JBL 705P Studio Monitor Review

[TimVG]

And if grandma had wheels she'd be a car. It's a moot point, the speaker is what it is. I find the on-axis problem, despite being correctable, a greater issue since that -will- be audible and in this price range it should not be up to me to correct a speaker that has a built-in DSP engine.
Besides, iIt's already proven that low Q anomalies are more audible than high Q ones as explained in Toole's book, based on real-world tests.
Having owned the 705i, of the things I found lacking - the imho inconsequential dip was not one of them. As I said: YMMV.

 

[Francis Vaughan]

If the bass reflex port had been moved from the front to the rear, the resonance at 750Hz might have been in phase instead of phase opposition and less pronounced.
In this case the resulting hump could have been reduced by means of a filter (the disadvantage is that wall mounting would no longer have been possible).

Wouldn't work. The phase relationship between the cone and the port output would remain the same, and by moving the port exit you are just adding an acoustic delay, making the final phase between the two heavily dependant upon listening angle. Actual total energy into the room would remain the same.

It is another just plain weird design thing. There are a number things that could be done to kill the resonance. If the speaker was used with a sub, just plug up the port and use a filter section to provide a Linkwitz transform to flatten the resultant bass alignment.

Whilst in use it is probably not all that noticeable, it would not be hard to create a piece of music that perfectly illustrated the issue, at which point it would probably be difficult to unhear the effect. A flute playing a scale passing through F# would be quite clear. Rock and roll: impossible to tell.

 

[ctrl]

off topic

Wouldn't work. The phase relationship between the cone and the port output would remain the same, and by moving the port exit you are just adding an acoustic delay, making the final phase between the two heavily dependant upon listening angle. Actual total energy into the room would remain the same.

You're right, by moving the bass reflex port to the rear it shows a delay in relation to the woofer. Which is nothing more than a phase shift.

For the very low frequencies this delay makes hardly any difference due to the large wavelength.
But at 750Hz the wavelength is only 0.46m, i.e. if the bass reflex port is shifted backwards in relation to the woofer the phase shift has a big influence.

This means that the "total energy" in the low bass hardly changes (again because of the large wavelength), but with increasing frequency the sound pressure change becomes bigger and bigger when the bass driver and the bass reflex port are shifted.

Shouldn't a phase shift of 90° bring a clear improvement (180° phase shift would only make a bump out of the notch)?

Update: In addition, the sound bundling at 750Hz, with a rear radiating bass reflex port, would further reduce the effect of port resonance.

 

[Deleted member 2944]

Thank you for that...been trying to sort out some aes issues...let me get that plug...

Cheers.

Unfortunately, you can't fix this issue without getting inside the speakers and removing the 110 ohm terminators of all the speakers in the chain. You could then attach an external terminator plug on the last speaker in the chain.

Dave.

 

[mac]

Same owner has a 708P so that is possible once I return these.

It's possible that the owner has my old set of 708P's. If he's local to Seattle.

I really enjoyed my time with the 708P. Best small speaker I've had in my listening room.

 

[amirm]

It's possible that the owner has my old set of 708P's. If he's local to Seattle.

He is not local.

 

[xyvyx]

I also really love my 708's.. I've got them wall-mounted in my office about head-high (for clearance above my PC's monitor), then tilted to keep them on-axis. I don't think they really needed a sub, but I added one to the mix, then attempted to smooth out the response w/ my RME's PEQ. I'm pleased with the results. If I had deeper pockets & were to redo my home theater setup, I'd love to have 3 of these across the front.

I'm a REW-noob and didn't do this w/ the MMM technique, but I tried to keep everything consistent between passes:

https://imgur.com/a/9t4h7b2

 

[BDWoody]

I also really love my 708's.. I've got them wall-mounted in my office about head-high (for clearance above my PC's monitor), then tilted to keep them on-axis. I don't think they really needed a sub, but I added one to the mix, then attempted to smooth out the response w/ my RME's PEQ. I'm pleased with the results. If I had deeper pockets & were to redo my home theater setup, I'd love to have 3 of these across the front.

I'm a REW-noob and didn't do this w/ the MMM technique, but I tried to keep everything consistent between passes:

https://imgur.com/a/9t4h7b2

That's got to surprise a few people when they come in...very nice way to indulge yourself!

