KEF Reference speakers - channel balance question

[Soniclife]

I directly went to disortion measurement option of the REW since it gives both frequency response and disortion in one shot. I did not change any settings between measurements.

Can you repost this with the axis changed so the distortion is visible on the graph, it might help.

Also both fundamentals on the same graph with a little smoothing would be clearer.

 

[AToMe]

However, I would like to know why the LF roll off for both speakers started at ~200 Hz.

I unlocked the binding posts of the speakers so that only the UNIQ's were playing. The woofers on them sounded absolutely fine for me and on listening tests as well as measurements, they were virtually identical. I love the sound of the speaker when my R speaker alone is playing music. I am not sure, if the intended sound from KEF is my Left speaker or the right speaker. The R has more air on top and vocals are bit "cold" but that's how I remember the Genelecs and Geithains. The L has bit of shoutiness to the sound, it's still a lovable sound, but the R is much more easy to listen to. Problem is when I play them both, especially when cymbals are panned from left to right, the tonality is bit different. also, mids are slightly justified to the left some times especially with vocals like Adele that at dead centre. When I slightly push the shadow flare on the left, the sound appears to become cool for come minutes and slowly it goes back to the shouty nature after few minutes. I am suspecting that the suspension of the L uniq is bit loose from the driver chasis itself as when I push the shadow flare its pushing suspensions part behind it back on the chassis place but eventually the force of the speaker after few vibrations is pushing it back to its new normal state. If it was not an expensive speaker I would have tried pushing the cone myself to see if it starts sounding "cool" like the other one, but here I am afraid!

I can be completely wrong, as when I pushing the shadow flare in, probably its still touching the UNIQ's suspension and the since the suspension is not free as before, it might be changing the impedance of the midrange as music plays through it.

By the way using multimeter is there anything that I can check?

 

[NTK]

Thanks. Not connecting the woofers explains FR early roll off.

Erin tested how the fitting of the R3 shadow flare affected FR. Looks like for the R3 it mostly affected the ~1 to 2 kHz region, unlike what your measurements showed.

Kef R3 Bookshelf Speaker Review

Kef R3 Bookshelf Speaker Review

www.erinsaudiocorner.com

I'd say your measured FR deviations between the L & R speakers are surely out-of-spec for KEF's TOTL Reference models. I know it is a lot of work, but to confirm the reliability of your acoustic measurements, I think you can double the number of measurements, e.g., test left speaker, test right speaker, test left speaker again, test right speaker again. That way, you can find out how repeatable your measurement method is. If the 2 left speaker tests match and the 2 right speaker tests match, and the deviations between the left and right speakers are consistent, you can confidently say that the deviations are real.

I doubt DMM measurements will tell us anything useful. It can only tell the DC coil resistance, and for the MF/HF Uni-Q, there should be a blocking cap in the cross-over, so the DC resistance should be infinite.

If you have the instrumentation, you can run an impedance sweep to measure each speaker's complex impedance vs frequency and compare. But I think your acoustic measurements, if you are confident that they are valid, should be sufficient evidence that at least one of the speakers is out-of-spec.

 

[AToMe]

Can you repost this with the axis changed so the distortion is visible on the graph, it might help.

Also both fundamentals on the same graph with a little smoothing would be clearer.

I will figure this out and post it here.

Thanks. Not connecting the woofers explains FR early roll off.

Erin tested how the fitting of the R3 shadow flare affected FR. Looks like for the R3 it mostly affected the ~1 to 2 kHz region, unlike what your measurements showed.

Kef R3 Bookshelf Speaker Review

Kef R3 Bookshelf Speaker Review

www.erinsaudiocorner.com

I'd say your measured FR deviations between the L & R speakers are surely out-of-spec for KEF's TOTL Reference models. I know it is a lot of work, but to confirm the reliability of your acoustic measurements, I think you can double the number of measurements, e.g., test left speaker, test right speaker, test left speaker again, test right speaker again. That way, you can find out how repeatable your measurement method is. If the 2 left speaker tests match and the 2 right speaker tests match, and the deviations between the left and right speakers are consistent, you can confidently say that the deviations are real.

I doubt DMM measurements will tell us anything useful. It can only tell the DC coil resistance, and for the MF/HF Uni-Q, there should be a blocking cap in the cross-over, so the DC resistance should be infinite.

