JBL M2 Reference Master Monitor Review

[synthetic]

Had to check after putting some more stuffing in my clones. Here's your after Klippel EQ vs my clones original Crown cdi M2;

Interesting that your clones don't have the 450hz dip. They use the 2262?

How did you measure the ports? Inside or at some distance? Between them?

I just measured the left port with the microphone flush with the mouth. The right one looked identical to me.

 

[Absolute]

Interesting that your clones don't have the 450hz dip. They use the 2262?
I just measured the left port with the microphone flush with the mouth. The right one looked identical to me.

I have the 2216nd woofers. I've added quite a bit of stuffing because I thought it sounded a bit hollow. Anyways, here's a nearfield of the left port. Red is yours, mine is not the original ports;

 

[synthetic]

I have the 2216nd woofers. I've added quite a bit of stuffing because I thought it sounded a bit hollow.

Very interesting. Might try adding some stuffing to mine.

 

[jtwrace]

If you're gonna compare it by measuring, it is of utmost importance that you make inner surface of the port really smooth. Depending on the nozzle diameter and the material you're using it can be more or less easy to do that (acetone or sanding).

Mine look like this
View attachment 150871

Yeah I used Onyx and Carbon so it’s very very good. 100 μm layers

 

[Zzzzz...]

It's disadvantage of 7 series, not advantage.

I'm saying it has a constant directivity over a larger set of frequencies because of that and also it's located low enough that most important (iirc) bands for spatial perception are played by a single driver. 7 series isn't even constant directivity - it's increasing (even if slightly), while m2 is constant until it starts to beam at 12 or so khz.

It's better in every area you can think of, sure it will sound better, if you want "graph" type competition. Are graphs really that representative if one will sound better than another? I don't think so.

Well, I believe in science and measurements. Spins won't capture everything but they are a good starting point. They can certainly indicate if a speaker will sound bad. Trying to distinguish a very good speaker from an even better one seems more tricky however.

Thanks for your comments.

Looking at the plots for the M2 and 708, it seems the M2 is roughly constant directivity from ~700 Hz to ~8 kHz and the 708 is roughly constant directivity from ~1.4 kHz to again ~8 kHz. It's a bit hard to say exactly where the cut-offs are (you mentioned ~12 kHz for the M2 for example) but it is clear that the M2 starts lower than the 708. I suppose this is a benefit of the M2 for the reason you mentioned. I wonder if Floyd Toole or Sean Olive for example ever talked about this specific issue? I'd be interested to know how much such differences contribute to listener preferences.

 

[Zzzzz...]

That would lower the unit-to-unit variations for sure, but is it necessary considering the intended market?
They'll have to be measured in-situ and equalized to suit the purpose anyway, so not sure there's much gain to be had for another 1 dB +- rating.
Considering the low number of measurements out there for the M2, we're really just guessing about the quality control.

Another thing we're guessing on is the notion that the perceived sound quality will improve with improved measurements.
I'm fairly confident that the 1-2 dB too hot tweeter in this sample would sound significant better if corrected due to my excessive experience with EQ'ing those particular drivers and toying with the tonality, but I'm not confident that the other wiggles would sound better corrected for.

I agree - the higher frequencies look like they need to be bumped down a little but as you say it's harder to tell how audible the other wiggles would be.

 

[Zzzzz...]

Anytime! Wish it was a perspective backed by double blind testing, but it's all I can offer at the moment. Listening distance was 11' 6".

From the specs and from what I've read online, it seems the 708s should used at a max distance of no more than about 8-10 feet. Perhaps they were a little far away in your tests to sound their best with regards to dynamics and such.

 

[Zzzzz...]

More directionality at low frequencies is an advantage because it reduces SBIR. This is also why cardioid bass is a desirable feature.

Hmm, I'm doing to having to think about this some more. Because of the way sound below the transition frequency interacts with the room, this is not completely clear.

 

[Zzzzz...]

That is what I suspect as well.
The final M2 EQ is said to be the result of a long objective/subjective process with a lot of listening sessions at JBL's facility. The official spinorama was probably taken at an early stage, when it was "perfect" measurement-wise, and then maybe modified to suit listener's feedbacks.
Having a raising on axis UHF response was also probably deemed problematic.

Measurements above a few kHz are easy to do with enough precision, and the issues measured in the HF such as the dip at 6.5kHz or the large bump above have been known and reported for a long time (check audioheritage threads).
The D2430K also appear to be a very consistent driver from unit to unit in it frequency response, so I am left to wonder if these "defects" where put there intentionally based on listeners's preferences?... I dunno, it might also be a case of some variations in the D2430K production, as the difference between the Vertec and M2 specific units have always been pretty unclear/shady...

Have you read the article attached below?
(I hope it is ok to post - it was available for free online before but I can no longer find it)

It's a great article that includes info about the development of the M2. The big change from the initial, not great sounding prototype was the introduction of the image control waveguide.

Do we have any evidence that the JBL engineers tweaked a "perfect" prototype? I can see the Revel guys tweaking their speakers from one type of "excellent but imperfect" to another such type simply for the reason that they never had a speaker that was as perfect as the M2 to begin with. Floyd Toole posts a lot on various forums and I don't think he has ever mentioned designers improving speakers by deviating from the principles he outlines in his book. He would have surely said something if he was aware of it.

