• 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!

Genelec 8351B Review (Studio Monitor)

Sebastiaan de Vries

Active Member
Joined
Sep 4, 2019
Messages
114
Likes
305
Location
China, Shenzhen
Would you consider an upward tilted on axis frequency response to be neutral?
I guess you imply that is the resulted on-axis response after a room-correction system attempts to correct the steady-state response. And no, that is not natural.

Let me nuance myself a bit more. You are correct that the majority of the room-eq systems I have tried (including Tact) didn't provide satisfying results when applying corrections higher than 300Hz. I was implying that Trinnov provides room optimization which can distinguish early reflections and steady-state response with the beam-forming mic. "How" they do it. I think non of us know since this is Trinnov proprietary. But I do know that this room correction (room optimization) just works. And a flat target curve doesn't sound up-lifted or unnatural, but flat, just as a very good flat headphones.

It has been in my planning to make some measurements with a second measurement system with waterfall response, amplitude, and phase. And compare the result with and without Trinnov room optimization.
 

Senior NEET Engineer

Addicted to Fun and Learning
Joined
Jan 6, 2020
Messages
538
Likes
591
Location
San Diego
I guess you imply that is the resulted on-axis response after a room-correction system attempts to correct the steady-state response. And no, that is not natural.

Let me nuance myself a bit more. You are correct that the majority of the room-eq systems I have tried (including Tact) didn't provide satisfying results when applying corrections higher than 300Hz. I was implying that Trinnov provides room optimization which can distinguish early reflections and steady-state response with the beam-forming mic. "How" they do it. I think non of us know since this is Trinnov proprietary. But I do know that this room correction (room optimization) just works. And a flat target curve doesn't sound up-lifted or unnatural, but flat, just as a very good flat headphones.

It has been in my planning to make some measurements with a second measurement system with waterfall response, amplitude, and phase. And compare the result with and without Trinnov room optimization.

If the steady state response looks flat, I can think of three possibilities.

1) omnidirectional speaker

2) nearfield in heavily damped room

3) upward sloping on axis frequency response

Any others you can think of? There is nothing wrong with preferring (3). That will sound similar to the tonal balance of an echo chamber. A lot of headphones are voiced like that.
 

Sebastiaan de Vries

Active Member
Joined
Sep 4, 2019
Messages
114
Likes
305
Location
China, Shenzhen
If the steady state response looks flat, I can think of three possibilities.

1) omnidirectional speaker

2) nearfield in heavily damped room

3) upward sloping on axis frequency response

Any others you can think of? There is nothing wrong with preferring (3). That will sound similar to the tonal balance of an echo chamber. A lot of headphones are voiced like that.

I am likely fitting option 2. My room is reasonable damped (RT60=0.27s) with around 2.5 meter distance still considered near-field.

Attached the measurement before and the predicted after response.
 

Attachments

  • F421BBB2-0DDC-4EE5-BFEA-0A7BD21A8110.jpeg
    F421BBB2-0DDC-4EE5-BFEA-0A7BD21A8110.jpeg
    107.1 KB · Views: 147
  • 425116AE-78F5-4372-84D1-E5EE75B033E8.jpeg
    425116AE-78F5-4372-84D1-E5EE75B033E8.jpeg
    105.7 KB · Views: 150
  • E6CEC088-705F-4FC4-8371-2555257CABDA.jpeg
    E6CEC088-705F-4FC4-8371-2555257CABDA.jpeg
    102.9 KB · Views: 149

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
I am likely fitting option 2. My room is reasonable damped (RT60=0.27s) with around 2.5 meter distance still considered near-field.

Attached the measurement before and the predicted after response.

I'm curious why you chose to have the magnitude equalized completely flat there... also is RT the same as RT60? I don't recall seeing anything like 27 before -- a value as low as 50, only ever once.

...

Haven't read most of this thread so I missed the page before where you said: "I belong to the minority who aim for a ruler flat curve, and to my ears, this sounds most accurate." Huh. The idea that we know really this is most "accurate" is odd to me considering the whole issue of the "circle of confusion" in audio production and reproduction, but, okay.
 
