Dynaudio Core 47 Review (Professional Monitor)

[dfuller]

The reason they 'love' them is not because they sound good, it's because it's a known quantity, its familiar. Some say they sound so shite that if you can listen to a bad recording through NS10's you can pretty much tolerate it anywhere else

Speaking to this, NS10s sound absolutely rotten, but they're a good "torture test" for a mix. Nobody in their right mind does all of their mixing on them - they're decidedly not enjoyable speakers to listen to long-term.

That said, I've been debating picking up a pair because if KH310s have a weak spot it's the upper mids being "soft", right where NS10s have a big hump.

 

[Somafunk]

Not really relevant to this thread but I’ve just ordered a pair of Dynaudio LYD 48’s, would have liked to get Core 47’s or Hedd type 20 mk2 but can’t justify £4K on speakers so I settled on these, should be here by the end of the month. Ive lived with my current KRK setup for a year and have been very happy with how it sounds but I’ve always wanted to hear a 3way, especially now my room is fully treated.
I‘ll try them with my KRK 10s sub and without and if I feel the sub is beneficial then I’ll get the matching Dynaudio 9s sub, or perhaps the 18s if my treated room/wallet can handle it.

 

[SpanishPassion]

I asked them to review the thread and comment

I hope they will comment. I don't expect them to say a lot of meaningful things, but I really hope they'll get the feedback and, perhaps, will consider to update their philosophy when designing new models.

 

[VenVile]

As a huge Dynadio fan, I'd be lying if I said I'm not disappointed by this. And I've been waiting on a LYD 48 review for the longest, as I've heard great things about it. Anyone with the LYD 48, please consider sending it to Amir so he can review it. It would make for a great comparison between the LYD 48 and Core 47.

 

[Erizo]

Hi Noel,

Thanks for your patience. We’ve been following the thread at Audio Science Review with interest as it develops and are glad to read from your comments in there that you’re happy with your Core 47s.

Having extensively studied the measurements given by Amir at the beginning of the thread and compared them with our own measurements (both those taken during the development of Core 47, as well as some made more recently) we’re left somewhat puzzled by his numbers and the methodology behind them.

The results presented unfortunately do not agree with our own measurements, or indeed with our real-world subjective perception (and that of countless other satisfied end-users) of the performance of Core 47 or its capacity to perform its intended function – ie, a premium professional studio monitor.

While this is of course disappointing, our main concern is that you are happy with your purchase of the Core 47s and continue to be so.

Any further commentary or discussion on our part of the measurements and interpretations presented will likely be taken directly with Audio Science Review, but we’d like to thank you for bringing this discussion to our attention.

Happy listening!




Med venlig hilsen / Best regards / Mit freundlichen Grüssen
Nicolas Cherencq

 

[dfuller]

Hi Noel,

Thanks for your patience. We’ve been following the thread at Audio Science Review with interest as it develops and are glad to read from your comments in there that you’re happy with your Core 47s.

Having extensively studied the measurements given by Amir at the beginning of the thread and compared them with our own measurements (both those taken during the development of Core 47, as well as some made more recently) we’re left somewhat puzzled by his numbers and the methodology behind them.

The results presented unfortunately do not agree with our own measurements, or indeed with our real-world subjective perception (and that of countless other satisfied end-users) of the performance of Core 47 or its capacity to perform its intended function – ie, a premium professional studio monitor.
Med venlig hilsen / Best regards / Mit freundlichen Grüssen
Nicolas Cherencq

I'm interested to see what comes of this, to be honest.

 

[VenVile]

I'm interested to see what comes of this, to be honest.

Likewise. Is it possible that the unit sent t Amir could have been faulty? I think that's possible. Is it likely? Maybe not.

Could Amir have fudged the measurements with error on his part? Possible, but not very likely, given the fact he's been doing this for quite some time.

I want to give Dynaudio the benefit of the doubt here, solely because of their reputation and these being high-end monitors. I'm waiting to see how this develops, and if Amir can get his hands on the Core 59 or maybe even the LYD 48, for comparison. If the Core 59 shows similar results, then maybe he might be on to something.

 

[monkeyboy]

Can't they just send a unit out for an independent 3rd party measurement...

 

[dominikz]

Having extensively studied the measurements given by Amir at the beginning of the thread and compared them with our own measurements (both those taken during the development of Core 47, as well as some made more recently) we’re left somewhat puzzled by his numbers and the methodology behind them.

I wonder what is it they find puzzling about the methodology? In any case, I'm really interested to see how this develops - kudos to them if they are willing to follow-up and investigate.

 

[xaviescacs]

Looks like you may have achieved quite something @Erizo! Congrats and thanks.

