JBL 708P Review (Professional Monitor)

[thewas]

I'm too stupid to figure out what that video should show me, but I measured my tweet before and after two high Q peqs, one were the response were flat and the other where there was a dip.
5000 hz and 9270 hz;

View attachment 110354
View attachment 110355

CSD shows the same, but this was easier on the eye. Ringing at 5 khz and no ringing where I just fixed the dip?

It is hard to tell because we don't know if the dip was minimal phase or not.
If you want to see the effect better in REW just take a measurement and at a frequency where you don't have any noticeable ringing add there a narrow band positive PEQ (like for example 10kHz, +10dB, Q=10) and the added ringing due to the filter both in the csd and impulse response will be nicely shown in the simulated responses.

 

[MZKM]

The thing that bugs me is that the JBL 708 costs ten times more than the JBL 308 and yet scores worse on the preference transfer function that the same company developed as a benchmark tool and implemented into their product design philosophy.

https://www.audiosciencereview.com/...review-professional-monitor.20006/post-659360

 

[Bartl007]

If anyone is interested in the passive 708i, I have a mint condition pair for sale with matching (required DSP) amplification if needed. Pm me here or see my classifieds post on AVS.

About the preference ratings not matching subjective impressions, I think it's important to remember that coorelation is not causation.

An example that helped this to sink in for me was the following fact: the sale of ice cream is positively coorelated with the rate of pediatric head trauma. If you look at a graph of ice cream sales compared to pediatric head trauma, it looks like this:

Common sense would tell you that buying Ice cream doesn't cause pediatric head trauma, but if all you looked at was the data, there could be no other conclusion.


Data is great... But don't forget to listen

 

[Thomas savage]

Sadly seems that, in common withe all the up market JBLs I have investigated in the last 20 years, it is pretty well unavailable here in the UK. The importer website is hopeless.

GAK sells them , thats where I got my 705p's . My Builder got those as he's more patient wrt sending back broken things, JBL aren't fixing them under warranty though, they were left it too long apparently and the warranty has expired ( 2 years ).

I won't buy anything from JBL over £300 again .

Edit , GAK seem to have ditched them but

https://www.gear4music.com/Recordin...MI246XkarO7gIVZGHmCh1fSwk7EAAYASAAEgJ-1fD_BwE

 

[pierre]

I'm too stupid to figure out what that video should show me, but I measured my tweet before and after two high Q peqs, one were the response were flat and the other where there was a dip.
5000 hz and 9270 hz;

View attachment 110354
View attachment 110355

CSD shows the same, but this was easier on the eye. Ringing at 5 khz and no ringing where I just fixed the dip?

i guess I have a vocabulary issue: ringing is for me oscillation in the time domain (phase) not on the freq domain.
if you want to see ringing look at the step respense in REW.

 

[MickeyBoy]

They're inherently problematic in 2-way designs, in which the port resonance frequency and the frequencies of internal standing waves in the enclosure are almost certain to be fall within the frequency range of the woofer (unless the port is made so small that it can't handle high SPLs without significant compression/distortion, which is the greater evil and best avoided).

In 3-way designs, ports are rarely problematic, or at least there's not usually any reason they must be so. The port can be made large enough to keep compression/distortion to within acceptable limits throughout the woofer's full range of linear displacement, while the upper frequency limit of the woofer tends to ensure that the frequencies of internal standing waves, and the port's own resonance frequency, all fall above the range of frequencies that the woofer reproduces.

Why not put a bung in the port and crossover to subs at 100Hz?

 

[Absolute]

It is hard to tell because we don't know if the dip was minimal phase or not.
If you want to see the effect better in REW just take a measurement and at a frequency where you don't have any noticeable ringing add there a narrow band positive PEQ (like for example 10kHz, +10dB, Q=10) and the added ringing due to the filter both in the csd and impulse response will be nicely shown in the simulated responses.

Measured 2 cm from the tweeter I'd guess that everything except break-ups is minimum phase, but perhaps that's just because I'm gullible and believe whatever sounds easiest for me to comprehend

So let me be spoonfed some truth;

If I want to correct the driver itself (dip) then I can do that as long as it's minimum phase without introducing ringing? According to my measurements it seems like a dip is a resonance in reverse that can also be turned back to perfection.

i guess I have a vocabulary issue: ringing is for me oscillation in the time domain (phase) not on the freq domain.
if you want to see ringing look at the step respense in REW.

Let's see what happens in the time domain when I correct a dip;

GREEN is predicted after EQ, GOLD is actual measured after EQ while BLUE is before EQ

Is it safe to conclude that I can correct driver related dips, at least as long as it's below break-ups?

 

[thewas]

If I want to correct the driver itself (dip) then I can do that as long as it's minimum phase without introducing ringing?

Correct, minimum phase behaviour can be fully corrected (invertibility) but

According to my measurements it seems like a dip is a resonance in reverse that can also be turned back to perfection.

how are you sure from that measurements? Because of the blue curve rise in the phase? That could be a good indication but

Is it safe to conclude that I can correct driver related dips, at least as long as it's below break-ups?

