Infinity Reference 253 Review (speaker)

[amirm]

Is the impedance measurement performed at 2.83V nominal? Is it possible to run this test at, say, 10V instead to see what gives?

Well, anything is possible but I am not lugging this heavy thing a third time up and down the stairs to measure it.

 

[amirm]

Hello thanks for the review, Amir. I'm a novice at this and the only EQ I have on my Onkyo receiver is a graphical EQ that I can do manually. From what I'm seeing it's best to boost the bass and mid range around 2khz and lower the peaks in the higher range to get a more satisfying sound?

Unfortunately no. Those equalizer controls are very inexact and too wide for this kind of work. That said, yes, if you change things and it sounds better to you, then by all means stay with them.

 

[restorer-john]

The highs are higher than what Harman shows so we still have some variation but not remotely not as bad as I had it before.

So the amplifier used in the Klippel rig had no problem with driving the speaker to high levels and yet the Levinson shut down. Perhaps there is a component in the crossover such as a parallel connected capacitor in a second order arrangement that is breaking down over a certain drive voltage and temporarily presenting a low impedance to the amp- enough to trigger the short circuit protection. It wouldn't show on the low level impedance sweep either.

 

[amirm]

So the amplifier used in the Klippel rig had no problem with driving the speaker to high levels and yet the Levinson shut down.

The Klippel measurements are smooth sweeps. It appears the problem here is on transients. The protection circuit in Mark Levinson is quite exotic. It checks for a million things and even has a web interface to tell you what it got unhappy about! With so much power on tap, it is a good thing that it makes sure everything is to its liking.

If I did not care about my amp (!), I could troubleshoot it but every time it shuts down, I get a mini scare. This ain't an amp to try to get into and fix!

 

[Doodski]

So the amplifier used in the Klippel rig had no problem with driving the speaker to high levels and yet the Levinson shut down. Perhaps there is a component in the crossover such as a parallel connected capacitor in a second order arrangement that is breaking down over a certain drive voltage and temporarily presenting a low impedance to the amp- enough to trigger the short circuit protection. It wouldn't show on the low level impedance sweep either.

There's only one way to know. Frequency sweep at max SPL on the Levison gear and see where the speaker fails.
If the Fr sweep can't find it then nothing will.

The Klippel measurements are smooth sweeps. It appears the problem here is on transients. The protection circuit in Mark Levinson is quite exotic. It checks for a million things and even has a web interface to tell you what it got unhappy about! With so much power on tap, it is a good thing that it makes sure everything is to its liking.

If I did not care about my amp (!), I could troubleshoot it but every time it shuts down, I get a mini scare. This ain't an amp to try to get into and fix!

The voice of reason. Best to take a properly measured approach.
I still think this bad boy should go in for warranty service or replacement.

 

[restorer-john]

If I did not care about my amp (!), I could troubleshoot it but every time it shuts down, I get a mini scare.

I agree. Protection events can go either way. The amplifiers are too valuable to use to troubleshoot a $160 speaker's vagaries.

 

[MZKM]

I’ll update my original comment, but:

Preference Rating
SCORE: 4.9
SCORE w/ sub: 6.7

 

[MZKM]

So the amplifier used in the Klippel rig had no problem with driving the speaker to high levels and yet the Levinson shut down.

Would be interesting to see at high voltage, but here is the EPDR at the given 2.83V:

Stereophile’s measurements of the amp show tons of wattage into 2ohm, but that’s at 1kHz, not ~120Hz. But I doubt this is the most difficult speaker in this regard that Amir has tested.

 

[peanuts]

this speaker is most likely matched with a cheap receiver, and it stomps a mark levinson. great

 

[Doodski]

this speaker is most likely matched with a cheap receiver, and it stomps a mark levinson. great

I'll bite; sigh.* Cheap and stomPZ? What does who how?

 

[restorer-john]

The Levinson has never shut down according to Amir into any speaker, likely at any level in 100 speaker tests. It squarely points the finger at the Infinity speaker as having the issue, unless there is a developing issue in the amp. It's a pity he didn't try the other monoblock to see if it behaved the same, but trust me, when your pride and joy power amp goes into protection, it makes you sweat bullets until it resets and appears OK. Even then, you feel bad for triggering the protection event.

These are seriously expensive power amps too. Not even me in cowboy restorer mode would subject such gear to multiple protection events.

 

[ROOSKIE]

this speaker is most likely matched with a cheap receiver, and it stomps a mark levinson. great

Well, thousands have been sold over the last 7-8 years and it has often been posted on deal sites where very normal people buy a set and connect to whatever they have. Complaints of this issue have not come up.
Looks like a bad unit vs a bad design.
It happens.

 

[Francis Vaughan]

The ML amp, as has been noted, has a comprehensive set of protection mechanisms. This is not just protecting the amp, but also monitoring for things that might damage the speaker. More than your average AVR does. And likely with more precision. So that it sees potential problems where other amps doesn’t isn’t a huge surprise. Something like a crossover capacitor with a voltage dependent fault might be convincing the amp that it is about to vaporise the speaker, at which point it decides to bail out.

