Genelec 8351B Review (Studio Monitor)

[JustJones]

After reading so many interesting DACs tests on ASR, I must have missed something because nobody seems to raise this question here. We're talking about a DAC + preamp + amp + speakers system. I'm a bit lost without measurements for those parts, which is why I'm a little hesitant.

Of course everyone has different priorities but I never worried about the electronics just the measurements from the speakers which is usually the weakest point in any system.

 

[Sancus]

This raises the question of sensitivity. In a surround application could they go to sleep and then wake up but with a delay to the output?

Just as a data point, my Genelecs(8351B/8341A) wake up/sleep and stay awake beautifully when connected to an AVR. In fact they do so better than my Neumann KH80s, which can be a little bit finicky when connected to certain channels for some reason.

Maybe I'm wrong looking at the measurements though. From what I've gathered, wider, smooth off axis response and bass extension boost a speakers preference score. Not SPL, newness, or being used in a studio (PMC and ATC come to mind.)

The preference score is a tool for filtering out bad speakers, that's all it is. It's of limited use with speakers as good as the 8351B. It's very unlikely there were any good(by modern standards) coaxials included in the study, and there were exactly 0 speakers predicted by the formula to score >7.1 and only 4 speakers that were rated by testers >7.1 in the entire study. The score also ignores many variables that have been reported to affect preference in real blind testing, like IMD. And of course SPL capability/distortion is a factor in preference, IF YOU TEST IT. The speaker tests in the study were deliberately held at moderate levels so that it wouldn't be tested.

wider,

Wideness is actually penalized by the preference score, amusingly, though it wasn't initially intended. Perhaps this will communicate to you that the preference score is poor for evaluating excellent speakers. The idea that wideness is preferred by most is inconclusive. It's a hypothesis advanced by Toole(and others), but it hasn't been studied in isolation, and there seem to be plenty of cases where people legitimately prefer narrower dispersion.

Additionally, this is a stereo-fixation-issue. Wide dispersion is not a meaningful advantage for multi-channel, and multi-channel is flat-out stated to be superior by Toole as well. It puzzles me why people are so quick to jump on the wide dispersion bandwagon, but just skip Toole's statements that multi-channel is technically and subjectively superior to stereo and that the good upmixing algorithms work well.

It's almost as if we all personally pick and choose the arguments we prefer

smooth off axis response

The 8260A isn't smoother than the 8351B, in fact it has a pretty abrupt woofer transition. You can see this in the power response where it sharply increases, and in the contour map there is bulging just below the crossover rather than perfectly even and gradual transition from controlled directivity to omnidirectional. Also, I think this is a very important frequency range for smoothness, as many key vocal frequencies are in the 100-500hz range. This holds for the 8361A as well, though it is not quite as smooth as the 8351B, it still has a better woofer transition.

Contour Maps:

Spoiler

Basically what I'm saying is, if you like the 8260A, great! But don't say that generalizes to everyone. You put priority on wideness and extension, but if you're using a sub then extension doesn't matter...and smoothness, SPL, and vertical dispersion are relevant attributes. You may not care about SPL and vertical dispersion, but that's not the same thing as the science saying they don't matter. It does not say that. The preference score is very far from a summary of all the science.

The actual speaker designers, themselves accomplished audio researchers, disagree with you on what attributes they wanted to improve. I'm not saying you should always trust speaker designers, but in Genelec's case, I do think they probably know what they're doing... don't you??

I don't want to turn this into a huge back and forth about the 8260A. I think it's a great speaker. But it isn't good enough to justify saying that Genelec made their modern speakers worse somehow. That's just not the case.

 

[MBI]

Of course everyone has different priorities but I never worried about the electronics just the measurements from the speakers which is usually the weakest point in any system.

But isn't that sort of position exactly what objective measurements support, as they do suggest that electronic price increase is not all that associated to higher performance and that speakers usually are the weakest point?

In other words, stating that electronic measurements aren't as crucial as speakers measurements does require many electronic and speakers measurements, doesn't it?