 

[blutfink]

Impulse response and frequency response are the same thing. One is measured in time domain and the other in frequency. They are directly transformed to each other via Fourier Transform. A certain FR will always have the same IR, or visa versa.

Some misunderstandings here. What you say is only true for complex-valued responses. The frequency responses we are considering here are magnitudes only, so any phase information has been removed. Thus, the impulse response is not given at all, it is completely ambiguous.

Secondly, what Exprymer likely meant to say is that optimizing for a certain frequency response may imply sacrificing phase behavior, which is a reasonable assumption, true for many filter designs.

 

[amirm]

Any further commentary or discussion on our part of the measurements and interpretations presented will likely be taken directly with Audio Science Review, but we’d like to thank you for bringing this discussion to our attention.

I welcome any communications with them on this. Please pass it on to them. They can email me using Contact Us link at the bottom or join the forum and create a conversation with me. Thanks.

 

[audio2920]

Don't suppose anyone has one of these in a studio or reasonable listening room they would be good enough measure in-room? (Even if you don't want to show us your room issues lower down the spectrum, ha!) Curious to see if FR happens to come out better or worse than predicted from like 1k to 7k ish.

I know normally the anomalies often track pretty well from "measured on the bench" to "measured in an arbitrary room" but ya know, just curious on this particular speaker...!?

 

[audio2920]

I have another question; about waterfall plots... (Picking on Core 47 and 8351B and posting in this thread as they're in families of speakers I'm still considering as a purchase!)

Dynaudio Core 47:

Genelec 8351:

(1) Is 0 on the Z axis where the impulse would be? Or are these not aligned like that? (i.e. time isn't absolute)

(2) Am I right in assuming the levels are arbitrary and it's just relative decay we should be looking at? Looks like the Genelec starts at around 95dB and the Dynaudio around 85dB (but maybe that's because on the Dyn's graph the first slice is "later" even though the time value is earlier, as per (Q.1)?)

(3) In any case, I'd say while the Dynaudio does have more resonances than the Genelec, the "fat-ness" of the graph below about 800Hz is radically different. Assuming more is worse, the Genelec looks worse. Assuming scaling isn't throwing me, does anyone have any thoughts on whether this might mean the Genelec isn't as "tight" as the Dynaudio around 120 to 800Hz?

I imagine at low frequencies having a few ms of decay isn't an issue as the wavelengths are long compared to such decay, and I'd expect decay to increase with wavelength, but in that upper bass / low-mid band I'd have thought it may be audible, smearing in the time domain?

Thanks!

 

[Flak]

Some misunderstandings here. What you say is only true for complex-valued responses. The frequency responses we are considering here are magnitudes only, so any phase information has been removed. Thus, the impulse response is not given at all, it is completely ambiguous.

Secondly, what Exprymer likely meant to say is that optimizing for a certain frequency response may imply sacrificing phase behavior, which is a reasonable assumption, true for many filter designs.

Let me quote this!
While Sarumbear is 100% correct when stating that the frequency response can be directly transformed via Fourier Transform, what we are seeing here is the magnitude response instead which completely removes any information about the behavior in the time domain... in other words, a certain frequency may even arrive 10 seconds later than another but that will not show in the magnitude response.
It would be interesting to see if the behavior of this speaker is better or worse than competitors...

 

[sarumbear]

Let me quote this!
While Sarumbear is 100% correct when stating that the frequency response can be directly transformed via Fourier Transform, what we are seeing here is the magnitude response which completely removes any information about the behavior in the time domain... in other words, a certain frequency may even arrive 10 seconds later than another but that will not show in the magnitude response.
It would be interesting to see if the behavior of this speaker is better or worse than competitors...

10 seconds is delay, less than 20mS is timbre change. Where do you see that long delay?

 

[Flak]

10 seconds is delay, less than 20mS is timbre change. Where do you see that long delay?

Nowhere of course
I was just outlining that the magnitude response completely removes the information about the time domain... in practice, we'll see miliseconds as you mention

 

[sarumbear]

Nowhere of course
I was just outlining that the magnitude response completely removes the information about the time domain... in practice, we'll see miliseconds as you mention

Then why have you said this:

...what we are seeing here is the magnitude response which completely removes any information about the behavior in the time domain...

 

[Flak]

Then why have you said this:

I'm sorry, I don't follow you... my two sentences state the same, am I missing something?

 

[sarumbear]

I'm sorry, I don't follow you... my two sentences state the same, am I missing something?

You said what you seeing here (which I presume the waterfall graph) "the magnitude response which completely removes any information about the behavior in the time domain." In other words the waterfall chart above shows that FFT is not working. That is what I understand.