 

[sam_adams]

@amirm this is a most unfortunate outcome. Possibly dropping a line to Infinity with the review mentioned could get a new unit for you to remeasure and perhaps they might send along Djambi to schlep the speaker up and down the stairs to save your back.

 

[beagleman]

The ML amp, as has been noted, has a comprehensive set of protection mechanisms. This is not just protecting the amp, but also monitoring for things that might damage the speaker. More than your average AVR does. And likely with more precision. So that it sees potential problems where other amps doesn’t isn’t a huge surprise. Something like a crossover capacitor with a voltage dependent fault might be convincing the amp that it is about to vaporise the speaker, at which point it decides to bail out.

Which, makes me wonder, could those complex protection circuits be the issue, or do they never fail or act oddly?

 

[Trouble Maker]

Which, makes me wonder, could those complex protection circuits be the issue, or do they never fail or act oddly?

I'm saying this as a controls engineer, granted not in the audio world, and not knowing any technical detail about this amp. In my mind as just a shot in the dark it's about as equally like there is something 'wrong' with the speaker as there is the amp tripping a false positive. Amirm has possibly put more real speakers on that amp than the makers do, and possibly more than any other customer. Everyone who makes products for consumer use has blind spots or simply misses some use case from time-to-time. Then has customers come to them and say 'hey, this widget is acting strangely'. If they are any good at what they do, they act to figure out why and how to prevent it in the future, if not retroactively through recalls or software update campaigns if this issue is bad enough. With the increase of simulation and even simulation with real environment or plants (speaker load in this case) that can be varied across the spectrum becoming more common, it can be easier to avoid these kinds of situations. It's also possible that something like this is a known false positive and they made an engineering decision that it's close enough to the real thing to warrant it. Maybe if they didn't detect at this threshold then by the time they could it would cause some damage. In my mind this is a less likely situation, it usually takes real work to make thees kinds of fringe cases and the likelihood Amirm found that, even after 100 speakers, is not likely.

I think it's easy to intuitively blame the speaker; the price and engineering complexity from an electrical/controls standpoint is magnitudes less. It's actually converse to this in my opinion, it's much harder to get the control systems in the amplifier right due to the complexity. How many of us have had speakers that are working 10+ years later VS an AVR or TV that some piece of electronics lost it's magic smoke and didn't work anymore?

But finally any talk here is all just speculation without any real root cause analysis without hands on the products and a way to test them. It's at least fun to run through the scenarios.

 

[beagleman]

I'm saying this as a controls engineer, granted not in the audio world, and not knowing any technical detail about this amp. In my mind as just a shot in the dark it's about as equally like there is something 'wrong' with the speaker as there is the amp tripping a false positive. Amirm has possibly put more real speakers on that amp than the makers do, and possibly more than any other customer. Everyone who makes products for consumer use has blind spots or simply misses some use case from time-to-time. Then has customers come to them and say 'hey, this widget is acting strangely'. If they are any good at what they do, they act to figure out why and how to prevent it in the future, if not retroactively through recalls or software update campaigns if this issue is bad enough. With the increase of simulation and even simulation with real environment or plants (speaker load in this case) that can be varied across the spectrum becoming more common, it can be easier to avoid these kinds of situations. It's also possible that something like this is a known false positive and they made an engineering decision that it's close enough to the real thing to warrant it. Maybe if they didn't detect at this threshold then by the time they could it would cause some damage. In my mind this is a less likely situation, it usually takes real work to make thees kinds of fringe cases and the likelihood Amirm found that, even after 100 speakers, is not likely.

I think it's easy to intuitively blame the speaker; the price and engineering complexity from an electrical/controls standpoint is magnitudes less. It's actually converse to this in my opinion, it's much harder to get the control systems in the amplifier right due to the complexity. How many of us have had speakers that are working 10+ years later VS an AVR or TV that some piece of electronics lost it's magic smoke and didn't work anymore?

But finally any talk here is all just speculation without any real root cause analysis without hands on the products and a way to test them. It's at least fun to run through the scenarios.

That was my thought, a speaker such as this is quite simple in design in relation to a microprocessor controlled, amplifier protection circuit.

Once a basic impedance plot is made (which would show many issues) issues at higher levels might be related to obvious speaker physical damage, but in my experiences speakers are far more robust and reliable than electronic circuits.

 

[Ralferator]

This is a review and detailed measurements of the Infinity R253 tower speaker. I purchased this months ago with one of those crazy sales prices JBL often has. Right now is one such occasion with the cost dropping to US $160 from list price of $400.

Considering what you get, the retail cost is hard to believe:

View attachment 96235

Dual woofers, mid-range, and tweeter in a nicely finished box for just $200? Stunning.

