Revel M16 Speaker Review

[QMuse]

I would personally be interested in knowing what causes the 5khz bump; if it's something caused by the geometry of the tweeter/waveguide or there’s something in the woofer here (doubtful unless the crossover slope is shallow).

This is Listening Window, Early Reflections and Sound Power responses magnified. What 5kHz bump are we speaking about?

 

[hardisj]

Despite your doubts, I suspect that the bump is coming from the woofer cone-breakup.
This also corresponds very well with the increased third order harmonic distortion HD3 at 1.5-1.8kHz.

With cone-breakup the bump can be 10-15dB and depending on the driver it can even increase under angle.
That's probably the reason why the bump under angle hardly gets weaker.

This bump must be specially treated during crossover to avoid negative effects.

Don't say that this chassis was used, but the Dayton DSA175 is a good example of such a case:

View attachment 52967

The increased HD3, caused by the heavy cone-breakup, is the reason why metal cone chassis have such a bad reputation. All too often, the crossover is too high, causing high HD3 - as in this case.

https://www.parts-express.com/dayto...designer-series-aluminum-cone-woofer--295-528
https://audioxpress.com/article/test-bench-dayton-audio-dsa175-designer-series-6-5-woofer

1) Are you certain that driver is the one used here? In other Revel models SB Acoustics drivers are used. I googled "revel m16 dayton" but didn't come up with any results that lend itself to proving this is the case.

2) My doubts are predicated on the fact they state the crossover point is at 2khz. Assuming the slope is LR2, even then that would put the breakup (which is only about 5dB above the mean SPL) at around 12dB down from the mean at 5khz. IOW, I don't see the bump in the above graphic being the culprit. I don't know. I just don't see it.

 

[ctrl]

Oh.. So Harman engineers made a beginners mistake which most DIY manage to avoid by choosing a woofer that would experience cone breakup within it's working range? Is that really what you believe is the case?

Yep!
It is not a beginner's mistake, but a result of making the loudspeaker as inexpensive as possible.

 

[hardisj]

This is Listening Window, Early Reflections and Sound Power responses magnified. What 5kHz bump are we speaking about?

View attachment 52971

The one on-axis. Even your various measurements above show a broader Q bump centered around 5khz, though.

 

[amirm]

What needs to be measure is the new JBL HDI line. In fact I’d argue that it is the most important speaker in 2020 to measure.

If it is, you and a few others need to get together to fund the purchase of one and send it in for testing. If it is not worth it to you to do that, then please don't complain about what I test. I get to decided to do that with my own money, and others do with what they buy to send me, or loan. That selection will be whatever it will be.

Keep in mind that to attract much higher end speakers to test, the value of what we do needs to spread to many more people and companies. The more speakers I test, the more it helps in this regard.

 

[amirm]

Also wish you could compare Ascend Sierra 2 in the same room.

The Ascend has been returned to its owner already.

 

[BurgerCheese]

Oh.. So Harman engineers made a beginners mistake which most DIY manage to avoid by choosing a woofer that would experience cone breakup within it's working range? Is that really what you believe is the case?

They certainly knew what they were doing, but it's a matter of cost. More advanced (expensive) 6" metal woofers are able to push the cone break up higher in frequency, but this is a relatively inexpensive speaker.

 

[q3cpma]

That breakup is quite bad, and is there a reason that ~300 Hz peak isn't mentioned other than as a resonance? 2% THD is a bit worrying. Too bad, it seemed almost perfect; well, it should be for that price.
Looks like using aluminium for a 2-way isn't a good idea.

 

[QMuse]

The one on-axis.

SInce when is on-axis more relevant than listening window, early reflections and sound power? It doesn't even contribute to the predicted in-room response.

Even your various measurements above show a broader Q bump centered around 5khz, though.

I don't see it on the magnified graph I posted. Can you please circle it on the graph?

 

[Morpheus]

Looks great, though I would like to see it do better on the distortion side. Is it ok to compare to the Adam S2V? Please note different y-scale.
View attachment 52934View attachment 52936

Distortion audibility is a controversial topic, but I am a bit puzzled such an abrupt rise at our most sensitive range, with 3rd order harmonics too, didn't raise an eyebrow on the measurements front, or draw attention on the listening test, especially when voices in particular were so highly praised...Also, bass extension is a bit limited for 900 dollars, although well judged, as well at that great tweeter/waveguide/crossover directivity management..So, all in all, nice to see Amir is no robot and does let a bit subjectivity creep in

 

[amirm]

sidetrack a bit are there any plan to measure the new Adam audio T5V or T7Vs?

