How did you managed to get measurements so clean in the sub-150Hz region?
How did you managed to get measurements so clean in the sub-150Hz region?
Ian, the large format drivers on the 4367 is really very nice behaved. There is a humb around 1,5kHz and the normal CD fall off from about 4-5kHz. I put a wide "minus" Bell to lower the hump and a normal boost High Shelving to compensate for the fall off. That cleared most of it. On top of that I took the liberty to correct some minor peaks & troughs. I also gave some boost under 1kHz to get a decent curve to at least 600Hz (drops from 550Hz) to be able to XO at 700Hz.
Rather basic EQ points in the end, but it takes a long time to find a combo that achieves your goal with minimal intervention. I can post the EQ settings later today, if of interest.
jmpsmash, the last two graphs does not contain info below 300Hz and may look like "nice curves" but it is really a too short measuring window to display any data there. The first two graphs are composite graphs that use data from both near field (<0,5cm from cone and in port, @85dbA, 170ms window.) measurements, that are reasonable free from room reflections, and gated far field measures (at least 1m out for a 15" DUT).
After measuring the cone and the ports near field I then add all 3 of them into a combined "near field curve" with little to no room reflections visible. Ports will play "louder" due to smaller radiating area compared to the woofer so that need to be compensated for when adding them.
With a 15" woofer near field data can be used to about 350Hz from where it is less accurate. Above 300Hz I merge far field measures. To get reasonable resolution I use 3 separate measures for each composite curve, 9ms 300-600Hz, 6ms 600-1200Hz and 3ms above. So the second of "my" curves is really a combination of 6 measures (2x ports, 1x near field cone and 3x far field)
Its a bit time consuming to do the process for each EQ correction but I don't know how to get clean data in the lower frequency area in any other way. You can measure outdoors but to elevate the 4367 high enough to remove floor/ground bounce would be a challenge due to size and weight. There are some doing ground plane measurements and that may be a good alternative. I did play around with that earlier but did not find that it gave any major improvement compared to the above described method.
Kind regards
//Rob
The solution to the problem changes the problem.
-And always remember that all of your equipment was made by the lowest bidder
Ian, please see attached the measurement of the 2450SL/475ND driver on the 4367 wave guide without any EQ applied. Not too shabby :-) .
The large format 1,5" driver cores really loves the newer JBL wave guides (VTX, M2 & 4367) .
Measurement is done 90 degrees straight in front and 67cm out. The time window is 9ms so the ripple is mostly room reflections and are much less pronounced with a 3ms window. Especially above 5kHz. Please note that there is no smoothing applied. Mic is corrected Earthworks M50.
Kind regards
//Rob
The solution to the problem changes the problem.
-And always remember that all of your equipment was made by the lowest bidder
Any ideas what is causing the ripples above ~10kHz? It looks like comb filtering from a tweeter, but since there is no tweeter, perhaps diffraction within the horn throat?Originally Posted by sebackman
Mostly just a curiosity as I doubt we can perceive it.
Widget
You see that in all of the 4 " coated/uncoated Ti s. It's secondary resonances in the diamond surround.
It's essentially controlled break up modes in the surround in the last octave. Page 3 attached Technote
Rob
https://www.audioheritage.org/vbulle...ume-1-Number-8
"I could be arguing in my spare time"
Of course... I haven't played with diamond surrounded diaphragms in years (decades).You see that in all of the 4 " coated/uncoated Ti s. It's secondary resonances in the diamond surround.
It's essentially controlled break up modes in the surround in the last octave. Page 3 attached Technote
Rob
https://www.audioheritage.org/vbulle...ume-1-Number-8
Widget
Hi,
Its not there in a 3ms window so it is room or cabinet.
Btw the 475nd dia has a coated diamond pattern.
Will post 3ms measure tonight.
Rob
The solution to the problem changes the problem.
-And always remember that all of your equipment was made by the lowest bidder
Good Job,
That non normalised horn curve is almost a linear (straight) mass roll off. Shouldn’t be to difficult to EQ .
By the way are you subjectively preferring the conventional drivers over the D2 annular motor? I was looking at some updates on the JBL Luxury Group page and it looks like the V shaped single annular polyester diaphragms are creeping into the product ranges. The phase plug is shaped like a cone. They are reasonably cheap at The Loudspeaker Exchange. On a small wave guide they would make a nice diy super tweeter….!
“At the core of the 2410H-2 device is a lightweight, polymer, annular diaphragm that improves high-frequency extension by reducing diaphragm mass”. BMS have gone the same direction with their hi end compression drivers. There are noises in the Voice Coil Loudspeaker Industry Journal that beryllium diaphragms could on the way out as new man made materials emerge.
