Sigberg Audio SBS.1 Active speakers development thread

[sigbergaudio]

Thanks to everyone who have followed our other threads about the Inkognito subwoofer as well as the upcoming Sigberg Audio 10D

At the moment we're also working on another, perhaps even more interesting product that we will present in this thread, and we will share measurements, pictures and progress here as well!

Sigberg Audio SBS.1 will be a small, active speaker with a big sound. So it may be possible to guess what SBS is short for. Many interesting, active speakers have emerged recently, utilizing DSP to magically create deep bass from relatively small enclosures. With SBS.1 we're going in the opposite direction. But however you do it, the limiting factor of small speakers will always be their ability to reproduce bass at realistic levels. And this challenge will hold the speakers back, limiting their overall SPL capability. This is why we have designed SBS.1 explicitly to be paired up to a subwoofer. By design, the SBS.1 -3dB point is 80hz (exact specifications subject to change).

Design and design goals
The concept and idea is to build a speaker that generates a large sound stage, and that sounds larger than the physical dimensions would imply. It can play very loud while at the same time reproduce music that is clean, full of detail and sparkle, with a wide and precise stereo perspective. And while that may sound like something I read in a Hifi magazine, it's actually my own listening impressions from the prototype.

Tentative dimensions: 190x420x200mm (WHD). The loudspeakers are fitted with one 5.5" midbass driver, and one 5.5" coax. It's 2.5-way speaker. Each speaker has a 3 channel Hypex amp supporting XLR/RCA/Digital in.

A pair of SBS.1 is expected to be able to play 110dB@1m, and thus be capable of SPL that is more than reasonable in a living room situation. This also means distortion will be very low even when you're enjoying live recordings at concert level. We also expect it to sound good both if your favorite artist is Norah Jones and if it's Five Finger Death Punch. In other words we hope to strike the balance where you can enjoy an open, clear sound that sounds great with vocals, realistic high frequency percussion, while not coming across as "unforgiving". SBS.1 should be pleasant to
listen to even though you have a broad and varied record collection, while still producing goosebumps when you put on a track with great production. When you get this speaker, you should want to rediscover your record collection because you want to listen to it all again.


Concept illustration

Finish / Enclosure
The loudspeakers will have clean design with a high quality cabinet, planned finishes: Satin black and Satin white, similar to our Inkognito subwoofers. The speakers should blend easily in any living room situation.

SBS.1 + Inkognito = TRUE
For those who have seen our Inkognito subwoofers, can probably guess that these speakers will be the natural companion to those (or our upcoming 10D). Our subwoofers will come with a preset that ensures perfect crossover integration with the SBS.1 (can be retrofitted for those who have purchased one of our subwoofers already). You may of course also pair SBS.1 with subwoofers from other manufacturers.

Now what?
Prototyping is in progress, and it looks very promising both on paper and based on listening sessions. We use both measurements and ears extensively during testing, and will keep tweaking through the winter and spring. If everything goes according to plan, SBS.1 will be available this fall, possibly opening up for preorder this summer. We will be sharing pretty detailed measurements in this thread quite soon, and of course general progress.

We appreciate any feedback, comments, suggestions and questions!

 

[sigbergaudio]

A small teaser with a couple of preliminary measurements. This will be dialed in further in the coming days.

Nearfield on-axis, 1/24th smoothing:

Listening position / Sweetspot in our test room. 1/12th smoothing, no room correction, just the speakers natural response in the room. This is more or less +/-3dB from 65-16,000hz, which I think is pretty decent in-room accuracy. The drop from 100hz-10khz is roughly 5dB which is mostly on target as well. This is a moderately damped room, so we may need to drop a bit more to avoid it being too bright in regular rooms. There's also a bit too much energy from 1-10khz due to the fact that the coax is more sensitive than the second driver, something we haven't fully compensated for yet. The low end extension is almost a bit much in-room since they're designed to be used with a subwoofer, so will probably roll it off a bit earlier.

More to follow.

 

[q3cpma]

Any reason for making it 2.5-way instead of 3-way? That would "solve" the problem of coaxial IMD and allow for less distortion in general; at the cost of one more amp, of course.

 

[sigbergaudio]

Any reason for making it 2.5-way instead of 3-way? That would "solve" the problem of coaxial IMD and allow for less distortion in general; at the cost of one more amp, of course.

The speakers are designed to play with a subwoofer, so it's essentially a 3-way or 3.5-way design. The speaker is rolling off early, so excursion is low to begin with, and the fact that we're running a 2.5 way, means we reduce the excursion by an additional 50% compared to a 3-way. The result is very high SPL capability, and both IMD and THD will be low at any reasonable SPL.

 

[Snickers-is]

Since these are made for using with subs with high pass filter, I would say these are 3,5-way systems.

