ELAC DF63 Floor standing Speaker Review

[amirm]

Why no double check? and a search for error if there is any? Isn't it this what makes out engagement of EE?

You are seeing that "double check." Erin tested one sample. I could have just decided to not test my sample and have that be. But I told the owner I would. And here you are. Most of the time you only have one review sample measured. Here, you have two, sourced differently.

And why is it up to us to do any "searching?" Why not ask the manufacturer to drive a sample to Klippel to be measured there? Or as a minimum, produce their own measurement data so we are not sitting here guessing? And further, have them play the test clip and see what they hear?

To be clear, to re-test another sample would cost nearly $700 for the speaker and has an opportunity cost of $2,000 and two days lost out of my life. On what basis do folks think they are justified in asking me to do more instead of the company???

 

[Power Pop 23]

ELAC specifies a thickness of 16MM for the MDF used in the enclosures for the Debut 3.0 loudspeakers.

https://elac.com/download/db63/debut-3-sales-sheet_1.0.0.pdf

Is this spec possibly affecting the performance of the sample DF63 measured by Amir ?

 

[Davey]

If this speaker/driver(s) was malfunctioning in some way Amir would have caught it and noted it.

Impertinent questions and comments are not constructive.

 

[amirm]

Is this spec possibly affecting the performance of the sample DF63 measured by Amir ?

The cabinet certainly resonates and a lot of it. But for the price class, it is hard to ding it. It is however possible that it contributed more to measurements like distortion.

 

[dfuller]

If this speaker/driver(s) was malfunctioning in some way Amir would have caught it and noted it.

Impertinent questions and comments are not constructive.

The sheer difference in distortion behavior between the two samples (never mind that socking great hole in the response around 1k on Amir's) that have been klippel'd with public data leads me to believe that 1.) Erin was sent a "golden sample", or 2.) Amir's speaker does not meet spec and is defective.

It's one of the two. Neither are particularly great looks, but for my money it seems significantly more likely this was a QC or shipping oops more than Elac trying to game the review system - because other than the one other speaker that showed some wacky behavior, Elacs are generally (mostly) well designed speakers.

 

[dfuller]

Good find. Assuming it is some kind of in-room, gated measurement, it more closely matches my estimated in-room measurement:

Than Erin's:

I hate to be the scaling police, but these are actually almost identical when you match the sizes.

 

[vroad]

Looking more closely at the dips and peaks on Erin's and Amir's graphs, there seem to be both similarities and differences.

For reasons I'm unsure of, while Erin's On-Axis and Front Wall Bounce graphs are clean around 1kHz, both Ceiling Bounce graphs show issues in that same frequency region.

Focusing on those Ceiling Bounce graphs, below 1kHz, the dip in Erin's measurement is shallower but broader (spanning 800-1000Hz). In contrast, the dip in Amir's measurement is deeper and narrower (covering 800-900Hz).

Furthermore, just above 1kHz, Erin's data shows a smaller but broad peak, whereas Amir's measurement has a deep, narrow dip in the same region.

 

[amirm]

I hate to be the scaling police, but these are actually almost identical when you match the sizes.

They are not the same in bass. There is about 2 dB more energy in mine, better matching the other review.

I should note that by default Klippel NFS under-reports bass for tall speakers where there is multi-radiating sources (woofers and ports). I spent a few hours correcting/optimizing for this. If you don't fully optimize, you will see less energy. Since this is a manual process, it may be the difference between my measurements in bass and Erin's.

 

[Thomas_A]

I am almost confident that this is "avalanche distortion" with lots of higher order harmonics from mid/bass/enclosure. I've measured and heard "hard and nasty" sound in my own experiments which are related to enclosure/damping. Blue below 19 mm MDF (unbraced), red is 9+9 mm MDF with Swedac damping glue MDF (constrained layer). Non-linearities of this kind is very audible on certain music and sound really bad and once you've heard it at high SPL, you also hear it at lower SPL. I posted a thread with sound files here. It is quite dependent on music type and in other situations it becomes masked.

'Speaker enclosure vibrations - part II - ABX wanted!'

Aug 2, 2025

Hi all,

based on a thread at SHF and my previous thread here I posted two recordings of speakers with the same form (see my Avatar). One is built with the usual 19 mm MDF and the other pair with 9+9 mm MDF and damping glue. The recording was made using a 540 Hz signal, where the cabinet vibrations are worst, and ≈80 dB at LP. I also posted an ABX with the files with positive result. Now, it is always difficult to make such tests when files are recorded in-room, due to small variations in the file. There is no response though on the files at SHF and since I know there is larger interests...

