measurement of Kef 104/2

[edavey]

I'm posting some room measurements of my 'new' Kef 104/2 speakers, recently aquired from 'Spider'.  Unfortunately the previous thread in 'classified' has been archived so I'm now starting a new thread to discuss their measurement and performance in my room.

It took me a while to get hold of a mic, but I've been now been able to borrow a calibrated miniDSP measurement mic and have done some measurements using the REW software package. The first three are waterfall charts of the 104/2:

a) without the Kube
b) with the Kube 
c) with the Kube, with 'extend' on

and then there's a similar chart of my previous speakers (B&W DM2a) which I perceived as less boomy.

I'm a novice at room measurement and I haven't had time to read the REW notes properly so here goes my attempt to interpret the above:

1) the Kube gives the 104/2 a flatter bass response
2) the Kube with the 'extend' gives an even flatter bass response
3) the 104/2 has a significant hump around 60Hz
4) the 35-75Hz range has a much slower decay rate than the other frequencies
5) the slow decay rate of the 35-75Hz range increases as the bass response is flattened by the Kube. i.e. is worse in [2] and worse again in [3]
6) the 'old' B&W DM2a speakers don't have the low bass, have a similar bass hump around 60Hz but the decay is faster.

Is what I'm describing what's referred to as a 'room mode'?

I'd be very grateful for your comments. 

Thanks

Ed

 

[SergeAuckland]

I find those waterfall plots rather harder to interpret . I'm much better with a straight SPL plot, with third octave smoothing. Third octave approximates to the way we hear, so also stacks up better with what we hear the loudspeakers  sound like.

The 60Hz peak  may well be a room mode, especially as it affects both the KEF and the DM2s,  and it shows energy storage effects by virtue of the time over which it goes on. 

S

 

[insider9]

Interesting, thanks for posting. Lovely to look at these. Based on these alone I'd say B&W are doing rather well. Not only decay times look better but they seem to behave better in midrange. It may of course be that they have been measured at lower volume which would make decay times more difficult to compare (It looks as they were).

Obviously these are only a fraction of measurements but very useful. Also looks like you have a room mode around 150 Hz. Is this your usual listening position? Have you tried to move it forward/back a little to see if it helps? It may be as little as 10-20cm that can make all the difference.

One thing I cannot tell is the volume these have been taken at and can't see the Y axis values. Are these screen shots? Did you know REW can produce pictures of graphs?

 

[danilo]

Speaker measurements is an Art Form. So much so that even Manufacturers typically get it wrong.

 A  cheapy (not calibrated that morning) mic  placed 1 meter (why?) from one's home setup driver  is IMO.... same as using an Ouija board: Meaningless.

 Real issue being how do they sound ?? to YOU?

 Human ears are far better measuring devices than any electro mechanical contraptions yet invented.

 Trust to them

 

[edavey]

I find those waterfall plots rather harder to interpret . I'm much better with a straight SPL plot, with third octave smoothing. Third octave approximates to the way we hear, so also stacks up better with what we hear the loudspeakers  sound like.

The 60Hz peak  may well be a room mode, especially as it affects both the KEF and the DM2s,  and it shows energy storage effects by virtue of the time over which it goes on. 

S

Thanks Serge,

I will have a go at producing SPL plots with 3rd octave smoothing as you suggest.

 

[edavey]

Interesting, thanks for posting. Lovely to look at these. Based on these alone I'd say B&W are doing rather well. Not only decay times look better but they seem to behave better in midrange. It may of course be that they have been measured at lower volume which would make decay times more difficult to compare (It looks as they were).

Obviously these are only a fraction of measurements but very useful. Also looks like you have a room mode around 150 Hz. Is this your usual listening position? Have you tried to move it forward/back a little to see if it helps? It may be as little as 10-20cm that can make all the difference.

One thing I cannot tell is the volume these have been taken at and can't see the Y axis values. Are these screen shots? Did you know REW can produce pictures of graphs?

thanks Insider:

- yes they were all done at pretty much the same volume, with REW reporting around "-12dB headroom"

- I did take the measurements at my normal listening position. I'm not very flexible with either that or the speaker placement as it's a busy family room, but I did try one other experiment: measuring just one speaker from a distance of about 2m, and the shape of the waterfall was very similar

- thanks for the tip about exporting graphics direct from REW - will give that a go -- or at least ensure I include the Y axis in future!

 

[edavey]

Speaker measurements is an Art Form. So much so that even Manufacturers typically get it wrong.

 A  cheapy (not calibrated that morning) mic  placed 1 meter (why?) from one's home setup driver  is IMO.... same as using an Ouija board: Meaningless.

 Real issue being how do they sound ?? to YOU?

 Human ears are far better measuring devices than any electro mechanical contraptions yet invented.

