ARCHIVED FORUM -- March 2012 to February 2022READ ONLY FORUM
This is the second Archived Forum which was active between 1st March 2012 and 23rd February 2022
It was about 4 years ago I auditioned the Beolab 90. Notwithstanding the sound quality I was a little surprised that the wide/narrow sound format was one or the other.
Subsequently I purchased the Beolab 50's and of course it is the same.
I have had the 50's nearly 3 years now and can appreciate the broadness and spaciality of the speakers in wide mode along with the detail and finess in narrow.
But I cannot help feeling that it would be a huge benefit to have the option of varying between the two?
Surely this would not be a difficult task and could be acheived via a software update?
I am always chasing the best of worlds I know, but this seems quite simple? Or is it?
New: Beovision Harmony, Beolab 50's, Beolab 28's, Beolab 18's, Beolab 17's, Beosound Stage & LG, Beosound 2, Beoplay M3, Beoplay A1, Beoplay Portal, Beoplay H4 gen 2, Beoplay E8 3.0
Mikipedia on YouTube: https://www.youtube.com/c/Mikipedia
Mikipedi4 on Twitch: https://www.twitch.tv/mikipedi4
Mikipedia on Intagram: https://www.instagram.com/mikipedi4/
Old: Beosound 9000 mk3, Beolab 3's, Beovision Eclipse, Beolab 1's, Beolab 2, Beovision 10-46, Overture 2300, beolab 8000's, Beolab 4000's, Beovision avant 32" etc. etc.
Not a mixture but the ability to move proportionally from one to another.
Hi,
The short answer is
"That's impossible. It's like asking why my Honda Civic doesn't have an infinitely variable gearbox that changes smoothly from 2nd to 3rd - therefore obviating the need of a clutch". (I know I know, there are some belt-drive cars that do this - but my Honda is not one of them....)
The long answer, the following, explains why what I just said is pragmatically true, but technically false:
In order to explain why this would be extremely difficult to achieve, maybe it would help to explain what's happening when you switch modes, and what we had to do to make that happen.
Let's start by pretending that Beolab 90, in its first prototype, was just a "normal" 3-way loudspeaker comprised of the front woofer, the front midrange, and the front tweeter, and let's pretend that we have crossovers at 350 Hz and 3.5 kHz (I just made these numbers up - they're not the real values).
First, we find the directivity (a.k.a. "Beam Width") by measuring each driver individually on-axis, and then with a rotation of 5º, resulting in 72 measurements around the speaker in the horizontal plane, (I'm keeping things in 2D to keep it simple).
Let's focus on the midrange drive alone:
What we want is a directivity that is as narrow as possible, but as SAME as possible at all frequencies. So, we put an identical midrange driver on the side of the loudspeaker to help to control this.
At 350 Hz, the side driver is attenuating the front driver by playing out-of-phase and just a little lower in level (we don't want to cancel completely). At 3.5 kHz, the side driver is doing nothing (because the front driver's directivity is already too narrow). The problem then is that the side driver is also omnidirectional at 350 Hz, so it's also cancelling out its side - which is the front of the loudspeaker. Therefore, it not only reduces the level at 90º, it also reduces the level at 0º. So, we add another identical driver in the front (just under the "main" front driver) to cancel the sound of the side driver towards the front (which is cancelling the sound of the front driver towards the side).
The nice thing is that the second front driver (the one just under the main front driver) can not only cancel the side-effects of the side driver at 350 Hz, but also to widen the beam width at 3.5 kHz.
So far so good...
What I've described so far is just 3 of the midrange drivers' behaviours at only 2 frequencies. However, the principle applies to all 18 drivers at all frequencies.
This means that each driver has a customised filter that controls not only the level at different frequencies (the magnitude response - which many people incorrectly call the "frequency response") but also the phase at different frequencies. That set of 18 filters works only with the drivers in the positions they're in on the loudspeaker. So, if you swapped out all the midrange drivers for a different brand (but that could be screwed into the housing) then it wouldn't work. This is because the frequency dependent levels and phases are built on the natural characteristics of the drivers in the housing.