• Thomas_A
• Replies: 41
• Forum: Speaker Reviews, Measurements and Discussion

• 

 

[Juhazi]

I'm still learning about measurements, but looking at the dip at 1000hz that shows up in Amir's measurements and not Erin's, I noticed that the resonance doesn't show up in the impedance. Is this because of smoothing?

Acoustic interferences of several drivers and ports of the speaker show up in response plots, these don't show up in impedance (electric domain). Only interferences/resonances inside the cabinet and drivers' motors show up in impedance.

Remember that Klippel NFS scans all around the speaker in nearfield (two layers) and then calculates farfield responses and directivity. Most likely Amir's and Erin's setup use different resolution and calibration, and perhaps also some calculation parameters or coefficients are different.

But differences overall are so huge that Erin's speaker must be fundamentally different from the mass production sample that Amir tested. I think that problems come from the midrange driver.

 

[DanielT]

I only have one speaker. These are sold individually. The owner purchased the speaker and donated it to be tested for the membership. It would have made little sense for him to spend twice as much money.

I suspected that was the case. If not, you would have already tested and compared the two.

 

[amirm]

If someone is interested in screwing around with this speaker in exchange for some donation to the forum, I am happy to ship it to you.

 

[Thomas_A]

Erins measurements show distortion peaks in the same region, but are ≈20 dB lower at the 96 dB level (not so at 86 dB). So it is baked into the model. Why the 20 dB difference is however not possible to find out without more testing and some speaker surgery.

 

[Sokel]

If someone is interested in screwing around with this speaker in exchange for some donation to the forum, I am happy to ship it to you.

Send it to @Ageve !

 

[Thomas_A]

Send it to @Ageve !

Sweden is a long way!

 

[Sokel]

To put things straight, "Fading Sun" ,s cowbells cover about 2kHz to 12kHz (with their harmonics) .

So, it affects both mid and tweeter and if you ask me it's either some crappy cheap driver or there's something amiss with the crossover there.
What Amir describes sound like a dented (by pushing) classic, tweeter damage also.

 

[DanielT]

If someone is interested in screwing around with this speaker in exchange for some donation to the forum, I am happy to ship it to you.

Do what you feel like and have time for. There is so much else you can test and measure than performing time-consuming troubleshooting on a speaker. But that doesn't stop ARS members from speculating in this thread about what you heard and what could be the reason for that. You too are wondering and curious about what it could be.

 

[DanielT]

I am almost confident that this is "avalanche distortion" with lots of higher order harmonics from mid/bass/enclosure. I've measured and heard "hard and nasty" sound in my own experiments which are related to enclosure/damping. Blue below 19 mm MDF (unbraced), red is 9+9 mm MDF with Swedac damping glue MDF (constrained layer). Non-linearities of this kind is very audible on certain music and sound really bad and once you've heard it at high SPL, you also hear it at lower SPL. I posted a thread with sound files here. It is quite dependent on music type and in other situations it becomes masked.

'Speaker enclosure vibrations - part II - ABX wanted!'

Aug 2, 2025

Hi all,

based on a thread at SHF and my previous thread here I posted two recordings of speakers with the same form (see my Avatar). One is built with the usual 19 mm MDF and the other pair with 9+9 mm MDF and damping glue. The recording was made using a 540 Hz signal, where the cabinet vibrations are worst, and ≈80 dB at LP. I also posted an ABX with the files with positive result. Now, it is always difficult to make such tests when files are recorded in-room, due to small variations in the file. There is no response though on the files at SHF and since I know there is larger interests...

• Thomas_A
• Replies: 41
• Forum: Speaker Reviews, Measurements and Discussion

• 

Maybe there could be an explanation but with such a small cavity as the midrange is in:


Plus ELAC specifies a thickness of 16MM for the MDF used in the enclosures for the Debut 3.0 loudspeakers.,#83 with, presumably, damping material in that midrange box. Do you think so? I don't know, maybe your explanation could be true. Nothing we know until it is tested in and of itself.

 

[Thomas_A]

To put things straight, "Fading Sun" ,s cowbells cover about 2kHz to 12kHz (with their harmonics) .

So, it affects both mid and tweeter and if you ask me it's either some crappy cheap driver or there's something amiss with the crossover there.
What Amir describes sound like a dented (by pushing) classic, tweeter damage also.

Was planning to record itand look at the spectrum of this section. Perhaps you have it and can upload a picture

 

[Sokel]

Was planning to record itand look at the spectrum of this section. Perhaps you have it and can upload a picture

REW's RTA mode, actual playback.


before the cowbells



after the cowbells