Sigberg Audio MANTA dual cardioid active speaker development thread

[Sigberg Audio]

I can't argue with what you say, but the wall behind the speaker is obviously part of the room. However if you are going to design a speaker to be used near a boundary, then it requires adjustments in design to minimize it. (..)
One of the huge advantages of DIY, is to be able to customize designs to the architectural considerations.

So yes, I believe it is a room effect, but still one that needs correction if the intent is to have an against wall design.

As you point out above (my bold / emphasis), you can't expect a commercial speaker to perfectly mitigate any room problems in any room. Not trying to be defensive here, it's just simply not possible. That's also not what we claim to do. We claim this design will reduce SBIR effects compared to a traditional speaker, which it will. Look at any "share your frequency response" thread on this forum or elsewhere, and as I'm sure you know you will see the 100-300hz range typically look wildly uneven and with both more and wider dips than this response.

How this looks in any given room is not only dependent on the speaker and the distance between speaker and wall, but also where the listener is situated.

Here is the same speaker in a different room (still no smoothing). As you can see the 100-200hz range is still what I'd say is significantly better than what I typically see, but completely different than the room I posted earlier. This time with only one dip, but a wider and more shallow one.

 

[Sigberg Audio]

Here are two graphs from a different speaker (not the Mantas). Both graphs are of the same speakers in the same room, but taken from different listening distances. Here we see a more typical and pretty wide dip between 110-140hz. We can also see that there are similar trends in both graphs, but they're not identical. What's also interesting is that we see a dip similar to the one you mentioned at 180hz in one of the graphs, but it's completely gone in the other. The speaker is the same and the speaker distance to the wall is the same, the only thing that moved is the listener.

Just another example to show how it's not really realistic to expect perfect response out of the box from a commercial design when whoever designed it (whether it's me or someone else) has zero control over the situation where the speaker is used by the person who bought it.

 

[TLS Guy]

Here are two graphs from a different speaker (not the Mantas). Both graphs are of the same speakers in the same room, but taken from different listening distances. Here we see a more typical and pretty wide dip between 110-140hz. We can also see that there are similar trends in both graphs, but they're not identical. What's also interesting is that we see a dip similar to the one you mentioned at 180hz in one of the graphs, but it's completely gone in the other. The speaker is the same and the speaker distance to the wall is the same, the only thing that moved is the listener.

Just another example to show how it's not really realistic to expect perfect response out of the box from a commercial design when whoever designed it (whether it's me or someone else) has zero control over the situation where the speaker is used by the person who bought it.

View attachment 58084

I completely understand that. That is why I'm a committed DIY guy. I do think though, due to the frequencies of those nulls, that the likely cause is from the rear wall, cardioid not withstanding. I really applaud you efforts to brake new ground and move the state of the art forward. That is part of the reason my main speakers are designed the way they are, so that reflections are all over the place with no frequency dominant.

This is the FR of one of my mains, obtained in my old room at Benedict Lake. Blue is axis, and the black is out close to 90 degrees. Distance 2 meters.

This is the axis FR in the new room here at Eagan, showing that 50 Hz peak was a room effect.

The droop below 50 Hz is purposely drooped in this location to curb room gain.

This is the center channel in the old room on and off axis. The blue is axis, and as is common with coaxials, the axis trace is usually the worst, because the cone reflections from the tweeter are symmetrical.

This is the axis response in the new room.

I had to significantly change the BSC signal for the change in venue.

This is the room response at the MLP all speakers driven, and you can see why I have that taper in the last octave of the mains.

The impulse response shows that all speakers are timed to the MLP. At the end, you have to determine the lower bass contour by ear to a large extent. The system is never boomy, and it is not too warm and not too dry.

Yes, you are correct that in a DIY situation, you absolutely do have luxuries not afforded to the commercial designer. Although being able to control the contour and level of the BSC signal, is a big advantage I believe. I described that first, and I'm pretty sure JBL pinched it, for their flagship speaker. I don't mind, and if you want to, please feel free to have at it. I believe that to be the dominant determinant of being able to voice a speaker to the room.

