• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Revel PerformaBe F226Be Floorstanding Speaker Review

Chromatischism

Major Contributor
Forum Donor
Joined
Jun 5, 2020
Messages
4,765
Likes
3,703
That's always been the selling point of electrostatics--their ability to be able to resolve the lower volume details on the recording. Kind of analogous to the resolution an OLED video display has for low light, shadow level details.
What does that really mean, though? Is it just the super narrow dispersion avoiding interference?
 

Beave

Major Contributor
Joined
May 10, 2020
Messages
1,329
Likes
2,728
I had an hour-long skype convo with their CEO back in July and he was ready to send me a set. They apparently have a couple out for review that haven't gotten back to them yet. One of their guys contacted me on FB last month but the situation with COVID is still causing delays. But, it's in the pipeline. I just hope I can get a pair before winter sets in because once that happens, my ability to conduct outdoor testing (which is a priority) diminishes drastically.

Good luck dealing with those brutal Alabama winters! ;)
 

yourmando

Active Member
Joined
Aug 14, 2020
Messages
150
Likes
178
Really top notch review as usual!

One thing that got me curious is this observation:

Using Dirac Live in my home theater resulted in poorer performance. I ran DL full range which corrected response from 20Hz to 20kHz. I believe, in the case of these speakers, it is best if DL is kept to a minimal passband (below Schroeder frequency).

I’d be interested to hear any thoughts you or others here have about why DL room correction seems to hurt performance here.

I get that you generally want a speaker with well controlled directivity and impeccable on and off axis frequency response, leaving room correction for below Schroeder frequency.

And that correcting above Schroeder can make things worse because it can’t fix directivity problems, making one axis flat but another worse.

But, F226Be seems to have excellent directivity control and is already pretty much within +-1.5 dB of flat. So I’m struggling to understand how room correction correction could make things worse. For one, it wouldn‘t have to do much since the curve is already pretty flat, unless the target curve is different. Then it would be a matter of adjusting the target curve to match, revealing it’s just a house curve preference issue? Or could it be something else?
 

Absolute

Major Contributor
Forum Donor
Joined
Feb 5, 2017
Messages
1,084
Likes
2,125
Great review, it makes me happy to see such an effort just to provide us ungrateful bastards with more nice graphs to bicker about! Well done, @hardisj !

On another note, I'm with @tuga on the matter of Be vs non-Be in terms of tuning and linearity. The Be is tuned brighter, and according to my own personal preferences I'm convinced I'd like the non-Be tuning better, at least over time. There's also a resonance smack bang in the most sensitive area of the ear at around 4 kHz that makes me uncertain whether or not the Be is actually an upgrade. Could it be responsible for that ringing behavior that tuga demonstrates?

Revel F226Be Horizontal FR.png



Other than that, wow! What a speaker!
 

Icboschert

Active Member
Forum Donor
Joined
Aug 2, 2018
Messages
139
Likes
197
Location
Wisconsin
@hardisj not to be the comparison guy but I imagine these floor standers throw a bigger wall of sound than the Selahs you reviewed? I'm trying out some NHT C3 speakers now crossed over to subs and I'm just not getting large scale sound like I've heard from a brief Revel F36 demo. Wondering if I need to step up to floor standers or maybe a bookshelf with a more powerful driver like the Purifi might do it.
 

Chromatischism

Major Contributor
Forum Donor
Joined
Jun 5, 2020
Messages
4,765
Likes
3,703
Really top notch review as usual!

One thing that got me curious is this observation:



I’d be interested to hear any thoughts you or others here have about why DL room correction seems to hurt performance here.

I get that you generally want a speaker with well controlled directivity and impeccable on and off axis frequency response, leaving room correction for below Schroeder frequency.

And that correcting above Schroeder can make things worse because it can’t fix directivity problems, making one axis flat but another worse.

But, F226Be seems to have excellent directivity control and is already pretty much within +-1.5 dB of flat. So I’m struggling to understand how room correction correction could make things worse. For one, it wouldn‘t have to do much since the curve is already pretty flat, unless the target curve is different. Then it would be a matter of adjusting the target curve to match, revealing it’s just a house curve preference issue? Or could it be something else?
My opinion, and with some experience to back it up, is that the better the speaker is, the easier it is to screw it up with room correction.

It's easier to make a straight line not straight.
 

ElNino

Addicted to Fun and Learning
Joined
Sep 26, 2019
Messages
557
Likes
722
My opinion, and with some experience to back it up, is that the better the speaker is, the easier it is to screw it up with room correction.

It's easier to make a straight line not straight.

This is also a very low nonlinear distortion speaker. It's possible that room correction artifacts are more audible with this type of speaker. Dirac is a mixed phase system; it would be interesting to see whether the same is true for pure minimum phase room correction systems and/or pure linear phase room correction systems.
 

Absolute

Major Contributor
Forum Donor
Joined
Feb 5, 2017
Messages
1,084
Likes
2,125
My opinion, and with some experience to back it up, is that the better the speaker is, the easier it is to screw it up with room correction.

It's easier to make a straight line not straight.
Indeed. Room correction is a field riddled with trip mines. I'll show an example from my living room with JBL M2 with and without Audiolense measured at the listening position with MMM method.

RED is Audiolense, BLUE is no room correction;

MMM original M2 (Blå) vs MMM Etter Audiolense (Rød).jpg


Why does Audiolense put so much energy between 100-1kHz? Because the correction is based on one single-point measurement. And that can be ok, and it can be not so ok. If we look at the single point measurement and the correction below, we can clearly see why the software put so much energy in that area. Moral of this demonstration is that any room correction needs knowledge and multiple measurements to make the best of it.
Each case will differ.


Skjermbilde (11).png
 
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
@hardisj not to be the comparison guy but I imagine these floor standers throw a bigger wall of sound than the Selahs you reviewed? I'm trying out some NHT C3 speakers now crossed over to subs and I'm just not getting large scale sound like I've heard from a brief Revel F36 demo. Wondering if I need to step up to floor standers or maybe a bookshelf with a more powerful driver like the Purifi might do it.

I'm terrible at comparing speakers months apart. I really only rely on my remembrance of the impression I had. I do remember really enjoying the Selahs but I prefer the Revels. Maybe it's the vertical dispersion. Maybe it's the overly compensated baffle step of the Selahs (in their raw form; can be fixed by Rick or by a shelf filter). The Revels can certainly play louder without ever bottoming out the woofers even without a crossover whereas the Selah showed its mechanical excursion limits closer to the 100dB (if I recall correctly). I hate when the speaker gives up before I do... but I can't really fault a bookshelf speaker. But that doesn't mean I can't be happy with a tower speaker that doesn't exhibit this problem. The Revel also has 5dB sensitivity over the Selah. But... the Revels are also >2x the cost (depending on finish of the Selah).

Personally, I prefer tower speakers. Not little bitty skinny ones. But something along the lines of these Revels (or their bigger brother) from an aesthetics POV. Acoustically, aside from everything else the towers may best bookshelves in, the one issue that remains is the room; the benefit of a large tower with large woofers begins to wane depending on how your room is and your feelings about multiple-arrayed subwoofers to help distribute the modal issues.

All that said, IMHO, it's really comparing apples and oranges. I'd prefer the Revel for the above reasons but the Selah Purezza is a fantastic bookshelf speaker with incredible midrange detail that is in the same ballpark of the Revel F226Be and the best distortion figures in its size class, I believe.
 
Last edited:

yourmando

Active Member
Joined
Aug 14, 2020
Messages
150
Likes
178
My opinion, and with some experience to back it up, is that the better the speaker is, the easier it is to screw it up with room correction.

It's easier to make a straight line not straight.
Yeah, I suppose it’s possible that it made some of axis curves worse. But it seems less likely with a speaker with this well controlled directivity. I’m thinking the reason DL sounds worse for this speaker might be explained by something else—and I’m curious to learn what that might be.

DL uses multiple measurements and is conservative about changes. It shouldn’t have touched the curve much at all and any minor changes shouldn’t effect the other curves much because of good DI

Which is why I wonder if it’s just a different target curve issue—DL default target curve might have been used and would definitely effect tonal balance. But maybe @hardisj edited the DL target curve to match the speaker’s ‘native’ in room curve and the issue is something else.
 
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
I don't remember the curve I used, TBH. I think it was the default suggested by DL... whatever that is/was.
 
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
My biggest fault with DL in this test was the lower midrange... 300-500Hz. It really made that area worse, which was a bit surprising. Made it sound boxy.
 

yourmando

Active Member
Joined
Aug 14, 2020
Messages
150
Likes
178
Indeed. Room correction is a field riddled with trip mines. I'll show an example from my living room with JBL M2 with and without Audiolense measured at the listening position with MMM method.

RED is Audiolense, BLUE is no room correction;

View attachment 88020

Why does Audiolense put so much energy between 100-1kHz? Because the correction is based on one single-point measurement. And that can be ok, and it can be not so ok. If we look at the single point measurement and the correction below, we can clearly see why the software put so much energy in that area. Moral of this demonstration is that any room correction needs knowledge and multiple measurements to make the best of it.
Each case will differ.


View attachment 88021
This is a great example. And the M2 is exactly the Harman speaker I was thinking about where some prefer SOTA room correction for the entire curve.

For example, @dallasjustice uses AL for his M2s as well:

https://www.audiosciencereview.com/...se-digital-crossovers-w-subs.2369/#post-66198

Because DI is so good on the M2, it works well.

@mitchco also uses AL for his JBL fridges across the whole FR to correct and also set his preferred 10db Harman house curve.

Yeah, I agree using multiple measurements is better. DL uses ~8-17 measurements, depending on the whether optimization is for 1 seat vs a whole couch, and takes care to only optimize for common differences across all the points.

I know @mitchco argues for 1 measurement point w/ AL being good enough, and has shown measurements in a grid covering his whole couch area being spot on. But AL also supports multiple measurements so I don’t see how using that would hurt—I would go with multiple as well.
 

yourmando

Active Member
Joined
Aug 14, 2020
Messages
150
Likes
178
I don't remember the curve I used, TBH. I think it was the default suggested by DL... whatever that is/was.
Ah, that could have been the reason, and would be very audible. The default DL curve slopes down about 2db, and might not match the curve of the speaker.
 

yourmando

Active Member
Joined
Aug 14, 2020
Messages
150
Likes
178
My biggest fault with DL in this test was the lower midrange... 300-500Hz. It really made that area worse, which was a bit surprising. Made it sound boxy.
Any idea what the difference in frequency response in this region was w/ and w/o DL?

Depending on the change, it’s possible it’s a curve mismatch issue, or if not, is editable by tweaking the DL target curve.
 

Absolute

Major Contributor
Forum Donor
Joined
Feb 5, 2017
Messages
1,084
Likes
2,125
Personally I believe the answer to boxy sound lies in our failure to measure adequately in the vertical space before correction. Usually there's quite a bit of difference in the lower frequencies from our butts to top of the head, and depending on the distance to back wall there's likely huge variances between the front and back of the seating pillow as well.

You can ignore that and try to get lucky or you can measure it all and find a decent compromise, but I don't think you'll ever get the same subjective quality as you will with treatment in the 100-300ish range due to said differences.

Regardless, this range has next to nothing to do with the speaker and everything to do with room acoustics and should perhaps be discussed elsewhere? Given the time constraint of a review setting it's next to impossible to get room correction set up to perfection, and I'm happy that Hardisj didn't spend too much time with that sidetrack in his review.
 
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
Any idea what the difference in frequency response in this region was w/ and w/o DL?

Depending on the change, it’s possible it’s a curve mismatch issue, or if not, is editable by tweaking the DL target curve.

I'll have to go back and look at my results. I measured the before and after. Didn't include it in my review because I already had the predicted in-room response as well as 2 actual in-room responses and thought people would just get confused with the DL result thrown in there.
 
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
Personally I believe the answer to boxy sound lies in our failure to measure adequately in the vertical space before correction. Usually there's quite a bit of difference in the lower frequencies from our butts to top of the head, and depending on the distance to back wall there's likely huge variances between the front and back of the seating pillow as well.

You can ignore that and try to get lucky or you can measure it all and find a decent compromise, but I don't think you'll ever get the same subjective quality as you will with treatment in the 100-300ish range due to said differences.

Regardless, this range has next to nothing to do with the speaker and everything to do with room acoustics and should perhaps be discussed elsewhere? Given the time constraint of a review setting it's next to impossible to get room correction set up to perfection, and I'm happy that Hardisj didn't spend too much time with that sidetrack in his review.

To date, I have measured most speakers with DL using the same target curve default. This is the first speaker which had the result of things actually sounding worse. Same room. Same seat. Same speaker positions. Same measurement method with DL (9 positions in a single seat reference).

I agree... not worth much energy. But just pointing this out as it is relevant. I attribute the issue to room as well. I just find it interested that this is the first time the results were not pleasing when using DL under the exact same conditions as all previous tests. YMMV.
 

yourmando

Active Member
Joined
Aug 14, 2020
Messages
150
Likes
178
I'll have to go back and look at my results. I measured the before and after. Didn't include it in my review because I already had the predicted in-room response as well as 2 actual in-room responses and thought people would just get confused with the DL result thrown in there.
Great, thanks!
 
Top Bottom