 

[Snickers-is]

Sorry, Mr Sigberg beat me to it!

I made some quick sims (since I have a bit of inside info). The drivers will stay within 2mm at 100W, crossed at 100Hz. That equals a 1m single tone performance of 119,6dB. For music content, it would be much more. So using these speakers one would very rarely exceed 0,5mm in real life situations. That should not be regarded as enough for the shape change effect of the waveguide to be audible.

 

[sigbergaudio]

at the cost of one more amp, of course.

For the record we are using a 3-channel amp, so we have full control of the .5 way driver as well.

 

[q3cpma]

For the record we are using a 3-channel amp, so we have full control of the .5 way driver as well.

I see, then I don't really understand the cons of a true 3-way with the woofer crossing at a few hundred Hz, in this case. A narrower vertical dispersion due to being like an oval driver in the overlapping range (though I suspect there'll be some interference in said vertical directivity)?

 

[sigbergaudio]

I see, then I don't really understand the cons of a true 3-way with the woofer crossing at a few hundred Hz, in this case. A narrower vertical dispersion due to being like an oval driver in the overlapping range (though I suspect there'll be some interference in said vertical directivity)?

Again, the second driver would have to work twice as hard where the work is hardest (in the lowest frequency range).

 

[q3cpma]

Again, the second driver would have to work twice as hard where it matters most (in the lowest frequency range).

Hmmm, true. Still not completely convinced, but it makes sense.

 

[Snickers-is]

You must not forget that a single midrange crossed over at 140Hz will have about the same excursion at XO as this setup, but with significantly higher cone speed.

I did an investigation of the doppler effect once. It appeared that this effect is more dependent on the cone speed than the cone position. At a cone speed of 1m/s is within "extremely hard to hear" (this is a type of frequency modulated intermodulation distortion, and it is less audible than amplitude modulated IMD). XO@100Hz and max 2mm excursion means about 0,56 m/s. Skipping one mid and moving up to 140Hz means the cone speed is close to 0,8m/s instead. Both are limited to 2mm, but the latter will have a higher likelyness of doppler being audible.

Here is an example made by Purifi:
https://purifi-audio.com/2019/12/07/doppler-distortion-vs-imd/

They used x-max of 12,5mm @ 43Hz. It shows that doppler is audible at this level (just above 1 m/s). However, it is muc harder to hear when we are not dealing with 2 simultaneous pure sines. In more realistic scenarios, even 1 m/s is something you can rarely register.

 

[HooStat]

Finally, some who is addressing the right problems. Leave the bass to drivers that can handle it, and maximize the performance of the drivers in the monitors themselves. I wanted to post something a while back, asking if anyone knew of a monitor that didn't make any trade-offs for bass extension, but I was pretty sure the answer was "no one". I mean, if we are "supposed to have" 2-3 subs for proper bass in a room, why try to design bass extension in a monitor?

I like this, and it fits your company's strengths. One thing to consider is whether it will work well on its side as a center channel. Very likely to be a rare use case, but that is the only issue I can think of. How tall do you expect this to be? (I realize nothing is final, of course.)

 

[sigbergaudio]

Finally, some who is addressing the right problems. Leave the bass to drivers that can handle it, and maximize the performance of the drivers in the monitors themselves. I wanted to post something a while back, asking if anyone knew of a monitor that didn't make any trade-offs for bass extension, but I was pretty sure the answer was "no one". I mean, if we are "supposed to have" 2-3 subs for proper bass in a room, why try to design bass extension in a monitor?

I like this, and it fits your company's strengths. One thing to consider is whether it will work well on its side as a center channel. Very likely to be a rare use case, but that is the only issue I can think of. How tall do you expect this to be? (I realize nothing is final, of course.)

Thank you for your positive interest!

Vertical and horisontal dispersion are of course different, as with any speaker, and many center speakers are actually a compromise in this regard. It's not explicitly designed as an LCR speaker, we considered this initially, but we have ended up targeting the hifi crowd to a larger degree than the home cinema people. That being said, this speaker would likely be a very effective monitor in most home cinema setups as well, due to its relatively high SPL capability. I will look into testing it as a center channel as some point as well, but it won't be a major design goal.

The current prototypes are 190x420x200mm (WHD), and so far this seems to work well, so I don't expect major deviations from these dimensions.