In closing, I really applaud your decision to publish all of this. I knew the late Peter Walker of Quad very well, over many years. Like you, he was an absolute open book about his designs and had no secrets. That way he said that you got a lot of help with your designs for free!

 

[Sigberg Audio]

@TLS Guy We may be misunderstanding each other here. I don't challenge the suggestion that the dips are from the back wall, I challenge the idea that something can be done with it (without causing something else instead). As mentioned it's not just the speaker and the walls that are part of the system, also the listener. This measurement is specific to my listening room and will look different if I move the microphone closer or further away from the speakers. It wil even change if I move it closer to the floor or closer to the ceiling. And if I do nearfield measurement (0.5-1m) there will be no dips at all.

As an end user of our products you will not have individual control of the drivers, but you will have access to a 9-band EQ that includes both boost/cut and shelf filters. You have access to this for each individual speaker and for each individual subwoofer. If you are person who have access to and competence to use measurements gear, you are also free to move your speakers around (including moving them further or closer to the wall) to maximise the effect of the cardioid pattern in your specific room. Or of course do this by ear if measurements are not available. If you in addition to that have the luxury of moving your listening position around, there should be a pretty significant chance of getting excellent sound. Most do not, but they will still benefit from the cardioid system and get better and more even midbass response than with most speakers.

In summary It's unclear to me specifically what you think should be done differently, but I'd be happy for you to share that feedback.

Also again please note that the response shared was with zero smoothing, while you present graphs with significant smoothing.

Here's the same graph with 1/6th smoothing:

 

[TLS Guy]

@TLS Guy We may be misunderstanding each other here. I don't challenge the suggestion that the dips are from the back wall, I challenge the idea that something can be done with it (without causing something else instead). As mentioned it's not just the speaker and the walls that are part of the system, also the listener. This measurement is specific to my listening room and will look different if I move the microphone closer or further away from the speakers. It wil even change if I move it closer to the floor or closer to the ceiling. And if I do nearfield measurement (0.5-1m) there will be no dips at all.

As an end user of our products you will not have individual control of the drivers, but you will have access to a 9-band EQ that includes both boost/cut and shelf filters. You have access to this for each individual speaker and for each individual subwoofer. If you are person who have access to and competence to use measurements gear, you are also free to move your speakers around (including moving them further or closer to the wall) to maximise the effect of the cardioid pattern in your specific room. Or of course do this by ear if measurements are not available. If you in addition to that have the luxury of moving your listening position around, there should be a pretty significant chance of getting excellent sound. Most do not, but they will still benefit from the cardioid system and get better and more even midbass response than with most speakers.

In summary It's unclear to me specifically what you think should be done differently, but I'd be happy for you to share that feedback.

Also again please note that the response shared was with zero smoothing, while you present graphs with significant smoothing.

Here's the same graph with 1/6th smoothing:
View attachment 58085

I'm not sure what should be done differently. The issue is that boundaries are a problem for speakers, especially when they are close to them, or actually in a boundary. That just presents an added problem. That is not to say that I disagree with your plans to pursue more elegant and architecturally friendly designs for audio in the home. Quite the opposite in fact, I applaud them. I think actually this is a more common problem than realized, and leads to subs being overdriven and an unnatural balance.

However, that trace with the smoothing is actually more revealing then the previous ones. This is because it gives more information about the energy liberated into the room from a crucial power band. You can see there is an 8db. drop between 90 and 135 Hz. The smoothing reveals this to be not so narrow in scope. So it reveals a drop in output between 90 and 165 Hz, which is the best part of an octave. I personally think that will be audible, and make the bass foundation of an orchestra, or even a piano or string quartet weak in its foundation.

So to answer your question more specifically, I think the first step is to repeat these measurements in more rooms, and see how peculiar the problem is to your room.

If it is not particularly specific to your room, then I think the cause needs further investigation. I understand the system has an equalizer, however correcting that anomaly will add significant power to the bass driver. I understand that there is lots of power at hand, but adding that much power to the driver is of concern. This gets into the area of use and abuse of equalization.

That is all a round about way of saying the issue is worthy of further investigation, and if possible correction.

