Revel F328Be Speaker Review

[Lorenzo74]

I would give up on cardioid. It's an interesting idea, but the trade-offs aren't worth it IMO.

I'm sorry but I disagree. could you pls elaborate what trade off you are referring to? the pro audio is based on cardio subs for live reproduction

 

[McFly]

There's definitely something going on with the measurements and floorstanders on this site. Every floorstander here that I see measured that I have personally heard goes deeper IMO than measured. It seems that all the bookshelves here measure better in the bass than the floorstanders. Three 8" woofers should easily hit 40hz, no matter how the designer intended the rest of the speaker sound. Maybe its the line array of all the woofers vs 1 mic or something to that effect.

 

[Lorenzo74]

From what I remember when I researched this a year or two ago, you lose a lot of SPL making it work. I guess if you aren't doing movies you may not mind.

At any rate, we have a better way to "make the room disappear" with subs: placed strategically and equalized together (ala Sound Field Management, Audyssey, Dirac, Multisub Optimizer, etc).

to me the only that works is Dirac Unison (only available in Volvo XC90) or its byproduct Dirac Live bass management. cardioid patter address the problem at the origin.

 

[andreasmaaan]

I'm sorry but I disagree. could you pls elaborate what trade off you are referring to? the pro audio is based on cardio subs for live reproduction

I think the advantages of cardioid subs are more obvious when it comes to live reproduction (i.e. avoiding drenching the performers in bass) than for home audio.

Not saying it's a bad idea in a home audio context btw, but just that one of the main advantages of it when it comes to live sound is not applicable in the home audio context.

to me the only that works is Dirac Unison (only available in Volvo XC90) or its byproduct Dirac Live bass management. cardiodi patter address the problem at the origin.

It can be relatively easily done without any kind of automated software.

EDIT: actually, maybe it would be better to ask instead of trying to answer: What do you mean specifically by "the problem" that cardioid bass addresses?

 

[Lorenzo74]

I think the advantages of cardioid subs are more obvious when it comes to live reproduction (i.e. avoiding drenching the performers in bass) than for home audio.

Not saying it's a bad idea in a home audio context btw, but just that one of the main advantages of it when it comes to live sound is not applicable in the home audio context.



It can be relatively easily done without any kind of automated software.

EDIT: actually, maybe it would be better to ask instead of trying to answer: What do you mean specifically by "the problem" that cardioid bass addresses?

from undisputed gentlemen...
He explains better than I can ever think of...
jump here if you don't have time 2:15-2:30 and 4:30. my recommendation is to see it entirely since very much interesting

enjoy
best
Lorenzo

 

[Robbo99999]

A glowing review of this speaker, and I can see why with the listening tests and the bass performance heard in those tests (a product of the low distortion too). I guess we can disregard what we're seeing in the spinorama re the bass, because it looks very bass shy there....Amir did say that it might be 1dB out, but I'm thinking it's probably more than that given that it looks very bass shy in the spinorama and the listening tests were so complementary of the bass. It's that age old problem with measuring towers on the Klippel then!

Ha, I wouldn't mind the distortion levels of these speakers for my JBL 308's that have just been measured a few days ago by Amir!

Nice speakers, I don't think I'd ever buy anything like this though, I think I'd go with a cheaper option of some quality monitors combined with a sub or two....as my hypothetical pinnacle system of the future. I think one advantage of this would be subwoofer bass all the way down to 20Hz flat in contrast to these tower speakers.

 

[andreasmaaan]

from undisputed gentlemen...
He explain better than I can ever think of...
jump here if you don't have time 2:15-2:30 and 4:30. my recommendation is to see it entirely since very much interesting

enjoy
best
Lorenzo

I agree 100% in terms of loudspeakers, but when it comes to subwoofers, I don't see the advantage of cardioid radiation for domestic rooms (since a non-cardioid subwoofer with 4pi radiation is best placed against adjacent room boundaries, restricting its radiation to 1pi or 0.5pi in the first place).

If your only option was to place a subwoofer away from the walls near the centre of the room, then I agree a good argument could be made for cardioid radiation.

 

[st379]

If there is a bass problem with tower speakers measurements it should be investigated.
Because people here are being hypocrite about this speaker. If the name of the speaker was Kef f328be, you would open up a facebook group "Don't buy Kef f328be" and tell everyone how it is overpriced with bad performence.

 

[amirm]

I would be really interested in a proper blind test between the JBL 308 mkII and this speaker! Even though I assume the 328be would come out on top, I wouldn't bet my house on it.

Well, maybe you should. Harman performs double blind tests of all of their speakers as a condition of release. That comparison involves both their own products and competitors. Here is how an F208Be (prototype) did:

I was listening to F208Be at a show and asked my friend and colleague Kevin Voecks how it compares to Salon 2 as the sound I was hearing from the F208Be was fantastic. His answer? "Salon 2 smokes it in blind tests."

