Sigberg Audio Saranna (fullrange, cardioid active floorstander) development thread

[Absolute]

I guess it would be interesting to build a version of the Manta without all the holes in it, so a completely traditional enclosure, and measure it in an anechoic chamber, but I never did that. And even the initial prototype had cardiod slots in the upper chamber so I don't even have rudimentary measurements of such an enclosure. But my theory is that the cardioid system makes for a somewhat more chaotic off-axis response. At least the way I ended up doing this. The Dutch&Dutch on the other hand is pretty clean.

The SBS.1 is a sealed speaker with a quite narrow baffle, so it's much easier to make "clean" from a measurement standpoint. But the fact remains that while the SBS looks cleaner and has a higher preference score, the Manta is clearly superior from a subjective listening perspective. Don't get me wrong, the SBS.1 sounds great too, but it's not the same.

With regards to rounded/chamfered edges my experience is that the effect of this has been greatly exaggerated. As an example to support it with other than my opinion; Look at for instance the clearly engineer driven 20,000€ Kef Reference 5. A great speaker in most respects. Not very rounded edges.

I'm not sure if I agree that the effect has been greatly exaggerated, I guess it depends on if we're talking about measurable effects or audible ones. I've not been a big believer in the "clearly audible" camp when it comes to these things. Could be, I don't know.
In terms of measurements, my DIY JBL M2's have sharp edges instead of the chamfered edge on the original and at far off-axis it's clearly measurable diffraction artifacts compared the original.
KEF have the "shadow flare" to fight the diffraction effect. It seems to work well.

To clarify: Are you implying that the rising tendency from 150-400hz is undesirable?

Yessir! In my opinion this is the area that separates normal sound from truly spectacular sound. Getting the bass/lower mids right is the greatest struggle, and that's the area where clever design of speakers, sane placement of arse, shoutboxes and effective acoustical treatment have the greatest effect.
It's where line sources, CBTs, infinity baffles, cardioide and horns have most of their advantages imo.

Note that I'm not arguing to tune the speaker in such a way as to make the measured in-room response look better, that will sound horrible! I'm simply implying that aiming for good in-room performance as well as anechoic performance is worth the effort.
I mentioned a woofer behind the speaker just because I figured the placement close to the front wall and floor would perhaps be a worthwhile trade-off.
If the Hypex amp can do FIR filters, it could even be used for cardioide in the higher range where the cancellation with the front woofer could be a problem without changing the timing/phase. This is how Kii does it.

Just throwing out ideas until somebody ignores me.

In Mantas defense I'd also like to point out that everything is relative, and also that the graphs at spinorama.org is pretty squished up, so they don't do anyone any favors.

There's not actually a lot of "wiggles" to fix. On-axis is +/-2.5dB, and if you give the graphs a bit more space, it doesn't look half bad off-axis either, including the famed 1-2khz area. (0-30-60 from the Klippel measurements below):

First of all, the Manta is truly fascinating! It's a balls to the walls attack on everything that sucks with regular speakers and how they work in the room.
Separated bassboxes, high capacity, cardioide, coax mid/tweet and if I'm not mistaken, with phase linear crossover filter?

Compared to the constant directivity design of JBL M2 your speaker looks exceptional on that graph. Well done!

Secondly, in terms of diffraction and wiggles, it's really hard to tell the effect of it without testing. I would guess the difference between 30 degree and 60 degree plot could be diffraction artifacts. At least in my own speakers it's beyond 45 degrees it really starts to show, and there it's around 8 khz it really starts to peak. How that would look in a complete Klippel NFS I wouldn't know. I have been thinking of doing outside measurements for a few years to find out how bad it really is compared to the original. But I'm afraid I won't like the speakers no more

 

[sigbergaudio]

Yessir! In my opinion this is the area that separates normal sound from truly spectacular sound. Getting the bass/lower mids right is the greatest struggle, and that's the area where clever design of speakers, sane placement of arse, shoutboxes and effective acoustical treatment have the greatest effect.
It's where line sources, CBTs, infinity baffles, cardioide and horns have most of their advantages imo.

Note that I'm not arguing to tune the speaker in such a way as to make the measured in-room response look better, that will sound horrible! I'm simply implying that aiming for good in-room performance as well as anechoic performance is worth the effort.