 Trust to them

Thanks Danilo,

As the previous thread was archived you may have missed the original context, which is that I got these 'new' speakers home and whilst I (subjectively) liked much about them:

- imaging (sense of music coming from between speakers)

-  sense of 'scale' / depth of soundscape

- detail in mid-range

I felt that the bass was boomy and spoiling the overall effect to an extent.

So what I'm doing now is attempting to confirm the problem -- with a view to exploring possible remedies, of course! If I were to just 'trust my ears' then I don't think there's much to do except conclude that I don't like the way they work in the room and pass them on...

 

[PuritéAudio]

Will the forum allow you to post the MDAT file, then we could all open it and Lok at the various aspects of the measurements?

The room looks pretty typical with as Serge said some standing waves at the lower end, you could experiment with a little EQ, JRiver has PEQ built in, or you could download DIRAC from correction software, and see what they think your room should sound like!

The microphone you are using is more than capable.

Keith

 

[SergeAuckland]

Speaker measurements is an Art Form. So much so that even Manufacturers typically get it wrong.

 A  cheapy (not calibrated that morning) mic  placed 1 meter (why?) from one's home setup driver  is IMO.... same as using an Ouija board: Meaningless.

 Real issue being how do they sound ?? to YOU?

 Human ears are far better measuring devices than any electro mechanical contraptions yet invented.

 Trust to them

Whether one likes them or not is irrelevant if the purpose of the exercise is to understand any problems, and then work out a solution.  That's the purpose of in-room loudspeaker measurement, and as such very important in improving the final result.

I agree with Keith that the microphone used is perfectly adequate for the task, as is the software.

S

 

[henchard]

 Human ears are far better measuring devices than any electro mechanical contraptions yet invented.

Not true.

 

[SergeAuckland]

Not true.

Exactly. The problem with ears, is that usually they have a brain attached, and it's that which is so fallible. 

The brain tries to make sense of what it's being given by the ears,  and that results in all sorts of 'colorations' dependent on the listener's world view.

The hearing has evolved to provide for the continuance of the species, not as an analytical tool, and it's very good at making quick judgements on little evidence, essential for survival, but very bad at deep analysis.

That's why instruments do the job of measurements so much better.

S

 

[henchard]

Exactly. The problem with ears, is that usually they have a brain attached, and it's that which is so fallible. 

And the equipment (ears) starts to fail fairly quickly as well

 

[Lawrence001]

I've got 2 points to add. Firstly I've owned 3 pairs of DM2s and 2a's, and 4 pairs of 104/2s over the years. I'm not trying to show off or anything, but it's important as aging components mean you can come to the wrong conclusions if you base your conclusions on a vintage speaker on a single bad pair. Firstly the overall sound of the DM2 is phenomenal given their age, in my opinion they represent the peak of loudspeaker design of their time. Up until getting my first pair I had thought that Kef were at the forefront of domestic speaker design, at least in the UK. With regards to bass, I found it sounded subjectively more natural than the several pairs of quarter wave TL speakers I have owned,if not as deep. I don't know what it is about the patented 1/8 TL of the DM2s but it works very well.

Having said that, I never had an issue with the bass from any of my 104/2s, treble yes, probably a function of the ferrofluid drying and ageing capacitors. I thought the coupled cavity approach was another good solution to producing natural if not extended bass. I would imagine that any issues are definitely room related, which brings us on to..

"If I were to just 'trust my ears' then I don't think there's much to do except conclude that I don't like the way they work in the room and pass them on..." Don't give up yet, you might find that moving them back and forward, and side to side by just a few inches can have surprisingly beneficial effects on room interaction. As can moving the listening seat back and forward slightly as you can find yourself sitting in a peak. If you're against the back wall, this is often the case, but either way try moving forwards (and back if not against a wall) a foot and see what happens.

Sent from my NEM-L51 using Tapatalk

 

[insider9]

Ears... is not just that we all hear different. It's that the same person perceives what's happening massively different due to volume alone. That's my biggest issue with listening tests. Worth adding I'm not saying here, don't trust your ears. But if you need to measure using an actual tool helps. There's no better way to understand what's happening in the room than to measure it, so good start there.

I'm currently working on a room treatment for my listening room so quite happy to contribute what I've learned. We'd need an extensive info about the room itself and what physically you'd be prepared to do. Especially dealing with bass if you can't move the listening position and/of speaker will not be small in size.

I would say this though. Like mentioned above, if you're able to play .mdat file that would be ideal. If not, if you could please repost the waterfalls using REW taken pictures and adjust the Time Range in ms to maybe 500 ms that would be also helpful.