Last edited:

Sebastiaan de Vries

Active Member
Joined
Sep 4, 2019
Messages
114
Likes
305
Location
China, Shenzhen
I'm curious why you chose to have the magnitude equalized completely flat there... also is RT the same as RT60? I don't recall seeing anything like 27 before -- a value as low as 50, only ever once.

...

Haven't read most of this thread so I missed the page before where you said: "I belong to the minority who aim for a ruler flat curve, and to my ears, this sounds most accurate." Huh. The idea that we know really this is most "accurate" is odd to me considering the whole issue of the "circle of confusion" in audio production and reproduction, but, okay.

I use high-pole carpet, big book shelves behind me and dampening to dampen the room.

I like a flat curve because to my ears this sound most accurate. It reminds me most of the sound of good control-rooms, or quality headphones. Nothing is in the way, sound is clean and I can hear all layers of sound and mix decisions.
 

Thomas Lund

Member
Technical Expert
Industry Insider
Joined
May 15, 2018
Messages
75
Likes
342
Location
Aarhus, Denmark
Some problems I have with it:

- Messes too much with direct sound by default, unless you are nearfield in heavily damped room...

In monitoring, there is but one reasonable goal, in 7 words: Flat frequency response of perceived-direct sound.

Outside anechoic reproduction, this poses a challenge, because what exactly is perceived as direct sound? That depends on the room, the loudspeakers, listening distance, the content and even the listener him/herself. ITU-R BS.1116 specifies requirements for the room, and some monitors control directivity, further promoting neutral reproduction not just by chance.

To the GLM comments, one should not accept its auto-calibration of frequency response without listening also, especially in rooms not acoustically designed for monitoring. However, flat in GLM = flat anechoic response, and you can adjust any auto-compensation, if your ears and brain don’t agree, e.g. at a particular frequency.

Regarding long-term support, 15 year old models can still be controlled and calibrated using the latest Mac and Windows OS. Monitors keep improving as GLM is refined, the next example being right around the corner.
 

Sebastiaan de Vries

Active Member
Joined
Sep 4, 2019
Messages
114
Likes
305
Location
China, Shenzhen
When I was 18, I had my first dental cavity. I hated it because otherwise, I had good teeth. I figured the cause was the sugar in my coffee. So I decided to quit the sugar and drink my coffee black. The first few days, I had to force myself to drink it raw. I couldn't stand the bitter taste. However, after a week, I started to actually appreciate the pure bean taste of excellent coffee and begin to enjoy it. Nowadays, if by accident anyone put any sweetener in my coffee, I cringe and can't stand the taste.

The point I try to make is, I do believe (without any scientistic proof other than experience) that our human brains adapt to a reference. With sound, that isn't any different.

We get accustomed to a sound reference we are familiar with, just as I was a customized with the way I reference coffee should taste.

I have worked exclusively in pro-audio environments for the first better half of my life (control rooms, broadcast van's, radio stations, etc.) And despite there is lack of a universal standard, generally, "flat" has always been the target. Those days, digital room eq wasn't an accepted tool yet and hardly available. In most of the environments where I worked, getting it "flat" was obtained by the following efforts.
  • In most cases, the room has been designed by an acoustic engineer. That doesn't guarantees an acoustically flat room, but it brings you closer than most domestic environments.
  • The big in-wall monitors were measured and tuned by their active crossovers with PEQ's (BSS and SPL were popular those days), time-align the drivers, and flatten the overall response at the sweet spot behind the desk.
  • 90% of the mixing be done on smaller monitors, such as the Yamaha NS10M and (that time) popular Genelec 1031's. Those are around 60 to 70cm from the engineers' ears and don't portray a downfall slope at that distance. Combined with the dampened room, most you hear is direct sound, and flat on-axis response.
It is fair to say that a majority of mix engineers ARE accustomed to a flat sound. Whether this flat response is "right" or "wrong" is under debate by some. Still, there must be some standard, and without officially disclosed, flat is. When control rooms or mix facilities are consistently flat, it is easier for mix engineers to work in different mix environments.

