• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Sigberg Audio SBS.1 Active speakers prototype/build thread

OP
sigbergaudio

sigbergaudio

Major Contributor
Manufacturer
Forum Donor
Joined
Aug 21, 2020
Messages
1,479
Likes
2,628
Location
Norway
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. :)

Absolutely. It will be listened to (and measured) in different rooms to make sure we iron out the kinks. The peak is what I'd call a "known issue" at the moment, so it will be adressed when/if necessary! Note that the response was quite close (50cm), so it looks different further away as well. Already at 1m it's less pronounced on-axis too.

EDIT: We're also measuring full 360 around the speaker to see which frequency ranges might have excessive energy off-axis, and still need damping despite what the on-axis or early off-axis measurements indicate.
 
OP
sigbergaudio

sigbergaudio

Major Contributor
Manufacturer
Forum Donor
Joined
Aug 21, 2020
Messages
1,479
Likes
2,628
Location
Norway
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).

Did some adjustments including shifting the crossover a bit, and ended up with this (green) as opposed to the purple which was the previous one. The new one is a bit more uneven as it's taken indoors. Both are nearfield on-axis. Will need some listening sessions now. :)

sbs1-prototype-onaxis-89.jpg
 

HooStat

Addicted to Fun and Learning
Joined
May 11, 2020
Messages
816
Likes
849
Location
Calabasas, CA
Do you have plans to "coordinate" the DSP between your sub(s) and the monitors? Cross-over and phase?
 
OP
sigbergaudio

sigbergaudio

Major Contributor
Manufacturer
Forum Donor
Joined
Aug 21, 2020
Messages
1,479
Likes
2,628
Location
Norway
Do you have plans to "coordinate" the DSP between your sub(s) and the monitors? Cross-over and phase?

The plan is for the subwoofers to ship with a preset that is matched with the speakers in terms of crossover. There won't be any automatic phase match. We will however choose crossover slopes on both the subwoofers and the speakers in order to minimize phase issues.
 

Mnyb

Major Contributor
Forum Donor
Joined
Aug 14, 2019
Messages
1,735
Likes
2,298
Location
Sweden, Västerås
Hey what's the smoothing in your graphs ? could the graphs be made to have the same scale as other speakers measured at ASR or do you plan to send a non prototype to Amir when your done designing :)

Curious about the coaxial you use, we are probably spoiled by KEF's and genelecs struggle to perfect coaxial drivers . Other coaxials have shown more artifacts in measurement than KEF or genelec .

Your happy with your choice of driver and thinks it's worth the effort and tradeoffs to use a coaxial ? i don't design speakers and realize that it's a 1000 tradeoffs that needs to be balanced to each others , but most speakers are not coaxial .

In your mind when is it right to go for the coax solution ? juggling all possibilities . have the aviable coaxial drivers become good enough now ?

I'm personally very interest to try a coax as the concept seem intriguing and the have got much better lately
 
OP
sigbergaudio

sigbergaudio

Major Contributor
Manufacturer
Forum Donor
Joined
Aug 21, 2020
Messages
1,479
Likes
2,628
Location
Norway
Hey what's the smoothing in your graphs ? could the graphs be made to have the same scale as other speakers measured at ASR or do you plan to send a non prototype to Amir when your done designing :)

Curious about the coaxial you use, we are probably spoiled by KEF's and genelecs struggle to perfect coaxial drivers . Other coaxials have shown more artifacts in measurement than KEF or genelec .

Your happy with your choice of driver and thinks it's worth the effort and tradeoffs to use a coaxial ? i don't design speakers and realize that it's a 1000 tradeoffs that needs to be balanced to each others , but most speakers are not coaxial .

In your mind when is it right to go for the coax solution ? juggling all possibilities . have the aviable coaxial drivers become good enough now ?

I'm personally very interest to try a coax as the concept seem intriguing and the have got much better lately

I'd say these coax drivers are insanely good, and it doesn't come through as a tradeoff or an effort at all, it's more like "Why doesn't everyone use coaxes?" :)

I hope to push out some more measurements today, will try to align the graphs to a more similar scale as the Amir reviews.
 
OP
sigbergaudio

sigbergaudio

Major Contributor
Manufacturer
Forum Donor
Joined
Aug 21, 2020
Messages
1,479
Likes
2,628
Location
Norway
Here are a few more measurements, more similar to those typically found in the reviews here on ASR. Please note that the graphs here are typically more uneven than you are used to seeing as they are partially gated, in-door measurements, and thus have more noise from the room than what Amir has in his fancy Klippel measurements. Due to the gating, the graphs loose accuracy below around 200-250hz. These graphs indicate roll-off starting at 300hz, while in reality roll-off doesn't start until 100hz.

