I just started to study and delve into the topic of acoustic construction, I did not expect so many subtleties, but I'm already starting to understand why when I replaced all the components several levels higher, the sound did not improve at all. 🙂
I don't quite understand the process you are describing. But I will study it and ask the right questions in the future. 🙂 Give me some time.
I have read this article. Conclusions I made:
1. Select speakers and subsequently make crossovers to get Ka = 2. 🙂
Where ka is a dimensionless number equal to the circumference (in mm, m, inches, feet, etc.) divided by the wavelength (in the same units).
But the wavelength at each frequency is different ... At, for example, 300 Hz, the wavelength will be 1.14 meters. Do I remember physics correctly?
In general, I do not fully understand this meaning of Ka. I will study further.
Well, i don't know if i'm respected but i see people i do have respect for giving you advice in this thread this is sure.
Let me tell how i see things: this is a huge amount of works to be done to be able to be successful in a build like you plan. Not that it is impossible but this will require you to gain knowledge in a lot of different fields if you start from zero ( let's suppose that even if it is untrue).
This with the help of members here can be relatively fast for theorical side of things. But you'll need some tools which are mandatory.
This include learning simulation software, taking measurements being able to interpret them, etc,etc,... All this takes time and comitment as well as money.
And to be successful i think this have to be planed.
And first and foremost you have to define a list of defined design objective. And to rank them by order of preference. From there you can start to think about some design choice.
Eg:
1 Aesthetic
2 cost
3 target spl and bandwidth requirement @ listening position
4 sd area in low end
5 ribbon tweeter for high end
6 systeme must be multi amplied
7 should accept hybrid approach for filtering dsp/ passive
8 overall size
9 ergonomics
etc,etc,etc,...
This is very important as you'll have to compromise on things and it'll help to not loose focus with the choices that'll have to be made.
And each choices will implies other ones...
Eg: you insist on multiway vertically aligned driver layout. This is perfectly fine for aesthetical reason let's say, but this choice will imply you understand the outcome from the acoustic behavior to see if it suits your preferences, expectations and room.
If it does then next step, with how to reach target spl and bandwidth in a given size and cost constraint,... this will gives you a list of potential drivers which will then give cues on where to put your xovers freq ( or not if you use Hornbach Keele approach as they'll be defined by this design criteria alone as well as size and physical location...).
Well you see the point.
It takes time.
That said if you commited to it it can be done. But not in a hurry.
Lojzek advice is wise and from a respected member which helped a lot of people here. And i can only advice the same as it is what really get me into all this:
Trying to 'better' my existing loudspeakers.
I learned a lot this way, experienced the theorical things i was reading, experienced limitations ( i've redone passive filter with better parts too), some things worked other not...
Meanwhile i kept accumulating knowledge up to the point i get 'the whole picture' ( the important parameters at play and their interaction) to be less blurred. Then only i designed my first basic loudspeakers. I won't tell you how long it took me though...
About directivity i will post some things to help you understand what is at play.
Here is an exemple of a given design criteria list:
https://www.diyaudio.com/community/...ligned-constrained-layer-construction.218437/
You write everything correctly. I already realize that the deeper I immerse myself in knowledge, the more I understand which path I need to go through, and how far the end point is located (is the end point correct? ... In general, the place where I in the end, if I live. 🙂 ). 🙂
There is no rush, it cannot be... After all, at first I did not understand where I was getting involved. 🙂 I thought that acoustics is a box and there are speakers + crossover in it. 🙂 I dismantled my Diatone DS-90, everything inside seemed elementary to me. But now I'm reading articles and I understand that if I want to do WOW, I need to understand in all aspects.
So far I have understood one thing: the D'Appolito circuit is not ideal, it is easier to implement, and more forgiving of speaker placement.
Yes the microphone pick the distortion but what did you thoughts of your system before being aware of it? I'm sure you liked it and wasn't bothered at all. 😉
What i mean is those are items from an area/period in time and they reflect what the designer had in mind at that time ( or not if it works by luck!), the trend of the moment.
It's a rendering amongst others, newer approach being differents but not mandatory better as it's all about compromise in the design process.
I'm a fan too. As i am for the designers i gave link to their works.
Gaining this much knowledge as those guys have require a lifetime of dedication.
This interview of Dunlavy give you all the hints to determine what design goals he had and how to reach them. He talks about FIR processing and dsp and had a clear view of what would come later imho. Too sad the illness took his brain before he was able to complete the Magnus.