The video world is ahead of the audio side by more consistent and clear standards. However, there isn't a standard for an audio curve in studios, as far as I am aware. One reviewer of a Trinnov Pro unit opted that now room EQ becomes more accepted, there should be a standard for every recording studio regarding a target curve.

Probable by these influences and the joy I had listening to my familiar music in control rooms and simply heard so much more details than on my home "high-end" system, I am also accustomed to this flat sound.

I don't dismiss the work Harman has done in any way, and as a matter of fact, I do believe the research outcome is accurate. Most consumers are accustomed to an expected downslope curve because this is how speakers sound under domestic circumstances. The same is true for modern TWS in-ear phones. Those who are measured flat(ter) are generally perceived as sounding worse (also by reviewers) than the ones having a Harman curve applied. This proves again, yes, this is what consumers expect and perceive as good sound.

That doesn't implicate that this IS (the only way of) good sound.

Referring back to myself. I am open-minded, and I am aware that what I believe in today can change tomorrow if it turns out I had it wrong all along.
Related to this topic, I programmed a downsloping curve in my system with the Genelecs. I listened to it for a prolonged time without switching back. I did so mildly at 3dB down at 20Khz. I didn't like it. Too much bass, which muffles the mix, reduces tiny details on attacks such as snares, etc. Especially with movies, this curve removes that "magic" bubble it portraits otherwise with a flat target curve. Please see attached the target curve and predicted corrected response.

The Genelecs (8351B) doesn't sound bright, harsh, or anything like that when giving it a flat target curve. (neither this curve gives me listening fatigue). They just sound clear and more upfront but not "in-your-face" without being aggressive. Neither does this flat curve change the excellent character (or lack off) of the speaker. It is unmistakable the 8351B but with a reduction of room influences with a more direct sound even at a 2-meter distance (my sweet spot)

Tonight, I go back to my flat curve again. I may be doing the "wrong" thing totally, but this is what I am accustomed to, and to me sounds like an uncolored window into the content you provide to them.

Note: The subwoofer and mains are not overlapping. Trinnov only shows the raw predicted correction without incorporating the crossover. Mains and subs are crossed at 85Hz, with an 8th order L/R filter. The predicted response isn't "ruler flat" because I have limited the EQ-ing of dips, because I believe that is a wrong thing to do if been done to extreme.
 

Attachments

  • Tilt curve.jpg
    Tilt curve.jpg
    101.5 KB · Views: 178
  • amplitude.jpg
    amplitude.jpg
    103.4 KB · Views: 172
Last edited:

mkawa

Addicted to Fun and Learning
Joined
Sep 17, 2019
Messages
788
Likes
695
i got an absolutely insane price on a pair of demos and am blown away by these, even more than in my listening sessions. my girlfriend and I thinking they do need a sub to really fill out the sound, but i do like the way GLM chases flat response regardless. the analog in is also so good that it's hard to imagine how the AES input could be measurably better, THD is already so low. anyway, really enjoying these and finding that my floorstandards have some THD issues in the midrange (but better low-end because that's where my excellent sub is set up right now) that i'm not crazy about.

i was originally going to pick up an all-digital minidsp dirac interface for the floorstanders in the living room, but space issues after the move and some doubts that dirac will fix the issues with that speaker, dsp and amp combo mean i'll probably put my 8040s in the bedroom and the 8351s in the living room. and leave it at that. now that the office needs to fit two of us, i think it will be headphones only.
 

q3cpma

Major Contributor
Joined
May 22, 2019
Messages
3,060
Likes
4,416
Location
France
i got an absolutely insane price on a pair of demos and am blown away by these, even more than in my listening sessions. my girlfriend and I thinking they do need a sub to really fill out the sound, but i do like the way GLM chases flat response regardless. the analog in is also so good that it's hard to imagine how the AES input could be measurably better, THD is already so low. anyway, really enjoying these and finding that my floorstandards have some THD issues in the midrange (but better low-end because that's where my excellent sub is set up right now) that i'm not crazy about.

i was originally going to pick up an all-digital minidsp dirac interface for the floorstanders in the living room, but space issues after the move and some doubts that dirac will fix the issues with that speaker, dsp and amp combo mean i'll probably put my 8040s in the bedroom and the 8351s in the living room. and leave it at that. now that the office needs to fit two of us, i think it will be headphones only.
How much, by curiosity?
 