Some electronics may limit speakers potential still, but how and by how much in this situation, I don't know. Wouldn't it be nice if Genelec disclosed some information still?

 

[FrantzM]

i'm in

Me too !

 

[pierre]

Merci Pierre. RPi4 with an AES output hat sounds such a cheap, handy and simple solution.

But should we just hope that "The Ones" internal DACs will at least be good enough not to make any audible difference on their active speakers, or does Genelec specify an estimated SINAD for the DACs or whatever it is they mount? (Can't seem to find anything about it, but I may have not google researched properly)



Not sure what to understand here, did you actually test it, or are you going to, Pierre?



That's what I keep reading, and it sounds pretty logical too, but doesn't this beg further questions as embarked DACs may or may not limit Genelec's speakers output quality?

After reading so many interesting DACs tests on ASR, I must have missed something because nobody seems to raise this question here. We're talking about a DAC + preamp + amp + speakers system. I'm a bit lost without measurements for those parts, which is why I'm a little hesitant.



Well yes and no, as all speakers may have the "ability" to degrade some of the signal, but not all of them force that signal into their own built in DAC beforehand (ie. passive speakers).



Thank you for this great explaination. Wouldn't that one loop loss be easy to calculate just knowing the SINAD of the embarked DAC?

@Jason K open one in this thread https://www.audiosciencereview.com/...-8351b-teardown-2nd-disassembled.22785/page-2

it looks like the ad/da are from akm: the associated Data sheet tell s/n around 115db and thd around -102dB so it gives an idea. The distortion from the drivers is order of magnitude higher. since it is the last ad/da before the speaker I would not care too much.

 

[Spocko]

JIC there's someone who doesn't know.

OMG I forgot how big these things will be! I'm getting the low slung ST version (HxWxD 28"x58"x21")

 

[Spocko]

In which use case don't you need controlled directivity?

LOL considering that every subjectivist review leaves out directivity plots it's safe to say that for such an audience, directivity issues are addressed as discussions about toe-in angles and carefully curated amp & cable matching so they don't "need" controlled directivity when a douse of snake oil can fix things.

 

[Spocko]

They're not that big( with the 8040 and 8010 in these photos) View attachment 130194View attachment 130192View attachment 130193)

OMG that's huge - I'd probably go with the D&D if I needed bass without subwoofer support

 

[stevenswall]

But it isn't good enough to justify saying that Genelec made their modern speakers worse somehow. That's just not the case.

The modern speakers would be better for those that need more spl and less bass extension, and worse for those using them without a sub that don't need the SPL. So worse in the latter case, and better in the former.

I am glad that the power response has been pointed out, I'll have to look into that more and what causes irregular response there when the off and on axis measurements seem to be better.

Thanks for the points about Toole and stereo. I appreciate you bringing that up, I didn't know. I've only seen bits of his and Sean Olive's research, and often stuff ends up behind an AES paywall.

It seems fair to say that "Genelec made the 8361 worse than the 8260 when it comes to measured power response, on axis accuracy, and off axis accuracy," while holding Genelec's position that, "The Genelec 8361 is better than the 8260 due to its higher SPL capabilities, increased DSP filter capacity, and better vertical directivity."

I can live with both of those, and the idea that the power response of other models of The Ones is smoother than the 8260, making them better there as well, and now that you point it out, that perhaps the wider dispersion of the 8260 isn't an advantage in stereo, and that there isn't enough evidence to support wider dispersion generally being better than narrow.

Thanks to everyone who helped and pointed to objective data. I'm glad I've learned a few things and don't have to rely on what I've seen as "appealing to authority," and newer=better arguments, though admittedly Genelec is a great authority to appeal to, and I do think The Ones are incredible, just didn't have the information to show why, and some information suggested they were worse from my level of understanding.

 

[Spocko]

Will you be using a pair of 7370 or just one?

Just one to support my L/C/R (2x 8360a, 1x 8351b) down to 30Hz with the JTR doing the heavy lifting down to 10Hz.