 

[Sigberg Audio]

@TLS Guy With all due respect you're getting this all wrong. The target curve here is not flat from 20-200hz. The dip between 100-160hz is mostly ~2dB below the target curve and ~5dB at most between a narrow range of 130-140hz. Yes this may be audible, but that's not the point.

We're discussing the response @ listening position in one room. This cannot be taken account when designing the speaker itself, it's cruicial to differentiate between tuning the speaker and tuning the room. I cannot account for individual room issues in one particular room when tuning the a speaker for commercial production. The point of sharing the graph was to show that the issues between 100-200hz in this particular room (and also in a different room that I shared response from) is less than what you typically see. Individual problems shown in these graphs (again taken at the listening position) cannot be used to evaluate the speaker or the commercial tuning of the speaker. I explicitly leave these types of problems be while testing and developing speakers to avoid falling into the trap of correcting for the room.

The proper way to evaluate if the cardioid effect is working properly and evenly is to measure this in an anechoic chamber, which we have done with the previous prototype and will with the new one as soon as that arrives. It has not yet been done with this prototype because we're waiting for the next, but since that has been delayed I'll consider doing that some time before Christmas.

 

[Sigberg Audio]

I realize sharing in-room responses in these types of threads often lead to misunderstandings, so I will try to do so sparingly and with detailed explanations of what we're looking at / what they're supposed to convey.

 

[TLS Guy]

@TLS Guy With all due respect you're getting this all wrong. The target curve here is not flat from 20-200hz. The dip between 100-160hz is mostly ~2dB below the target curve and ~5dB at most between a narrow range of 130-140hz. Yes this may be audible, but that's not the point.

View attachment 58088

We're discussing the response @ listening position in one room. This cannot be taken account when designing the speaker itself, it's cruicial to differentiate between tuning the speaker and tuning the room. I cannot account for individual room issues in one particular room when tuning the a speaker for commercial production. The point of sharing the graph was to show that the issues between 100-200hz in this particular room (and also in a different room that I shared response from) is less than what you typically see. Individual problems shown in these graphs (again taken at the listening position) cannot be used to evaluate the speaker or the commercial tuning of the speaker. I explicitly leave these types of problems be while testing and developing speakers to avoid falling into the trap of correcting for the room.

The proper way to evaluate if the cardioid effect is working properly and evenly is to measure this in an anechoic chamber, which we have done with the previous prototype and will with the new one as soon as that arrives. It has not yet been done with this prototype because we're waiting for the next, but since that has been delayed I'll consider doing that some time before Christmas.

Thanks for your reply. It begs the question though as to what was the rationale for that target curve? I raise this issue as being weak in this crucial power band is far from an uncommon vice among speakers. For classical music in particular, it is a crucial band, that when deficient really compromises orchestral sound. When I look at the power frequency distribution in WaveLab on my DAW, you can see how important it is to have "heft" in that region.

All I can say, is that having a target curve like that is contrary to my experience, and therefore my long held design goals for a speaker. So it might just be that we have different design intent, which is fine. To me though, in my room orchestral balance sounds exactly correct to me, under all circumstances. This is where tastes may differ.

In closing, I would say I have posted before in quite a number of threads about the importance of plenty of power in that band, and the deficient resources of many designs, that even cross to small mids in that region, which I regard as a disaster.

 

[Sigberg Audio]

Thanks for your reply. It begs the question though as to what was the rationale for that target curve? I raise this issue as being weak in this crucial power band is far from an uncommon vice among speakers. For classical music in particular, it is a crucial band, that when deficient really compromises orchestral sound. When I look at the power frequency distribution in WaveLab on my DAW, you can see how important it is to have "heft" in that region.

All I can say, is that having a target curve like that is contrary to my experience, and therefore my long held design goals for a speaker. So it might just be that we have different design intent, which is fine. To me though, in my room orchestral balance sounds exactly correct to me, under all circumstances. This is where tastes may differ.

In closing, I would say I have posted before in quite a number of threads about the importance of plenty of power in that band, and the deficient resources of many designs, that even cross to small mids in that region, which I regard as a disaster.