Sebastiaan de Vries

Active Member
Joined
Sep 4, 2019
Messages
114
Likes
305
Location
China, Shenzhen
i got an absolutely insane price on a pair of demos and am blown away by these, even more than in my listening sessions. my girlfriend and I thinking they do need a sub to really fill out the sound, but i do like the way GLM chases flat response regardless. the analog in is also so good that it's hard to imagine how the AES input could be measurably better, THD is already so low. anyway, really enjoying these and finding that my floorstandards have some THD issues in the midrange (but better low-end because that's where my excellent sub is set up right now) that i'm not crazy about.

i was originally going to pick up an all-digital minidsp dirac interface for the floorstanders in the living room, but space issues after the move and some doubts that dirac will fix the issues with that speaker, dsp and amp combo mean i'll probably put my 8040s in the bedroom and the 8351s in the living room. and leave it at that. now that the office needs to fit two of us, i think it will be headphones only.

Congrats with your purchase. As overwhelmed you currently are with your purchase (and you should really enjoy all facets of it) I can guarantee you that:
  1. Feeding directly with AES/EBU will provide you this very last edge of precision and resolution. Keep in mind, despite that the analoge input stage in the 8351B is very well executed, this is only half of the story. Essentially you also eliminate the DAC and analogue output stage of the source when going digitally. From a chain point of view you will bypass one DAC, one ADC and two pre amps stages by going digitally.
  2. Extending the system with a well integrated subwoofer solution ALWAYS improves the overal fidelity and system performance.
    1. You can (and should) place the subwoofer where it performs best, which is mostly not where the monitors performs best for mid-high
    2. You take away most of the strain from the drivers and power electronics from the 8351’s which generally reduces the overal distortion and makes the system breathe.
    3. You transform the system into full range, which has merits in any source reproduction or production.
 
  • Like
Reactions: kuf

HooStat

Addicted to Fun and Learning
Joined
May 11, 2020
Messages
856
Likes
933
Location
Calabasas, CA
Feeding directly with AES/EBU will provide you this very last edge of precision and resolution. Keep in mind, despite that the analoge input stage in the 8351B is very well executed, this is only half of the story. Essentially you also eliminate the DAC and analogue output stage of the source when going digitally. From a chain point of view you will bypass one DAC, one ADC and two pre amps stages by going digitally.
I understand what you are saying, and I would probably do my best to avoid an analog stage myself. But I find it hard to believe that this could really be audible in any meaningful way. I would love someone to actually do some blinded, level-matched testing on this. This is not directed at you specifically -- this comment comes up for any monitor that can handle both analog and digital. I appreciate all your comments in this thread because they are very helpful, and I am still hoping to get myself a pair of these.
 

dominikz

Addicted to Fun and Learning
Forum Donor
Joined
Oct 10, 2020
Messages
803
Likes
2,626
I don't dismiss the work Harman has done in any way, and as a matter of fact, I do believe the research outcome is accurate. Most consumers are accustomed to an expected downslope curve because this is how speakers sound under domestic circumstances. The same is true for modern TWS in-ear phones. Those who are measured flat(ter) are generally perceived as sounding worse (also by reviewers) than the ones having a Harman curve applied. This proves again, yes, this is what consumers expect and perceive as good sound.
In my understanding Harman/Toole/Olive research on loudspeaker preference actually confirms that anechoically flat-measuring loudspeakers are typically prefered by most listeners. So the direct sound coming out of the loudspeakers should indeed be flat - and in the nearfield (and/or very dead rooms) this is what we will also measure at the listening position.