 

[dfuller]

The preference score is a tool for filtering out bad speakers, that's all it is. It's of limited use with speakers as good as the 8351B. It's very unlikely there were any good(by modern standards) coaxials included in the study, and there were exactly 0 speakers predicted by the formula to score >7.1 and only 4 speakers that were rated by testers >7.1 in the entire study. The score also ignores many variables that have been reported to affect preference in real blind testing, like IMD. And of course SPL capability/distortion is a factor in preference, IF YOU TEST IT. The speaker tests in the study were deliberately held at moderate levels so that it wouldn't be tested.

Yes - I think people overlook the influence of distortion and max SPL a lot. This to me is what separates a good speaker from a great one - assuming similar FR and directivity behavior, the distortion behavior is what separates the good from great. A speaker like a JBL 308 is good in terms of FR and directivity behavior (and to be honest, I think constant directivity is somewhat overrated at some level), but the distortion performance firmly places it in a low-tier.

 

[Spocko]

Hello Amir,

For some reason your measurement above has ten times the bulk delay there is in reality. Correct bulk delay for 8351B is about 3.1 ms (for analogue input >1 kHz). Could you please amend your measurement?

Thank you @Ilkka Rissanen and your team for creating an amazing line of speakers and mounting accessories. Your ceiling and wall mount solutions are awesome.

 

[richard12511]

Choice of speaker does depend how you intend to use it, the 8351b works really well as a nearfield monitor, but for me once you sit more than two metres away you are looking at the sound rather than being immersed in the sound.
Home comparison is always best.
Keith

My impression kinda differs from yours here I think(depends what is meant by "rather than being immersed"). I actually found the 8351 to be more enveloping and immersive than the 8C. 8C was cleaner and more in front sounding. Maybe we're talking about different things, though. I'm talking about that sense of being surrounded by the sound that Toole says comes from the 200-800Hz region. The wider the dispersion of the speaker is within that range, the more enveloping it should sound. Maybe this is something that's room dependent?

I think the best way to see this is by looking at the horizontal curves that @MZKM posts.

The 8351 has considerably wider dispersion in that circled region, so it makes sense that it sounds more enveloping. You can also look at the CEA2034s side by side and compare the areas under the curve of the DI lines

That includes, vertical, though. To me it's easier to see in the first comparison.

This was actually the first time I've ever looked at the 2 side by side, but something that really sticks out to me is just how even they are from 2-10kHz, which is what seems to be what determines the soundstage width and image width. They should(based on measurements) image and soundstage very similarly, and that aligns very well with what we heard.

Those red lines for each image are the same length. Crazy how close they are in terms of width. 8C looks a little smoother in that region(to my eye).

Also, I have to disagree with your impression that the 8351 is only suitable for near field. For fullrange use, we strongly preferred the 8C, but with subwoofers in play(crossed at 100Hz) we both actually slightly preferred the 8351(difference was tiny, though). Our listening distance was somewhere between 3.5 and 4m. One difference between my comparison and yours is the room correction. We used DiracLive 3.0(below 200Hz) for both.

You noted that the 8351 image is closer in space. I've recently started using GLM(instead of DiracLive), and I do notice that the image moves closer to me when I engage it. I'm super curious what the objective reason for that is.

 

[Spocko]

Is feeding "The Ones" with a DAC-less digital AES/EBU input from a simple streamer (such as an AES/EBU hat onto a RoPieee running Rpi4, for example) instead of a high SINAD DAC's analog output expected to give similar results or not?

In other words, does anybody know how "The Ones" embarked digital to analog conversion compare with top rated DACs feeding them?

(Trying to find the best quality, simplest and most direct way to connect 8351B to a Nucleus.)

P.S. Very many thanks for the review!

If you're going direct like that, you should get the volume control.

 

[Spocko]

Choice of speaker does depend how you intend to use it, the 8351b works really well as a nearfield monitor, but for me once you sit more than two metres away you are looking at the sound rather than being immersed in the sound.
Home comparison is always best.
Keith

Would you agree that much of this immersion is due to mid-bass presence between 70Hz and 100Hz where the 8351b will definitely drop off beyond 2 meters?