 

[aac]

Well, I believe in science and measurements. Spins won't capture everything but they are a good starting point. They can certainly indicate if a speaker will sound bad. Trying to distinguish a very good speaker from an even better one seems more tricky however.

Thanks for your comments.

Looking at the plots for the M2 and 708, it seems the M2 is roughly constant directivity from ~700 Hz to ~8 kHz and the 708 is roughly constant directivity from ~1.4 kHz to again ~8 kHz. It's a bit hard to say exactly where the cut-offs are (you mentioned ~12 kHz for the M2 for example) but it is clear that the M2 starts lower than the 708. I suppose this is a benefit of the M2 for the reason you mentioned. I wonder if Floyd Toole or Sean Olive for example ever talked about this specific issue? I'd be interested to know how much such differences contribute to listener preferences.

Harman speakers perform rather poorly on "Olive score".
This is a sign for me that we don't really know how to read spinorama, it's a trade secret most likely, or just a tool for acoustic engineers with a fancy name.

 

[Zvu]

We know how to read spinorama just Olive score isn't as relevant as some guys think.

If i like how spinorama looks, i couldn't care less what is some score for that loudspeaker. Algorithm takes some things into acount i don't really care about (like if the speaker doesn't do 20Hz and 20kHz score goes down).

From spinorama i derive frequency response smoothnes on axis in regard to off axis, i look for resonances, look for directivity errors and how wide the dispersion is. Not all what i'd want to know but close enough.

What I'd like to see in future are averaged response +/-90° horizontal and +/-90° vertical in at least 10° increments on separated graphs. I also don't take too much faith in standard on how individual frequency response curves at certain axes are weighted in spinorama.

 

[Sancus]

Harman speakers perform rather poorly on "Olive score".

This isn't even close to true. Most Revels reviewed so far have great scores, only a small handful of actives with built in EQ beat them. Some JBLs have quite good scores as well, like the HDI-1600.

And, as usual, the score is far from authoritive especially with speakers on the margins of capability. Stop fixating on it. It's a fantastic rough sorting method, and that is all it is.

 

[Bugal1998]

From the specs and from what I've read online, it seems the 708s should used at a max distance of no more than about 8-10 feet. Perhaps they were a little far away in your tests to sound their best with regards to dynamics and such.

It's certainly possible. And it's worth repeating that I consider the 708s to be excellent speakers.

 

[Absolute]

Question for the intelligent people here;

Is there a way to simulate how the spinorama would look if we changed the crossover to, say, LR 24 dB/oct at 700 hz?
I'm curious to see how that would affect the directivity curves.

Been experimenting a bit with crossover slopes and frequencies and thought I liked my example above better than the original.
Completely different tweeter, though, so keep that in mind.

 

[abdo123]

Question for the intelligent people here;

Is there a way to simulate how the spinorama would look if we changed the crossover to, say, LR 24 dB/oct at 700 hz?
I'm curious to see how that would affect the directivity curves.

Been experimenting a bit with crossover slopes and frequencies and thought I liked my example above better than the original.
Completely different tweeter, though, so keep that in mind.

No you can’t.

 

[ctrl]

I have the 2216nd woofers. I've added quite a bit of stuffing because I thought it sounded a bit hollow. Anyways, here's a nearfield of the left port. Red is yours, mine is not the original ports;

If the green curve is a near field measurement of the BR port, then you have a problem. Near field in this case means measured directly at the BR port output (no distance from the BR-Port to avoid room resonances).
You should not get a 15dB drop around 42Hz.

 

[Koeitje]

Nothing stops a hobbyist from doing that if they want to enough.

What no hobbyist can do is match the Harman turntable The American Way: throw money at an off the shelf product from a fancy Euro company.

Then struggle to match the parent company measurements and proclaim themselves the more accurate one — while still technically on a learning curve.

Seems like you don't really understand what Klippel near-field scanner actually does, nor understand the limits of anechoic chambers.

This isn't even close to true. Most Revels reviewed so far have great scores, only a small handful of actives with built in EQ beat them. Some JBLs have quite good scores as well, like the HDI-1600.

And, as usual, the score is far from authoritive especially with speakers on the margins of capability. Stop fixating on it. It's a fantastic rough sorting method, and that is all it is.

Indeed, the value of these scores is to separate the wheat from the chaff. If I have spins and/or scores I can far more easily make a shortlist of loudspeakers that should be worth listening to.

 

[Absolute]

No you can’t.

I know, I'm stupid. Which is why I'm asking for guidance

If the green curve is a near field measurement of the BR port, then you have a problem. Near field in this case means measured directly at the BR port output (no distance from the BR-Port to avoid room resonances).
You should not get a 15dB drop around 42Hz.

Good catch. That's the result of a PEQ point aimed at a room mode

 

[abdo123]

I know, I'm stupid. Which is why I'm asking for guidance

Perhaps if you send whatever you have to Erin with a sizeable donation he would put it on the Klippel for you and measure it the way you want?

 

[Absolute]

Perhaps if you send whatever you have to Erin with a sizeable donation he would put it on the Klippel for you and measure it the way you want?

It would cost about the same as buying my own Klippel.
Anyways, I figured that VituixCAD might have some magic that would help simulate directivity index. I've downloaded it to see for myself, but being horrendously impatient as well as horrifyingly stupid, I figured I'm better off asking instead.

If not for simulating such things, what is it for?