Amir, Owner of a pair of Revel Salon 2.

At the point where we think Harman is ignoring audio science, then all bets are off. I think F208Be became F228Be (?). Here is a picture I took of the prototype at CES 2017:

 

[st379]

Dude. Focus. No one on here has said KEF offers bad performance. Some people have had issues with the frequency response (which is easy to eq) and Amir had an issue with a room mode. More, Amir prefers wider directivity. That's his opinion to have and he has made the case for it based on science. If you are coming to this based on the preference score, please stop. It has nothing to do with Amir's reviews. Again, the score is frivolous.

Is English your first language? This is a sincere question and may help in how we can approach your, frankly, ridiculous statements.

English is not my first language but I can read minus 15 db at around 35hz just fine, thank you.

 

[amirm]

If there is a bass problem with tower speakers measurements it should be investigated.

I have done a lot here. I worked Neumann to dial it in almost perfectly to their anechoic measurements. On these towers, I ran the full scan twice and a number of quicker scans to optimize the results. It is outside of my physical ability to keep messing with it with no reliable reference to shoot for. Remember, no anechoic chamber is correct in these low frequencies. Companies test a speaker outdoor and then use that to calibrate the indoor anechoic measurements. So neither measurement is precise.

Ultimately, this is academic anyway because in room variations of bass can be as much as 20 dB. As much as I like to get this right if I can, as a real issue, it does not exist or justify me spinning my wheels on these tall and heavy speakers.

In the future we may get a more willing manufacturer than Harman to get to the bottom of this with another tower speaker.

 

[andreasmaaan]

Is English your first language? This is a sincere question and may help in how we can approach your, frankly, ridiculous statements.

I'm sorry, but I don't believe this is an appropriate response. Firstly, @st379's comment, while indeed a little antagonistic, was written in perfectly good English.

Secondly, your response - while I'm sure this wasn't your intention - does not help make the forum a welcoming place for people of non-English speaking backgrounds.

 

[st379]

I have done a lot here. I worked Neumann to dial it in almost perfectly to their anechoic measurements. On these towers, I ran the full scan twice and a number of quicker scans to optimize the results. It is outside of my physical ability to keep messing with it with no reliable reference to shoot for. Remember, no anechoic chamber is correct in these low frequencies. Companies test a speaker outdoor and then use that to calibrate the indoor anechoic measurements. So neither measurement is precise.

Ultimately, this is academic anyway because in room variations of bass can be as much as 20 dB. As much as I like to get this right if I can, as a real issue, it does not exist or justify me spinning my wheels on these tall and heavy speakers.

In the future we may get a more willing manufacturer than Harman to get to the bottom of this with another tower speaker.

Thank you for your explanation and efforts you put on this review.

You said that your measurements are better than anechoic chamber it mean that the bass measurements is very accurate?
You also said in this review that the bass of this speaker is excellent. So if the measurements in the klippel are better than anechoic chamber how come your bass description and the measurements do not correlate?.

Edit:
From your review :"What surprised me though was the deep bass. I had not heard such clean and deep bass out of any speaker I have tested so far. "

 

[andreasmaaan]

I have done a lot here. I worked Neumann to dial it in almost perfectly to their anechoic measurements. On these towers, I ran the full scan twice and a number of quicker scans to optimize the results. It is outside of my physical ability to keep messing with it with no reliable reference to shoot for. Remember, no anechoic chamber is correct in these low frequencies. Companies test a speaker outdoor and then use that to calibrate the indoor anechoic measurements. So neither measurement is precise.

Thanks @amirm. I'm pretty confident your measurement setup is reasonably accurate (as accurate as any method) in the bass. I'm just wondering, though, how you arrived at the approximate 1dB margin of error that you mentioned in the OP?

Ultimately, this is academic anyway because in room variations of bass can be as much as 20 dB.

According to the Olive research, bass extension counts for around 30% in terms of listener preference. While I agree that in-room variations may render mild irregularities in a speaker's bass output of little more than academic interest, low-frequency extension appears to be of more than academic interest according to the science, I would have thought.

 

[amirm]

We shouldn’t have to explain again and again in an objectivists audio science forum, why multisub+room Eq is mandatory for proper bass reproduction in rooms.

It has already been well explained and scientifically documented in chapter 8, pages 230-251, of @Floyd Toole
Sound reproduction : the acoustics and psychoacoustics of loudspeakers and rooms.

I am not saying it’s easy to get it right, but it’s the only way.

No one here is using the technique described in those pages. Namely, Harman's proprietary Sound Field Management (SFM). The system only exists in JBL Synthesis ARCOS DSP which we used at our old theater at Madrona. And I am setup to use in my theater. I have written a full published article on it with tons of data: https://www.audiosciencereview.com/...s/computer-optimization-of-room-acoustics.12/

In a nutshell, the system varies level, phase and frequency response of each sub independently to optimize its response against every other speaker in the room. The computational system can iterate millions of settings to arrive at its correct value. No way, no how you are doing that manually with PEQ, or any other automated system.