 

[RobL]

My very picky wife was looking over my shoulder and said, "what are those, they look really cool!" So, go figure.

Yep, everyone has different likes I suppose. To me,
these are “function over form”...Genelec’s priority was superlative performance and aesthetics were well down the list. These are precision tools meant for a specific purpose after all.
I think it would be great if Genelec were to produce a more consumer oriented line...something more than just rebadging their pro audio equipment.

 

[richard12511]

Maybe I'm wrong looking at the measurements though. From what I've gathered, wider, smooth off axis response and bass extension boost a speakers preference score. Not SPL, newness, or being used in a studio (PMC and ATC come to mind.)
.

Actually, I think(from what others have said) that the preference score actually prefers more narrow dispersion.

Toole is the one that advocates for wider dispersion(based on some blind tests they performed). I think dispersion width is one of those preference things. Geddes for example advocates for even narrower dispersion than what the 8351/8361 show. Much of it (imo) depends on how you value image precision vs ASW(apparent source width). Geddes seems to value imaging more, whereas Toole/Amir seems to value ASW more. Also might depend on how many channels one has. Even if one would prefer the slightly wider soundstage/images of the 8260 in stereo, that might change when going 5.1+. IME, Auro3D gives me that soundstage width(actually wider) that I get from my Revels, without messing up the imaging as much.

So I think it depends on what you want and what your situation is. For someone who prefers a more narrow dispersion and is using subwoofers, the 8361 has many advantages: better dispersion, better power response, more headroom/dynamics, and better vertical performance.

My initial response to your post was more about calling the 8260 the flagship. Regardless of whether or not some users might actually prefer the older model under certain circumstances, it seems weird to call a discontinued product the flagship. It seems to me that Genelec thinks the 8361 is superior.

 

[Spocko]

Yep, everyone has different likes I suppose. To me,
these are “function over form”...Genelec’s priority was superlative performance and aesthetics were well down the list. These are precision tools meant for a specific purpose after all.
I think it would be great if Genelec were to produce a more consumer oriented line...something more than just rebadging their pro audio equipment.

Which takes us to the biggest cost for the "home audiophile" - aesthetics contribute greatly to cost. For speakers it's the additional cost of adapting a proven acoustic design to a more furniture grade shape that doesn't compromise performance; add to that the cost of textures, veneer and paint plus matching stands. Looking at the KEF LS50 I can tell much was invested into the overall design - how much more to adapt The Ones to look like the LS50?

 

[MBI]

If you're going direct like that, you should get the volume control.
View attachment 130268

Thanks for the tip! Would Roon's volume control not work or be of poor quality?

 

[echopraxia]

Thank you for this great explaination. Wouldn't that one loop loss be easy to calculate just knowing the SINAD of the embarked DAC?

Yes it should be quite easy to calculate. I really doubt you'd hear any difference even after may such loops, and to be honest this is probably true even for the worst performing of the 'ASR recommended' DACs, let alone the super high SINAD ones we buy for the simple reason that we can (and because we appreciate the engineering).

I have yet to see any blind test show that there is any audible difference between a 95db SINAD DAC and a 120db SINAD DAC.

End-to-end performance is always what really matters. When the distortion from the physical speaker drivers is orders of magnitude more than that of the DACs/amps, the latter really don't matter as much as some audiophiles might want to believe (because it feels good to think your tinkering and purchase of expensive components makes a difference).

Of course every bit counts to some degree, but if -40db or so of distortion at medium/loud SPL is state-of-the-art from some of the best, lowest distortion speakers in the world, then the DAC/amp distortion barely make a dent in the overall end-to-end distortion picture. Perhaps we don't have an easy intuition that chaining a -40db distortion device through a -100db device yields a total distortion result of... still -40db (the difference is so small my calculator won't even show it, but it's probably something like -39.9999999999db)!