Which mounts did you use?

 

[xyvyx]

That's got to surprise a few people when they come in...very nice way to indulge yourself!

Which mounts did you use?

Haha.. yeah, they're kinda chunky & stick out more than I like.

Those are the Allen Products MM-060... overpriced, but well-built. I mean, I thought I was getting a pair of 'em when I first ordered! One of the threaded inserts (M8s I believe!?) required a longer screw to get any grip...

 

[gr-e]

Someone on MiniDSP forums posted the filters that are supposed to be used with the passive versions (705i and 708i):

Spoiler: 705i single wire

Spoiler: 705i biamped

Spoiler: 708i single wire

Spoiler: 708i biamped

The HF delay is 0.073ms

 

[richard12511]

@amirm How would you rank these speakers in comparison to the Ascend Sierras? The objective measurements look about equal to me, or maybe slightly worse(harder to eq?), but your subjective impressions seem way more positive than they did in the Ascend review. I was interested in the Ascends for a bit, but I've never heard a RAAL ribbon. Also, I've been interested in picking up a pair of 708p for a few years. I have 305ps that I use in my office and I think they sound wonderful.

 

[jhaider]

Ceiling and floor reflections. The score and the massive crossover null and poor vertical dispersion don't seem to be at the cost of anything other than engineering time, and the IN-8 doesn't seem to have poor horizontal dispersion at the cost of vertical... Just seems to be a cheaper.

Massive crossover null? The vertical dispersion is a function of geometry, namely center-to-center spacing.

Possible but I have no idea. I don't think I've ever seen an oval coaxial. I have no idea how that might work.

KEF sold an Andrew Jones designed 6x9 car audio Uni-Q for KEF.

The JBL 70XP speakers don't properly configure a termination scheme where the last speaker (and only that speaker) has the proper 110 ohm AES terminator.

Wow. It's amazing what lurks behind the curtains even in gear from "good" brands.

For professional use the speaker is out of question at all. Nobody would buy an instrument where the F#5 would only play almost half as loud as all other tones.

If this speaker is "out of the question at all" for studio use, then every recording should be destroyed. Every recording in history was probably mixed on loudspeakers with far worse performance.

The decay problem at 750Hz that Amir showed in the CSD diagram is not even considered.

Yes, that is correct. It is not worth considering.

We don't even know if the error can be completely eliminated by closing the bass reflex port.

You may not. The physics however are known to others.

The bump between 800 and 1700Hz on the other hand, is relatively unproblematic and can easily be eliminated by a suitable filter.

The distinction you're not grasping is that the port resonance is documented in JBL's factory measurements. The bump thereafter deviates from spec. Everyone who buys them who cares to find their measurements will know that's there. Likewise, people who chose not to buy them because they listen with their eyes (but their eyes are unskilled listeners) will keep harping on that port resonance. However, Amir's measurements deviate pretty substantially and in a material way from the published specifications. That is a real problem.

 

[amirm]

@amirm How would you rank these speakers in comparison to the Ascend Sierras?

From memory, the JBLs are far better.

 

[stevenswall]

1. Massive crossover null? The vertical dispersion is a function of geometry, namely center-to-center spacing.


2. If this speaker is "out of the question at all" for studio use, then every recording should be destroyed. Every recording in history was probably mixed on loudspeakers with far worse performance.

3. You may not. The physics however are known to others.

1. My mistake, the port issue seems to be causing the null. I'm used to saying crossover null and mis-typed.

2. I wouldn't destroy recordings made with these or other bad speakers, but I do think a lot of music is ruined, whether it's The Who album I have that will never sound better than a cassette tape, The Beatles stuff that just isn't very clear, or the smashed modern music whose dynamics will never be recovered. I wonder if there are any recordings that obviously were mastered on colored speakers? Seems like the NS10 was fairly common and things turned out fine.

3. If only JBL and a lot of other companies knew those physics, and how to design ports so that this wouldn't be a conversation. If this is an issue an not a unit specific defect, then I'm glad people are saying it's inaudible. Curious though if you would do that every 50hz or so and if it would still sound right with that "comb filter."

