SoundStage’s review of Dutch&Dutch’s 8C

[oivavoi]

Looking at the on-axis response, I’d put the difference at 1.5-2 dB, not 3-4 dB:

View attachment 25070

I too suspect it’s a deliberate voicing decision, since of all measurement methods, true anechoic is least likely to exaggerate the bass response, ie if the designers used ground plane or near field measurements in the design process you’d expect any measurement error to push them the other way.

Interesting.

I agree. In as sophisticated DSP speakers as the 8C's it would obviously be easy to make them totally flat. This aligns with my subjective impression of them btw. On the continuum between "meaty/organic" and "airy/detailed" they lean more towards the meaty/organic side of things, for my ears. This kind of eq/voicing can explain that, as increased treble often creates perceived air and detail, and increased upper bass and low-mids creates an increased perception of weight or "meatiness".

 

[dc655321]

if the designers used ground plane or near field measurements in the design process you’d expect any measurement error to push them the other way.

Can you explain this assertion in a little more depth, please?
Not sure my reasoning on those techniques ends up at the same conclusion as yours...


As @oivavoi says, given the control afforded by DSP here, an anechoically flat output is trivially achievable.
Based on the reviewer's own words, I think one reason for the measurement outcomes could be simple operator error: I got the sense the reviewer did only a minimum of adjustment to the speakers, out of the box.

 

[svart-hvitt]

I don’t fully understand the measurement of NRC (see frequency response of Dutch & Dutch 8C in NRC’s anechoic chamber below).

According to the specifications of Dutch & Dutch 8C, we should expect this:

Source: https://dutchdutch.com/wp-content/uploads/2018/11/8c-Spec-Sheet-2019.pdf

In plain numbers, Dutch & Dutch state this: FREQUENCY RESPONSE 30 HZ – 20 KHZ ± 1 DB (Source: https://dutchdutch.com/8c/).

So the question arises: How did Dutch & Dutch come up with the chart and the frequency response figures on their web site?

Have Dutch & Dutch invented a new way of stating specifications, where they use an estimated room response compensation, especially for the lower frequencies?

Have Dutch & Dutch made a speaker with a big dip in the low frequencies to take into account that the speaker will compensate for this in many rooms (i.e. many rooms, not all rooms, for example damped rooms)?

There has been some confusion among professionals in setting up the Dutch & Dutch speakers, confusion that may not have arisen if people knew about the NRC frequency response measurement, which is so different from Dutch & Dutch’s own version. The confusion arose among users with damped rooms (damped wall behind the speakers).

Again, I also noticed that the frequency response error, even over 100 Hz, is higher than the FR error of +/-1dB that the producer stated. Which begs the question if the specifications from some of the newer speaker producers come from computer calculations, not real measurements.

 

[oivavoi]

I'm 99 percent sure that the dip between 100 and 40 hz is due to some measurement error at the NRC - where the mics didn't capture the output from the rear-facing woofers or something. It makes zero sense that the output drops like a stone below 100 hz (thr crossover to the subwoofers) and picks up again at 40 hz. Having listened to them a lot I can say with very high confidence that they don't display any such below-100 hz dip.

But the response above 100 hz is interesting. I can't see any reason to doubt that those measurements are accurate.

 

[andreasmaaan]

Can you explain this assertion in a little more depth, please?
Not sure my reasoning on those techniques ends up at the same conclusion as yours...

Basically, nearfield and groundplane measurements tend to result in the measurement being slightly boosted in the bass and/or low-midrange.

This is because:

• Groundplane measurement effectively doubles the baffle width
• Nearfield measurement tends to effectively increase the baffle dimensions to infinity

This means that if a speaker’s bass or lower midrange response is off due to a measurement error on the part of the designers’ use of these measurement techniques, it will tend to be too lean rather than too pronounced (assuming flat was their target ofc).

in other words, their groundplane/nearfield measurements will be telling them there is more bass than there really is, so to hit a flat target they will end up dialling in too little bass.