If you have the instrumentation, you can run an impedance sweep to measure each speaker's complex impedance vs frequency and compare. But I think your acoustic measurements, if you are confident that they are valid, should be sufficient evidence that at least one of the speakers is out-of-spec.

I would repeat the measurements multiple times then, to see if I get consistent results. My UMIK is also new, and I have not tried many measurements with it before. Also, I quickly drew a conclusion fast, as soon as I saw a deviation in measurement, as my reasoning to buy UMIK itself was, that I could hear a difference on listening to them to confirm what I was hearing. I tried recording both speakers with my iPhone at the same spot but eventually though the recording sounds slightly different on replaying with my AirPods, its somehow not having the same level of contrast I am hearing in my room. Probably the difference lies in the place where iPhones mics are mostly normalised. Anyway I will try to post the recording here. Hope some of you might be able to hear it, but I don't have high hopes with it.

 

[BrokenEnglishGuy]

I don't see that you have done anything wrong here.
An SPL of around 70-75 dB is enough to see if the two speakers have a matched response or not, and as you say, you didn't move the microphone or anything else and just swapped out the speakers to the same position.

As your loudspeakers are heavy as hell, I hope you will be able to just send in both the UNIQ drivers to KEF, so that they can figure out what is wrong with one or the other and make sure you get a matched pair. No matter if this will be on you or if it's going under warranty, I guess you want it fixed either way. But you should of course reach out to the dealer first if you bought it from a store, it's just a bit unfortunate that it took some time before you realized there was something wrong with the speakers.

Sorry @BrokenEnglishGuy, but you are focusing on irrelevant things that have very little to do with what AToMe wants to investigate with his measurements.

Correct taken measurement is the first priority. Turn up the spl at 85dB to eliminate in-room evirorements noise. This is what i do, because noticed sometimes get some errors at 70dB that dissapiar at 85db~. As others members noticed the measurements was taken incorrectIy.

But, too much text.
Go ahead, and take a decent measurement. a measurements with 519% noisefloor as goat76 want to point '' normal '' is not.


Take care to remove near object that cause weird refection, point the 2 speakers poiting at the exact very exact pace. For make this easier, take the measurement at 1 meter with the speaker straight, no toe in. with the speaker straight, shoud be easier to no make mistakes.

This is reaI ? i have never seen a speaker behavior the same in the both channeIs? or is just the Iack of spI?.
--------------
EDIT: I read your post carefuIIy and i noticed this weren't main Iistening position
<and I carefully placed both the loudspeakers at the marked-out spot on the floor. So... don't pay too much attention to the frequency response in the below measurement as it's highly affected by the in-room response, just look at how the responses match each other. >
so aII i wrote about that measurement is incorrect
-------------------

Bass aIways Iooks Iike this: room aren't empty and most of houses don't are symetricaI ( right vs Ieft ) pIus the objects in the room. This make amost impossibe to Iooks even simiIar in the bass region, but... your bass is Iike a enochoic chamber? what?
If that is your in-room measurement at main Iistening position... Iooks Iike someone did a wrong measurement too , or you have the most symetricaI Iistening position i'ever seen.

Iets wait for measurement from the OP, in the end is far easier to take more measurements than think much about it

 

[BrokenEnglishGuy]

I will figure this out and post it here.

I would repeat the measurements multiple times then, to see if I get consistent results. My UMIK is also new, and I have not tried many measurements with it before. Also, I quickly drew a conclusion fast, as soon as I saw a deviation in measurement, as my reasoning to buy UMIK itself was, that I could hear a difference on listening to them to confirm what I was hearing. I tried recording both speakers with my iPhone at the same spot but eventually though the recording sounds slightly different on replaying with my AirPods, its somehow not having the same level of contrast I am hearing in my room. Probably the difference lies in the place where iPhones mics are mostly normalised. Anyway I will try to post the recording here. Hope some of you might be able to hear it, but I don't have high hopes with it.

Dont waste time with ur iphone, you have the UMIK my guy and don't waste time posting the recording. Not necessary and not usefuII.

 

[goat76]

Correct taken measurement is the first priority. Turn up the spl at 85dB to eliminate in-room evirorements noise. This is what i do, because noticed sometimes get some errors at 70dB that dissapiar at 85db~. As others members noticed the measurements was taken incorrectIy.

