

Jdunk54nl

Addicted to Fun and Learning
Joined
Aug 5, 2020
Messages
968
Likes
1,048
Location
Arizona
Looking at the first post, the 90 hz dip corresponds to the back wall bounce. By adjusting your speaker positioning you have moved that up to the point where you have nulls just above 100 hz. Try pulling the speakers out so the speaker cones are 1.35 meters from the front wall. That should make the room mode at ~64-65 hz and the bass null the same frequency and should sound phenomenal. Of course, that is far to pull them out into the room.

Man, I wish I had a room that I could pull the speakers 4.4 ft (1.35m) from the wall. I am sure that is not realistic for most people.
 

TurtlePaul

Addicted to Fun and Learning
Joined
May 19, 2020
Messages
631
Likes
1,027
Location
New York
Man, I wish I had a room that I could pull the speakers 4.4 ft (1.35m) from the wall. I am sure that is not realistic for most people.

Due to WAF or not a big enough room? The provided room dimensions are about a 17 ft square. That is no small room. In such a big room having the speakers 3 ft. from the wall shouldn't be that crazy (but may start to push spousal boundaries). The speakers are over a foot deep, so the back of the speaker would be a little over 3 feet from the wall.
 

Jdunk54nl

Addicted to Fun and Learning
Joined
Aug 5, 2020
Messages
968
Likes
1,048
Location
Arizona
Due to WAF or not a big enough room? The provided room dimensions are about a 17 ft square. That is no small room. In such a big room having the speakers 3 ft. from the wall shouldn't be that crazy (but may start to push spousal boundaries). The speakers are over a foot deep, so the back of the speaker would be a little over 3 feet from the wall.
My room is 18ft x about 18ft (that then starts the kitchen in an open concept area), but walk ways and other seating areas prevent losing 4ft behind the speakers just to have them off the wall. Nothing will be able to go behind them or it defeats the purpose.

Unless I made a dedicated listening area, I don't see a lot of people having 4ft just to "give up" to have the speakers that far out in the room. That gives the OP 13 ft to play with in instead of like ~16ft. That is ~50 sq ft of space given up just for moving the speaker out into the room.
 

TurtlePaul

Addicted to Fun and Learning
Joined
May 19, 2020
Messages
631
Likes
1,027
Location
New York
Depending how deep the TV cabinet is, it could look really out of place or not. They won't have 16 ft to play with due to the cabinet between the speakers. The OP isn't showing any other furniture in that room other than an entertainment room.
 

EEE272

Active Member
Joined
Jan 4, 2022
Messages
128
Likes
61
Going back to what @EEE272 wrote - should I be worried that something is wrong with my speakers or amplifier?
I didn't mean to scare you ! Sorry if I did.
I do not think that something is broken, but I still think something is strange about the measurement.
Regarding the mic, you should use the 90degree and point the mic at the listening position to the ceiling. Otherwise reflections are underestimated. Maybe that is what throws me off.

To check your direct sound, you can try a a near-field measure. Here, you can indeed use the 0 degree calibration file and point the mic directly to the speaker. If you get very close, you can also lower the volume. Hereby, you can eliminate some of the room influence to see only how the direct sound looks. That line should be flat. It is also shown in Amir's To be more precise, you can can choose a time window in REW that will then eliminate sound that arrives late and you can focus only on the first arrival. This should also be flat.

@abdo123 regarding your questions. Sorry if I was not clear. The more your room is treated with absorbers, the flatter the line becomes because the reflected sound is eliminated and therefore you have more direct sound. The distance to the speaker has an effect as well due to the shift of direct vs. indirect. Further, above 6K, air attenuation plays a role. This is also why often the slope up to 10K is looked at, as there is some significant drop off afterwards. In consequence, if you draw the line further out, you get the overestimation of the falloff due to the air effect. The 0.4db/Octave are from Floyd Toole's book (he draws it not up to 20K but rather 10K).

But again, I don't think anything is broken! The measurement still puzzles me a bit though but if you could try the above things, I would be very curious to see the results!