However, I don’t think that would be the case here, hence my suspicion that the slightly elevated bass is deliberate.

Does that make sense?

PS ofc I’m not talking about the apparent dip under 100Hz here - that is a result of the rear-firing woofers being more directional at the top of their passband and is merely an artefact of their orientation combined with the anechoic measurement method, and should therefore be ignored as others have pointed out.

 

[Soniclife]

I don’t fully understand the measurement of NRC (see frequency response of Dutch & Dutch 8C in NRC’s anechoic chamber below).

View attachment 25080

According to the specifications of Dutch & Dutch 8C, we should expect this:
View attachment 25081
Source: https://dutchdutch.com/wp-content/uploads/2018/11/8c-Spec-Sheet-2019.pdf

In plain numbers, Dutch & Dutch state this: FREQUENCY RESPONSE 30 HZ – 20 KHZ ± 1 DB (Source: https://dutchdutch.com/8c/).

So the question arises: How did Dutch & Dutch come up with the chart and the frequency response figures on their web site?

Have Dutch & Dutch invented a new way of stating specifications, where they use an estimated room response compensation, especially for the lower frequencies?

Have Dutch & Dutch made a speaker with a big dip in the low frequencies to take into account that the speaker will compensate for this in many rooms (i.e. many rooms, not all rooms, for example damped rooms)?

There has been some confusion among professionals in setting up the Dutch & Dutch speakers, confusion that may not have arisen if people knew about the NRC frequency response measurement, which is so different from Dutch & Dutch’s own version. The confusion arose among users with damped rooms (damped wall behind the speakers).

Again, I also noticed that the frequency response error, even over 100 Hz, is higher than the FR error of +/-1dB that the producer stated. Which begs the question if the specifications from some of the newer speaker producers come from computer calculations, not real measurements.

Go back to the start of the thread and reread, it's all there.

P.S. When you see a measurement with an * go and read the qualification of that *.

 

[svart-hvitt]

I'm 99 percent sure that the dip between 100 and 40 hz is due to some measurement error at the NRC - where the mics didn't capture the output from the rear-facing woofers or something. It makes zero sense that the output drops like a stone below 100 hz (thr crossover to the subwoofers) and picks up again at 40 hz. Having listened to them a lot I can say with very high confidence that they don't display any such below-100 hz dip.

But the response above 100 hz is interesting. I can't see any reason to doubt that those measurements are accurate.

According to Soundstagenetwork, NRC have measured 233 speakers on the magazine’s behalf (https://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=16&Itemid=18). One can only speculate how many more measurements NRC have done.

So gross incompetence by NRC wouldn’t be the first thing that struck my mind, even if incompetence can never be ruled out.

Wouldn’t you agree that the NRC measurement explains, too, why some engineers at Gearslutz had problems setting them up due to unexpected room interaction behaviour (due to damped wall behind speaker).

 

[Soniclife]

Basically, nearfield and groundplane measurements tend to result in the measurement being slightly boosted in the bass and/or low-midrange.

This is because:

• Groundplane measurement effectively doubles the baffle width
• Nearfield measurement tends to effectively increase the baffle dimensions to infinity

This means that if a speaker’s bass or lower midrange response is off due to a measurement error on the part of the designers’ use of these measurement techniques, it will tend to be too lean rather than pronounced (assuming flat was their target ofc).

in other words, their measurements will be telling them there is more bass than there really is, so to hit a flat target they will end up dialling I’m too little bass.

Does that make sense?

With this speaker you can tell if that it's being used if free space, and it was measured with this setting, so it's down to what they want free space measurement to look like.

At what point do anehonic chambers start to reflect bass and not totally absorb it?

 

[oivavoi]

According to Soundstagenetwork, NRC have measured 233 speakers on the magazine’s behalf (https://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=16&Itemid=18). One can only speculate how many more measurements NRC have done.

So incompetence by NRC wouldn’t be the first thing that struck my mind, even if incompetence can never be ruled out.

