• 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!

Dutch & Dutch 8c Review

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,243
Likes
11,469
Location
Land O’ Lakes, FL
If I understand you correctly, IMHO it should not be provided then. No reason to “penalize” the speaker due to the equations not taking a high-performer in to account.
Do you have SPL data down to 15Hz, or is ~20.5Hz the lowest you went?
Because ~15Hz is what the “w/ sub” score uses for LFX.
 

Kvalsvoll

Addicted to Fun and Learning
Audio Company
Joined
Apr 25, 2019
Messages
878
Likes
1,643
Location
Norway
I read some concerns about the measured distortion levels, and questions regarding what causes this, and to what extent this affects sound quality.

As I have my own speaker designs using not the same, but somehow similar technology with acoustic ports to control directivity, I have an interest in explaining how this works - what causes this distortion, how does it affect sound quality.

The D&D speaker tested here looks like a very well engineered design, where they made good choices for the final product. What they did, how they did it, and why they ended up choosing they way they did, does not necessarily match any of my own observations or solutions.

My designs are different form the D&D, but they share some common principles, such as the directivity control in the upper bass - lower midrange frequency range using acoustic ports with acoustic resistive damping in those ports, as well as inside the cabinet.

My speakers also show the same behavior with increased distortion with lower frequency, levels may be different, but same tendency.

As for sound quality, see post #1 where the reviewer @hardisj described the sound as excellent, and that there were no indications of high levels of distortion, even when playing very loud.

This corresponds well with my own findings - distortion is not a problem, even at very loud volume. A clue to understanding why this distortion is not as huge a problem as one can initially suspect from looking at the graphs, can be found when looking deeper into the data. Then you see that the distortion is not caused by some hard limit - similar to amplifier clipping or a cone bottoming out, it just gradually increases with increasing spl. And the distortion is (mostly..) dominated by low order harmonics. At higher frequencies this increase in distortion disappears (clue - volume flow velocities decrease..).

As to what causes this, it is a combination of several factors. The distortion from the drivers is not the only source of this distortion, and choice of acoustic damping materials, their placement and internal design of the enclosure all affects the end result. In such small speakers, it may prove to be quite difficult to get rid of completely, and I stopped working on improving this after getting the distortion levels down to something reasonable. After all - there was indeed nothing wrong with the sound.

Future state-of-the-art speakers will all have some sort of directivity control like this, the impact on sound quality is quite significant, especially in rooms with less than ideal acoustic properties.
 

Pritaudio

Active Member
Joined
Nov 7, 2020
Messages
220
Likes
81
Good question which I have no answer for, other than speculation. But you can get an RME ADI-2 DAC fs + Hypex NCore NC502MP based amp for around 2k usd and that leaves some 8k usd for speakers, which is a lot of speakers to buy, until it reaches the price of the D&D 8c.
I suppose adding an external sub to the main stereo channels won’t allow software to produce an eq which will integrate well into a room.
it would be better to use a stereo three way for the eq to be more effective in its objective.
maybe even floorstanders.
 

Absolute

Major Contributor
Forum Donor
Joined
Feb 5, 2017
Messages
1,084
Likes
2,125
Great stuff, Hardisj!

With a special speaker like this I believe we are at a point where the spinorama no longer can be used reliably to determine subjective speaker quality, and the same goes for the preference rating. The horizontal and vertical dispersion plots are likely better suited to tell speaker quality apart when we're dealing with special dispersion patterns.

The lack of vertical cardioide will likely account for a larger degrading factor in the spinorama plot/preference rating than is the case subjectively. In other words the speaker is not rewarded enough for perfect horizontal dispersion and lack of SBIR because it's a novel thing that wasn't tested during the data collection they did.

I believe that this speaker is nearly the best that can be done outside of coaxials for dispersion and I believe the benefits of the cardioide is greater than a preference rating developed for normal forward-firing speakers can tell us.

Acoustic interference from the vents gives raised distortion much like the Kii Threes in the low mids? It would be nice to see how much dsp is required to balance the frequency response of that (if any). @Kvalsvoll ?
 

theyellowspecial

Active Member
Forum Donor
Joined
Sep 21, 2020
Messages
253
Likes
274
Possible to show more real-world distortion measurements maybe? That's my only gripe with ASR reviews. Who listens to their bookshelves at 86dD free-field in their medium-sized rooms (although I do understand the need to have 86dB+ measurements)? ~70dB in-room is loud to me. Maybe I'll go ~75dB for a laugh once in awhile.
 

Alexanderc

Addicted to Fun and Learning
Forum Donor
Joined
Jun 11, 2019
Messages
640
Likes
1,011
Location
Florida, USA
Thanks, this is a great review and an excellent way to make use of your new toy.

Check the 12th bullet point under “parting thoughts.” You’re missing a word I think.
 

Kvalsvoll

Addicted to Fun and Learning
Audio Company
Joined
Apr 25, 2019
Messages
878
Likes
1,643
Location
Norway
Acoustic interference from the vents gives raised distortion much like the Kii Threes in the low mids? It would be nice to see how much dsp is required to balance the frequency response of that (if any). @Kvalsvoll ?

No general rule for how much eq is required, because that depends on the chosen design - cabinet, radiation pattern, drivers used, low cut-off frequency.
 
