The low mids are plagued by boundary interference as well 
I have no idea what kind of system would be more preferable than others, as small as possible or deep or wide.
Also reflex port and panel resonanes happen somewhere on the lower mids (referring to augerpro test thread). Many small aspects to optimize on.
Might be so that it all averages out nicely in the hearing system. There probably are some papers available that study the phenomenon.
I have no idea what kind of system would be more preferable than others, as small as possible or deep or wide. Also reflex port and panel resonanes happen somewhere on the lower mids (referring to augerpro test thread). Many small aspects to optimize on.
Might be so that it all averages out nicely in the hearing system. There probably are some papers available that study the phenomenon.
The low mids are plagued by boundary interference as well
Also reflex port and panel resonanes happen somewhere on the lower mids (referring to augerpro test thread). Many small aspects to optimize on.
Might be so that it all averages out nicely in the hearing system. There probably are some papers available that study the phenomenon.
Regarding this specific aspect of port pipe resonances, I have attached a picture of the woofer and port nearfield responses I saw in one of my previous attempts to build a 2 way with an Sb17MFC35-8 woofer.
I don't know if I have ever heard that pipe resonance while playing music (since I don't know how it will sound like specifically). But looking at that graph of port response was terrifying enough for me to try sealed alignment for the woofer in this project
(even though I wanted to really try out a good bass reflex design in this one).
Attachments
Yes Vituix models the baffle step which is the dominant feature. When you measure or simulate 360 degrees around the speaker, there is a second step or transition depending on the geometry which has more directivity than the baffle step alone would predict.
That is not so easy to answer. If the width and depth are quite similar you tend to get reinforcement and cancellation from the dimensions occurring at similar frequencies. This is why a cube is often considered a bad enclosure shape due to all sides being equal.
In this case the woofer could be modelled in a more simple rectangular box with the width and depth varied to see the effect. This would take some time but it is interesting enough to try.
If you retain the dimensions as they are you will get something like the response predicted. A flat disc is a better representation at higher frequencies and adding the depth of the cone is more accurate at low frequencies, but if too deep ABEC runs into issues with some of the basic assumptions made in the simulation and it becomes less accurate.
This is in the transition zone and whatever your speaker measures like in free space will not be what it measures like in room. The transition zone is the hardest to fix with equalization. So for me it seems worth the effort to try and minimize the issues in this range.The other question I have is, audibly, what could be kind of impact
of this kind of peaks in the response since it looks like the kind of frequencies below which the room takes over. Also I have heard people saying that the 200-400 Hz frequencies are very important perceivably in setting audible levels for the rest of the spectrum etc. I don't know much about any of these yet.
The level of treatment and position of the speakers and listening point that can be used will determine how much effort would be worth putting into the acoustic design to minimize problems created from difficult placement.
In a three way speaker it should be much easier to avoid midrange leakage by putting the port and damping in the right place, crossing the woofer below it and sealing the mid in it's own enclosure.I don't know if I have ever heard that pipe resonance while playing music (since I don't know how it will sound like specifically). But looking at that graph of port response was terrifying enough for me to try sealed alignment for the woofer in this project
(even though I wanted to really try out a good bass reflex design in this one).
Thanks a lot fluid.
This is a lot of information I didn't know much about, earlier. I really happy that I am learning a lot. Regarding design of the woofer enclosure, I am not fixed on this design yet. I am very interested in making the design better, if it is doable. I can try modifying current design more in VituixCAD, but I am not sure how much more I will progress with that given that the interaction of the depth and overall box geometry may not be modeled in it. I can try playing around with making differences in baffle width and depth dimesions to make them more dissimilar.
I will also try more in the direction of the bass-reflex design by reading more about the work that augerpro has done and see if I can accommodate it in my project. If you have any more suggestions/papers/books/anything that you can point me to regarding enclosure geometry and its effects on directivity in the lower frequencies, please let me know. I can pursue learning more in this direction.
Even from a physical construction of the enclosure point of view, I think it would be easier if the midrange and woofer are in a separate boxes compared to what I currently have, as you suggested. So I like this idea and can pursue modifying the current box more in this direction.
I will wait for your simulations and suggestions about the woofer enclosure part and the remaining part of the overall design like the tweeter and mid interaction to see if changes need to be done in that area also..