I don't see any in-room environment noise problems in the operating range of the UniQ driver in AToMe's measurements. I have not seen anyone else but you who has said that the measurements were done incorrectly.

But, too much text.
Go ahead, and take a decent measurement. a measurements with 519% noisefloor as goat76 want to point '' normal '' is not.

Maybe you missed it, but AToMe said he disconnected the bass drivers and only measured the UniQ drivers in the speakers. The UniQ drivers operate down to about 200Hz and then they pretty much nose-dive under that, so it's not strange at all if the noise floor happens to be up at 519% at 19.8 Hz which we see the numbers of, as the UniQ hardly makes any noise at all that low in the frequency range.

For the measurement for the other speaker, the noise floor at 19.8 Hz is down at 87%. This change in level is not very strange either as it can be from a passing truck, a neighbor who walked on the floor, or a change in the air conditioning. But no matter what it was, it's not necessarily important as it likely didn't affect the operating range of the UniQ that was measured.

Take care to remove near object that cause weird refection, point the 2 speakers poiting at the exact very exact pace. For make this easier, take the measurement at 1 meter with the speaker straight, no toe in. with the speaker straight, shoud be easier to no make mistakes.

Any "weird" reflection point will be the same for both speaker measurements as long as you are positioning the speakers at the same spot while measuring them, and nothing else is changed between the measurements.

View attachment 383642
This is reaI ? i have never seen a speaker behavior the same in the both channeIs? or is just the Iack of spI?.

Yes, that is the real.
That is how a well-matched pair of speakers will measure when measured at the same spot in the room, and it will not make a difference if they both are measured at 70 dB, 75 dB, or 85 dB and if they can both handle the same SPL.

Here are the measurements of my ATC SCM11s from almost the same spot in the room where the SCM40s were measured from. The smaller speakers are also a well-matched pair except for some larger deviations above 10kHz. I think most speaker manufacturers are referring to the range from 100Hz to 10kHz when it comes to matched speaker pairs.

--------------
EDIT: I read your post carefuIIy and i noticed this weren't main Iistening position
<and I carefully placed both the loudspeakers at the marked-out spot on the floor. So... don't pay too much attention to the frequency response in the below measurement as it's highly affected by the in-room response, just look at how the responses match each other. >
so aII i wrote about that measurement is incorrect
-------------------

Sorry, but it's not clear to me what you mean by your EDIT. Did you write something you decided to erase, if so I missed it.

Bass aIways Iooks Iike this: room aren't empty and most of houses don't are symetricaI ( right vs Ieft ) pIus the objects in the room. This make amost impossibe to Iooks even simiIar in the bass region, but... your bass is Iike a enochoic chamber? what?
If that is your in-room measurement at main Iistening position... Iooks Iike someone did a wrong measurement too , or you have the most symetricaI Iistening position i'ever seen.
View attachment 383645

The measurements of the two loudspeakers will of course look different if you don't measure them from the same spot in the room.

The measurements you see of my loudspeakers in this thread are just done to see how they match up pair-wise, that is the only purpose of these measurements. Therefore, the two speakers have taken turns being positioned at the same spot in the room to eliminate any environmental differences that would otherwise occur if they were positioned at different spots in the room.

You can easily try this yourself.
Measure one of your speakers, and mark that exact position on the floor so that you can measure your other speaker from that same spot without changing anything else. If you see large differences between the measurements (no matter where in the frequency response you see them) it means that the two speakers don't match very well, as the only parameter that is changed is the swapping between the two speakers taking turns being measured from the same spot in the room.

Iets wait for measurement from the OP, in the end is far easier to take more measurements than think much about it

The only thing I expect to change in the new measurements is minor changes to the noise floor, and that will not affect the differences in the frequency response of the two UniQ drivers, not even the slightest as the level of the noise floor is way below the level of the fundamental signal.

 

[BrokenEnglishGuy]

I don't see any in-room environment noise problems in the operating range of the UniQ driver in AToMe's measurements. I have not seen anyone else but you who has said that the measurements were done incorrectly.



Maybe you missed it, but AToMe said he disconnected the bass drivers and only measured the UniQ drivers in the speakers. The UniQ drivers operate down to about 200Hz and then they pretty much nose-dive under that, so it's not strange at all if the noise floor happens to be up at 519% at 19.8 Hz which we see the numbers of, as the UniQ hardly makes any noise at all that low in the frequency range.

