Revel F228Be Review (Speaker)

[richard12511]

This is probably because revels are expected to be used in untreated rooms and use room gain to make up for the deficiency in ~60hz bass. KEF's are also tuned this way.

This is what I used to think, but unless I'm misreading the "Simulated Room Gain" thread, room gain isn't really a thing, unless your room is the size of a car. It seems like the ideal bass response should probably be a a flat line(with no drop at all). Seems like active design or passive plus EQ are only ways to achieve this(by limiting output)?

 

[Rottmannash]

Would a Purifi 1ET400A amplifier with 425 W @4 Ohms and 227 W @8 Ohms be sufficient to drive these speakers at sufficient volume without degrading the low end?

It drives my F208's to ridiculous levels.

 

[phoenixdogfan]

It drives my F208's to ridiculous levels.

Good to hear, mine is in the shop, I went DIY and apparently there were issues with the SMPS. should drive my LS 50 Metas quite well.

 

[Francis Vaughan]

I would think the tuning of the port is very low. This sacrifices sensitivity but the extension is lower. With the Revel you can have usable output to 30Hz wheras with a higher Tuning (like 50Hz) you would have a better sensitivity but you could not get any meaningful output below that even with EQ. Group delay is also lower for the midbass region and the speaker acts more like a sealed cabinet from 80Hz on.

I think this is part of it.

The impedance plot, Amir's rough port output measurement and the overall response all put the port resonance at 28Hz. Which is low. I'm mostly puzzled by the drop off point. Indeed I would expect the response to mirror that of the F208. An here is the clue. The 1.5dB lower sensitivity in the F208 will help it, and that does indeed seem to pretty much cover the difference. I overlaid the two, and they track almost identically, until the higher sensitivity of the F228Be takes it higher in level still tracking the overall slope. So, a slightly odd choice by the designer, but brought upon the speaker by the higher sensitivity. The higher overall sensitivity and sound output in the Beryllium series results in a less flattering bass rolloff. At least from the point of view of the simple measurements. The final in-room result is only loosely dependant upon this, and the ultimate result seems more than satisfactory.
Which is an interesting lesson in worrying about the raw bass measurements. Pundits dismissing the speaker based upon these bass measurements need to take note.

Harmon's actual published figures remain curiously out of whack, but it doesn't really matter here. Odd however.

 

[beefkabob]

I remember reading a report on the 705p, 708p, and M2 at a show booth. All three were setup, and music was playing loud. Then the a talk began, and everyone thought the M2 had been playing but it was the 705p. I say this because I wonder how much the impressions of size are just psychoacoustics and not actual acoustics. It's a big speaker so it sounds big. Hopefully you can do some blind testing to figure it out. Try them with the smaller speakers at different heights, maybe matching the tweeter height, to see how it matters. Can it be measured at a distance of 2m? That kind of thing. Our ears are points, more or less, not 1 meter in height.

Do your ears hang low...

 

[Deleted member 31750]

Mostly confirms the finding from Sterophile but got that’s an ugly looking speaker. At least in white.

 

[laudio]

The 6.2 rating speaks volumes. "Alas, the customer can no longer afford them so if you are interested, let me know " How much? They look fantastic, for most folks an end game speaker. Any discounts? Sorry, but you're a review site and also selling them I ignored all the techno babble in the thread.

 

[laudio]

Mostly confirms the finding from Sterophile but got that’s an ugly looking speaker. At least in white.

Is that the best you got?

 

[amirm]

The 6.2 rating speaks volumes. "Alas, the customer can no longer afford them so if you are interested, let me know " How much? They look fantastic, for most folks an end game speaker. Any discounts? Sorry, but you're a review site and also selling them I ignored all the techno babble in the thread.

Yes, significant discount. Start a conversation with me if you are in US and interested. Act quick though as we are trying to get one of them returned directly to Harman. On top of the great price, availability is horrible on Revel speakers due to high demand and low production. So if you want one, this is the time.

As to "selling them," I only respond to members reaching out with interest. Otherwise, there is no promotion or attempt to sell any products on this site related to Madrona Digital (our company).

 

[Deleted member 31750]

Is that the best you got?

What do you mean? are You interested what I make of these measurement? If that’s the case then I could say that the speaker performs well from measurements point of view. It has probably some cabinet resonance when the impedance curve crackles and it deeps way to low for my system and what I am looking for. Also is too flat for my taste. A well engineered design that is probably not my taste. the fact that I find it ugly is personal, like the sound signature of a speaker. i Wouldn’t like to have this speaker in my living room but all of the above doesn’t mean this is a bad speaker in any way shape or form. just not for me

 

[beagleman]

This is what I used to think, but unless I'm misreading the "Simulated Room Gain" thread, room gain isn't really a thing, unless your room is the size of a car. It seems like the ideal bass response should probably be a a flat line(with no drop at all). Seems like active design or passive plus EQ are only ways to achieve this(by limiting output)?