The correct / best target curve is indeed an interesting discussion, is there perhaps an existing thread on that topic here on Audioholics where we can continue that?

Note also again that it is important not to confuse my personal preference with "design goals" for the speaker. The in-room target curve is something every customer is free to set for themselves, it is not limited by the design of this speaker, on the contrary as mentioned before there are powerful built-in tools allowing you to do this according to your personal preference.

 

[TLS Guy]

The correct / best target curve is indeed an interesting discussion, is there perhaps an existing thread on that topic here on Audioholics where we can continue that?

Note also again that it is important not to confuse my personal preference with "design goals" for the speaker. The in-room target curve is something every customer is free to set for themselves, it is not limited by the design of this speaker, on the contrary as mentioned before there are powerful built-in tools allowing you to do this according to your personal preference.

I don't recall a thread about target curves, but there might be one.

Personally I dislike the concept of target curves. I'm hard core, and believe there is truth (precision), and error.

So I should tell you a little about myself. For some odd reason I became obsessed with sound reproduction from my earliest years. I had an obsession with it. I built my first speaker age 7. So I have seen very nearly 70 years of development.

So for most of that time, the home constructor had to use hand math calculations and fine tune by ear. 1984 was a turning point for me, when I obtained the floppy discs of Bullock and White's design software. I designed my first computer assisted design speakers. Those speakers are now my surrounds, having been my recording location monitors for many years.

As you know things progressed, and we now have measuring microphones and associated software.

In the late eighties Stereophile had a competitive "Audio sound off." I designed and submitted a pair of these speakers.

Much to my surprise they won first prize. The winner received assistance with production. So I organized a production run and had custom packaging made.

The small first production run sold out promptly. When I went to start the next run, I found out that Focal had stopped production of the bass/mid drivers.

That cured me of going into the loudspeaker production business for good. I sincerely hope you fare a lot better than I did!

Although that winner was a ported design, I have mainly concentrated on the development of transmission line speakers. The reason is that I know of no other alignment that produces such accurate and realistic bass with very uniform dispersion throughout the room, and with great efficiency with excellent control of driver excursion.

I think one of the issues a speaker designer has to really understand is where to put and distribute your power resources. I can not understand why this is not discussed more often, as it is crucial to making a robust design.

The starting point has to be the distribution of the fundamentals of musical instruments. Of course the fundamentals demand the largest portion of the power as the higher fundamentals demand progressively less of the power.

As you can see a huge number of instruments have fundamentals in that 80 to 400 Hz range.

Now this is in the area where with narrow cabinets BSC needs to be applied. So this greatly increases power demands in that, what I call, the foundational region.

This is not so much a problem for amplifiers, but loudspeaker drivers, especially those with drivers under 7" and some 8". These days there are a lot of speakers in that category.

Shortcomings in this region are NOT shown by speaker measurements we usually have for review, subs excepted. In my view the need for this is much greater in the other speakers. In my view few drivers produced today, have motor systems up to the task, which is why I have a very short list of acceptable drivers. It also requires more than one driver as a rule, and that extends into the midrange. The result if the drivers are not up to it, is serious dynamic compression and increased distortion.
If I am again suggesting there is far too much attention paid to subs versus the three to four octaves above you are absolutely right. When I do waste dealers time and hear other systems, I find that the sub is too loud to try and cover the above shortcomings, This is NOT accurate reproduction and the overall sound is not balanced.

Now before software help was available I decided that I would invest in accurate recording equipment and first class microphones. I also made the decision that is still unusual, to make phase coherent intensity difference recordings, rather then the ubiquitous phase difference recordings more often than not sprouting microphones growing like mushrooms.

Most of my recordings were made of live recordings for public radio station KFJM Grand Fork ND, and the the Dakota public broadcasting system. In the days prior to software and computer aids this was an excellent tool for loud speaker evaluation. Prior to 1984 recordings were made with a Revox A700 and a bespoke Brenell Mk VI with parabolic heads without pressure pads. Tape was Ampex Grand Master running at 15 ips stereo on 1/4 tape with dbx 1 noise reduction. The microphone was a Neumann SM69 FET usually set up in M-S mode, most often, or figure of 8 mode. Spot mics were used sparingly, if at all, and only cracked open. These recordings were extremely useful for loudspeaker evaluation, for multiple reasons.