So, in order to create the narrow mode in Beolab 90, we set a target for the directivity (e.g. -20 dB relative to on-axis at all frequencies within the band limits at 90º off-axis ). We then built the 18 FIR filters (one per driver) that control the frequency-dependent levels and phases of each driver that makes this happen.
In order to create the wide mode, we set a different target (e.g. -6 dB relative to on-axis at all frequencies within the band limits at 90º off-axis), and built a completely different set of 18 FIR filters.
Note that these FIR filters are also completely different in High Latency and Low Latency modes.
This means that, in Beolab 90, because we have 3 Beam Widths, one of which has 5 Beam directions (= 7 directivity settings) and two latencies, then we have 14 different sets of 18 FIR filters stored in the loudspeaker's memory.
The problem is that there is no real relationship between these FIR filters. in addition, if you interpolated the levels and phases from one to the other (say, 50% Narrow FIR and 50% Wide) then you would not get a result that does what you think. (To go back to my pithy analogy from the start, it would be like forcing my Honda's gearstick into the non-existent space between two gears... I wouldn't get a gear 2.5. I'd get a broken gearbox...)
So, in order to make a variable directivity, we would have to make a set of 18 FIR filters for each "step" between Narrow and Wide. Let's say that you had a slider with 10 steps... That means that, instead of the 14 filter sets, we just made a lot more... 140 filter sets, actually... (because there are 9 steps between each of the current 7 settings, and there are still two latency modes to multiply by...)
This means that the answer boils down to two things:
- we can't interpolate between the filters that are in there (that just wouldn't work)
- 140 filters (instead of 14) would require not only 10 times the memory in the loudspeaker (those filters are big) but 10 times the time to hand-tweak and test them individually (the final testing of the 14 filters in Beolab 90 was 3 months of my job, working full-time on only that... and I wasn't alone...)
This means that my initial statement of "that's impossible" isn't really true. It's just not pragmatic...
Hope that answers the 'question' "Surely this would not be a difficult task..." It would.
Have a nice weekend.
-geoff
BV Harmony 65 / Eclipse 55 / BL 50 / 19 / 18 / BS 2 all brass
BV Horizon 48 / V1 40 / BL 17 / Beoplay S3 all black
V1 / BL 3 / BS Essence MKII / Beoplay A3 all white
BS Moment / BS 3000 / BS 3200 / BL 400 all silver
Headphones / bluetooth speakers / BLC NL-ML / BLGateway
I know…B&O virus has grown on me !
Fansastic:Hi Geoff, - as the 90’s have all the magic programmed in them, how big a step down are the acoustical enhancements of the 50’s and then again the 28’s?
- as the 90’s have all the magic programmed in them, how big a step down are the acoustical enhancements of the 50’s and then again the 28’s?
Fansastic:Hi Geoff, - my brother in law happens to be a professional music producer and he is actually understanding all of this. I was discussing with him the working principals of room compensation capability of the BL’s. He was impressed but still a bit sceptical (and here comes my question); he mentioned that no matter how you enhance the sound beam, it will bounce back on the back wall of the room and still have impact of the sound beam. How does this apply for the 90’s for instance?
- my brother in law happens to be a professional music producer and he is actually understanding all of this. I was discussing with him the working principals of room compensation capability of the BL’s. He was impressed but still a bit sceptical (and here comes my question); he mentioned that no matter how you enhance the sound beam, it will bounce back on the back wall of the room and still have impact of the sound beam. How does this apply for the 90’s for instance?