In isolation I agree with everything here, but in context I disagree, since I am of the opinion that many speakers out there have to little energy in the 100-300hz area in-room. In my experience many rooms drain energy from this frequency range. All Sigberg Audio speakers have as slight, intentional lift in the ~100-500hz area, the alternative does not sound right.

So yes, this is the area where you can bring out spectacular sound, but not by holding back.

 

[Curvature]

The computed "in-room" response is just an approximation and average based on the loudspeakers, rooms, acoustics, placements, listening distances and measurement window width that Harman researches used and does in reality vary on all those factors so for me I am rather surprised when it matches well some other "random" real life experiments.

It matches really well above the transition region where the speaker dominates. Less well as you go down.

And it really is just a computation based on the anechoic data. It doesn't take into account all else you listed.

 

[sigbergaudio]

First of all, the Manta is truly fascinating! It's a balls to the walls attack on everything that sucks with regular speakers and how they work in the room.
Separated bassboxes, high capacity, cardioide, coax mid/tweet and if I'm not mistaken, with phase linear crossover filter?

Compared to the constant directivity design of JBL M2 your speaker looks exceptional on that graph. Well done!

Secondly, in terms of diffraction and wiggles, it's really hard to tell the effect of it without testing. I would guess the difference between 30 degree and 60 degree plot could be diffraction artifacts. At least in my own speakers it's beyond 45 degrees it really starts to show, and there it's around 8 khz it really starts to peak. How that would look in a complete Klippel NFS I wouldn't know. I have been thinking of doing outside measurements for a few years to find out how bad it really is compared to the original. But I'm afraid I won't like the speakers no more

View attachment 332090

Funny story, I was looking at the wrong data set, what I posted was actually the SBS. On the bright side I got the opportunity to show how well that speaker measures as well.

Here is the 0-30-60 of the Manta:

 

[Absolute]

In isolation I agree with everything here, but in context I disagree, since I am of the opinion that many speakers out there have to little energy in the 100-300hz area in-room. In my experience many rooms drain energy from this frequency range. All Sigberg Audio speakers have as slight, intentional lift in the ~100-500hz area, the alternative does not sound right.

So yes, this is the area where you can bring out spectacular sound, but not by holding back.

Oh, I agree completely that most speakers have too little energy there. What I meant was that using in-room response to decide speaker tuning is a gamble at best.
Resonances in that area have shown to be readily audible and detrimental in blind testing according to Sean Olive's interview with Erin. In my own experiments I've found that a small shelf in that region is beneficial. It looks to me like you, KEF and some others manufacturers have found the same.

Either way, the point I was making is that clever engineering to mitigate the problem is better than compensating for the problem because it will be far more universally true across all rooms. How you attack these things is a question about what you aim to achieve in terms of performance, cost, profitability, branding, uniqueness etc.
Those are things we as bystanders cannot tell you.
I like Duke's point about branding and doing what you preach, but like you I'm also pragmatic at heart. There's a bunch of different concerns you as a company need to balance, things I as a consumer care nothing about. Your time, investments and effort is completely free to me

Funny story, I was looking at the wrong data set, what I posted was actually the SBS. On the bright side I got the opportunity to show how well that speaker measures as well.

Here is the 0-30-60 of the Manta:

Thank you for clearing that up. That little bump in the 1-2 khz region off-axis reminded me of something I looked at a few years ago, namely the tuning JBL and Revel aim for.
Here's a bunch of their speakers that demonstrates this, the sum of early reflections;

So I had to take a look at the Manta's estimated in-room response compared to some of the Revels. It's eerily similar to a number of them, actually. The difference mostly lies below 1 kHz as I would expect since your speaker will give less reflections and more direct sound due to cardioide. Actual in-room response would look much better with the Manta's below 1-2 kHz.

Take a look at just one example;

Easily compare loudspeakers measurements

www.spinorama.org

For me this is a sign that all your listening, tweaking and going through the trouble of measuring east and west over years leads to an excellent place.
It's like top of the line Revels only with less room-havoc and the benefit of a coax driver.

Consider me convinced about your methods.

 

[Absolute]

This new speaker, how about giving the name some norwegian swung? Like SigbergAudio Stilk?
Or are you looking to keep it related in animal spirit to the Manta?