Finally, all that sounds well is fine... ;)
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,425
Likes
7,941
Location
Brussels, Belgium
@abdo123 regarding your questions. Sorry if I was not clear. The more your room is treated with absorbers, the flatter the line becomes because the reflected sound is eliminated and therefore you have more direct sound. The distance to the speaker has an effect as well due to the shift of direct vs. indirect. Further, above 6K, air attenuation plays a role. This is also why often the slope up to 10K is looked at, as there is some significant drop off afterwards. In consequence, if you draw the line further out, you get the overestimation of the falloff due to the air effect. The 0.4db/Octave are from Floyd Toole's book (he draws it not up to 20K but rather 10K).
You really have to pick a stance, does absorption removes output in higher frequencies (more aggressive tilt) or adds output (flatter response). you can't just juggle both opinions and throw one of them when it fits the situation ;).
 
OP
M

Marcin

Member
Joined
Jan 12, 2022
Messages
49
Likes
20
Location
Poland
Looking at the first post, the 90 hz dip corresponds to the back wall bounce. By adjusting your speaker positioning you have moved that up to the point where you have nulls just above 100 hz. Try pulling the speakers out so the speaker cones are 1.35 meters from the front wall. That should make the room mode at ~64-65 hz and the bass null the same frequency and should sound phenomenal. Of course, that is far to pull them out into the room.
Assuming it is doable where would the listening position be?

The OP isn't showing any other furniture in that room other than an entertainment room.
It is a living room however the furniture I showed in the drawing is mostly it. There's also some kids stuff which I didn't draw like IKEA Mammut table or boxes with toys ;) The area behind the sofa is currently empty. Sometimes I use it for a table when inviting guests. Maybe I will permanently place a table there at some point...
I didn't mean to scare you ! Sorry if I did.
I do not think that something is broken, but I still think something is strange about the measurement.
Regarding the mic, you should use the 90degree and point the mic at the listening position to the ceiling. Otherwise reflections are underestimated. Maybe that is what throws me off.

To check your direct sound, you can try a a near-field measure. Here, you can indeed use the 0 degree calibration file and point the mic directly to the speaker. If you get very close, you can also lower the volume. Hereby, you can eliminate some of the room influence to see only how the direct sound looks. That line should be flat. It is also shown in Amir's To be more precise, you can can choose a time window in REW that will then eliminate sound that arrives late and you can focus only on the first arrival. This should also be flat.
Thanks for clarification :) Tomorrow I will do measurements according to your directions. As for the near-field should the mic be placed in the middle of the twitter just a few centimeters away?
 

TurtlePaul

Addicted to Fun and Learning
Joined
May 19, 2020
Messages
631
Likes
1,027
Location
New York
Assuming it is doable where would the listening position be?
If your drawing is to scale, it looks like the sofa is centered from left to right and about 1/3 the room distance away from the back wall? That is a pretty good place to be. The speakers will just be somewhat close to you. The main thing to avoid is to not have the listening position right exactly at half way between the front and back wall, that area usually has some cancelations.
 

EEE272

Active Member
Joined
Jan 4, 2022
Messages
128
Likes
61
You really have to pick a stance, does absorption removes output in higher frequencies (more aggressive tilt) or adds output (flatter response). you can't just juggle both opinions and throw one of them when it fits the situation ;).
Did I say something else?
I will check - I am really not aware that I did. More absorption=flatter curve.
 

EEE272

Active Member
Joined
Jan 4, 2022
Messages
128
Likes
61
You really have to pick a stance, does absorption removes output in higher frequencies (more aggressive tilt) or adds output (flatter response). you can't just juggle both opinions and throw one of them when it fits the situation ;).
In the standard room, Toole describes it at 0.4 per octave up to around 6K-10K where you typically get a fall off that is stronger.

Was it this statement? I still stand by it. Above 10k air attenuation also affects reflections that travel a longer way.
There is a drawing in Toole's book. I don't have it here but can give you the page indication later.

Maybe it was not formulated well?
I meant from around 6-10k the fall off no longer follows a straight line.
 
Top Bottom