However what you will measure in a typical (reflective) room, some distance from the loudspeakers (i.e. out of nearfield), is not only the loudspeaker on-axis direct sound - it will also contain a significant amount of reflections of off-axis radiated sound. Since conventional forward-firing loudspeakers are increasingly directional with frequency, most of the off-axis radiated sound has a downward slope.

For forward-firing loudspeakers with flat on-axis/LW sound and good directivity, such in-room, steady-state measured response will therefore be downward-sloping, due to the contribution of these reflections.

If the measured in-room response of such loudspeakers was instead flat, that would effectively mean the loudspeaker's anechoic on-axis/LW response had to have an upward/bright tilt - I believe this is what some members were suggesting.
 
Last edited:

Sebastiaan de Vries

Active Member
Joined
Sep 4, 2019
Messages
114
Likes
305
Location
China, Shenzhen
In my understanding Harman/Toole/Olive research on loudspeaker preference actually confirms that anechoically flat-measuring loudspeakers are typically prefered by most listeners. So the direct sound coming out of the loudspeakers should indeed be flat - and in the nearfield (and/or very dead rooms) this is what we will also measure at the listening position.

However what you will measure in a typical (reflective) room, some distance from the loudspeakers (i.e. out of nearfield), is not only the loudspeaker on-axis direct sound - it will also contain a significant amount of reflections of off-axis radiated sound. Since conventional forward-firing loudspeakers are increasingly directional with frequency, most of the off-axis radiated sound has a downward slope.

For forward-firing loudspeakers with flat on-axis/LW sound and good directivity, such in-room, steady-state measured response will therefore be downward-sloping, due to the contribution of these reflections.

If the measured in-room response of such loudspeakers was instead flat, that would effectively mean the loudspeaker's anechoic on-axis/LW response had to have an upward/bright tilt - I believe this is what some members were suggesting.
I believe we both have the same understanding.
My point is, in the control room of a recording studio, the monitors placed on the meterbridge of the mixing console and very close to the mix engineer. Combined with the dampened control room acoustics what you get is mostly a flat response without the typical down sloping curve you get when place a anechoic flat speaker in a typical living room with typical far field distances.

I am used to this kind of flat response and that might explain my preference.
 

dominikz

Addicted to Fun and Learning
Forum Donor
Joined
Oct 10, 2020
Messages
803
Likes
2,626
I believe we both have the same understanding.
My point is, in the control room of a recording studio, the monitors placed on the meterbridge of the mixing console and very close to the mix engineer. Combined with the dampened control room acoustics what you get is mostly a flat response without the typical down sloping curve you get when place a anechoic flat speaker in a typical living room with typical far field distances.
Agreed. :)

However, I'd just like to add that the research also seems to suggests that, with a flat measuring speaker in a reflective room, even if the steady-state in-room response is downward sloping we 'hear through the room' and in effect our brain seems to be able to separate the reflections from direct sound and perceive the 'flat' tonality of the loudspeaker - i.e. we do not seem to hear the 'tilted' in-room response literally.

Have to say that this also tracks with my personal experience - if I measure my (very) nearfield desktop setup in-room at LP it measures relatively flat. If I measure my living room system in-room at MLP it measures downward sloping - but I perceive a similar tonality in both cases since both pairs of speakers have reasonably flat anechoic LW response. Of course I perceive different levels of envelopment, spaciousness, phantom image width, etc. in the two rooms - it is just the basic tonality that remains reasonably similar for me.

Early in my experimentation with room correction I initially tried EQ-ing my living room loudspeakers to a flat target and found the result excessively bass-shy (also compared to my nearfield setup) - but note that my living room is fairly reflective. This is where I realized that, as long as you know your loudspeakers are flat anechoically, tracking the natural in-room response with room correction targets may give more natural results. In my case the natural in-room response slope is somewhere around -0,7dB per octave.