Thinking about all that I have learnt so far, now I cant un see the kind of peaks around the 200-500 Hz and the smaller ones higher in frequencies in the directivity plots, once I look at polar measurements available online for many speakers.
Hence I am just pondering over some design aspects from whatever little I know so far. Just thinking about how diffraction hates less symmetry, smooth surfaces etc, Will it help if we create a woofer module where the woofer is asymmetrically mounted on the baffle (in vertical direction-i.e., woofer location shifted more down towards the bottom of the cabinet or more towards the top in the woofer module cabinet, with a separate smaller cabinet for mids and tweeter kept on top of the woofer module). This would be unlike the woofer location in my current design where it is almost at the center of the cabinet in vertical direction.
The other thought I have is regarding what would be the effect of a generous rounding of the overall cabinet itself (again thinking like diffraction doesn't like rounding much).
I couldn't help myself notice something like these aspects in these speakers (Even though these might be way beyond reach DIY options available to me).
Atria II - Rockport Technologies
Deep cabinets which look the least like a rectangular box. At least it looks aesthetically very pleasing to me.
I will try to create a woofer box with generous rounding but like a simplified version of above in fusion 360. But given my poor fusion360 skills, it might take a bit of time..
Hence I am just pondering over some design aspects from whatever little I know so far. Just thinking about how diffraction hates less symmetry, smooth surfaces etc, Will it help if we create a woofer module where the woofer is asymmetrically mounted on the baffle (in vertical direction-i.e., woofer location shifted more down towards the bottom of the cabinet or more towards the top in the woofer module cabinet, with a separate smaller cabinet for mids and tweeter kept on top of the woofer module). This would be unlike the woofer location in my current design where it is almost at the center of the cabinet in vertical direction.
The other thought I have is regarding what would be the effect of a generous rounding of the overall cabinet itself (again thinking like diffraction doesn't like rounding much
). I couldn't help myself notice something like these aspects in these speakers (Even though these might be way beyond reach DIY options available to me).
Atria II - Rockport Technologies
Deep cabinets which look the least like a rectangular box. At least it looks aesthetically very pleasing to me.
I will try to create a woofer box with generous rounding but like a simplified version of above in fusion 360. But given my poor fusion360 skills, it might take a bit of time..
You are welcomeThanks a lot fluid.
There is information on port placement and reducing midrange leakage as to the directivity effects from cabinets I don't know of any specific sources of information.
The tweeter should be OK and the Mid looks fine as long as it will be crossed over high enough. The same sort of issues are visible in it's response too but they are more likely to be out of the passband. Fixing the enclosure for the woofer would have the same effect for the mid too.I will wait for your simulations and suggestions about the woofer enclosure part and the remaining part of the overall design like the tweeter and mid interaction to see if changes need to be done in that area also..
This would have some effect but more so on the vertical directivity and overall power response. As the same sort of issue is visible in the mid it won't be fixed just by moving the woofer. Having the woofer closer to the floor is a good idea in general as you get more reinforcement that way.Will it help if we create a woofer module where the woofer is asymmetrically mounted on the baffle (in vertical direction-i.e., woofer location shifted more down towards the bottom of the cabinet or more towards the top in the woofer module cabinet, with a separate smaller cabinet for mids and tweeter kept on top of the woofer module). This would be unlike the woofer location in my current design where it is almost at the center of the cabinet in vertical direction.
The Rockport is basically a faceted baffle. Chamfers and round overs work much the same. Facets and round overs have more effect as frequency rises as the longer wavelengths just don't see the baffle edge like the mid and high ones do.The other thought I have is regarding what would be the effect of a generous rounding of the overall cabinet itself (again thinking like diffraction doesn't like rounding much
I couldn't help myself notice something like these aspects in these speakers
Here is an example of a different cabinet shape and driver layout intended to make the directivity in this region better.
Directiva r2 project: market requirements gathering | Page 23 | Audio Science Review (ASR) Forum
Beamwidth and Polar
Directiva r2 project: market requirements gathering | Page 22 | Audio Science Review (ASR) Forum
Hmm, nice footage thanks for the links fluid It would indeed be very interesting to see sims how interference changes with relation of the box width and depth. I guess there is a region / ratio where the interference is "optimized" compared to other ratios for the bandwidth the driver/box is responsible for.