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
Oh, and just to double check even though I asked a while ago, did you compute the Spinorama or did you manually calculate it? Because to my knowledge the Klippel software still does the incorrect calculation of the Early Reflections & PIR.


Glad you brought that up because I meant to mention it a couple days ago.


I was under the impression that Amir had already told them about the issue. Not sure if I imagined that or if someone said he did. I wondered why it was still not taken care of, though, as it seemed like a long time to not have it taken care of so I sent an email asking about it. Here is the email I sent Christian:
Question about the CEA-2034 Early Reflections/ERDI calculation. I've had a couple people mention that it appears these are incorrect in the Klippel template, based on Amir of ASR's published reviews. I double checked the script and I think I see the error. In the script for the calculation of ER/ERDI the angles used are all averaged together. However, CTA-2034 calls for an "average of averages". I have posted a screenshot below, taken from Appendix F (Post-Processing section). The frustrating thing is even the spec lists this incorrectly in another section (5.2) of the same document. Thus, the confusion.

1614996324162.png


I tried to see if there was a way for me to change this myself and just provide you with the updated template. But I didn't have any luck. I don't know how to take an average of other averages. It may be easy; I just couldn't figure it out.

Thoughts?



And his reply below shows they didn't even know there was an issue. So, I guess that explains why it hadn't already been fixed.

Also thanks for the hint with the early reflection curve. We read this standard so many time, but however we missed this little detail. Why they are not simply putting some equations into the standard? :)

However, we can fix it very easily by adding some weights to the average calculation. I will send you an update for the spread sheet.


I'll update the data once I get the updated template.
 
Last edited:
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
Hmm, Amir has said he has let them know, and they told him how to manually fix it every time.

Got me, man. So Amir’s are all fixed? If so I imagine it’s just the calculation in the template. Easy to access. I can change things for my drive levels (I.e., 0.30 volts instead of 2.83v for a powered speaker). But I wasn’t sure how to do an average of averages with it.
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,243
Likes
11,469
Location
Land O’ Lakes, FL
Glad you brought that up because I meant to mention it a couple days ago.


I was under the impression that Amir had already told them about the issue. Not sure if I imagined that or if someone said he did. I wondered why it was still not taken care of, though, as it seemed like a long time to not have it taken care of so I sent an email asking about it. Here is the email I sent Christian:




And his reply below shows they didn't even know there was an issue. So, I guess that explains why it hadn't already been fixed.




I'll update the data once I get the updated template.
The average of averages isn’t the issue (unless that too), it’s that for the Rear Wall Bounce, it only uses +90°, -90°, and 180°, when it should also be using all rear angles from +/-90° to 180°.

Oh, and no, Amir doesn’t do this manual fix. So all his ER & PIR curves are slightly off. Which is why I post mine.

Also, for powered speakers, are you able to make the graph scale referenced like how I do (-40dB to +10dB)? Like this one shows ~105dB, and no doubt someone gonna ask about that. I should have done this for passive speakers when I started, as then that would make comparisons a lot easier (@edechamps does it for his I believe), but kinda too late know (unless I love self torture).
 
Last edited:

Sancus

Major Contributor
Forum Donor
Joined
Nov 30, 2018
Messages
2,923
Likes
7,616
Location
Canada
Possible to show more real-world distortion measurements maybe? That's my only gripe with ASR reviews. Who listens to their bookshelves at 86dD free-field in their medium-sized rooms (although I do understand the need to have 86dB+ measurements)? ~70dB in-room is loud to me. Maybe I'll go ~75dB for a laugh once in awhile.

Listening levels vary quite widely, but dynamic range is a thing. If you listen at an average level of 70dB, there exists music which will peak at 90dB or even higher in some rare cases. You don't want distortion on peaks. Personally I listen at around 80dB, but there's plenty around here who listen louder. Reference level is 83dBC and you will find people who prefer much louder than that.

In any case, distortion should be at acceptable levels at 86dB for a good speaker. And that would imply lower levels will be fine as well.
 
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
Possible to show more real-world distortion measurements maybe? That's my only gripe with ASR reviews. Who listens to their bookshelves at 86dD free-field in their medium-sized rooms (although I do understand the need to have 86dB+ measurements)? ~70dB in-room is loud to me. Maybe I'll go ~75dB for a laugh once in awhile.

Remember, these tests are referenced to 1m. So, at 4m that’s 12dB down (anechoic). Equal to 74dB and 84dB at 4 meters.
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,243
Likes
11,469
Location
Land O’ Lakes, FL
But I wasn’t sure how to do an average of averages with it.
What is confusing about averaging a set of averages?

Ignoring the conversion of Pascal and Decibel, all you do is =average(average(1),average(2),etc.):
78C6CCD7-8075-4F2F-B163-EACF6806C9F5.jpeg

I already calculate all the Early Reflections, so it could be more compact.

The real pain in the ass is the Sound Power, as that‘s a weighted average. I’ve posted it before, but here is my calculation:
F0925762-8EA6-4F28-A8F1-1288E51345D6.jpeg
 
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
What is confusing about averaging a set of averages?

It’s not in a spreadsheet. I don’t even know what language it is. It’s probably easy for someone who uses it. But even Christian said they’d have to do some weighting (and not a simple average of averages like we know can easily be done in excel).
 

Vintage57

Senior Member
Forum Donor
Joined
Feb 2, 2018
Messages
412
Likes
595
Location
Ontario, Canada
Top Bottom