Geoff Martin:This is almost-as-complicated a question, however: In principle, the difference between the 90s, the 50s, and the 28s is the frequency range of the directivity control that I describe above. The 50s control the directivity of the high frequency band using the movable sides of the lens instead. This actually means that, at 20 kHz, the beam width of the 50s is more like the other bands than in the 90s, where the 1" tweeter has started beaming. This is noticeable in the measurements published in the Technical Sound Guides for the two loudspeakers. The 28s only have a single woofer, so it's impossible to do directivity control in the frequency band covered by that driver. However, the method of controlling the directivity is, in principle, the same. Of course, the filters that are calculated and implemented are completely different. Your brother in law is completely correct. However, the back wall reflection will have a primary impact on the timbre instead of the imaging. So, we use the beam width control (directivity) to reduce the side wall reflections, and the ARC (Adaptive Room Compensation) to look after the timbral effects of the back wall reflection. Tell your brother in law that the primary analysis and processing for the ARC is correcting the frequency-dependent phase, not the magnitude response, of the loudspeakers. So, we're not just attenuating the symptoms - we're trying to correct the problem. Cheers -geoff
Mikipidia: Btw maybe silly question, but could one sort of make one of these half way filters by measuring widely (like how the manual tells you to do for wide mode) in the narrow mode and vice versa or taking a bunch of narrow measurements and combining them in to a filter while the speakers play in wide mode? It’s still somewhat static, but could that be a half way between both modes?
Btw maybe silly question, but could one sort of make one of these half way filters by measuring widely (like how the manual tells you to do for wide mode) in the narrow mode and vice versa or taking a bunch of narrow measurements and combining them in to a filter while the speakers play in wide mode?
It’s still somewhat static, but could that be a half way between both modes?
I'm not sure that I understand the question, but you seem to be mixing up the ARC measurements that are done in the listening room with the individual loudspeaker driver measurements that we do in the Cube.
We only check the "total" output of the loudspeaker (using all 18 drivers at once, for example) to verify that the sum of the individual measurements produces the directivity result we expect.
However, the 18 individual driver measurements are done one at a time - they have to be. We don't even assume that the loudspeaker is symmetrical - we measure the left and the right drivers to make sure that this is, in fact, true.
Does that answer your question? Or have I misunderstood?
Cheers
A big thank you to Geoff for such a detailed response. I did not expect that. Whilst I love your analagy with gearboxes I immediately thought of the DAF Variomatic. That didn't help.
So my understanding is that a proportional variance would not be possible but many many "inbetween" positions would, in theory, bar the fact every one would need a new filter?
That does make sense to me and my original "simple software update" comment does seem a tad ignorant now.
But I'll stick my neck on the line and predict that in time B&O do something akin to this is the future?
I'll be 1st in the line for the royalties? :)
Mikipidia:Sorry for not being clear, my question was more inregards to the bl50’s. They sound quite different in wide vs narrow mode and the way you measure them in for both situations varies drastically. So what i guess what i was trying to ask is if we would reach this sort of middlepoint if: we were to calibrate the bl50’s in narrow mode, in 3 different positions( 3 times when you move the mic 3 times with in ~30cm), then combine those filters and have them play in wide? Or vice versa, where you have the speakers calibrate them in narrow mode, but put the mic in 3 different positions, like you would when setting them up in the wide calibration (one mic position per one seating spot), and then use that filter for narrow mode? I know it’s a bit odd, and let me know if it makes sense?! I have one filter setup in wide mode calibration where i calibrated them in like it says to do in narrow mode(with in that virtual box around your head at ~30cm appart) and that sounds more accurate to me than when it’s done along 3 positions along the couch. I also happen to like the sort of airy feeling wide mode give you on the 50’s compared to narrow mode with the same measurement place and method. Either way iam a happy camper New: Beovision Harmony, Beolab 50's, Beolab 18's, Beolab 17's, Beolab 3's, Beosound 9000 mk3, Beosound Stage & LG, Beosound 2, Beoplay M3, Beoplay A1, Beoplay H4 gen 2, Beoplay E8 3.0 Mikipedia on YouTube: https://www.youtube.com/c/Mikipedia mikipedi4 on Twitch: https://www.twitch.tv/mikipedi4 Old: Beovision Eclipse, Beolab 1's, Beolab 2, Beovision 10-46, Overture 2300, beolab 8000's, Beolab 4000's, Beovision avant 32" etc. etc.