I suspect best is to use minimal enclosure, it will inevitably have as small footprint in the acoustic domain as possible. Some sims could help to bust or confirm this, there might be configurations that are larger and still have minimal effect (like a sphere for example). Small enclosure reduces the bass capability but it is no problem on a multiway system if the eventual bass speaker operates wavelengths well greater than the enclosure dimensions required, in general something like < 200Hz. This is conveniently about the Schroeder frequency zone. Small enclosure internal resonances and problems could be above the passband depending on the system design or use passive cardioid box to reduce the resonances to minimum. Passive cardioid reduces the bass capability further but seems to promise cleanest sound with very easy manufacturing, although extensive prototyping seems to be required to achieve.
For the bass box, you can play with Hornresp software to see a bit more closely what happens in a ported enclosure. Use the MLTL mode to see how driver position affects the resonances etc. Basically you want to have acoustic center of the driver(s) at the center of the longest dimension of the box to kill the lowest internal standing wave. This leaves only the higher modes and the port pipe resonance which all are hopefully well above the bandwidth of the bass box and relatively easy to dampen with damping material compared to the lowest mode(s) you just got rid of with the appropriate driver positioning. Shallow enclosure and driver middle of the baffle gets the acoustic center of single driver system almost center of the box killing the problematic standing waves. Actually killing is wrong word, they just don't happen Four drivers top and bottom of front and back, or any other "symmetric" 2n driver configuration could achieve the same depending on which is most suitable for your system design.
A three way speaker can be optimized so much better than a two way, but still many people say two ways sound better for some reason. Two ways can be smaller and cheaper and easier, but there is no way they can be optimized to the max like a three way can. Still, it leaves a question what is audible and what is not. Anyway, interesting engineering challenge to optimize either.
edit. I've got to add that a DSP as crossover solution is a must with most of my ponderings. It would be very hard to do a passive xo for passive cardioid system for example. And parts get expensive for anything < 200Hz. I think this is one reason people use the 2 way boxes. They think 3-ways are too hard and expensive because they think only passive xo and don't even consider DSP thinking it is somehow evil even though DSP solves million problems when looking audio quality from engineering perspective. I'm not saying which sounds better, audio and speakers are very much aesthetics and feelgood luxury stuff as it is absolute audio quality and most part of the perceived sound quality comes from the visual cortex and other parts of the brain than the hearing system anyway Plus there is aspect rolling the amplifiers and opamps and other "meaningful" stuff that is a lot easier and cheaper with passive crossover speakers... What ever makes one happy.
It would indeed be very interesting to see sims how interference changes with relation of the box width and depth. I guess there is a region / ratio where the interference is "optimized" compared to other ratios for the bandwidth the driver/box is responsible for.I suspect best is to use minimal enclosure, it will inevitably have as small footprint in the acoustic domain as possible. Some sims could help to bust or confirm this, there might be configurations that are larger and still have minimal effect (like a sphere for example). Small enclosure reduces the bass capability but it is no problem on a multiway system if the eventual bass speaker operates wavelengths well greater than the enclosure dimensions required, in general something like < 200Hz. This is conveniently about the Schroeder frequency zone. Small enclosure internal resonances and problems could be above the passband depending on the system design or use passive cardioid box to reduce the resonances to minimum. Passive cardioid reduces the bass capability further but seems to promise cleanest sound with very easy manufacturing, although extensive prototyping seems to be required to achieve.
For the bass box, you can play with Hornresp software to see a bit more closely what happens in a ported enclosure. Use the MLTL mode to see how driver position affects the resonances etc. Basically you want to have acoustic center of the driver(s) at the center of the longest dimension of the box to kill the lowest internal standing wave. This leaves only the higher modes and the port pipe resonance which all are hopefully well above the bandwidth of the bass box and relatively easy to dampen with damping material compared to the lowest mode(s) you just got rid of with the appropriate driver positioning. Shallow enclosure and driver middle of the baffle gets the acoustic center of single driver system almost center of the box killing the problematic standing waves. Actually killing is wrong word, they just don't happen
Four drivers top and bottom of front and back, or any other "symmetric" 2n driver configuration could achieve the same depending on which is most suitable for your system design.A three way speaker can be optimized so much better than a two way, but still many people say two ways sound better for some reason. Two ways can be smaller and cheaper and easier, but there is no way they can be optimized to the max like a three way can. Still, it leaves a question what is audible and what is not. Anyway, interesting engineering challenge to optimize either.