For the measurement for the other speaker, the noise floor at 19.8 Hz is down at 87%. This change in level is not very strange either as it can be from a passing truck, a neighbor who walked on the floor, or a change in the air conditioning. But no matter what it was, it's not necessarily important as it likely didn't affect the operating range of the UniQ that was measured.



Any "weird" reflection point will be the same for both speaker measurements as long as you are positioning the speakers at the same spot while measuring them, and nothing else is changed between the measurements.



Yes, that is the real.
That is how a well-matched pair of speakers will measure when measured at the same spot in the room, and it will not make a difference if they both are measured at 70 dB, 75 dB, or 85 dB and if they can both handle the same SPL.

Here are the measurements of my ATC SCM11s from almost the same spot in the room where the SCM40s were measured from. The smaller speakers are also a well-matched pair except for some larger deviations above 10kHz. I think most speaker manufacturers are referring to the range from 100Hz to 10kHz when it comes to matched speaker pairs.

View attachment 383692


Sorry, but it's not clear to me what you mean by your EDIT. Did you write something you decided to erase, if so I missed it.



The measurements of the two loudspeakers will of course look different if you don't measure them from the same spot in the room.

The measurements you see of my loudspeakers in this thread are just done to see how they match up pair-wise, that is the only purpose of these measurements. Therefore, the two speakers have taken turns being positioned at the same spot in the room to eliminate any environmental differences that would otherwise occur if they were positioned at different spots in the room.

You can easily try this yourself.
Measure one of your speakers, and mark that exact position on the floor so that you can measure your other speaker from that same spot without changing anything else. If you see large differences between the measurements (no matter where in the frequency response you see them) it means that the two speakers don't match very well, as the only parameter that is changed is the swapping between the two speakers taking turns being measured from the same spot in the room.



The only thing I expect to change in the new measurements is minor changes to the noise floor, and that will not affect the differences in the frequency response of the two UniQ drivers, not even the slightest as the level of the noise floor is way below the level of the fundamental signal.

I mean your way to take in the same position the speaker - the mic is great.
I have some old measurements that i taked for EQ the left and right channel


But at MLP this just doesn't work, there is so many objects in the room. You measurement method for see the MF-HF is optimal, but don't take these at Main Listening Position like i did. Because as you see, the graph can change a lot just by breathing . I also noticed this error at MPL is reduced by using the 85dB output as i said previously. But if you have a quiet enviorement and can move the speaker in the center and then take the measurement very very carefully at 1 meter.. you can measure that correctly.

This is at listening position, i got A LOT of variancy at 2khz+

I dont have any idea why there is a peak at 3khz in the green-blue line, thats UMIK1 error.



For example this is the 2 channels at the same time:

This was for correcting the bass more (another profile), but the HF looks better lol

-------
In resumen you have to do what @goat76 said for seeying correctly the MF-HF behavior

 

[goat76]

@BrokenEnglishGuy

It doesn't matter if you measure the loudspeakers from your MLP or any other spot in the room, and it doesn't matter what part of the frequency response you see the differences in the measurements. As long as the microphone is not moved and the two loudspeakers are measured from the same spot in the room... if you see any differences in any part of the frequency range (bass, mid, or highs), it can only come from a difference between the two loudspeakers as long as that is the only parameter that is changed between the measurements.

 

[BrokenEnglishGuy]

@BrokenEnglishGuy

It doesn't matter if you measure the loudspeakers from your MLP or any other spot in the room, and it doesn't matter what part of the frequency response you see the differences in the measurements. As long as the microphone is not moved and the two loudspeakers are measured from the same spot in the room... if you see any differences in any part of the frequency range (bass, mid, or highs), it can only come from a difference between the two loudspeakers as long as that is the only parameter that is changed between the measurements.

If you are measuring... object in the room cause changes in the sounds, if your objects in the left side aren't the same to the right side, you just can't measure MF-HF at MLP.
I attached the measurements above just for show you the truth in MPL.
For example, due to the objects this part is just imaginary, the others measurements in L&R behavior different as i showed each time i measured them, what im trying to point, you can measure the speaker at your MPL, what you can't is not measure the L&R in differents positions in the room, must be the same position for both speakers in order to measure correctly the MF-HF behavior of both channels
These two measurements just tell you... Nothing.

