• 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!

JBL HDI-3800 Floorstanding Speaker Review

OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
FWIW, it took me (6) rounds of testing the center channel to get it to have proper LF measurements. I had to move the mic to the furthest r-axis extension so I could get it to give me a proper fitting error above 30Hz. Earlier measurements had issues as high (low?) as 50Hz.

This is the third test of said center channel (don't have the first 2 on my desktop):

1616699559155.png





This is the 6th test of the center channel:
1616699620904.png




As you can see, the difference in the fitting error between these two isn't huge but it's significant enough on the LF to make a difference. The HF error is again due to the complexity and is "allowable" given the design complexity (waveguide in the center flanked by a 2.5-way mtm design).
 
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
I suspect these issues are one reason (maybe the primary reason?) you rarely test floorstanding speakers. I have spent weeks measuring these two speakers (the HDI-3800 and HDI-4500). A bookshelf would take me 2 hours. And a bookshelf won't throw out my back. :D
 
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
and in case anyone missed my comment above...

And, realistically, even if we are concerned with that bit of data it's, what... 25Hz? The response here is so far down that I consider it inconsequential to the overall performance of the speaker itself. We are truly picking nits at this point. I just think we are going to cause people to "throw the baby out with the bathwater" here.

I never mentioned the 20Hz region in my review. Why? Because the response begins rolloff at 60Hz. And the fitting error is extremely low there, thus indicating no issues in the ability to measure. 20Hz is the last thing I'm concerned with. So, we are back to a (potential) area of concern at 25Hz. Much ado about nothing, IMHO. Especially when you consider anechoic chambers are inaccurate below 100Hz.

That's my $0.02 on the matter.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
I suspect these issues are one reason (maybe the primary reason?) you rarely test floorstanding speakers.
It is not. It simply is a supply issue.
 

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,109
Likes
8,420
Location
NYC
So the Klippel NFS might not be perfect? Buahahah! *Laughs as he clutches his Umik-1*

(meanwhile, struggled just to measure the teensy little HDI-1600).

By the way, not meaning to add more fuel to the debate, but here's the harman data (why so smoothed??) from Speakerdata2034:

1616699736228.png


I don't know where they got the spin from, but it's crazy smoothed though. That has to be like 1/3 octave. Sighs.

The comparator does show a lot more bass in harman's measurement, which tracks audioholics' ground plane fairly closely. *shrugs* I've come to believe bass measurements just depend on what the bass fairy feels like granting you with on a given day.

Snag_94c627a.png
 
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
It is not. It simply is a supply issue.

Noted.


Well, I will say that I didn't go looking to test this speaker. I got an email out of the blue from JBL's marketing team to ask if I would be interested in testing this pair and the 4500 center channel so I said sure. I knew full-well I may very well have to resort back to my "old ways" of measuring but luckily the NFS provided what I needed after a steep learning curve. I haven't heard back from JBL on the review (I didn't share with them any of this before I published it). I'll be interested to see what they have to say.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
And, realistically, even if we are concerned with that bit of data it's, what... 25Hz? The response here is so far down that I consider it inconsequential to the overall performance of the speaker itself. We are truly picking nits at this point. I just think we are going to cause people to "throw the baby out with the bathwater" here.
I made all these arguments in my reviews. They didn't work. :)
 
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
I made all these arguments in my reviews. They didn't work. :)

I know. I've been doing reviews for a long time now. I always had to take a step back from the data and talk about what makes sense. Sometimes we get so stuck in the data mindset that we forget to just use plain ol' logic.
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,240
Likes
11,462
Location
Land O’ Lakes, FL
Last edited:

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,240
Likes
11,462
Location
Land O’ Lakes, FL
ooooohhhhhh snap! Is that a knee I see in the Harman data at ~25Hz?!
Well, ladies and gentlemen, apparently my 25Hz knee is indeed correct. ;) :D

buuuahahahahahahaha

Now, let's talk about the rolloff rate ...
The mentioned Audioholics ground plane:
image


In reference to 100Hz:

From Audioholics:
30Hz being around -8dB and 40Hz being around -5dB.

From your Spin (using Sound Power) we get:
30Hz being around -17dB and 40Hz being around -8dB.

From the Harman Spin (using Sound Power) we get:
30Hz being around -8dB and 40Hz being around -4dB.

_________
In post #32 Amir shows before/after optimization and it’s not just the dip but the entire bass roll-off also improved. So while the 25Hz region is indeed not too important, reducing this issue will make everything <100Hz better.
 
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
Great then. No more floor standing speaker tests. Made my life a lot easier. :) :p
 
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
OTOH, both Harman and AH data above have their own issues. Both are heavily smoothed (not by choice in the case of AH as the response is gated to the low hundred hertz resolution).