edit. I've got to add that a DSP as crossover solution is a must with most of my ponderings. It would be very hard to do a passive xo for passive cardioid system for example. And parts get expensive for anything < 200Hz. I think this is one reason people use the 2 way boxes. They think 3-ways are too hard and expensive because they think only passive xo and don't even consider DSP thinking it is somehow evil even though DSP solves million problems when looking audio quality from engineering perspective. I'm not saying which sounds better, audio and speakers are very much aesthetics and feelgood luxury stuff as it is absolute audio quality and most part of the perceived sound quality comes from the visual cortex and other parts of the brain than the hearing system anyway
Plus there is aspect rolling the amplifiers and opamps and other "meaningful" stuff that is a lot easier and cheaper with passive crossover speakers... What ever makes one happy.
Last edited:
Thanks again fluid and tmuikku..
It looks like I have to take a real hard look at what I want to accomplish with this speaker build.
Obviously sound quality comes first, but l think looks of the speaker also matter to some extent, even if not for me, others at home..
Passive cardioid like designs, from the little I understand about them requires extensive prototyping and more knowledge about the domain, for both of which, I currently lack resources in terms of compatible drivers, more capable DSP and amplification.
Right now, all I have are a pair each of Satori WO24P-8 woofer, SB15CAC30-8 mids, and waveguided SB26CDC004 tweeters, all bought initially with the intention of trying a conventional DSP crossover-based 3 way design.
Now I am left wondering what is a good design I can achieve with what I have. The bass reflex woofer module design using Hornresp, I might be able to do. But solutions for completely active directivity control/passive cardioid-like approaches, my current level of knowledge, skills and financial resources won't suffice.
I have been following the directiva threads on ASR that fluid pointed out with interest from its earliest versions days. The discussions, proposals, knowledge sharing etc seems great on that thread. I have also been wondering about ctrl's specific cabinet shaping approach seeing the pic. But from what I understand, it seems like there the approach is to go as low as possible with the two mid woofers on the specific shaped box and well below in the bass range like probably <100 Hz cross it over to a capable subwoofer. I definitely liked his approach to the problem of directivity control but don't know if it is right away adaptable to my case with just a single midwoofer and a woofer capable of going upto > 500 Hz with ease.
One thing I am getting more and more attracted to is the separate bass module design with separate smaller shaped box for mid and tweeter like the
Directiva r2 project: market requirements gathering | Page 23 | Audio Science Review (ASR) Forum
Atleast it is easier to manufacture as separate pieces than one large box.
Again, I dont know whether this would be a solution in my case, But I will try to think about this more.. I am very confused about all this now
It looks like I have to take a real hard look at what I want to accomplish with this speaker build.
Obviously sound quality comes first, but l think looks of the speaker also matter to some extent, even if not for me, others at home..
Passive cardioid like designs, from the little I understand about them requires extensive prototyping and more knowledge about the domain, for both of which, I currently lack resources in terms of compatible drivers, more capable DSP and amplification.
Right now, all I have are a pair each of Satori WO24P-8 woofer, SB15CAC30-8 mids, and waveguided SB26CDC004 tweeters, all bought initially with the intention of trying a conventional DSP crossover-based 3 way design.
Now I am left wondering what is a good design I can achieve with what I have. The bass reflex woofer module design using Hornresp, I might be able to do. But solutions for completely active directivity control/passive cardioid-like approaches, my current level of knowledge, skills and financial resources won't suffice.
I have been following the directiva threads on ASR that fluid pointed out with interest from its earliest versions days. The discussions, proposals, knowledge sharing etc seems great on that thread. I have also been wondering about ctrl's specific cabinet shaping approach seeing the pic. But from what I understand, it seems like there the approach is to go as low as possible with the two mid woofers on the specific shaped box and well below in the bass range like probably <100 Hz cross it over to a capable subwoofer. I definitely liked his approach to the problem of directivity control but don't know if it is right away adaptable to my case with just a single midwoofer and a woofer capable of going upto > 500 Hz with ease.