I am used to this kind of flat response and that might explain my preference.
Of course, there is nothing wrong with any kind of personal preference!
 
Last edited:

Sebastiaan de Vries

Active Member
Joined
Sep 4, 2019
Messages
114
Likes
305
Location
China, Shenzhen
I
Agreed. :)

However, I'd just like to add that the research also seems to suggests that, with a flat measuring speaker in a reflective room, even if the steady-state in-room response is downward sloping we 'hear through the room' and in effect our brain seems to be able to separate the reflections from direct sound and perceive the 'flat' tonality of the loudspeaker - i.e. we do not seem to hear the 'tilted' in-room response literally.

Have to say that this also tracks with my personal experience - if I measure my (very) nearfield desktop setup in-room at LP it measures relatively flat. If I measure my living room system in-room at MLP it measures downward sloping - but I perceive a similar tonality in both cases since both pairs of speakers have reasonably flat anechoic LW response. Of course I perceive different levels of envelopment, spaciousness, phantom image width, etc. in the two rooms - it is just the basic tonality that remains reasonably similar for me.

Early in my experimentation with room correction I initially tried EQ-ing my living room loudspeakers to a flat target and found the result excessively bass-shy (also compared to my nearfield setup) - but note that my living room is fairly reflective. This is where I realized that, as long as you know your loudspeakers are flat anechoically, tracking the natural in-room response with room correction targets may give more natural results. In my case the natural in-room response slope is somewhere around -0,7dB per octave.


Of course, there is nothing wrong with any kind of personal preference!

I am in line with all you say. I think I got accustomed my tuned flat response at the sweet spot. When applied the sloped curve I felt it got muffled and I missed something.

One thing still keeps me thinking. The downwards slope curve varies per distance and listening. Would one always try to attempt to follow the downwards slope of the particular living room? Or should there be sort of a standard average Harman curve?
 

thewas

Master Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
6,741
Likes
16,174
One thing still keeps me thinking. The downwards slope curve varies per distance and listening. Would one always try to attempt to follow the downwards slope of the particular living room? Or should there be sort of a standard average Harman curve?
Because like you say the downward slope depends on loudspeakers directivity, room reverberation trend and listening distance, its best to aim for flat direct sound above modal region and not some fixed predefined targets, see more here https://www.audiosciencereview.com/...ut-room-curve-targets-room-eq-and-more.10950/
 

dominikz

Addicted to Fun and Learning
Forum Donor
Joined
Oct 10, 2020
Messages
803
Likes
2,626
Would one always try to attempt to follow the downwards slope of the particular living room? Or should there be sort of a standard average Harman curve?
In my understanding dr. Toole (and I'm assuming Harman) doesn't really advocate EQ-ing to any 'target curve' above the transition frequency - the curve(s) from dr. Toole's book often shared in this forum show the statistical average of in-room measurements of 'flat' measuring loudspeakers. So the 'Harman curve' is really a result of having a good loudspeaker measured in a room, rather than a target for EQ :)
 

mkawa

Addicted to Fun and Learning
Joined
Sep 17, 2019
Messages
788
Likes
695
Agreed. :)

However, I'd just like to add that the research also seems to suggests that, with a flat measuring speaker in a reflective room, even if the steady-state in-room response is downward sloping we 'hear through the room' and in effect our brain seems to be able to separate the reflections from direct sound and perceive the 'flat' tonality of the loudspeaker - i.e. we do not seem to hear the 'tilted' in-room response literally.
that is fascinating.
 

dominikz

Addicted to Fun and Learning
Forum Donor
Joined
Oct 10, 2020
Messages
803
Likes
2,626
that is fascinating.
It is :) But on second thought it may also be just a tiny bit less surprising if we remember that with real sound sources it is similar - e.g. if you speak with a friend outdoor and in a church their voice will in both cases have the same familiar tonality and timbre - the level of perceived 'spaciousness' surrounding it is what changes between the two settings.

There's some pretty advanced signal analysis and processing happening inside our heads, apparently :D
 
Top Bottom