Wouldn’t you agree that the NRC measurement explains, too, why some engineers at Gearslutz had problems setting them up due to unexpected room interaction behaviour (due to damped wall behind speaker).

Turns out it wasn't measurement error, only reader error! The NRC published additional measurements of the rear-facing woofers. So the first chart was on-axis from the front, and this is probably from the back.

 

[andreasmaaan]

With this speaker you can tell if that it's being used if free space, and it was measured with this setting, so it's down to what they want free space measurement to look like.

At what point do anehonic chambers start to reflect bass and not totally absorb it?

The NRC’s chamber is said to have a low frequency cutoff of 80Hz, so assuming the NRC is not mistaken about that, it alone doesn’t explain these results completely.

 

[andreasmaaan]

PS just let me clarify, none of what I’m saying here applies to anything below 100Hz. What’s going on down there is 100% to be predicted as a result of the facts that the woofers face rearward and that the speaker is being measured front-on in an anechoic chamber. It is nothing to worry about.

All I’m talking about is the approx. 2dB difference in level between the lower midrange/bass and the treble as shown by my hand drawn red lines in this graph:

 

[Soniclife]

All I’m talking about is the approx. 2dB difference in level between the lower midrange/bass and the treble as shown by my hand drawn red lines in this graph:

If you were designing a speaker where you tell it the distance to the rear and side walls, what cutoff frequency would you stop making changes up to? I haven't got a clue.

The up to 500hz small lift is coincidentally where the rear facing dip corresponds to. Related?

 

[andreasmaaan]

If you were designing a speaker where you tell it the distance to the rear and side walls, what cutoff frequency would you stop making changes up to? I haven't got a clue.

Taking D&D's stated (up to) 80cm boundary placement as a starting point, we would want rear output up to the frequency with a wavelength of up to 4 x 80cm, which is approx. 107Hz. Above that frequency any rearward output will combine destructively with the direct output when it is reflected back, and will be unwanted. No doubt this is why D&D chose 100Hz as the crossover frequency between the rear-facing woofers and the cardioid midrange (or conversely, why they chose 80cm as the maximum boundary distance).

The up to 500hz small lift is coincidentally where the rear facing dip corresponds to. Related?

I don't believe that's a dip in the rear-facing woofers' output per se. I believe that dip - or to be more precise, the lift above 500Hz - in fact reflects the imperfect cardioid behaviour of the front-firing midrange, i.e. that lift above 500Hz that you're seeing is beig splashed back by the front-firing midrange, not the rear-firing woofers.

 

[Juhazi]

svart-hvitt copied the smoothed response that doesn't rhyme with NRC. Manufacturer's directivity plot is very much like NRC (notice X scale)
https://dutchdutch.com/wp-content/uploads/2018/11/8c-Spec-Sheet-2019.pdf

Like said, published measurements are not directly comparable and in most cases measurement conditions and FFT parameters are not told. Open access to all dsp parameters helps final tuning a lot, but it requires a lot from the end user. Every manufacturer must set some presets and some constraints for options.

From my experience I can say that a partly cardioid speaker is not easy to "voice" right in a room, especially if you look at measurements simultaneously! In-room RTA measurement of a single speaker at intended position is the best measurement to evaluate voicing.

To me it looks like 8c's lowest bass is too loud, or there is a mode peak around 20Hz. Decay response shows modes easily. NRC chamber is not anechoic below 80Hz!

 

[dc655321]

Does that make sense?

Understood. Thanks.
Must dig out my copy of D'Appolito's measurement book and refresh my memory on the intricacies.

in other words, their groundplane/nearfield measurements will be telling them there is more bass than there really is, so to hit a flat target they will end up dialling in too little bass.

However, I don’t think that would be the case here, hence my suspicion that the slightly elevated bass is deliberate.

As I said before, with the adjustment capabilities of this speaker, one has a broad range of possible responses that could be dialed in.
These measurements reflect just one of many possibilities, so I'm gonna stick with my "it's a problem between the chair and keyboard" hypothesis (until I learn otherwise) .