One thing I am getting more and more attracted to is the separate bass module design with separate smaller shaped box for mid and tweeter like the
Directiva r2 project: market requirements gathering | Page 23 | Audio Science Review (ASR) Forum
Atleast it is easier to manufacture as separate pieces than one large box.
Again, I dont know whether this would be a solution in my case, But I will try to think about this more.. I am very confused about all this now
Yeah it is easy to get confused since there are multiple solutions to multiple problems I suggest you build the current project as is, at least a prototype. Measure it and optimize the stuff that needs optimizing and finish the project. I suspect the sound will be good as long as you have addressed an issue and either accepted it or bettered it one way or another. On the next project you have opportunity to redefine the goal and when you have very good plan (goal) and knowledge how to get there you can make better one.
Separate top and bass box is a fine solution sound quality wise since you can optimize dimensions of both to their respective bandwidths. More work involved in a way that now there are two boxes, but it could still be easier to build since both can be smaller than if they were combined as one. Big boxes are very difficult to manufacture. Small errors multiply, big assembly space required (I've assembled on floor.. ), huge clamps and what not. Pain in the back in general Have fun!
I suggest you build the current project as is, at least a prototype. Measure it and optimize the stuff that needs optimizing and finish the project. I suspect the sound will be good as long as you have addressed an issue and either accepted it or bettered it one way or another. On the next project you have opportunity to redefine the goal and when you have very good plan (goal) and knowledge how to get there you can make better one. Separate top and bass box is a fine solution sound quality wise since you can optimize dimensions of both to their respective bandwidths. More work involved in a way that now there are two boxes, but it could still be easier to build since both can be smaller than if they were combined as one. Big boxes are very difficult to manufacture. Small errors multiply, big assembly space required (I've assembled on floor.. ), huge clamps and what not. Pain in the back in general
Have fun!
Last edited:
I will run some variations, wait and see if the woofer position and width/depth ratio make an improvement. With the drivers you have I think this will turn out to be a decent solution.Thanks again fluid and tmuikku..
I am very confused about all this now
This is a popular shape for good reason. You were talking CNC before and that will alleviate some of the pain in construction if some thought is put into it.
Thanks fluid..
Yes. I will have access to CNC cutting for getting the cabinet panels cut.
I will surely wait for your results to see if we can make this design at least slightly better
Kimmosto says something about this speaker having a cardioid-like polar pattern from lower midrange..
KS-483
Attaching the picture of the box from above site with this post.
I also see that this approach looks similar to ctrl's approach
Speaker box is shallow in depth with very wide baffle and generously rounded box sides..
Maybe this is one way to go if we want directivity control into lower frequencies for free. He also says about the sound character of the speaker at the end, which looks good..
Just wondering that since all designs trade off some feature(s) for the something else, subjective sound character-wise, there may be perceivable difference between this type and narrow baffle type designs.
Also posting this thread about the analysis of another wide baffle speaker (Grimm LS1):
Analysis of the Pros/Cons of Wide Baffles, case Study: Grimm LS-1
KS-483
Attaching the picture of the box from above site with this post.
I also see that this approach looks similar to ctrl's approach
Speaker box is shallow in depth with very wide baffle and generously rounded box sides..
Maybe this is one way to go if we want directivity control into lower frequencies for free. He also says about the sound character of the speaker at the end, which looks good..
Just wondering that since all designs trade off some feature(s) for the something else, subjective sound character-wise, there may be perceivable difference between this type and narrow baffle type designs.
Also posting this thread about the analysis of another wide baffle speaker (Grimm LS1):
Analysis of the Pros/Cons of Wide Baffles, case Study: Grimm LS-1
Attachments
Last edited:
Since edit time is over for the last post, posting as a separate post here. Seems like the entire baffle (+depth also?) will act like a waveguide at lower frequencies trying to still maintain directivity before the radiation pattern collapses into omni. So wider baffles give us one variable for controlling the beamforming aspect. If we could understand the precise impact of depth also, we have two variables to control, but the weighing that can be applied to each variable may need more investigation..
I think fluid's simulations will help a lot in this regard if whatever pondering i posted above is true..
Fascinating.. All these things..
but the weighing that can be applied to each variable may need more investigation.. I think fluid's simulations will help a lot in this regard if whatever pondering i posted above is true..
Fascinating..
All these things..