Beovision 7-55 MK1 red, Beolab 10 red. Beolab 50, all black. Beolab 17 broken ice. Beolab transmitter. Apple tv4 and apple express 2.
Mikipidia:Iam very happy with them, the whole last paragraph is about that! I was just trying to think along the lines of what beofile7 was saying/trying to do New: Beovision Harmony, Beolab 50's, Beolab 18's, Beolab 17's, Beolab 3's, Beosound 9000 mk3, Beosound Stage & LG, Beosound 2, Beoplay M3, Beoplay A1, Beoplay H4 gen 2, Beoplay E8 3.0 Mikipedia on YouTube: https://www.youtube.com/c/Mikipedia mikipedi4 on Twitch: https://www.twitch.tv/mikipedi4 Old: Beovision Eclipse, Beolab 1's, Beolab 2, Beovision 10-46, Overture 2300, beolab 8000's, Beolab 4000's, Beovision avant 32" etc. etc.
Mikipidia:Sorry for not being clear, my question was more inregards to the bl50’s. They sound quite different in wide vs narrow mode and the way you measure them in for both situations varies drastically. So what i guess what i was trying to ask is if we would reach this sort of middlepoint if: we were to calibrate the bl50’s in narrow mode, in 3 different positions( 3 times when you move the mic 3 times with in ~30cm), then combine those filters and have them play in wide? Or vice versa, where you have the speakers calibrate them in narrow mode, but put the mic in 3 different positions, like you would when setting them up in the wide calibration (one mic position per one seating spot), and then use that filter for narrow mode? I know it’s a bit odd, and let me know if it makes sense?! I have one filter setup in wide mode calibration where i calibrated them in like it says to do in narrow mode(with in that virtual box around your head at ~30cm appart) and that sounds more accurate to me than when it’s done along 3 positions along the couch. I also happen to like the sort of airy feeling wide mode give you on the 50’s compared to narrow mode with the same measurement place and method. Either way iam a happy camper
we were to calibrate the bl50’s in narrow mode, in 3 different positions( 3 times when you move the mic 3 times with in ~30cm), then combine those filters and have them play in wide?
Or vice versa, where you have the speakers calibrate them in narrow mode, but put the mic in 3 different positions, like you would when setting them up in the wide calibration (one mic position per one seating spot), and then use that filter for narrow mode?
I know it’s a bit odd, and let me know if it makes sense?! I have one filter setup in wide mode calibration where i calibrated them in like it says to do in narrow mode(with in that virtual box around your head at ~30cm appart) and that sounds more accurate to me than when it’s done along 3 positions along the couch. I also happen to like the sort of airy feeling wide mode give you on the 50’s compared to narrow mode with the same measurement place and method. Either way iam a happy camper
Sorry... but I'm still a little confused...
Remember that the ARC measurement and resulting filters only work from about 2 kHz and down. Any differences in the timbral behaviour above that are going to be different from room to room (more precisely, as a result of your sidewall reflective characteristics and the relative locations of the loudspeakers, the walls, and you.
The ARC filters will also be different in Low and High latency modes, because they have to compensate differently for the different coupling of the loudspeaker to the room modes (primarily).
You say that the 50s "sound quite different" in wide and narrow - but this is to be expected, in my opinion. If they didn't, then there would be no need to have the two modes.
I guess that the "airy feeling" that you're enjoying in Wide mode is primarily the result of the sidewall reflections. The more obvious this is, then the more it indicates that you'll also get more high frequency content as a result of those reflections. You could attenuate this using the PEQ, however, if you prefer it, then this is a bad idea. :-)
I also don't know what you mean when you say "sounds more accurate" - with respect to timbral balance / overall spectrum? Phantom imaging locations and precision? Temporal control (i.e. lower levels of frequency-dependent ringing?)