The Whitepapers from E.Geddes and W.Parham expose theirs clearly too.
The interview of Dave Smith as well as the review of the Reference xp too. The review is worth a read as there is some very clear description of what a nice controled vertical directivity brings ( 'the pink noise test') and you could expect if implemented correctly.
Thats a nice euphemism! There is many ways to approach box construction with different schools of thoughts.
But for this we are lucky to be a community with experience about (probably) everything that could have been done. So it's a matter of finding the good infos about each approach and see the compromise to be made... then decide by yourself what make the much sense to you and your skills and limitations.
Mind you one thing about boxes in commercial design: this is the most costly and complex parts.
It plays a role as well as drivers, filters,... it's shape, volume and realisation play a role in the final outcome. It's not something arbitrarly shaped ( well you can but...).
My weak link is carpentry. I'm not good enough to reach the level of finish i would like ( objectively) and i'm a perfectionist too ( read a bear always unsatisfied!). So it is the frustrating part for me because: you'll need to prototype to gain knowledge about the 'real thing'. Simulation and theory is great but only the reality is the truth.
And despite that practice i don't get better.... well i'll have to consider help from a pro for my next loudspeakers to come...
Well anyway, you'll need prototyping to make your own judgement on things. For this no need for pricey drivers but some you have around good enough to try a concept of acceptable size without too much objectable issues....
You can do it this way yes.
Or you can have a global goal about your design and define the Ka you need over the frequency range you want and from then, define the diameter of drivers for the Ka value you specified for a given freq range. It'll then dictate the xover freq or some target area ( some other points have to be checked regarding other driver issue (eg: breakup freq).
You are right, wavelength vary with frequency so Ka is meaningfull at only one freq ( if you look at other freq it might be very different). But it give an idea of the directivity to expect at a given though. Not the whole story for sure but meaningful enough for some assumptions.
But let's look at that one instead ( scroll down to speaker directivity simulator 1.13 and look at picture):
http://www.tonestack.net/software/speaker-directivity-simulators.html
The directivity simulator display a polar map of a direct radiator driver. From there we can estimate the behavior of the driver ( and in 3dimension).
From quicly looking at it we can estimate the -6db @ +/- 45* to be around 1,8khz ( the color area correspond to level in upper left).
From this i can tell it is an 8" driver. What we see is there is narrowing happening in the high and this start from circa 800hz. Below 800hz the behavior is constant ( there is no significant variation in color shade) and it spread very wide at same level. It is in a low Ka range. Iow we can approximate the behavior in 3d as an hemisphere ( this isn't the case as it is an infinite baffle simulation but let's assume this it'll help better understand what comes next).
Then from 800hz it starts narrowing until becoming a laser beam past 8khz.
So in 3d it would look alike a water drop shape with the driver in center of the hemisphere and at start of the 'sting' part.
This is the typical behavior of any direct radiator whatever it's size. The overall shape will be the same only the frequency at which the narowing happen will change.
And this is where Ka is interesting as it'll tell you more or less where the narrowing will happen ( Ka bigger than 2 as a reminder Ka 0,5 is almost perfect omni, below the 800hz freq in our exemple).
Ok. But that is not all. Once you introduce another driver ( let's say a tweeter) there is chance there will be difference into their radiating freq: the woofer'll narrow past 800 with a 90* opening angle at 1800*, while at this frequency the tweeter will be in low Ka range, it'll spread very wide ( omnidirectionally).
There will be issue: some destructive interference are going to happen and then the loudspeaker could not radiate the same sound everywhere in the room: what is called 'lobing' happens. That's why we often see same drivers combination often used ( eg: 6,5"+1") and other not or less ( or with waveguide).
The main issues about this are that there will be 'hotspot' in space where sound will be louder than other, (so it'll have an influence on sweetspot, narrowing or widening it, or creating 'holes' in an area) and it means that the (wall) reflections we are going to listen to ( as well as the loudspeakers direct sound) might be 'unbalanced' ( like if a different eq was used for reverb than for direct sound).
Ok why all that? Because now you understand directivity of one driver alone we could see what happen when combining them:
http://www.mes-enceintes-acoustiques.com/2021/06/moniteurs-mtm.html
Here is two 'ballon' view of directivity ( in 3D) of MtM and an Mt. As can be seen there is significant difference between both behavior. At first, mtm looks more consistant in both horizontal and vertical. Mt have a main lobe located around the tweeter axis ( do you get why people insist to have tweeter at hear eight?).