Someone was going to loan me one but backed out due to shipping costs. I can buy one but I have bought so many already that I need to test first.

 

[hardisj]

SInce when is on-axis more relevant than listening window, early reflections and sound power? It doesn't even contribute to the predicted in-room response.

Yikes. I didn't say it was. I didn't say it wasn't. I said...

I would personally be interested in knowing what causes the 5khz bump; if it's something caused by the geometry of the tweeter/waveguide or there’s something in the woofer here (doubtful unless the crossover slope is shallow).

I don't see it on the magnified graph I posted. Can you please circle it on the graph?

I'm unable to do that at the moment. But look at 5khz. Bump in all your quoted measurements between 4.5 - 6.5khz.

Then go back and look at the on-axis response. It would seem that the peak around 5khz could be causing this broad bump. Or it could be that it's simply situated between a mild lull and a high-Q dip. Again, can't know without breaking it down in to components which is why I stated what I did about being "interested in knowing what causes the 5khz bump".

 

[amirm]

I get the impression by wording and the panther that you like these better than the genelecs? Why?/Why not?

The Genelec stopped me cold the moment I started the listening test with its protection light coming up and distortion heavily setting in. With the Revel M16, distortion comes in but is gradual and there is no cliff like there was with Genelec.

I also did not try to test the Genelec far field. Maybe it would have made a better impression then.

In general, perfection cannot be had when it comes to manual labor on my part, or subjective conclusions reached.

 

[amirm]

Perhaps. If Amir is planning to keep these speakers he can directly compare against a bigger Genelec or other speakers with comparable output capabilities to check if he's reacting to something.

I purchased this speaker myself and yes, I plan to keep it as a reference for the future.

 

[amirm]

Impressive results, 1k$ in the HIFI world is the price some people paid for their cables.

Just to add to this, Revel is a luxury speaker line. As such, it has healthy margins for the channel that does not exist when you buy a budget speaker from an online retailer. Compared to other speakers sold the same way this is some of the cheapest speakers you can buy.

 

[amirm]

I also have JBL Studio 530s and found them of similar quality and soundstage in my tight space. Would be interesting to see those measured given their reputation among users and frequent value price.

I bought a pair of 530s and it has been shipped already. So that review will be coming.

 

[ctrl]

This is Listening Window, Early Reflections and Sound Power responses magnified. What 5kHz bump are we speaking about?

That breakup is quite bad, and is there a reason that ~300 Hz peak isn't mentioned other than as a resonance? 2% THD is a bit worrying.

The resonance mainly causes second order harmonic distortions, which are perceived as less disturbing.

But without question, such a "bug" must not be allowed to appear in this price segment nowadays.

Looks like using aluminium for a 2-way isn't a good idea.

As others have already mentioned, there are many metal cone drivers that are two-way capable.
For example:
http://sbacoustics.com/product/6in-sb17nac35-4/

 

[amirm]

That's several speakers now that have trusted 3rd party measurements where your Klippel displays high frequency rolloff against their measurements which show a rise in high frequencies. It seems plausible that your Klippel may be attenuating high frequencies above 10 kHz? @amirm

Anything is possible but I already checked for one factor that may cause it and it was not (too few measurement points/expansion order). Keep in mind that other speakers of recent design have shown flat response:

That said, I too am surprised by the last few outcomes and have an eye out for it.

Ultimately, the only way we know who is right and who is not is to test the same speaker sample both in anechoic chamber and my system. Until that happens, you should assign similar probability of error to all measurements, not just mine.

 

[QMuse]

View attachment 52972

I don't think you are reading these curves right. All those curves, along with vertical reflections measurement curves, have been accounted for in Early Reflections curve with appropriate weighting, and in Early Reflections curve there is no peak at 5kHz.

 

[q3cpma]

As others have already mentioned, there are many metal cone drivers that are two-way capable.
For example:
http://sbacoustics.com/product/6in-sb17nac35-4/

Hmm, indeed. It's just that I don't see any point of not using well treated paper for woofers. What do you gain?