Here is part of my last analog recording with the Revox and Brenell machines on location. These machines are still in my system. After this a changed to a VHS based digital system.

I have YouTube video on my channel of parts of this recording with explanations. Thorbjorn, you may be interested to know this video contains a work by a much loved Norwegian Immigrant to Minnesota, F. Melius Christiansen. The work is O Beautiful Savior. This work is much loved in Minnesota and the Dakotas.

This recoding was made entirely with the SM 69 FET in M-S mode. The mixer was by Tapco/Electrovoice. I show the video of my DAW. In the lower right you can see the band by frequency power band intensity. Now remember this is only a choir and smallish Baroque orchestra, and large forces are even more weighted to that 80 to 500 and even above power band. The below 80 band is trivial by comparison. However one can not dismiss that power band as it is essential to realistic reproduction. For reasons that are still not clear, it has long been observed that the 20 to 80 Hz band contributes immensely to giving a recording a sense of space and ambience. The reason for this phenomenon has never been entirely clear. The recordings I made were always monitored by speakers, and did as I hoped, give me enormous insights as to what was really required of speakers.

Here is the video of which I speak.

The first 11 minutes or so, are my verbiage, which you are welcome, and may be encouraged to skip. If you watch this carefully, it will give you significant insights as to what is required of loudspeakers, at least excellent ones.

As usual there are caveats. I do not listen to music of the pop culture as a rule. However local engineers who work in that genre really like this system and have used it as a reference.

This system is used for movies and they contain music from the popular culture often, and it reproduces that well and with excellent intelligible speech quality.

So to cut to the chase. What is my opinion of Target Curves. I say quite unequivocally that they are nonsense. You don't need a curve, you need a flat straight line and the flatter the better. However there is some license on the last octave or so 20 Hz to 40 to 50 Hz, in the area of room gain. To get extremely "hair shirt" in this region can make reproduction excessively dry. So a small amount of leeway in this area is required. I think this is actually best judged by ear, on a room to room basis.
I absolutely do not agree with the hot sub and attenuation above sub range. That in my view has come about by pandering to speakers with inadequate power response on the two octaves or so above sub range. I will state this again for the record that this defect likely affects the majority of speakers, with the resulting lack of balance at power, limiting realistic reproduction. This became clear to me as a result of the invaluable experience of making live recordings.

So this has been a long winded way of saying we need more straight lines and less curves.

 

[ryanosaur]

we need more straight lines and less curves.

*blink, blink...

I think there is a joke in there somewhere.

 

[Sigberg Audio]

So this has been a long winded way of saying we need more straight lines and less curves.

I didn't really want this discussion in the Manta thread as it's not necessarily interesting or relevant to whomever just want to follow the development of the speakers. So I started a new thread, let's please continue the discussion there.

Link to new thread: https://forums.audioholics.com/forums/threads/what-does-an-accurate-in-room-response-look-like.126024/

 

[Sigberg Audio]

A quick progress update:

• New revision / prototype of the stands is in progress and will probable be done before xmas.
• Preproduction cabinets are expected in January (produced by the company who will make the production cabinets)

So if there are no big misunderstandings on either of the above, we're on track for launch summer 2023, hopefully accepting preorders before Easter next year.

 

[Sigberg Audio]

We had a long thread on what makes a good speaker for rock on a Norwegian audio forum, which I find very interesting since I think it's a difficult genre to get right, and many speakers fall short. I spend a lot of time making ours really master it, and the relatively "simple" formula is a balanced top end (a slight dip between 2-4khz may be favorable for some recordings), a slight focus on the upper bass / lower midrange, as well as plenty of capacity across the entire frequency range.

Sharing a few tracks that sound spectacular on the Manta.

 

[Sigberg Audio]

 

[TLS Guy]

That sounds very much like the old "BBC Smiley". That was very much in vogue forty or more years ago now, and persisted in many UK speakers, actually in some up until the present day. I must confess to using it in years past. The last speakers I built that had that smiley was the location monitors I built in 1984, now my surrounds.