I am personally missing the verified evidence of being always minimum-phase below break-ups but maybe others have more experience or even literature for it, which would be also very interesting for me.
According to what I have read till now for example a baffle reflections can also cause peaks and dips in the FR which are not minimum phase though and thus cannot be fully corrected by EQ.
To be more sure you could extract the excess phase and excess group delay (derivation of the phase) and see if the last is flat or not at that frequency, if yes it would mean that the unequalised system behaves there as minimum phase.

 

[Absolute]

Thank you @thewas , it makes sense to me now. Once you take the mic further away from the driver you might end up with all kinds of nastiness that you may or may not be able to EQ successfully. But here's the thing I'm thinking, if the driver itself is measuring perfectly flat without ringing and then you start to EQ the mess introduced by waveguide and diffraction etc you WILL introduce ringing to the driver itself even if what you measure from afar is minimum phase?

 

[thewas]

You are welcome @Absolute.
Since the mess from the diffraction would be non-minimum phase your far measurement would also start partially deviating from a minimum phase response, on the other hand what is most important for us, is to which amount EQ and ringing would be audible and from what point it would deteriorate the total perceived sound quality compared to the smoother FR, all questions where there still needs to be quite some research done...

 

[test1223]

Harmonic distortion isn't a good measure. Intermodulation distortion correlates much more with the perception of distortion. The jbl 708 will outperform the similar neumann and genelec speaker by a lot looking at the max spl with low imd. Jbl uses essentially pa chassis for the 708. The bigger genelec S360 are more alike, but with a narrower beam.

 

[thewas]

A 2-way implementation like the 708 will have at the acoustically most important and most audible mids high(er) IMD when the woofer at the same time does large excursions due to paying loud and deep bass compared to 3 way designs like Genelec Ones and Neumann KH310.

 

[test1223]

A 2-way implementation like the 708 will have at the acoustically most important and most audible mids high(er) IMD when the woofer at the same time does large excursions due to paying loud and deep bass compared to 3 way designs like Genelec Ones and Neumann KH310.

Yes, but this doesn't limit in general the max spl capacity with lower imd, if the bass provided some more linear movement.

Here the bass chassis of the jbl is more like a pa chassie with stiff suspension and higher resonance. This gives more max spl and lower imd you loose deeper bass.

 

[thewas]

Still the imd will be higher at the mids where it will be most audible. Also the here measured 96 dB HD measurements show that a relatively clean deep bass is not sufficient if the other distortions rise already too much.

 

[Xyrium]

Harmonic distortion isn't a good measure. Intermodulation distortion correlates much more with the perception of distortion. The jbl 708 will outperform the similar neumann and genelec speaker by a lot looking at the max spl with low imd. Jbl uses essentially pa chassis for the 708. The bigger genelec S360 are more alike, but with a narrower beam.

I would also love to see IMD, but from what I've read, it's a time consuming measurement, or perhaps requires a different module for the Klippel?

 

[infinitesymphony]

I'm confused by the idea that the 708P has effortlessly high SPL when the distortion graph shows that it's not happy playing at 96 dB.

JBL 708P (background, red) overlaid with Genelec 8341A (foreground, magenta), scaled:

 

[amirm]

I'm confused by the idea that the 708P has effortlessly high SPL when the distortion graph shows that it's not happy playing at 96 dB.

JBL 708P (background, red) overlaid with Genelec 8341A (foreground, magenta), scaled:

This may help from my review of 8341A:

All was not well. You may be wondering with measurements as good as posted, why the 8341 did not get the top honor panther and had to settle for the next grade down. I was quite surprised that as I turned up the volume, listening at just 1 meter or so from the speaker, it just would not get that loud. At first I heard a glitching/ticking sound which then moved into red LED coming up with much more distortion. The amplification is simply too low for the amount of bass this speaker produces.

Distortion measurements are not meant to show max SPL.

 

[infinitesymphony]

Distortion is excellent at low levels but becomes objectively an issue at elevated level but it was not something I noticed when listening.

So the takeaway is that the 708P can get much louder, and even though the distortion increases with volume, it is not enough to bring down the experience. Makes sense.

 

[richard12511]

If anyone is interested in the passive 708i, I have a mint condition pair for sale with matching (required DSP) amplification if needed. Pm me here or see my classifieds post on AVS.

About the preference ratings not matching subjective impressions, I think it's important to remember that coorelation is not causation.

An example that helped this to sink in for me was the following fact: the sale of ice cream is positively coorelated with the rate of pediatric head trauma. If you look at a graph of ice cream sales compared to pediatric head trauma, it looks like this:View attachment 110363
Common sense would tell you that buying Ice cream doesn't cause pediatric head trauma, but if all you looked at was the data, there could be no other conclusion.


Data is great... But don't forget to listen

Are we sure there's no actual causation there?

 

[richard12511]

Harmonic distortion isn't a good measure. Intermodulation distortion correlates much more with the perception of distortion. The jbl 708 will outperform the similar neumann and genelec speaker by a lot looking at the max spl with low imd. Jbl uses essentially pa chassis for the 708. The bigger genelec S360 are more alike, but with a narrower beam.

I would think it's the other way around, at least in comparison with the Neumann, given that the Neumann is 3 way, which usually shows much less imd.