 

[Crosstalk]

If you are measuring... object in the room cause changes in the sounds, if your objects in the left side aren't the same to the right side, you just can't measure MF-HF at MLP.
I attached the measurements above just for show you the truth in MPL.
For example, due to the objects this part is just imaginary, the others measurements in L&R behavior different as i showed each time i measured them, what im trying to point, you can measure the speaker at your MPL, what you can't is not measure the L&R in differents positions in the room, must be the same position for both speakers in order to measure correctly the MF-HF behavior of both channels
These two measurements just tell you... Nothing.
View attachment 383726
View attachment 383728

I think the OP measured the speakers at the same point as per his initial post. I don’t see any flaw in his method. If the second speaker is positioned at the same spot where the first speaker was measured, the room remains the same. I expect at this point, near identical measurements, given their price(and what I experienced recently when I had a chance to listen to them). To me the one I listened sounded similar to how his right speaker is measuring than the left speaker here ! Anyway OP mentioned they are under warranty so he can be at peace. KEF generally are superb with their service in Europe.

 

[Sokel]

Measuring in a room can be tricky.
This distortion we see down low can very well be some resonating surface other than the speaker.

I would bite the knife (55 Kg is not light but not very heavy either),take them out and measure them for peace of mind.
Or at least change R for L and see.

 

[goat76]

Measuring in a room can be tricky.
This distortion we see down low can very well be some resonating surface other than the speaker.

I would bite the knife (55 Kg is not light but not very heavy either),take them out and measure them for peace of mind.
Or at least change R for L and see.

As mentioned earlier in the thread, OP doesn't need to measure his speakers outside as he just has to make sure they are, one at a time, positioned at the same market-up spot on the floor, and make sure the microphone is not moved between the two measurements. When nothing has changed in the room except swapping out the loudspeakers to the same marked-up spot, any deviations seen between the measurements will indicate differences between the two loudspeakers, no matter what the room may add to the equation as any eventual resonating surfaces are the same between the two measurements.

Remember that these measurements don't have the goal of measuring how this particular speaker model measures, the goal is to determine if the two speakers measure the same as a speaker pair, and if a fairly large deviation is found it could indicate that something is wrong with one of the speakers.

 

[Sokel]

As mentioned earlier in the thread, OP doesn't need to measure his speakers outside as he just has to make sure they are, one at a time, positioned at the same market-up spot on the floor, and make sure the microphone is not moved between the two measurements. When nothing has changed in the room except swapping out the loudspeakers to the same marked-up spot, any deviations seen between the measurements will indicate differences between the two loudspeakers, no matter what the room may add to the equation as any eventual resonating surfaces are the same between the two measurements.

Remember that these measurements don't have the goal of measuring how this particular speaker model measures, the goal is to determine if the two speakers measure the same as a speaker pair, and if a fairly large deviation is found it could indicate that something is wrong with one of the speakers.

That's why I said to change left with right as a last resort,we're saying the same.

 

[goat76]

That's why I said to change left with right as a last resort,we're saying the same.

Ah okay, but as long as the same resonance in the bass shows up in both measurements it should be the room, as the two speakers unlikely have the same problem in the bass. As a last resort, I would drag the 55 kg speakers outside for measurements.

 

[Crosstalk]

Ah okay, but as long as the same resonance in the bass shows up in both measurements it should be the room, as the two speakers unlikely have the same problem in the bass. As a last resort, I would drag the 55 kg speakers outside for measurements.

That's why I said to change left with right as a last resort,we're saying the same.

As far as I can see OP swapped the left and right positions of the speakers during the measurements and this in my book is as good as it needs to confirm if left and right speakers sound the same - which is his problem. At the same spot both should sound identical, whatever peaks or valleys the left speaker produces at the measuring spot, should be replicated by the second speaker. This doesn’t have to confirm with the anechoic measurements of the speaker by the manufacturer. Any deviation between left and right is indeed a manufacturer‘s quality control issue and must be repaired if it’s in warranty. Outside warranty I feel sorry for the OP, that these must cost a fortune to get it fixed. I have first hand experience with KEF and I must say their support is superb, atleast in Europe.