 

[sigbergaudio]

Either way, the point I was making is that clever engineering to mitigate the problem is better than compensating for the problem because it will be far more universally true across all rooms. How you attack these things is a question about what you aim to achieve in terms of performance, cost, profitability, branding, uniqueness etc.
Those are things we as bystanders cannot tell you.
I like Duke's point about branding and doing what you preach, but like you I'm also pragmatic at heart. There's a bunch of different concerns you as a company need to balance, things I as a consumer care nothing about. Your time, investments and effort is completely free to me

Yes, and I consider all feedback in these threads to be opinions of an individual, as that is what they are. Some are completely ignored, others are beneficial to help further my own reasoning - and for that I am grateful. In addition it is interesting and fun to share the journey.

(...)

For me this is a sign that all your listening, tweaking and going through the trouble of measuring east and west over years leads to an excellent place.
It's like top of the line Revels only with less room-havoc and the benefit of a coax driver.

Consider me convinced about your methods.

Interesting comparisons. "It's like top of the line Revels only with less room-havoc and the benefit of a coax driver." <- I can live with that.

 

[sigbergaudio]

This new speaker, how about giving the name some norwegian swung? Like SigbergAudio Stilk?
Or are you looking to keep it related in animal spirit to the Manta?

I spend far too much time thinking about that. And yes, I'm actually trying to find a Norwegian word / use Norwegian spelling as I did with the Inkognito subwoofer (admittedly the word is almost identical in English). But it's pretty hard.

It's tempting to try to find a word connected to music or sound somehow, but strange as it may sound, it appears I am not the first person in history who wanted to do that, so several of my ideas have turned out to be names that have already been used on another loudspeaker (sometimes several ). It will probably be easier to use a name that is not directly connected, like the Manta.

Personally I have a great passion for music, and find it immensly satisfying to build great speakers that bring out emotions in others as they listen to their favorite music - so it would be great to have a name that brings out some kind of feeling or emotion in people as well. But since we're all different with different experiences, that is hard too.

With regards to Manta and animal kingdom, I'm not sure if I ever told this story in the Manta thread, but there's a duality in that name. One thing is the Manta ray and how the ports in the upper chamber may resemble the form of a Manta. But the main origin of the name is actually a car, the Opel Manta. And how on earth is that connected to anything? My late uncle was who originally got me into hifi back in the 80s, as he was a DIY speaker builder. He was also stuck in a wheel chair, but being a kid I didn't really think about that as anything other than objective fact, to me he was just the coolest guy ever. He built his own speakers, and he also had what in the 80s was a super cool car: A neon green Opel Manta.

So the Manta is a tribute to him and how he got me into all this, some 35 years ago - and it makes me cry just writing about it. I guess in part because it also surfaces feelings about how damn hard this it is, but also because it's worth it. He died many years ago, but I miss him dearly.

 

[thewas]

Thank you for clearing that up. That little bump in the 1-2 khz region off-axis reminded me of something I looked at a few years ago, namely the tuning JBL and Revel aim for.

Also most classic style STAX headphones have a little bump around 1.5 kHz which gives makes them sound more direct/fresh, exemplary:

Dropbox

www.dropbox.com

Edifier who owns now STAX even implemented a "STAX mode" preset on their current BT headphones Edifier STAX Spirit S3 with such.

 

[Absolute]

@sigbergaudio Wow. Now, that's a story worth telling. I would consider adding this backstory on your website, pictures and all. I think it's important to pay homage to the people who inspired us.

I'm sorry for your loss.

 

[Colonel7]

I spend far too much time thinking about that. And yes, I'm actually trying to find a Norwegian word / use Norwegian spelling as I did with the Inkognito subwoofer (admittedly the word is almost identical in English). But it's pretty hard.

It's tempting to try to find a word connected to music or sound somehow, but strange as it may sound, it appears I am not the first person in history who wanted to do that, so several of my ideas have turned out to be names that have already been used on another loudspeaker (sometimes several ). It will probably be easier to use a name that is not directly connected, like the Manta.

Personally I have a great passion for music, and find it immensly satisfying to build great speakers that bring out emotions in others as they listen to their favorite music - so it would be great to have a name that brings out some kind of feeling or emotion in people as well. But since we're all different with different experiences, that is hard too.