A material called Textreme has evolved with a new level of stiffness. See attached.
Rob, I think you are right but it is probably more or less neglectable today. I recently did a shoot-out with 9 JBL 1,5" drivers on VTX, M2 and 4367 wave guides with many different measures on each (+400 all-in-all) and they all show similar ripple of about +/- 1-2db above 10khz.
I do think the technical manual linked was from the time before the coherent phasing plug (from 2446) and modern day wave guides. You still see some of that but but I argue that it is now reduced. Some ripple is no doubt present but it does not, in my opinion, pose any audible issues.
Below is a driver graph with shorter time window, 3ms @67cm distance. Note that there is NO smoothing.
It is probably better than many other mid/tweeter systems on the planet, especially taking phase and distortion into account.
As we are seldom provided "raw data" from "High Street" speakers, we don't really know how they measure. There are some measurements on the ASR forum and but Erin but they are Klippel "system data", albeit as good as it gets. It would be really interesting to see anechoic data for each element in cabinet for a bunch of High End speaker to compare. I think our JBL's would come out with flying flags.
Most drivers JBL look the same regardless of diaphragm but the 475Nd dia seems to be slightly "calmer" as is has a coated suspension. I would argue that none of the +8kHz ripple is audible with coated or Mg/Be dias. The Ti ones tend to sound a bit harsh when EQ'd, IMHO.
Ian, yes they are easy to EQ but it takes time to reduce the number of EQ points and still get the desired result. Rather time consuming, but fun. DSP is a fantastic tool that lets you do all the measures and corrections in real time. Pandoras Box for us speaker builders.
I have not had a chance to try the new "super-drivers".
I personally prefer the 4" drivers to the D2's on most wave guides. The 4" has more power deeper, which in my experience is audible especially on piano and deep female voices. I did try the D2's on the 4367M's but prefer the 2451/475Nd combo.
Given a choice, I think a really good matched pair of SL's are preferable to the Be's as the Be's will need UHF support and the improvement compared to the technical complication and high price makes it less a compelling combo to me (I have both). I have not had any MG's and maybe they are the Holy Grail. If they were available and had no problem with corrosion.
Compromises, compromises. :-)
Kind regards
//Rob
The solution to the problem changes the problem.
-And always remember that all of your equipment was made by the lowest bidder
Hello Rob
I never said they were audible??? Just referenced the tech sheet as a reference and probable cause.
I took a look at the 2450 non SL data on the site. You can see impendence ripple and a decided rougher output from 9K and up.
I agree the techsheet is dated but the basic surround concept is still the same and mode of operation is still in use. And who knows what if any changes JBL has made to the surround's.
I have never seen an SL diaphragm with a coated surround?? That certainly makes the 475's a stand out. I also attached a some 2453SL data to review. Shows ripple and slower decay last octave in the waterfall.
Rob
"I could be arguing in my spare time"
Hi,
Rob, sorry did not mean to imply that you did say that the resonances are audible. Just wanted to make a point in that we, at least I do, sometimes chase to resurrect issues that are not there. I have spent ours chasing the "final" point and lost the "whole" on the way there. For me comparing to a reference is the salvation to keep me sane. :-)
Attached are the DATS3 impedance curve for the 2451/475Nd driver on the 4367 horn in my actual 4367M cabinet. The resolution is not fantastic but the resonances are limited from what I can see.
The 4" diaphragms come with coating on the dome only, coating on half the surround and coating on the entire surround. I have no idea why. The newer ones seem to only cover the dome. There are other subtle differences on diaphragms supplied in boxes with the same markings. JBL may be better on speakers that QC.
Best regards
//Rob
The solution to the problem changes the problem.
-And always remember that all of your equipment was made by the lowest bidder
Hello Rob
No harm no foul and frankly who hasn't been down the "rabbit hole" at one point or another.
Thanks for the coating info learn something everyday didn't know there were variations on the theme with the 4" diaphragm's.
Just for info the 476Mg has a smooth Imp curve as well as a clean CSD compared to the 2453 10K and up so go figure????
I started a new thread addressing the secondary resonance WRT materials and surround geometry in General to keep this on topic.
https://www.audioheritage.org/vbulle...resion-drivers
Hello Ian
I only have the above mentioned 4" drivers. As far Oswald's Mill no interest. Even if I could swing the prices I don't like the looks. It is also appointment only so why waste someone's time??
Rob
"I could be arguing in my spare time"
Okay,
I will edit / delete my above posts.
I would change that statement to "Even if they were the best measuring and sounding speakers extant, I have no interest due to their high school woodshop aesthetic."
Thanks Rob for starting those other threads. I think we should explore those topics more thoroughly and not hijack this wonderful project thread.