The whole Mt shape should remind you about one of the graph in the 'Biro technology' article( or remind the 'water drop shape' i talked previously)... as well as the Mtm shape and the Horbach Keele paper...
So yes, mtm have issues. Mt too and all other configuration too. These are compromise, tradeoff,...
One can choose an mtm for this behavior ( consistency of directivity) or because it mimics a coaxial arrangement, give a stable rendering if you don't move in the vertical axis ( for standard mtm, and why they made their way to studio control room /you work seated and without this much height movement), because they lessen floor bounce,... there is many others pro than just directivity ( and cons too!).
But you have to make educated guess about choices so need to understand it ( and for me it is a Key parameter of loudspeaker rendering: how the directivity behave....).
Last edited:
Hi, heroic act krivium, lots of infornation is hard to concentrate and deliver, well done 🙂
Vetusto, I'll add that if nothing else sticks to the
head currently, if it all gets overwhelming, just do one step at the time. Always remember to think the trade-offs, not just the options themselves one is weighting against each other but also their trade-offs. If there is clear goal what one wants to achieve with the project, or at least simple list of priorities, it is rather easy to determine if the trade-off would prevent achieving the goal 😀 and if it is, one should think alternatives, or ways to pass the trade-off forward to some other area that is less important for the project success. If you cannot figure out what trade-off(s) there is for option you are looking at you need to study more closely, there is always some trade-off, hardly anything comes without.
Priority list would be as simple as arranging these basics in order of importance: cost (time, money), size (can it be big), looks (shoebox or something more organic) and audio quality. If you sort audioquality first priority, you design is good when you don't have any compromises taken on audio quality 🙂 what ever that is, you define the list, try stick to it unless you find it unrealistic.
And most importantly, remember to have fun, it is a hobby afterall!
Vetusto, I'll add that if nothing else sticks to the
head currently, if it all gets overwhelming, just do one step at the time. Always remember to think the trade-offs, not just the options themselves one is weighting against each other but also their trade-offs. If there is clear goal what one wants to achieve with the project, or at least simple list of priorities, it is rather easy to determine if the trade-off would prevent achieving the goal 😀 and if it is, one should think alternatives, or ways to pass the trade-off forward to some other area that is less important for the project success. If you cannot figure out what trade-off(s) there is for option you are looking at you need to study more closely, there is always some trade-off, hardly anything comes without.
Priority list would be as simple as arranging these basics in order of importance: cost (time, money), size (can it be big), looks (shoebox or something more organic) and audio quality. If you sort audioquality first priority, you design is good when you don't have any compromises taken on audio quality 🙂 what ever that is, you define the list, try stick to it unless you find it unrealistic.
And most importantly, remember to have fun, it is a hobby afterall!
Last edited:
My weak links are in general all aspects. 🙂 But I decided for myself that the main thing is to correctly calculate and draw a model of the hull, and think over a system for fixing the walls, then I will give the drawings to production, they will make separately all the walls that I myself will assemble at home, I will also take them to painting in my car spouses for production. I think two meters will easily fit into the trunk of a car.
In general, the main thing is to think over the wall fixing system, either with the help of glue and special furniture ties, then, when the glue dries, remove the ties and install wooden chopsticks in their place. But it is not exactly. I'm still thinking about it.
In general, my brain is constantly working on all the options. I think when I make a model in SolidWorks, other options will come to me.
In general, the idea so far is this: to make the entire body completely collapsible. This will greatly facilitate the movement of the body and it will be possible to make improvements on the fly. Because I don't understand yet what the inner cavity should be.
It's the most important! I just started all this for the sake of satisfaction and pleasure.
By the way, yesterday I met a man who manufactures speakers in Russia (not a case, but only speakers, midrange and bass speakers).
I don't know why you need a 4 way or 5 way system to fill a room of that size up to 120 db. Look at this thread of a guy that built a party system for outdoors with two way tops with 8" woofer & one 18" sub. https://www.diyaudio.com/community/threads/3-way-pa.383424/ He states crossover @ 90 hz, which I find suspect with 8" woofers. Going to 4 way tops allows a lot more power in each driver, which is not necessary in a 18 sqm room to achieve 120 db. Even at a party with lots of sound absorbing clothes.
Dunlavy complained about passive crossovers messing up time alignment. I'm not sure how important time alignment is, but a 5 way has about 8 energy storage units (capacitors, inductors) minimum per channel in the passive crossover. If you don't use passive, you have to maintain a smart phone at $$ a month to program a DSP, and maintain 10 amp channels or 5 stereo amps, when consumer grade amps these days have MBTF of about 5 years. All that heat! anybody hear of global warming? I don't keep a smart phone.