Last edited:
Hi,
Check the gazillion Patrick Bateman's thread about this particular aspect ( width and depth of box. There is one with kef ls50 where there is some interesting points exposed).
Those loudspeakers in the picture you linked looks alot like A Capella's speakers from T.Gravesen rather than Grimm LS. At least to me.
troelsgravesen.dk/Acappela_WB.htm
Check the gazillion Patrick Bateman's thread about this particular aspect ( width and depth of box. There is one with kef ls50 where there is some interesting points exposed).
Those loudspeakers in the picture you linked looks alot like A Capella's speakers from T.Gravesen rather than Grimm LS. At least to me.
troelsgravesen.dk/Acappela_WB.htm
Last edited:
Thanks a lot for pointing Patrick Bateman's threads to me.
I just read the thread where ls50 is mentioned.. Fascinating..
I should have done this reading earlier instead of wasting time thinking about all this..
So it seems like a deep cabinet with WAF-wise good width may also help in getting results close to wide baffle + narrow depth, but interactions are hard to predict.
If Revel M16 can do that good, the Rockport-like deep and curved cabinet sides with chamfered front may also help with woofer directivity control at least to some extent. I have my speaker shape similar to Rockport in chamfers and all but depth of the cabinet may be not enough/proportion of depth and width may not be as good.
Now I will wait for fluid's simulation results with fingers crossed
Wide curved baffle comes from Sonus Faber Stradivari, later version is called Elipsa
Sonus Faber Stradivari Homage loudspeaker | Stereophile.com
Sonus Faber Elipsa SE – Reviews | TONEAudio MAGAZINE
Sonus Faber Stradivari Homage loudspeaker | Stereophile.com
Sonus Faber Elipsa SE – Reviews | TONEAudio MAGAZINE
Last edited:
....
I should have done this reading earlier instead of wasting time thinking about all this..
,,,
I disagree: better think about something, trying to figure out what it is about then have confirmation you were right or wrong, it help identify processus at play. At least for me it works.
'Wide' cabinet are there for a long time, VOTT A2 or A4 were standard for movie industry since 1955, 10 years after their first introduction. And in studio world infinite bafle is common ( inwall).
Anyway, it is interesting to read what some designers have to say about this:
Loudspeaker designer John Dunlavy: By the Numbers... | Stereophile.com
An SC-VI is 18" wide which make bsc center freq around 250hz. Earl Geddes's Summa was approximately same size. I own a pair of loudspeakers with ~same width and i share some of the comments made in T.Gravesen's 'Acapella' link ( in direct comparison to other 'thinner' loudspeakers i own too).
Maybe it is diffraction related, maybe not but the comments made by J.Dunlavy in the interview make me think it may be the reason i like the wide width more for acoustic material.
Like for every other choices to be made it is a compromise ( my girlfriend is not particularly happy about the look but as long they sound good...).
Last edited:
Thanks Juhazi and krivium for pointing out these resources. Really interesting material..
So wide baffle designs have been around for a while... @Krivium: Yes. You are right and I was wrong about thinking of taking the short cut and not thinking about all this. Earlier I was just thinking about me going down the "trying to reinvent the wheel" route.
Thinking about these design aspects helped me learn whatever little I know so far about all this which I may have never known otherwise
And I really liked reading J.Dunlavy's interview. Lot of insights, especially regarding subjective interpretations of spectral imbalance in the 100-300 Hz region. I also liked his approach to speaker design based on his background in antenna theory. I can see how it really helps if one already has solid background regarding impedance transformations and antenna radiation patterns..
Thanks fluid
Can you elaborate a little bit about how to interpret these pics. I am a little confused looking at them..
This is what I understood so far.
In general we are trying to look at how a 400Hz signal radiates into free space surrounding the speaker
Pic-1: We are looking at the radiation from the front. I see dips in the attenuated blue colored wavefront at the top of the speaker. So something has caused energy to dropin those dips. Is it secondary sources due to diffraction from the edges?
Pic-2: Side view of the radiation. I again see blue colored dips at the back. If the speaker had more depth, would the radiation have been more of a "focussed beam?
Pic-3: Top view of the radiation. Again the dips at the back. Would it have been better if those dips were not there or that pattern was more focussed forwards
Pic-4: We have a slice each of the directivity baloon in the horizontal and vertical directions comtaining the combined information in all 3 plots before
Last edited: