• 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!

Is it correct to correct the dip at crossover?

skyfly

Active Member
Joined
May 1, 2020
Messages
205
Likes
54
The amplitude-frequency response curves (near on-axis only, far off-axis only, or both on-axis and far off-axis) of a speaker may have a dip around the crossover frequency.

Is it correct to eliminate the dip by EQing?

As an example, let us see 2019 Audioholics B&W 603 review.

The frequency response curves have dips around 2 kHz both on-axis and off-axis.
image_preview2


Subjectively, however, the speaker sounded okay to the reviewer.

Another complaint I would list is that there is a dip in the response centered around 2 kHz, however, this might only be an academic complaint since I didn’t hear anything missing in this upper midrange band during actual listening.
(Audioholics Bowers & Wilkins 603 Tower Speaker Measurements & Conclusion )


If we EQed to flat response curve at 2 kHz, we might hear emphasized 2 kHz.
 
D

Deleted member 2944

Guest
You've worded your question interestingly enough. :)

In this particular case where you have a fairly consistent dip both on and off axis, I'd say the answer should be a qualified "yes." But realizing the corrected on-axis result won't necessarily translate into the off-axis measurement.
However, this dip is in a sensitive range and there's no guarantee the audible result would be preferable. It might be less preferable. :)

Dave.
 

puppet

Senior Member
Joined
Dec 23, 2020
Messages
446
Likes
284
The dip would indicate to me that the tweeter's polarity isn't correct. Try reversing the tweeters polarity and observe the dip.
 
D

Deleted member 2944

Guest
The dip would indicate to me that the tweeter's polarity isn't correct.
Doubtful. The xover frequency of this system is listed at 4khz, so you'd expect to see a dip near that frequency if the polarity were incorrect.

Dave.
 

Soniclife

Major Contributor
Forum Donor
Joined
Apr 13, 2017
Messages
4,500
Likes
5,417
Location
UK
My speakers have a similar dip.
https://www.stereophile.com/content/sonus-faber-guarneri-evolution-loudspeaker-measurements

At the listening position, measured using MMM the dip looks bad, I corrected it so it wasn't there with measurements, it wasn't a massive difference, but after a short time of listening I thought it was too much, so I backed it off till I found the balance I preferred. Ended up slightly less than half the full correction. You need a feedback loop of measuring, correcting, and listening.
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,244
Likes
11,479
Location
Land O’ Lakes, FL
SoundStageNetwork.com | SoundStage.com - NRC Measurements: Bowers & Wilkins 603 Loudspeakers
Saw this as well. So much for 48hz-28khz +-3db.
Something isn't right with the xo. Unless B&W used a mid with an almost 10db dip right in its passband.
... and it gets good reviews to boot.
The “W” is their signature sound (maybe not anymore now that Sound United owns them, their recent gen states crossover improvements).

But yes, if the on & off-axis are similar to a degree, then feel free to EQ. If they are not, then you gotta play a balancing game.
 

dualazmak

Major Contributor
Forum Donor
Joined
Feb 29, 2020
Messages
2,823
Likes
2,951
Location
Ichihara City, Chiba Prefecture, Japan
I usually (sometimes) check frequency response in digital level (of course after digital software crossover), and then line level in output from the DACs, and also in line level (ex. pre-out and/or headphone out) from the amplifiers, and then finally the real sound from the SP drivers using measurement microphone; just like as I recently shared here in my multichannel multi-amplifier project.
 
Last edited:

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,728
Location
Prague
I think the room acoustic will determine how this speaker would sound. I do not think it is a good idea to try to change speaker directivity by using the EQ. If you like the speaker sound, keep it, if not, get another one. But do not fix the directivity by EQ.
 

dualazmak

Major Contributor
Forum Donor
Joined
Feb 29, 2020
Messages
2,823
Likes
2,951
Location
Ichihara City, Chiba Prefecture, Japan
I think the room acoustic will determine how this speaker would sound. I do not think it is a good idea to try to change speaker directivity by using the EQ. If you like the speaker sound, keep it, if not, get another one. But do not fix the directivity by EQ.

I fully agree and like your stance and approach. I also believe "speaker plus room/environment" is our HiFi "music instrument".

One analogy would be like this; if you have large enough listening room (or a hall, and enough money also) in your home, we can bring Boesendorfer Concert Grand Piano, Steinway & Sons Concert Grand, and/or Yamaha CFX Concert Grand in the room for your decision of selecting one of these. You may choose whichever you "like the sound" not by how differently the sound of the pianos are objectively measured; and never to tune the piano by yourself to change its own unique characteristics (unless you were a world top ranked piano tuner).  
 
  • Like
Reactions: pma
D

Deleted member 2944

Guest
SoundStageNetwork.com | SoundStage.com - NRC Measurements: Bowers & Wilkins 603 Loudspeakers
Saw this as well. So much for 48hz-28khz +-3db.
Something isn't right with the xo. Unless B&W used a mid with an almost 10db dip right in its passband.
... and it gets good reviews to boot.
Whether the crossover could be better optimized is beside the point.
The OP's query was a general one regarding EQ. If a dip like this is consistent and the primary objective is to get the speaker to 'measure' better in free-field conditions, then EQ'ing the dip would be perfectly fine.

For this particular speaker we're assuming that B&W didn't design in this dip intentionally. They might have. :)

Dave.
 
Last edited by a moderator:

thewas

Master Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
6,755
Likes
16,214
Correct, the B&W presence dip on their after 2000 loudspeakers is a part of their characteristic euphonic voicing, if someone doesn't like it he is usually better off with different loudspeakers as just flilling it with a PEQ can make it sound worse due to the directivity jump.
 

MrPeabody

Addicted to Fun and Learning
Joined
Dec 19, 2020
Messages
657
Likes
942
Location
USA
That dip certainly does look like phase mismatch between the two drivers. It may or may not be relevant, depending on how restrictively you interpret the question.

EQ is generally fine so long as it is done well and so long as it is not being done in the vain expectation of fixing a directivity problem. In this particular case, the dip at frequency just below 2 kHz is present in both the on-axis response and the off-axis response. As such it is entirely appropriate to use EQ to level the on-axis response. But only the on-axis response. The 3D plot indicates that this dip is much more severe in the off-axis response. As such, a dip will remain in the off-axis response, after the on-axis response has been made flat.

But there is another question that also needs to be considered, this being the obvious question of whether the filtering will be sharp enough to make things much better.

To answer the question of whether it makes good sense to use EQ to fix this, I think it is essential to look also at whether there may be another approach that might be easier and more effective. If I were faced with a case similar to this one, where there is a fairly severe notch at the crossover point, I would certainly experiment with reversing the tweeter polarity before taking the EQ route. I don't see any reason why this advice would be missing the point. When the question is, "Is it prudent to do this?" the question implicitly solicits alternatives that might be preferable.
 

dualazmak

Major Contributor
Forum Donor
Joined
Feb 29, 2020
Messages
2,823
Likes
2,951
Location
Ichihara City, Chiba Prefecture, Japan
That dip certainly does look like phase mismatch between the two drivers. It may or may not be relevant, depending on how restrictively you interpret the question.
...
...
To answer the question of whether it makes good sense to use EQ to fix this, I think it is essential to look also at whether there may be another approach that might be easier and more effective. If I were faced with a case similar to this one, where there is a fairly severe notch at the crossover point, I would certainly experiment with reversing the tweeter polarity before taking the EQ route. I don't see any reason why this advice would be missing the point.

Thank you, really interesting; just for your (our) reference, in my YAMAHA NS-1000, also in YAMAHA-NS1000M, they (YAMAHA people) intentionally built them with "reverse (inverted) polarity" connection to Be-tweeters, as I shared here.
 

MrPeabody

Addicted to Fun and Learning
Joined
Dec 19, 2020
Messages
657
Likes
942
Location
USA
Thank you, really interesting; just for your (our) reference, in my YAMAHA NS-1000, also in YAMAHA-NS1000M, they (YAMAHA people) intentionally built them with "reverse (inverted) polarity" connection to Be-tweeters, as I shared here.