The first one is what's called "listening window", which is an average of on-axis, and 15 degree off-axis in all directions. Relatively smooth and flat, which is what we're looking for.
sbs.1.listeningwindow.jpg


This is on-axis, 15 degree off-axis and 30 degree off-axis 1/12th smoothing. Again more noise here due to the room, but it shows the off-axis roll-off.

sbs.1.0-15-30.jpg



[email protected] fundamental + THD. This shows one speaker easily displaying less than 1% THD across the board. Remember that these are small, dual 5.5" bookshelf speakers.

sbs.1.96dB.jpg


Posting Impulse response again for convenience
sbs1-prototype-impulse.jpg


Posting Step response again for convenience
sbs1-prototype-step.jpg
 
OP
sigbergaudio

sigbergaudio

Major Contributor
Manufacturer
Forum Donor
Joined
Aug 21, 2020
Messages
1,479
Likes
2,628
Location
Norway
OP
sigbergaudio

sigbergaudio

Major Contributor
Manufacturer
Forum Donor
Joined
Aug 21, 2020
Messages
1,479
Likes
2,628
Location
Norway
Hey what's the smoothing in your graphs ? could the graphs be made to have the same scale as other speakers measured at ASR or do you plan to send a non prototype to Amir when your done designing :)

Curious about the coaxial you use, we are probably spoiled by KEF's and genelecs struggle to perfect coaxial drivers . Other coaxials have shown more artifacts in measurement than KEF or genelec .

Your happy with your choice of driver and thinks it's worth the effort and tradeoffs to use a coaxial ? i don't design speakers and realize that it's a 1000 tradeoffs that needs to be balanced to each others , but most speakers are not coaxial .

In your mind when is it right to go for the coax solution ? juggling all possibilities . have the aviable coaxial drivers become good enough now ?

I'm personally very interest to try a coax as the concept seem intriguing and the have got much better lately

To expand on my answer a bit (or actually answer all your questions):

The reason for choosing a coaxial in the first place was to be able to have the benefit of two woofers / midbass drivers in the same size and form factor as a traditional 2-way. This enables us to create a speaker with more power and headroom than would otherwise be possible. This driver is very well behaved, and since these are active speakers, it's also relatively easy to do minor adjustments if necessary.

I'm not sure why not more speakers use coax today, I suspect it's down to marketing and brand image as much as it is technical reasons. Based on the positive experience with this driver so far, I'm inclined to explore coaxes in future designs as well.
 
OP
sigbergaudio

sigbergaudio

Major Contributor
Manufacturer
Forum Donor
Joined
Aug 21, 2020
Messages
1,479
Likes
2,628
Location
Norway
The loudspeakers have now been moved from the test room into a living room space that is significantly larger and also have less dampening. We're also testing a slanted baffle, currently leaning back around 4 degrees.

That means we had to adjust the top end a bit to end up with roughly the same nearfield on-axis response. This means the end result is a bit more top end energy in the room, and a more lively sound.

Here is 50cm on-axis (green curve is new, now with the loudspeaker tilted 4 degrees), looks almost the same.

1614413775322-png.700737


In addition to increasing the top end, I've also adjusted the crossover and treble gain a bit to lift the 2-8k area. This isn't as easily spotted in the nearfield measurement above, but when we move into the room and to the listening position, the change is apparent (red is old curve):

1614413993046-png.700740


Since the change is across such a broad frequency range, the difference is quite audible.


1614414184067-png.700742



So was the change a bit too much or just right? That remains to be seen, and will be evaulated not just by eyeballing frequency curves, but through extended listening. :)

In other news the first pre-production model with proper finish has been ordered, and will probably arrive in roughly a month. Can't wait to show pictures of it! It will also come with some acoustical improvements, so it will come with not only improved visuals, but also improved sound. :)

Finally we've decided to increase the depth of the loudspeaker a bit. Most 2-way speakers are almost as deep as they are tall to get enough volume for bass. This means many speaker stands are simply too deep for this speaker. We don't really need that extra volume for bass in these speakers, but will nonetheless increase the depth a bit to make it more compatible with speaker stands. This means we can reduce the height slightly.

Current size (size of prototype): 190x420x200mm.

New size: 190x410x230mm.

Which looks roughly like this:
1614421900261.png


--
 
Last edited:
OP
sigbergaudio

sigbergaudio

Major Contributor
Manufacturer
Forum Donor
Joined
Aug 21, 2020
Messages
1,479
Likes
2,628
Location
Norway
Looking nice. Good idea about reducing height vs depth, the difference between top and bottom margin in your pictures made my mind itch a bit.
By the way, black screws would really fit nicely.

Do you mean difference in space between the drivers and the top edge vs bottom edge? Does it look wrong to you in the actual picture of the prototypes (on each side of the TV) as well?