Ian, you have posted some interesting and useful posts above. I think it would be great if you would repost them on Rob's new thread(s).
Widget
Hi all,
I have now had the 4367M (M for Modified 😊operational for some time and had the chance to develop some quasi anechoic and in the room DSP curves. The room is pretty ok except for a nasty modal at about 30Hz. The curves are nice and smooth from about 30hz to 20k. Even phase looks decent.
The first observation is that they are very neutral and revealing. Not very sensitive to placement compared to my former M2’s, but do benefit from some toe-in. IMHO they work better on some stands to get the driver exit in listening height.
I did extend the ports which demanded some DSP support, but I think it works just fine and does provide that low end quite effortless. F0 is 27Hz.
I was happy.
Now that is of course not enough so a few days ago I invited some seasoned JBL friends for a shoot-out. They all have trained ears and excellent systems at home. At least 2 are in their prime so their ears are still intact. For the older crowd time is probably taking its toll, albeit I can still hear 15k on a good day 😊.
We added a pair of LSR6332 monitors with LSR6212 subs, a pair of LSR708i and my active standard system with SUB1500/1400PRO/476Be/M2/045. Very different systems but to some extent related to one and other. 708i’s are well known to be excellent. The 6332 were the monitor reference of its time and still no slouch. The go-to system that I normally use is very neutral I have spent a ridiculous amount of time on the active DSP setup. They sound and measure very behaved.
Would the fully active, 2451/475 DSP enhanced 4367M clones measure up? Scary.
To add to the competition there was a high end HiFi trade show in town the same weekend and that we visited. They had speakers there priced over $100k.
It was by no means a blind test and does not aspire to be scientific in any shape or form.
We did assemble some decent electronics and I would even go so far to suggest that most was SOTA parts. My friend “Beatelund” managed to get us a brand spanking new $10k Lyngdorf MXA8200 8-channel amplifier, right from the HiFi show in Barcelona where it was shown for the first time, so we could run all speaker on the same amp. We used an RPi (running Roon endpoint SW) for each speaker set feeding digital signal (USB to AES/SPDIF) to a fully digital (in and out) BSS BLU160 and then via an Okto DAC8 DAC to the Lyngdorf. All controlled by a Roon master unit on an iPad. Music from Tidal and Qobuz. Neat
I have been an advocate for analog amps most of my life. Having had the fortune to try many amps, including digital such, I always came back to traditional class A/B amps. However, it is with a small tear in my eye I now have to confess that the digital amps have cached up and probably surpassed the analog units. The Lyndorf provided plenty of power while regardless of material was completely free from distortion. Almost too clean, if there is such a thing. Pure High Fidelity.
To be able to compare these very different speakers, we tried to equalize the bass output meaning reducing it on the bigger speakers using REW and a good measuring microphone in the room. We also set the level at listening position at the same level. Sweetspot was averaged over 3 seats in the couch.
Then the feast began. We played all sorts of material and being able to shift speaker in real time was really an interesting experience. Surprisingly the sound timbre is reasonable similar but the compression drivers do deliver a sound that is experienced as “cleaner”.
The big surprise was the 708i’s that really had some big shoes to fill but did so surprisingly good. They are really exceptional speaker for the size and price.
I do of course have a strong bias in that I’m used to the sound of my standard system. It did take some listening to get to grips with that. The feedback from the others is much more valuable in that they did not have that specific bias, albeit they came from excellent systems (M2, K2 aso) with a bias for them.
At the end of the day and after spending the following full day at the show, yes I know that you cannot direct compare different rooms/music aso, I’m please to note that all held the 4367M as good as the best in the show, or potentially better. We did spend some time with the 4367M’s the same day after the show to compare. Maybe the guests were just being polite…
However, after countless hours the 4367M clones are now officially finished. My personal view is that using a DSP combined with large format driver provides for a lower XO which the 4” delivers on in a way I don’t see in the original version.
I also think the 4367 wave guide is easier to integrate in a residential room than many other horns. I like piano and to my ears the 4” does really shine there. -And on deeper voices. With good material the speakers disappear, only music remains. Odd experience.
I have no idea if that is due to the driver, my reasonable sturdy cabinets, the passive vs active XO or anything else but after having had the luxury to spend some time with 4367 original on several occasions my clones have by no means destroyed that concept.
If there are any improvement that is in the behold of the listener, but for me I choose to believe that is the case.
I’m done. I’m home.
Kind regards
//Rob
PS thanks to the friends helping out trough the process and especially member “Beatelund” who made this day possible. An understanding wife also helps.
The solution to the problem changes the problem.
-And always remember that all of your equipment was made by the lowest bidder
There are currently 1 users browsing this thread. (0 members and 1 guests)
Posting Permissions
Reply With Quote