With regards to Manta and animal kingdom, I'm not sure if I ever told this story in the Manta thread, but there's a duality in that name. One thing is the Manta ray and how the ports in the upper chamber may resemble the form of a Manta. But the main origin of the name is actually a car, the Opel Manta. And how on earth is that connected to anything? My late uncle was who originally got me into hifi back in the 80s, as he was a DIY speaker builder. He was also stuck in a wheel chair, but being a kid I didn't really think about that as anything other than objective fact, to me he was just the coolest guy ever. He built his own speakers, and he also had what in the 80s was a super cool car: A neon green Opel Manta.

So the Manta is a tribute to him and how he got me into all this, some 35 years ago - and it makes me cry just writing about it. I guess in part because it also surfaces feelings about how damn hard this it is, but also because it's worth it. He died many years ago, but I miss him dearly.

Thanks for sharing your inspiration. He was a great influence. Sounds like your next speaker should be called the Revenant

 

[Colonel7]

Back to the cabinet looks. I agree with Keith that the veneer band on the baffle like the Manta would make it stand out and add a touch of class. Otherwise the plain black looks utilitarian or home theater hide it-ish. I personally like the veneer on the full sides and black or white baffle too.

 

[sigbergaudio]

Hm.

 

[Colonel7]

Yes, I know it was your original feedback request and it is classic!. Wanted to chime in bc there was some criticism of it but I like it!

 

[sigbergaudio]

Yes, I know it was your original feedback request and it is classic!. Wanted to chime in bc there was some criticism of it but I like it!

Notice any other differences from previous illustrations?

 

[boXem]

Notice any other differences from previous illustrations?

s
You adapted the speaker to the age of the average high end hifi customer and suppressed the tweeter. Makes sense.
/s

 

[sigbergaudio]

s
You adapted the speaker to the age of the average high end hifi customer and suppressed the tweeter. Makes sense.
/s

 

[sigbergaudio]

Since this is ASR and this is a development thread, I guess we can't just talk about names and show pictures, so here's a few more (very) preliminary measurements.

Current in-room response in a pretty normal European living room, around 450ms decay in the room. I must say it's surpisingly tidy, not least at 100-300hz.

A measurement of the 20-200hz area measured in-room @1m (1/12 smoothing) just to give an impression of the bass response/performance. That's what you get when you build a Sigberg Audio subwoofer into a floorstander.

0/30/60/90@40cm, 1/6 smoothing. 300hz-20,000hz. Pretty neat.

~listening axis / 15deg@40cm, 300hz-20,000hz - gotta love this coax. Will increase top end a bit.

Going back to the room, here's an average of the right speaker from three different seats. Naturally there's variance below 100hz, but pretty consistent otherwise.

 

[SeanAtx]

Wow @sigbergaudio you don't sleep, do you? Your passion for audio is inspiring, thanks for sharing a bit of history about the Manta. I'm almost tempted to book a flight to London to stop by Purite Audio just to hear your speakers! They're a ways outside of my budget for now, but a guy can dream

 

[Mnyb]

I guess it would be interesting to build a version of the Manta without all the holes in it, so a completely traditional enclosure, and measure it in an anechoic chamber, but I never did that. And even the initial prototype had cardiod slots in the upper chamber so I don't even have rudimentary measurements of such an enclosure. But my theory is that the cardioid system makes for a somewhat more chaotic off-axis response. At least the way I ended up doing this. The Dutch&Dutch on the other hand is pretty clean.

The SBS.1 is a sealed speaker with a quite narrow baffle, so it's much easier to make "clean" from a measurement standpoint. But the fact remains that while the SBS looks cleaner and has a higher preference score, the Manta is clearly superior from a subjective listening perspective. Don't get me wrong, the SBS.1 sounds great too, but it's not the same.

With regards to rounded/chamfered edges my experience is that the effect of this has been greatly exaggerated. As an example to support it with other than my opinion; Look at for instance the clearly engineer driven 20,000€ Kef Reference 5. A great speaker in most respects. Not very rounded edges.



To clarify: Are you implying that the rising tendency from 150-400hz is undesirable?

With regards to alternative woofer placement;
I suspect it won't make a dramatic difference (it may reduce SBIR effects somewhat), but I'm contemplating moving one of the woofers down.

View attachment 332063

This look with wood sides and spaced woofers are not that unusual I like it you already have the slots and coax but i’m Weird . Maybe you need grills/cloth in the front ? Weirdos prefer the drivers visible but evidently it’s “ugly” according to some ?