All of that being said, we often see when doing listening tests where we ask for a listener's preference between "A" and "B" is that they want something in-between... I'm starting to form the opinion that this is human nature: given a choice between two things, people will want something else.
Geoff Martin: All of that being said, we often see when doing listening tests where we ask for a listener's preference between "A" and "B" is that they want something in-between... I'm starting to form the opinion that this is human nature: given a choice between two things, people will want something else. Cheers -geoff
When I give my dog the choice between two things, it always wants both.
Also - I remember - when my kids were young, they too wanted both…….instantly.
MM
There is a tv - and there is a BV
Millemissen: Also - I remember - when my kids were young, they too wanted both…….instantly.
Good point... but the average is just half of both... :-)
Cheers-geoff
OK, since we're talking about what's possible, and not possible, with the 50s, I have a question for Geoff, should you be so kind as to reply--as my 50s will be arriving at port any day now, and will be installed sometime in July.
I recently developed a case of tinnitus from an ear infection, and though the infection has healed sadly the tinnitus hasn't gone away. According to my ENT/audiologist, my hearing is completely normal without any discernible loss (at least within the range they test), but it may persist for some time.
I've I've been able to pinpoint my tinnitus to a ringing tone in both ears at exactly 10Khz at a volume of approximately 7%. Is there some way, in the setup process for the 50s when we calibrate them with ARC, that I could compensate for that and create a (hopefully temporary) setting to negate the tinnitus while listening to the 50s?
--Beovision Harmony 77, Beovision Harmony 65 x2, Beovision Contour 48, LG GX 65/Beosound Stage/STB StandBeolab 50s, Beolab 28s, Beolab 18s x3, Beolab 19, Beoplay A9, Beosound Edge, Beosound 2 x2, Beosound 1 x2, Beosound Level x2, Beosound Core x2, Beoplay M3, Beosound 9000, Beogram TX2, Beogram 6500 White Edition, 4000c (on order), Beoremote One x8, Beoremote Halo x3, BLC NL/MLBeoplay H95 x3, Beoplay EQ x2, Beoplay E8, Beoplay E8 Sport, Beosound A1 x2
BeoJosh: Is there some way, in the setup process for the 50s when we calibrate them with ARC, that I could compensate for that and create a (hopefully temporary) setting to negate the tinnitus while listening to the 50s?
Is there some way, in the setup process for the 50s when we calibrate them with ARC, that I could compensate for that and create a (hopefully temporary) setting to negate the tinnitus while listening to the 50s?
Sorry, but as far as I know, there's no way to negate tinnitus. If there were hearing loss, then it's quite easy to use the parametric equaliser in the Beolab 50 to make some kind of compensation for this - although it would be difficult / impossible to do it "correctly" whatever that means.
To my knowledge, the only external way to deal with tinnitus is to mask it, which means making the external sound simply loud enough to "down out" the internal sound. However, this means turning the entire signal up - not just one frequency band - which is not necessarily a healthy option in the long term.
Actually, at the last visit to an ENT/audiologist, I learned that many / most / all people have tinnitus - the only differences are the frequency and the level. For many / most people, the level is so low that normal, everyday noises mask the internal sound(s). It's only when sitting quietly in a quiet room that you find out you have it. This is similar to suddenly learning that you're colour blind when having a conversation in a paint shop over the choice between "white", "eggshell", "ivory", "vanilla", and "alabaster" for painting the hallway.
Thanks, Geoff. That's what I figured, but it was worth a shot--and I appreciate your quick reply. Regardless, I'm very much looking forward to getting the 50s properly set up in my system. Can't wait!
Hi Geoff,
I was wondering the other day if it is theoretically possible to make the Beolab 90's beam width even narrower through DSP? Or would it require additional side firing drivers to decrease the energy sent to the sides? And would there theoretically have additional benefits or disadvantages to an even narrower beam?
Thanks,
Ray