They were designed with the old Bullock and White software on floppy discs on an Apple II E. The program slightly underestimated the bass ripple from that sealed design as you can see. The "Smiley" does create a not unpleasant recessed sound stage, that is not unpleasant. I think this all came about as measuring techniques and modelling were not as good as they are now. What I think really happened is that it avoided even the slightest rise in FR in that region, which is absolutely lethal to good sound. That is a pass band where the ear is totally intolerant of any rise.

I think it is virtually impossible to judge the merits of a speaker with rock music, as you have no point of reference, the tonal sound scape is set largely electronically and so no one has the least clue what it is supposed to sound like it. I would absolutely NEVER use it to assess the tonal accuracy of a loudspeaker. This is all aside from the fact that I can't abide any of it.

I personally believe that it is the classical audio listeners that support the high end speaker market. This is all for the reasons above and the fact that even the most aberrant speakers seem to be able to satisfy the rock aficionados.

This is obviously not such a huge problem for you, as I would suppose a purchaser could easily cancel the aberration.

 

[Sigberg Audio]

I think it is virtually impossible to judge the merits of a speaker with rock music, as you have no point of reference, the tonal sound scape is set largely electronically and so no one has the least clue what it is supposed to sound like it. I would absolutely NEVER use it to assess the tonal accuracy of a loudspeaker. This is all aside from the fact that I can't abide any of it.

Completely disagree, how a speaker handles complex material like rock is a great way to judge the merits of a speaker. And most struggle with this test.

 

[TLS Guy]

Completely disagree, how a speaker handles complex material like rock is a great way to judge the merits of a speaker. And most struggle with this test.

Well, that I fail to understand as there is a zero reference point. A studio recording is totally different from a live concert, and then if it is a live concert the recording mix will be nothing like the PA mix. I know what the Minnesota Orchestra sounds like in Orchestra Hall, but I have no idea what a rock row is supposed to sound like. I would have no clue as to whether a speaker reproduced the racket faithfully or not, and I fail to se how anyone else could either.

I would never, and never have used rock music as any sort of reference point in any of my speaker designs.

 

[Sigberg Audio]

Well, that I fail to understand as there is a zero reference point. A studio recording is totally different from a live concert, and then if it is a live concert the recording mix will be nothing like the PA mix. I know what the Minnesota Orchestra sounds like in Orchestra Hall, but I have no idea what a rock row is supposed to sound like. I would have no clue as to whether a speaker reproduced the racket faithfully or not, and I fail to se how anyone else could either.

That we have zero reference point (as in we don't have a reference to an actual live event, and one that we actually heard no less, that represents what's on the track) is a fact for almost everything we listen to, regardless of genre. So in your opinion it's no point listening to studio albums in order to evaluate speakers.

I beg to differ.

I would never, and never have used rock music as any sort of reference point in any of my speaker designs.

Yes, you mentioned that.

 

[TLS Guy]

That we have zero reference point (as in we don't have a reference to an actual live event, and one that we actually heard no less, that represents what's on the track) is a fact for almost everything we listen to, regardless of genre. So in your opinion it's no point listening to studio albums in order to evaluate speakers.

I beg to differ.



Yes, you mentioned that.

I don't use studio recordings, and certainly not those with electronic instruments. I use recordings made in halls, many of them live recordings. The BPO now have live loss less streams. I use recordings made in churches and many locations. I know the Albert Hall well and others. Only a collection of my chamber music and items like that were made in the studio, and not all of them. In any event they are the smaller ensembles like the Beaux Arts Trio for instance, and a studio is a good intimate space in any event, with good mics, and I doubt any Eq. was used. I never used Eq. in my recordings.

So I have hundreds of recordings that I know what they should sound like including my own recordings. Actually I got into doing recordings, for the evaluation of loudspeakers. I could not possibly evaluate a speaker listening to a rock band, mixed with all sort of effects units thrown in for good measure.

The market for high end expensive accurate speakers has always been from the classical music fraternity, and I suspect always will be. That part of the market has been the driver for accurate audio from the inception of Hi-Fi.