Back panel shows dual ports:
View attachment 96236

Measurements that you are about to see were performed using the Klippel Near-field Scanner (NFS). This is a robotic measurement system that analyzes the speaker all around and is able (using advanced mathematics and dual scan) to subtract room reflections (so where I measure it doesn't matter). It also measures the speaker at close distance ("near-field") which sharply reduces the impact of room noise. Both of these factors enable testing in ordinary rooms yet results that can be more accurate than an anechoic chamber. In a nutshell, the measurements show the actual sound coming out of the speaker independent of the room.

The system performed over 1000 measurements which resulted in error rate of more or less 1%.

Temperature was 61 degrees F.

Measurements are compliant with latest speaker research into what can predict the speaker preference and is standardized in CEA/CTA-2034 ANSI specifications. Likewise listening tests are performed per research that shows mono listening is much more revealing of differences between speakers than stereo or multichannel.

Reference axis is the center of the tweeter. I also tried it at mid-point between midrange and tweeter but it made no difference.

EDIT: something had gone wrong physically with my measurement mic in the original review. The cap on it had come loose and then proceeded to rotate 45 degrees, hanging by a thread! It was not that way when I started the measurements but somehow ended there. That changed the high frequency resonances of the mic, resulting in peaking that we had. I have remeasured the speaker and edited the data in this review now.

Infinity 253 Measurements
Acoustic measurements can be grouped in a way that can be perceptually analyzed to determine how good a speaker is and how it can be used in a room. This so called spinorama shows us just about everything we need to know about the speaker with respect to tonality and some flaws:

View attachment 96849

The on-axis response is pretty good from 100 Hz to about 1.5 kHz. Then it becomes a bit messy followed by shelving up starting at 4 kHz. Deviation is not huge but likely going to make the speaker sound bright without the delicious detail I like to hear around 1 to 2 kHz.

Directivity (how similar on and off axis response are) is good so equalization should work.

Because directivity is good, early window reflections show same issues:

View attachment 96850

Putting the two together tells us what we already know about predicted in-room frequency response:
View attachment 96851

We like to have more of a tilt than what we have here as otherwise the speaker will sound bright. Reduction of bass energy between 50 and 100 Hz will accentuate this brightness.

Here is the near-field response:

View attachment 96240

The tweeter level is higher but doesn't peak the same away. Perhaps that bowtie guard keeps it from doing this when measured in near field?

Distortion levels were disappointing as well:

View attachment 96241

View attachment 96242

Inclusion of midrange provides good directivity:

View attachment 96852

View attachment 96859

Other than that dip around 2 kHz, that is excellent response.

Here is our vertical response:

View attachment 96865

You have much more degrees of freedom as compared to 2-way bookshelves. Still, best to stay at or below tweeter axis as otherwise you get more of that 2 kHz dip.

Impedance graph shows a couple of resonances although I don't see a sign of the first one in the spin:

View attachment 96246

Infinity R253 Speaker Listening Tests
In the original review, I failed to get rid of the brightness in this speaker despite using my measurements as a rule. The new measurements show why I went wrong as we just need a broad shelving filter to get rid of the brightness:
View attachment 96866

I then boosted up the mids around 2 kHz and that provided a bit more detail in that critical region.

I really liked the sound so I started to click to my playlist only to be greeted with my Mark Levinson No 53 shutting down again. Powered back on and bam, it happened again. Switched in an old Infinity bookshelf speaker. No problem whatsoever. Put the R253 in, turn up the volume and shut down again. Switch to my Revel Salon 2 speakers. No problem with even ear bleeding levels. Something in this speaker is causing the protection in my amplifier to turn on. Having tested nearly 100 speakers with the same amplifier, it definitely is not due to the amp being overly sensitive.

If it had stayed running, I would have likely added in some bass to level the 40 to 100 Hz as that worked well before.

Conclusions
The Infinity Reference 253 is one hell of a package. You couldn't build one quarter of it for what they are selling it for right now. The new measurements show excellent performance, sans two bits: a dip around 2 kHz and too much treble energy. Both were easily fixable with EQ. Alas, I could not listen for long due to it not getting along with my amplifier. How much this issue translates to others using the speaker, I don't know.

The faults above keep me from recommending the Infinity R253 even though it can be a killer speaker if you use EQ and it doesn't shut down on you. You have the data to decide for yourself.

New spin data is enclosed.

------------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

Could it be that Soundandvision had similar problems when measuring the Infinity R162?

 

[Laserjock]

The Klippel measurements are smooth sweeps. It appears the problem here is on transients. The protection circuit in Mark Levinson is quite exotic. It checks for a million things and even has a web interface to tell you what it got unhappy about! With so much power on tap, it is a good thing that it makes sure everything is to its liking.

If I did not care about my amp (!), I could troubleshoot it but every time it shuts down, I get a mini scare. This ain't an amp to try to get into and fix!

It shut down on the second sample speaker also?

 

[amirm]

It shut down on the second sample speaker also?

I only have one speaker to test.