It looks like the OVERALL response is not taking into consideration the Ports output at all.

The frequency response seems to roll off below about 80 Hz, which would be almost impossible. (as to how they sound in person in a room)

In the nearfield measurements the port adds quite a bit down to around 30 hz or so, which would be more likely about where they would be down maybe 3 or 6 db.

 

[Xyrium]

This is what I used to think, but unless I'm misreading the "Simulated Room Gain" thread, room gain isn't really a thing, unless your room is the size of a car. It seems like the ideal bass response should probably be a a flat line(with no drop at all). Seems like active design or passive plus EQ are only ways to achieve this(by limiting output)?

Unless he's referring to room modes, if they are planning to tune a very precise seating position to take advantage of them.

 

[sarumbear]

I am still puzzled by the measured bass response. Mostly because it doesn't look like a proper bass-reflex response. It isn't clear what is going on, but the fundamental physics of how one would assume it has been designed doesn't add up.

I think I can see what’s going on but why it’s done I have no idea.

The port is tuned to 30Hz but the woofer has a F3 at around double that number. Too high for a standard reflex tuning. Hence, the speaker acts like a closed box speaker until the port takes over. However, as by then the energy inside the box is low the port cannot compensate the roll-off fully. The result is like a small speaker with a subwoofer working at a low level.

Interesting design, but why? What was the benefit?

 

[pma]

I think I can see what’s going on but why it’s done I have no idea.

The port is tuned to 30Hz but the woofer has a F3 at around double that number. Too high for a standard reflex tuning. Hence, the speaker acts like a closed box speaker until the port takes over.

Yes and the measured frequency response supports this explanation. Seems that the sacred Harman engineers made a mistake?

 

[Spocko]

I have Revel F36 and are/were dreaming of a xxxBe model for better directivity and avoid the aluminium tweeter break-up ringing at 22k.
But I was hoping Revel would fix the 7-9khz dip in the treble for 228, I am not too interested in buying a premium speaker with such deviations. I do not want a worse treble...Revel should to improve the tweeter/waveguide , beryllium only is not enough. Maybe I should dream of a KEF instead...

By the way, you need to go to the Be range to avoid electrolytic capacitors in the crossover...

Here is my F36 in room, with and without "room correction DSP". Mooving Mic Method Pink noise RTA 1/48. DSP Target curve in bottom of plot

View attachment 131722


Single point measurement:
View attachment 131724

F36 Tweeter check. Singe point sweep 1/6
View attachment 131726


The Be tweeter in 228 does not seem to be worth the money

Wait, can you hear the tweeter break-up ringing at 22Khz, or rather, how does this flaw affect the perceived listening experience? I understand how it looks on the chart but what does this really mean in terms of subjective real world listening impressions, ie reduced imaging, soundstage not as expansive, high hat rings longer than it should, etc.

 

[sarumbear]

Yes and the measured frequency response supports this explanation. Seems that the sacred Harman engineers made a mistake?

I doubt it’s a mistake. I thought on this a bit more and I can think of a reason. One can argue that it’s not possible to get a flat frequency response in the 30-70Hz range in a room, hence you can sacrifice frequency linearity in favour of extended bass response.

 

[Francis Vaughan]

I think there is a mix of things happening here. Little doubt the low port frequency is deliberate, and the alignment is not intended to be maximally flat into the bass. Harman quote different number for the bass roll off than measured, and there is still a curious discrepancy here. Moreover, thinking about it, earlier I wrote I had laid the F208 over the F228Be, and got a good match. Both of those measurements were by Amir's Klippel. So any common issue will remain common.
I need to sleep on this, and maybe do some modelling later. It is past midnight here in Oz. Maybe tomorrow.

 

[FeddyLost]

The port is tuned to 30Hz but the woofer has a F3 at around double that number. Too high for a standard reflex tuning.

Stereophile measured even 25 Hz
Port is frontal, so if it would be equal to woofers' output, some artifacts may be heard.
And they needed to have 23 Hz (-10 db) number in datasheet without sacrificing too much of sensitivity, i think.

 

[sarumbear]

...Port is frontal, so if it would be equal to woofers' output, some artifacts may be heard.

What can those artefacts be do you think?

 

[B4ICU]

For a $10k the low band is poor. I would not use EQ for that kind of price, and also a sub is kind of weird addition.
$10k? I don't know now, but for that kind of money, you could buy a Klipsch Horn! A speaker that sounds as live music and not like a speaker.