There will definitely be black screws (M4 bolts similar to in the picture below) on the production model, this is an early prototype. :)
1614585183213.png
 
Last edited:

q3cpma

Major Contributor
Joined
May 22, 2019
Messages
3,062
Likes
4,348
Location
France
Do you mean difference in space between the drivers and the top edge vs bottom edge? Does it look wrong to you in the actual picture of the prototypes (on each side of the TV) as well?

There will definitely be black screws (M4 bolts similar to in the picture below) on the production model, this is an early prototype. :)
View attachment 115595
Ah, it did look "wrong" on the TV picture (final design). I do expect the rounded corners and logo filling the void to change my perception, though. The bare space just made the drivers seem very close to me.
 
Last edited:

HooStat

Addicted to Fun and Learning
Joined
May 11, 2020
Messages
816
Likes
849
Location
Calabasas, CA
Why tilt the baffle? I assume you would use DSP to do any time alignment. And with concentric drivers, it seems even less important. But clearly it was important.
 
OP
sigbergaudio

sigbergaudio

Major Contributor
Manufacturer
Forum Donor
Joined
Aug 21, 2020
Messages
1,479
Likes
2,628
Location
Norway
Why tilt the baffle? I assume you would use DSP to do any time alignment. And with concentric drivers, it seems even less important. But clearly it was important.

First of all I'd like to point out that this is a prototype thread, and the point of prototyping is of course to test how things work. That means you may see things in this thread (and our other prototype threads) that won't make it to the production models. :)

When it comes to tilting of the baffle, you are of course correct that the most common reason (or at least stated reason) for this is time alignment. I also suspect some do it just to make their speakers look interesting. :)

As you indicate, these drivers are basically time aligned by default, since the upper driver is a coax, and the lower driver is an identical size driver from the same manufacturer with almost the exact same dimensions. The timing is also less critical to the lower driver since it's rolling off relatively early. Finally we're using 1.order / async slopes so the phase is linear throughout both crossover points.

So then why tilt the baffle? We're currently testing 4 degrees and 7 degrees. If we choose the latter, on-axis will be roughly half way between a sitting and standing position. So if we optimize the frequency response for 7 degrees off-axis in the vertical plane, we will actually get roughly similar response wether you sit or stand. Based on the testing so far I'm leaning (no pun intended) towards the 4 degrees variant, which will not give the same effect, but nonetheless still leave the standing position closer to the intended listening axis than if it wasn't tilted at all.

(Disclaimer, the following paragraph has some subjective / creative descriptions of the sound, not sure how to put it any other way :D)
But the primary effect is more about the sound energy as well as imaging and sound stage in the room. These speakers will be designed to have very little toe-in. The recommended toe-in will probably be zero degrees, so none. Thanks in part to the coax drivers, with these speakers you get the benefit of the wide soundstage that little to no toe-in gives you, while still having really good imaging and stereo perspective. At no point does it feel like the sound is coming from the speakers, the different instruments are painted across the entire width of the room and are somewhere in between, or even to the side or in the vertical plane. Slightly tilting them so the tweeter on-axis is slighlty above ear level gives an even wider impression, and also gives an added impression of height, giving for instance vocals a more lifelike sound - you don't want the singer to sound like she's 90cm tall, right? :) So all of these things adds up to making the speaker sound "larger than life" which was one of the design parameters of these speakers.

A secondary effect is that when designing for off-axis listening, we have to increase the high frequency energy to get the correct frequency response in the listening position. This subtly increases the total high frequency energy in the room, adding a bit of sparkle and clarity without making them too bright. Tilting them a bit requires us to do that again, further increasing this effect. The result is quite immersive and fascinating - small speakers that sound really large, and they also sound smooth and without a hint of listening fatigue, without losing clarity or attack.
 
OP
sigbergaudio

sigbergaudio

Major Contributor
Manufacturer
Forum Donor
Joined
Aug 21, 2020
Messages
1,479
Likes
2,628
Location
Norway
Tested more with baffle tilt, will probably end up with 4 degrees.

Here's updated 0-15-30 degrees off-axis, pretty happy with the dispersion now. 1/12th smoothing. Since the speakers are designed to have minimal toe-in, the 15 degrees off-axis graph is closest to what will be the listening axis.

sbs-1-0-15-30-4deg-tilt-jpg.701806
 
OP
sigbergaudio

sigbergaudio

Major Contributor
Manufacturer
Forum Donor
Joined
Aug 21, 2020
Messages
1,479
Likes
2,628
Location
Norway
And left+right channel in current setup at listening position. Obviously getting room information as well, but you can still see an indication of overall response (below 80-100hz can be ignored as there's a subwoofer enabled here too).

1/12th smoothing.

1614793230513.png
 
Top Bottom