JBL HDI-3800 Floorstanding Speaker Review

[Joecarrow]

Thank you very much for the review! I have to say I had extremely high hopes for these based on my experience with the much cheaper LSR 308 mkII. It’s a shame to hear that the lows may not live up to the waveguide implementation.

 

[daftcombo]

Very nice review Mr. Floortanders!

 

[dfuller]

That upper treble (above 10k) response is quite something, and not in a good way. Is that tweeter breakup or just a resonance? The directivity heat maps show some strange behavior up there, too.

I can't tell if I'm seeing things or not (what smoothing is this?) but the midrange has some weird jaggedness going on in the FR. Port problems, perhaps?

 

[MZKM]

That upper treble (above 10k) response is quite something, and not in a good way. Is that tweeter breakup or just a resonance? The directivity heat maps show some strange behavior up there, too.

I can't tell if I'm seeing things or not (what smoothing is this?) but the midrange has some weird jaggedness going on in the FR. Port problems, perhaps?

 

[dfuller]

Yeah, I wasn't quite able to parse that out from the nearfield per-driver graphs. It looks like there are some resonances, but it's rather hard to tell.

 

[Sonny1]

Great review Erin! These didn’t perform nearly as well as I was expecting. I’m a little surprised because I expected better performance from JBL, especially in this price range.

 

[hardisj]

Great review Erin!

Thanks!

 

[MZKM]

Yeah, I wasn't quite able to parse that out from the nearfield per-driver graphs. It looks like there are some resonances, but it's rather hard to tell.

Lots of port noise.

 

[Mnyb]

Excellent review like always, thank you!

I think it could be also due to the large vertical directivity rise around 500 Hz due to the all 3 woofers still active there. It can be also seen in your in room response which is too low between 300-700 Hz

View attachment 120241

compared for example to the D&D 8c

View attachment 120242

Imho a 2.5 way is an unnecessary compromise (especially with 3 woofers) at the price class and would be the reason for me not to buy those.

Yes a pity they did not made it a three way

 

[Juhazi]

This is a 2.5 way speaker, "mid" plays for 0 to 2khz and obviously all woofers share same airspace. No wonder there is wiggles at midrange output with all those separated drivers puffing. Add port resonances to that...

Erin said Revel sounds better in midrange and how come it is a real 3-way! In a 3-way port resonances and midrange noise through large cones is minimal and midrange plays in solitude and independence! Old shool rules!

 

[amirm]

This makes sense. The Bass Response did look suspect to me. Is this the same problem you had with the Revel F328be? With that speaker, your measurements showed very poor bass extension, but it was obvious from the subjective listen that it was wrong. If so, do you know how to fix it?

I do indeed. Let me show you my trials before producing the measurement that I showed:

Notice how the scan on the left has the trough as Erin's measurement of HDI-3800 shows:

Some speakers have no response there so having that notch is OK but in this case, I don't think it is. As you see in my measurements with F328Be which has similar port configuration, I almost managed to remove it and with it, show more bass response.

The issue is the distance from speaker. The larger and more spread apart the sources of low frequency radiation, the more distance you need to have for NFS to properly compute the energy in that region. Since posting the F328Be measurements, I have squeezed a bit more distance out of the setup but did not want to spend a day measuring that speaker again.

 

[hardisj]

Lots of port noise.

Dude, where you at with the preference score?! Come on, man!

 

[amirm]

The speaker was measured on its side. I used the r-axis extension and the mic distance was somewhere close to 0.90 meters from the DUT. I have to send the data to Klippel to get it 'signed' via the new beta software and as part of this process I have been in talks with them to make sure the measurement data looks correct as the speaker was measured on its side. This was their feedback when they sent me the signed data file:

As I just noted, you need to have larger distance from the speaker. To the extent you have turned it sideways, you now need to have much more clearance vertically. This is not possible below the speaker due to limitations of the jig. You could try setting a much higher vertical distance to see if it helps.

Again, I apologize bringing up this issue. After testing the F328Be, I almost regretted testing any large towers like this! Dealing with the massive speaker on top of member complaints about accuracy of the measurements was a bit much. In my case, I had even purchased the speaker myself so was stuck with the cost as well! Then again, this is one weakness of NFS (in a home setting as opposed to a large space) that you just have to learn about and figure out how to deal with (not testing such speakers, optimizing even more, etc.).

 

[pozz]

The issue is the distance from speaker.

If you don't mind, how is this set for the NFS? Are you referring to the distance between the mic and the speaker or a setting for the farfield calculation? Or something else, like the maximum distance the measurement arm can move outwards?

 

[hardisj]

but in this case, I don't think it is

Did you see where I quoted the response from Klippel themselves regarding the LF response in my reply to you earlier?


FWIW, the error was below 1% all the way through other than the 5kHz which is explained by Ruben in the above reply.

I'm not sure what the fitting error was (is) in your data (feel free to share if you'd like to do some group investigation). Maybe that's what you were having issues with. But as far as Klippel is concerned and as far as the data shows, the measurements are accurate in the LF.

 

[amirm]

If you don't mind, how is this set for the NFS? Are you referring to the distance between the mic and the speaker or a setting for the farfield calculation?

Physical distance between the mic and speaker. Software optimizations, more scanning points, etc. don't help it.

 

[amirm]

I'm not sure what the fitting error was (is) in your data (feel free to share if you'd like to do some group investigation).

Fitting error is not helpful here as you are surely in the near-field for bass frequencies. Ruben is commenting on the general scheme, not the issue I am discussing. It is the same response I received a year ago when I went to them with similar issues I had with F35 measurements.

Erin, please, please don't get defensive on this. I am not accusing you at all of not being truthful. Please don't go there. I have suffered through this issue and I am sharing with you knowledge I had to develop to help mitigate it. From day one that you had your system up, I have been talking about measurement distance and its importance. It is for this reason.

I have been remiss in not closing the loop with Klippel on my findings. I should do that.

 

[hardisj]

As I just noted, you need to have larger distance from the speaker. To the extent you have turned it sideways, you now need to have much more clearance vertically. This is not possible below the speaker due to limitations of the jig. You could try setting a much higher vertical distance to see if it helps.

Again, I apologize bringing up this issue. After testing the F328Be, I almost regretted testing any large towers like this! Dealing with the massive speaker on top of member complaints about accuracy of the measurements was a bit much. In my case, I had even purchased the speaker myself so was stuck with the cost as well! Then again, this is one weakness of NFS (in a home setting as opposed to a large space) that you just have to learn about and figure out how to deal with (not testing such speakers, optimizing even more, etc.).

Amir, no worries. I get it. I do. I'm not trying to be defensive (text doesn't do a good job of this). I'm just providing you with what I know, what I had discussed with Klippel before I considered the data final.

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. Nits that neither Klippel nor the data show as having merit. I just think we are going to cause people to "throw the baby out with the bathwater" here.

If the people of this forum want to talk to Klippel and address their concerns with their own assessment and approval of my results then they are certainly welcome to. 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.

 

[amirm]

But as far as Klippel is concerned and as far as the data shows, the measurements are accurate in the LF.

Share with them ground plane measurements and get their reaction. I showed them the Harman anechoic data and we were all puzzled by the large gap between what NFS had generated and their results.

 

[amirm]

If the people of this forum want to talk to Klippel and address their concerns with their own assessment and approval of my results then they are certainly welcome to. 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.

Thanks for the permission. I will do that.