Yes, I realize my hypothesis applies equally to myself as well ;-)

 

[HammerSandwich]

PS ofc I’m not talking about the apparent dip under 100Hz here - that is a result of the rear-firing woofers being more directional at the top of their passband and is merely an artefact of their orientation combined with the anechoic measurement method, and should therefore be ignored as others have pointed out.

PS just let me clarify, none of what I’m saying here applies to anything below 100Hz. What’s going on down there is 100% to be predicted as a result of the facts that the woofers face rearward and that the speaker is being measured front-on in an anechoic chamber. It is nothing to worry about.

The 1st quote implies that 8" woofers are significantly directional with >100" wavelengths, which seems unlikely. And the cabinet's 15" depth shouldn't allow the drivers to be significantly out of phase with such long waves, unless the NRC had an inappropriate delay setting in place.

IOW, don't these sub-100Hz issues most likely arise because the chamber is NOT anechoic that low?

 

[Dialectic]

I'm 99 percent sure that the dip between 100 and 40 hz is due to some measurement error at the NRC - where the mics didn't capture the output from the rear-facing woofers or something. It makes zero sense that the output drops like a stone below 100 hz (thr crossover to the subwoofers) and picks up again at 40 hz. Having listened to them a lot I can say with very high confidence that they don't display any such below-100 hz dip.

But the response above 100 hz is interesting. I can't see any reason to doubt that those measurements are accurate.

I don't have reason to doubt their accuracy, but I strongly suspect that some apparent problems in the measurements can be explained away by the unusual design of the speakers. I haven't seen those issues completely explained away in any post above.

Based on all other evidence I have seen and heard, however, I don't think there's any chance that the 8Cs have higher distortion than, e.g., passive KEF LS50s, yet that is what the NRC measurements indicate.

 

[andreasmaaan]

The 1st quote implies that 8" woofers are significantly directional with >100" wavelengths, which seems unlikely. And the cabinet's 15" depth shouldn't allow the drivers to be significantly out of phase with such long waves, unless the NRC had an inappropriate delay setting in place.

IOW, don't these sub-100Hz issues most likely arise because the chamber is NOT anechoic that low?

That's possibly the better explanation, yeh. OTOH, it's not the diameter of the woofer that's the only factor, it's also the cabinet itself, which is no doubt wider and most certainly taller than 8". I'm not sure exactly how tall the cabinet is, but assuming it's about 18-20" which looks about right at a glance, some directivity in this range and possibly some dips due to diffraction effects are a plausible partial explanation IMHO.

But you're right, the dip shown in that graph looks more pronounced than you'd expect.

Conversely, I find it hard to believe the NRC is incorrect in their assessment of their anchoic chamber....

 

[andreasmaaan]

I don't have reason to doubt their accuracy, but I strongly suspect that some apparent problems in the measurements can be explained away by the unusual design of the speakers. I haven't seen those issues completely explained away in any post above.

Based on all other evidence I have seen and heard, however, I don't think there's any chance that the 8Cs have higher distortion than, e.g., passive KEF LS50s, yet that is what the NRC measurements indicate.

The NRC measurements indicate that the LS50s do indeed have higher distortion (LS50 is on the left in the graphs below):

We see similar distortion in the 100-200Hz range, ie the lower range of the 8C's midrange.

This is explicable simply by the fact that the LS50's woofer is in a ported box, which will always give reduced bass distortion vs what is effectively an open-baffle design for the 8C. That is of course balanced out by the fact that the 8C's midrange is larger than the LS50's midwoofer, and no doubt various other factors that we have no knowledge of in these measurements.

But the general point is that it's not surprising that a 5.5" woofer in a ported box has similar upper bass distortion to an 8" woofer in an open baffle.

In terms of what's going on at around 50Hz, this is about the port tuning frequency of the LS50, which is another frequency range where far lower distortion in a ported box vs a sealed box design is to be expected (the 8C's subs are sealed and in a very small box for two woofers that size).

 

[Shadrach]

Seems like an awful lot of money for a mid field active studio monitor to me.