My understanding is that when tweeters are wired in polarity opposite to the woofer, the reason is to compensate for a 180-degree difference in phase between the two drivers, caused by the crossover slopes. If Yamaha used reverse polarity with these speakers, I expect that the reason was that phase mismatch was introduced by the crossover slopes and that reverse polarity was needed to correct the phase mismatch so as to avoid wave cancellation in the horizontal plane in which the speaker rests (and for a good amount of degrees above and below that plane), thereby avoiding the sort of steep notch that is evident in this B&W speaker.
 

dualazmak

Major Contributor
Forum Donor
Joined
Feb 29, 2020
Messages
2,823
Likes
2,951
Location
Ichihara City, Chiba Prefecture, Japan
Hello MrPeabody,

Thank you again for your nice insights. Yes, I agree with you.

As you may well aware, NS-1000 and NS-1000M has woofer, Be-squawker and Be-tweeter. I believe the still amazing Be-squawker of magically low distortion and the "inverted" Be-tweeter are/were really nicely aligned and cross-overed.

In my multichannel multi-amplifier project with NS-1000, therefore, I have been trying to preserve/simulate Yamaha's miracle design as much as possible while eliminating the LC-network and driving the SP drivers directly/dedicatedly by each of the multiple amplifiers.

I am sorry that my posts here are now becoming somewhat out of the scope from the owner of this thread... After having these nice communication with you, I assume better to stop posting here now.
 

puppet

Senior Member
Joined
Dec 23, 2020
Messages
446
Likes
284
Whether the crossover could be better optimized is beside the point.
The OP's query was a general one regarding EQ. If a dip like this is consistent and the primary objective is to get the speaker to 'measure' better in free-field conditions, then EQ'ing the dip would be perfectly fine.

For this particular speaker we're assuming that B&W didn't design in this dip intentionally. They might have. :)

Dave.
EQ wouldn't be my first move. I've done some DIY loudspeakers and that dip is characteristic of phase misalignment. If that was intentional, which I doubt, I'd like to better understand the reasoning for it 'cause it sounds a little like "sales speak" :). Accentuating the range around 4khz brings with it pronounced issues with vocal rendering IMO.

The measurements in the link I posted were done in an anechoic environment. They back up the link in the OP. Maybe somebody ordered the wrong parts for the XO at the factory or assembled them incorrectly. Either way, these wouldn't be on my must have list.

Regards, Phil
 
D

Deleted member 2944

Guest
I understand EQ wouldn't be 'your' first move, but that's not what the OP asked about.

You seem to be having a discussion with yourself on how this particular speaker (B&W 603) could be 'fixed'. (Assuming it needs fixing.)
Or that maybe it emerged from the factory defective in some way and it needs to be troubleshot. You might be right in either or both cases.

But regardless, the OP's question was not primarily about this speaker, but rather a more general question. But since he hasn't returned to elaborate, the simplest answer to his question is Yes.

Dave.
 

MrPeabody

Addicted to Fun and Learning
Joined
Dec 19, 2020
Messages
657
Likes
942
Location
USA
I do not see any problem with puppet's advice that in a case similar to this one, experimentation with the tweeter polarity is indicated.

Any question of the sort, "Is it prudent to do this?" implicitly elicits alternatives that might be preferable. I can't think of any reason why anyone would think otherwise. Even if the question had been worded thus: "Is there any reason not to apply EQ to correct this notch in the ON-axis response?" the most reasonable answer would not be simply "No", because this answer would be tantamount to wrong advice given that there is manifest reason for the suggestion that it would be better to first see what effect is had by reversing the tweeter polarity.
 
Last edited:
D

Deleted member 2944

Guest
I do not see any problem with puppet's advice that in a case similar to this one, experimentation with the tweeter polarity is indicated.
I didn't say anything about not trying reversing the tweeter polarity, did I? :)
What I did say is that I don't believe this particular dip is associated with a possible wiring issue.....because it's not at the crossover frequency. The info I found on this speaker indicates a xover frequency of 4khz. (Which could be wrong I suppose.)

I do think it would be handy if when folks test speakers like this with separated crossovers they test each section on its own. In that way you get a better insight on the type of crossover design, the xover frequency, etc, etc.

Dave.
 
Top Bottom