 

[edavey]

Will the forum allow you to post the MDAT file, then we could all open it and Lok at the various aspects of the measurements?
The room looks pretty typical with as Serge said some standing waves at the lower end, you could experiment with a little EQ, JRiver has PEQ built in, or you could download DIRAC from correction software, and see what they think your room should sound like!
The microphone you are using is more than capable.
Keith

Thanks Keith,

I'm travelling at the moment but yes I should be able to make the REW file available and it would be great to get your input.

Regarding correction systems I do use Jriver but also lots of other analogue sources so if DSP is possible / necessary, I would be wanting something I can put in between pre and power amps.

I think miniDSP offer a box like that. I will look into Dirac too.

 

[edavey]

I've got 2 points to add. Firstly I've owned 3 pairs of DM2s and 2a's, and 4 pairs of 104/2s over the years. I'm not trying to show off or anything, but it's important as aging components mean you can come to the wrong conclusions if you base your conclusions on a vintage speaker on a single bad pair. Firstly the overall sound of the DM2 is phenomenal given their age, in my opinion they represent the peak of loudspeaker design of their time. Up until getting my first pair I had thought that Kef were at the forefront of domestic speaker design, at least in the UK. With regards to bass, I found it sounded subjectively more natural than the several pairs of quarter wave TL speakers I have owned,if not as deep. I don't know what it is about the patented 1/8 TL of the DM2s but it works very well.

Having said that, I never had an issue with the bass from any of my 104/2s, treble yes, probably a function of the ferrofluid drying and ageing capacitors. I thought the coupled cavity approach was another good solution to producing natural if not extended bass. I would imagine that any issues are definitely room related, which brings us on to..

"If I were to just 'trust my ears' then I don't think there's much to do except conclude that I don't like the way they work in the room and pass them on..." Don't give up yet, you might find that moving them back and forward, and side to side by just a few inches can have surprisingly beneficial effects on room interaction. As can moving the listening seat back and forward slightly as you can find yourself sitting in a peak. If you're against the back wall, this is often the case, but either way try moving forwards (and back if not against a wall) a foot and see what happens.

Sent from my NEM-L51 using Tapatalk

Thks Laurence,

The ferrofluid was recently replaced and I will have to assume that the speakers are working properly -- which I think the measurements support.

I will have a go at moving the listening position and speaker placement but I don't have too much flexibility there in terms of permanent positioning!

 

[SergeAuckland]

Thks Laurence,

The ferrofluid was recently replaced and I will have to assume that the speakers are working properly -- which I think the measurements support.

I will have a go at moving the listening position and speaker placement but I don't have too much flexibility there in terms of permanent positioning!

If you can get one 'speaker into the middle of the room, raised up on a stand of some sort so that the tweeter is about half-way between floor and ceiling, you can use REW or ARTA to measure the loudspeaker's pseudo anechoic response down to about 200Hz with 2.4m ceiling, lower if the distance to any boundary, (wall or floor or ceiling) is more than 1.2m. 200Hz is good enough to show the response of the mids and tweeter, and show that the tweeter is operating correctly. 

Repeat for the second loudspeaker and you can check that they match at MF and HF. 

LF is much more room and room position dependent, and that can be checked as an in-room response rather than as a specific loudspeaker measurement independent of the room.

That's how I set up mine, adjusted the equaliser to get the loudspeakers' response flat anechoically, then checked that the in-room response is acceptable. The in-room response will always be worse than the loudspeaker's own response.

S 

 

[edavey]

If you can get one 'speaker into the middle of the room, raised up on a stand of some sort so that the tweeter is about half-way between floor and ceiling, you can use REW or ARTA to measure the loudspeaker's pseudo anechoic response down to about 200Hz with 2.4m ceiling, lower if the distance to any boundary, (wall or floor or ceiling) is more than 1.2m. 200Hz is good enough to show the response of the mids and tweeter, and show that the tweeter is operating correctly. 

Repeat for the second loudspeaker and you can check that they match at MF and HF. 

LF is much more room and room position dependent, and that can be checked as an in-room response rather than as a specific loudspeaker measurement independent of the room.

That's how I set up mine, adjusted the equaliser to get the loudspeakers' response flat anechoically, then checked that the in-room response is acceptable. The in-room response will always be worse than the loudspeaker's own response.

S 

Thanks Serge,

I will have go at the anechoic measurement as you suggest, though it sounds like a recipe for a hernia!

In the meantime here are the earlier measurements presented as SPL with 1/3 octave smoothing as suggested. What do these reveal in your opinion?

Thanks,

Ed

 

[edavey]

Will the forum allow you to post the MDAT file, then we could all open it and Lok at the various aspects of the measurements?

for Keith and anyone else who is interested at having a look at these measurements in the raw, you can download the REW (mdat) file here:

http://www.veryreal.co.uk/measurements_june_17_edited.mdat.zip

Thanks a lot!

Ed