 

[ctrl]

For professional use the speaker is out of question at all. Nobody would buy an instrument where the F#5 would only play almost half as loud as all other tones.

If this speaker is "out of the question at all" for studio use, then every recording should be destroyed. Every recording in history was probably mixed on loudspeakers with far worse performance.

That's a "refreshing" logic, because in the past recordings were mixed on crappy loudspeakers, in the present you should use loudspeakers with obvious faults as a tool for professional recordings too... yeah!

We don't even know if the error can be completely eliminated by closing the bass reflex port.

You may not. The physics however are known to others.

... and will you share your knowledge with us "ignorants"?
Simply saying the opposite without justification does not help anyone (especially since the forum is called "audioscience...").

By closing the bass reflex port the standing wave inside the cabinet does not disappear, only the part of the wave that is transmitted through the bass reflex port disappears.

There are near-field measurements from Sound&Recording that show this fact.

In red the overall frequency response of the 705P is shown. In green the bass reflex port alone, in blue the frequency response of the mid-bass unit.

The overall frequency response of the 705P drops continuously because the diagram shows the nearfield measurements without baffle step correction (with correction the bass would be reduced by 6dB, which would then correspond to Amir's measurement).

In the overall frequency response of the 705P the notch at 750Hz is about 10-11dB. The part that is only radiated by the mid/bass driver is about 7-8dB.
After these measurements closing the bass reflex port would only "fill up" the notch a bit.

Shouldn't a phase shift of 90° bring a clear improvement (180° phase shift would only make a bump out of the notch)?

I have to correct myself, this statement is nonsense. The measurement of S&R suggests that a phase shift of 180° (of the port frequency response) by cleverly moving the bass reflex port backwards or downwards would reduce the notch.

 

[Francis Vaughan]

The port would appear to be simply resonating as a pipe. Energised by the driver when there is signal at that frequency, this resonance will affect the driver as well, and is why it is apparent in the driver response. Closing up the port would eliminate the entire effect. Which is why I mentioned closing the port and using a subwoofer would solve it.

It is clear JBL have been balancing a number of competing factors. They want a small speaker, one that can be in-wall and near-wall mounted, and one with a small face. This all leads to a rectangular port on the front. Then they have put a lot of effort into making the speaker play loud. This means they need to pay a lot of attention to port noise. Clearly the port's geometry is directed to avoiding turbulence at what are likely to be quite high air velocities. The unwanted result is a port with a strong resonance at 750Hz. But balanced against the advantages the design has, they clearly decided to go with it. Very hard to blame them.
One cn be sure this design was subjected to lots of unsighted listening tests using the swapping robot, comparing with competing options for the design. It almost certainly doesn't make the sort of difference the nasty looking graphs suggest.

 

[ctrl]

The port would appear to be simply resonating as a pipe. Energised by the driver when there is signal at that frequency, this resonance will affect the driver as well, and is why it is apparent in the driver response. Closing up the port would eliminate the entire effect.

So you assume that the resonance is not a standing wave inside the cabinet, but a standing wave in the bass reflex tube.

For the first natural oscillation inside the cabinet and at the bass reflex port (open tube) L = c/2*f (L length, f resonance frequency) applies.

This results in a length of 0.23m (9.1 inch). This corresponds quite well to the internal height of the cabinet.

If the resonance at 750Hz would be the bass reflex tube resonance, the bass reflex tube would also have to be 0.23m (9.1 inch) long - is there any indication for this, on the pictures I can hardly estimate the bass reflex tube length.

But if you're right, the resonance would really disappear completely when the bass reflex port is closed.

 

[QMuse]

The port would appear to be simply resonating as a pipe.

As resonance is happening at 750Hz where wavelength is 44cm it is easier for me to imagine it is happening inside the box than inside the port pipe.

 

[Francis Vaughan]

The port is tapered/flared, and one would need to add end effect corrections (usually about 1/3 the diameter if the port was circular). Estimating the port resonance of a flared square port isn't going to be easy, but 750Hz isn't unreasonable. The Q must be quite high to get the effect seen, and that would tend to point towards the port.

A simple circular diameter pipe of say 5cm diameter would be about 19cm long to have a resonance at 750Hz. Tapering it could cause all sorts of interesting changes.

 

[TimVG]