 

[Snickers-is]

I visited Mr Sigberg today. So far he has really pulled this off. Not too long ago I heard the B&W 800D3 on an Ncore based NAD amp. It was a rather well designed room and placement within the room. I heard some of the same tracks, and it sounded surprisingly similar. However, the 800s has this tendency to be a bit harsch in the upper octaves. It also tends to be a bit bright and miss some of the nuances in the midrange that are far better exposed when the midrange is kept in balance. I feel Sigbergs speakers does much of the same, but they add a layer of diversity between recordings that I do not often hear. They are also able to deliver a physical dimension that I do not often hear from finished speakers, and especially not from stand mount speakers. One speaker that comes to mind is the KEF Blade 2, and the reason for this is that it is probably the KEF-speaker (except from Muon) that is best at connecting the high and low midrange.

Since I know his pricing estimates, I have to say I am a bit stunned.

I also have to mention that besides being a speaker driver and finished loudspeaker engineer by profession, I do not have any kind of commercial connection to this project. All the drivers I am working on at the moment are really far from anything that could fit into Mr Sigbergs cute shoeboxes. So, by letting you know I am working on some pretty bad ass 15 inch bass/midrange drivers for studio applications, as well as some pretty cutting edge large format compression drivers and matching horns, you get an idea where I come from.

 

[sigbergaudio]

Thank you for both the visit and your kind words @Snickers-is !

I'm sure most (in part even me) will find it absurd that the Blade 2 or 800D3 is mentioned after auditioning what at first glance look like a pair of compact, measly 2-way bookshelf speakers. And even at a second glance they're a pair of compact (perhaps not that measly) 2.5-way bookshelf speakers.

What sets them apart is that they're not actually 2.5-way bookshelf speakers. They're designed to be what's essentially a 3.5-way system by tightly integrating them with one or more subwoofers. During our listening and tuning session yesterday, the SBS.1 was paired with our Sigberg Audio 10D prototype. The speakers are (subjectively speaking) amazing in their own right, but they do naturally rely on a high quality subwoofer for the magic to happen.

 

[sigbergaudio]

For those eager to see some measurements - we're still waiting for slightly warmer weather, hoping to be able to do some measurements outside. If all else failse we can do some nearfield inside as well, but it will be a bit noisy, despite the state of the art test room (see photos below).

Early prototype SBS.1 as well as a 10D Prototype in the corner. The finish still need some work, but the sound is already really really promising!

 

[sigbergaudio]

It's still quite cold here, but I couldn't help myself, so in the name of science and music, I've been outside and almost died. Some quick outside measurements helped verify the groundwork we've done inside so far. It also gave me an opportunity to do some quick measurements with less reflections and noise.

Please note that these are all preliminary measurements and not the final response of the speaker. In part because the tuning of the speaker is still in progress, and partly because the production model will have a different / better cabinet.

So first, here's on-axis and 15deg off-axis. These are measured quite close, so there's more boost in the low end in this curve than in reality. So in reality the speaker is currently quite flat on-axis, which is what we want to see. There's a peak at 3khz, but this goes away off-axis, which is closer to the actual response in the listening position. Thus, this has not been damped so far. We can also observe a small dip at 200hz, this is a reflection off the ground - not the actual response of the speaker. Finally, we can see the off-axis rolloff at the top looks quite nice, which means we avoid too much energy due to reflections in the room.

For those interested in step response and impulse response, here you go - these look pretty good as far as I can see.


Step response:

Impulse response:

 

[thewas]

Nice progress, personally I would EQ that almost 5dB on-axis peak on 3,5 kHz and rather prefer a "BBC dip" there in the sound power as its in a region which can be quite fatiguing.
A quick way to get an approximation of the the important listening window and sound power is to place the loudspeaker away from walls and do MMM measurements at a limited frontal "window" and and at several arcs covering approximately the whole sphere around it, respectively (both at appox. distances of 30-50cm to the loudspeaker).

 

[sigbergaudio]

Nice progress, personally I would EQ that almost 5dB on-axis peak on 3,5 kHz and rather prefer a "BBC dip" there in the sound power as its in a region which can be quite fatiguing.
A quick way to get an approximation of the the important listening window and sound power is to place the loudspeaker away from walls and do MMM measurements at a limited frontal "window" and and at several arcs covering approximately the whole sphere around it, respectively (both at appox. distances of 30-50cm to the loudspeaker).

I agree on principle, and we also have configurations where that have been done. But there will be some differences in baffle etc in the production speaker, so we're a bit careful to EQ out things that might disappear anyway, and/or isn't a real problem in-room.

See the most current in-room measurement below, as you can see there's no pronounced peak, and the sound is not fatiguing at all. But we're playing around with several different configurations, we need to listen as well, not just measure. If that peak ends up being a problem in-room, it will of course be damped.

 

[thewas]

Yes, mind you depending also on the listening distance and room reverberation an on-axis peak can be still be perceived even if it doesn't really appear at an ungated LP measurement as our brain perceives above the transition frequency more the direct sound.

But as you say listening is in the end what matters, especially also at different distances and room configurations.