I don’t believe James does vertical response measurements for tower speakers. So the AH data is missing that component. And thus no SPIN data.

So, there’s that.
 

ctrl

Major Contributor
Forum Donor
Joined
Jan 24, 2020
Messages
1,616
Likes
6,086
Location
.de, DE, DEU
Erin, this measurement indicates to me that the full response of the bass is not captured. Whenever you see that inversion in the sub-bass output, it indicates a problem. Did you measure this speaker vertically and if so, what was the mic distance?
FYI here is the ground plane measurements from Audioholics:
I, however, will trust the engineers who developed the NFS and take their input as the final word. So, you guys won't find me going in circles on this topic any longer. I've already moved on to the next set of measurements.

Yeah, catfight! ;):eek:

All the information is there to decide who is right.
We can take Erin's near field measurements and calculate the HD3800's low frequency response the old fashioned way.

We take Erin's three woofer and two BR port measurements, scan them, generate the minimum phase, sum the three woofer and the two ports, calculate the baffle-step correction for both, lower the sound pressure of the summed BR ports by about 4dB* compared to the summed woofers, delay the BR ports by the depth of the speakers cabinet, invert the phase** of the summend BR ports and sum the whole thing up - it looks like this:
* based on the ratio of the surface areas of the three woofers (about 630cm²) and the two BR ports (about 230cm²), we have to scale the sound level of the summed BR ports down (correction = 10*log(230/630)).
** Of course, only if the measurement does not contain the phase frequency response information, as in this case, where only the frequency responses without phase were available.

1616711048625.png



Now we just have to compare the measurements of the HD3800 from Audioholics ground plane measurement, the one from Erin's NFS measurement and the calculated frequency response from the near field measurements:
1616711061344.png


Ad the winner is...
ground plane measurement of the HD3800 by Audioholics.com. Erin's NFS measurements are clearly off the mark.
The GP measurement agrees very well with the calculated FR from the near-field measurement***.
IMHO, Klippel has clearly made a mistake in the evaluation of Erin's NFS measurement.

*** Depending on how good and accurate Erin's near-field measurements are, there will be more or less differences in the addition of woofer and BR-ports. From my point of view, the BR port measurement of -8dB is almost a bit low compared to the woofers.
 
Last edited:

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,335
Likes
6,700
Yeah, catfight! ;):eek:

All the information is there to decide who is right.
We can take Erin's near field measurements and calculate the HD3800's low frequency response the old fashioned way.

We take Erin's three woofer and two BR port measurements, scan them, generate the minimum phase, sum the three woofer and the two ports, calculate the baffle-step correction for both, lower the sound pressure of the summed BR ports by about 4dB* compared to the summed woofers, delay the BR ports by the depth of the speakers cabinet and sum the whole thing up - it looks like this:
*based on the ratio of the surface areas of the three woofers (about 630cm²) and the two BR ports (about 230cm²), we have to scale the sound level of the summed BR ports down (correction = 10*log(230/630)).
View attachment 120357


Now we just have to compare the measurements of the HD3800 from Audioholics ground plane measurement, the one from Erin's NFS measurement and the calculated frequency response from the near field measurements:
View attachment 120358

Ad the winner is...
ground plane measurement of the HD3800 by Audioholics.com. Erin's NFS measurements are clearly off the mark.
The GP measurement agrees very well with the calculated FR from the near-field measurement**.
IMHO, Klippel has clearly made a mistake in the evaluation of Erin's NFS measurement.

**Depending on how good and accurate Erin's near-field measurements are, there will be more or less differences in the addition of woofer and BR-ports. From my point of view, the BR port measurement of -8dB is almost a bit low compared to the woofers.

Yeah, it's clear to me here that the NFS is wrong here in the bass measurement. The fact that Klippel "signed off" on the correctness of their own machine is of no significance to me. It's the same situation as when Amir measured the F328Be with the NFS. The NFS was clearly wrong there too. Supposedly, Amir has found a way to correct this, but has he shared that?
 

ctrl

Major Contributor
Forum Donor
Joined
Jan 24, 2020
Messages
1,616
Likes
6,086
Location
.de, DE, DEU
Yeah, it's clear to me here that the NFS is wrong here in the bass measurement. The fact that Klippel "signed off" on the correctness of their own machine is of no significance to me. It's the same situation as when Amir measured the F328Be with the NFS. The NFS was clearly wrong there too. Supposedly, Amir has found a way to correct this, but has he shared that?

The easiest way to avoid errors with large floorstanding loudspeakers would be to simply make careful near-field measurements. Then you can compare the result of the NFS with the calculated frequency response from the near field measurements.

The near field measurements are done in five minutes, the calculation of the FR from the near field measurements in 10-20 minutes, depending on the complexity of the loudspeaker and experience.
 
Top Bottom