Imagine how a pair of sub would be optimally crossed to not just one speaker but two or more. If your room is irregular then even identical main speakers will have different response. As will identical subs in different locations.

Sadly JBL discontinued its only consumer standalone box that had SFM in (BassX) years ago. The closest to may be Dirac bass management but I am not sure that even that gets close.

Here is Dr. Toole in the very pages you listed:

And the motivation:

Yes, you can get usable results with equalization let's not think that is the same as optimization done inside of a speaker.

 

[Rockfella]

Congrats Amir

 

[Lsc]

The 328 features a newer and slightly improved waveguide. I don’t remember if it uses the same tweeeter.

The 328 has an improved tweeter as well. Also the 8” base drivers were improved. The only driver that is the same is the midrange.

Seems like Revel really went out of their way to create a superior speaker. The F226Be might be better for some people (even me whose man-cave is 16 x 11 x 8) but I always thought what separates the “ultra high end” speakers is their bass performance. I don’t think just adding a sub is the same.

I have a sub with my F228Be and it definitely adds a lot even for music, but I’m not sure if it’s “just as good”.

 

[amirm]

Thanks @amirm. I'm pretty confident your measurement setup is reasonably accurate (as accurate as any method) in the bass. I'm just wondering, though, how you arrived at the approximate 1dB margin of error that you mentioned in the OP?

As I explained at the outset, I went through a series of steps to optimize the low frequency accuracy. This reached an asymptotic level and hence my guess of 1 dB or so which applies to that shelving.

There is also another tool here which I can share with you. Namely, Klippel system is self-checking and provides a comparison of measured response versus computed one (both projection of soundfield and removal of reflections). Here is that result for reference axis:

The room reflections are quite high (much higher than other speakers I have measured) which naturally causes the irregularity in red/green curve. Eyeballing that and comparing it to the computed response in solid black, it seems to more or less match. No? Look for example between 40 and 50 Hz where there is little effect from room mode. We match that almost perfectly.

The only error I see there is down below 30 Hz where our response falls lower than real measurement. So maybe that should be 65 dB vs 60 dB.

The above also shows the challenges in measuring in an anechoic chamber that is not reflection free below 80 Hz (which describes most of them including those that Harman uses). Reflections become quite significant as we see so the location of the speaker and mic make a big difference even in anechoic chamber. Calibration won't be accurate for them either as there are so many radiating surfaces exciting the room modes (slightly) differently.

 

[andreasmaaan]

Going back, I'd say that that last comment was not just a little antagonistic but made in bad faith. And I mentioned the English thing because, going back and looking at all his posts--beyond just the ones in this thread--the relationship between the reviews and the preference score has been explained to him again and again and again and again and again. More, the important parts of the KEF review thread--horizontal directivity vs. vertical, frequency response, and room modes--are either ignored or misunderstood in his comments. Amir does not simply dislike those KEF speakers. I just don't think dialogue has been working and he is spreading misinformation. I apologize if my tone was off, but in a thread where the preference score can confuse readers, we must be clear about its status.

Ok, thanks for the explanation I appreciate that you didn't intend to alienate anyone with your comment. But as a migrant myself, I'm aware of how comments like this can be (unintentionally) unwelcoming, and potentially quite painful to someone who has put a lot of effort into learning a second language.

I haven't been following whatever background there has been in other threads, so I can't comment on that. In terms of the question in this thread, I do think @st379 has a point about the more open-minded reception this speaker seems to have got on ASR than certain other speakers with arguably comparable issues, although I would have expressed it differently myself, and I agree with you that the characterisation of attitudes towards KEF was an oversimplification.

 

[amirm]

You said that your measurements are better than anechoic chamber it mean that the bass measurements is very accurate?

For many speakers, yes, the NFS computation is more accurate than anechoic chamber. Indeed the system is also sold to be used in anechoic chamber to improve its accuracy.

For these speakers, until we have a gold standard (speaker measured outdoor away from any reflections for countless feet), we don't know. Both systems have accuracy errors. I just don't know enough to say for sure.

I have asked Klippel about this and they don't have an answer but stick by NFS working correctly and producing the right response.

You also said in this review that the bass of this speaker is excellent. So if the measurements in the klippel are better than anechoic chamber how come your bass description and the measurements do not correlate?.

Simple: laws of physics. Measurements are showing anechoic response, not what a speaker does in a room. As soon as you place a speaker close to any walls, those walls focus the bass energy forward and you get more bass. Then other walls add to it causing destructive and constructive interference which we call room modes. Take your speaker outdoor and listen to it. The amount of bass it produces will shrink way down and will not sound anything like what you hear in your room.