I'm running stereo two way speakers in a 47 sq m room, with one 2 channel amp. I have response +-3db 54 hz to 17.5 khz. It is down 10 db at 40 hz due to bass boosting room effect. I could buy a sub to 54 hz to 20 hz if I wanted that response and already own an electronic crossover that merges the two channels below an adjustable cutoff frequency, that I paid $50 for. I have 4 crossover parts per cabinet. My 2 ways will produce 125 db pink noise @ 500 W rms. I limit them to 90 w/ch inside because I value the preservation of my hearing. Time misalignment is specified at 0.48 ms for highs. About the dreaded "beaming" of 15" drivers, the 6 db down beam width is 90 deg horizontal and 40 deg vertical. 10 deg above horizontal and 30 deg below, for mounting on stands at the top of the room.
So why a 5 way, or 4 way tops. Eye candy?
Last edited:
Hi,
Well i'm not sure you can compare Dunlavy to your loudspeakers Indianajo.
They are 5 way yes but 6db/octave filters so don't expect what you get from yours. With 91db efficiency and 250w max allowed they satisfy requirements for 'standard' level practice ( 85dbspl rms at 3meters with 20db headroom allowed, iow 115dbspl at 1m).
Those where developed with quality in mind, not spl and this is one of the few ref from audiophile world which made it's way as a reference loudspeakers in mastering rooms.
Some guys use them in small rooms with no real drawbacks:
https://www.inner-magazines.com/audiophilia/the-mighty-dunlavy-sc-vi-powered-by-asr-brinkmann/
I disagree with your analysis of multiamp systems too but tbh i don't see the point here: where have you seen Dunlavy are multiamped?
Horbach-Keele approach is though and if a Diyer plan a kind of Dunlavy's style loudspeaker there is less chance to waste money using this approach rather than the one used by Dunlavy ( remember, 6db/octave filters, all boxes sealed,...).
That said i'm not convinced by the fact multiamps 'eats' more energy than passive approach with a 'big amp'. In fact at same total power a multiamp will be +3db louder ( at most), so more efficient by application of principle only ( not optimising the kind of amplifier used).
Anyway, i suggest you read the Dunlavy's interview to get what his design criteria was and why he made the choices he did.
Well i'm not sure you can compare Dunlavy to your loudspeakers Indianajo.
They are 5 way yes but 6db/octave filters so don't expect what you get from yours. With 91db efficiency and 250w max allowed they satisfy requirements for 'standard' level practice ( 85dbspl rms at 3meters with 20db headroom allowed, iow 115dbspl at 1m).
Those where developed with quality in mind, not spl and this is one of the few ref from audiophile world which made it's way as a reference loudspeakers in mastering rooms.
Some guys use them in small rooms with no real drawbacks:
https://www.inner-magazines.com/audiophilia/the-mighty-dunlavy-sc-vi-powered-by-asr-brinkmann/
I disagree with your analysis of multiamp systems too but tbh i don't see the point here: where have you seen Dunlavy are multiamped?
Horbach-Keele approach is though and if a Diyer plan a kind of Dunlavy's style loudspeaker there is less chance to waste money using this approach rather than the one used by Dunlavy ( remember, 6db/octave filters, all boxes sealed,...).
That said i'm not convinced by the fact multiamps 'eats' more energy than passive approach with a 'big amp'. In fact at same total power a multiamp will be +3db louder ( at most), so more efficient by application of principle only ( not optimising the kind of amplifier used).
Anyway, i suggest you read the Dunlavy's interview to get what his design criteria was and why he made the choices he did.
I read the dunlavy review. He maximizes time alignment of impulse response & step response. He says this maximizes simulation of actual musical instruments. There is no evidence for this in the interview. He does say frequency response is <+-1 db.
I bought the speaker that maximizes simulation of musical instruments in my market. I tested mine with piano CD track, comparing to sound of actual Steinway grands. I own a Steinway console in next room to the speakers. No Dunlavy speaker for sale here, obviously. The only other speaker here that sounds that good is a Meyersound at Brown theater. A speaker the size of a Fiat Urbana car.
See HD chart for my speaker @ 1% full power, or 5 watts. 2nd & 3rd harmonic <20 db of emitted sound. That is louder than I use them 99.9% of the time. My measured voltage at normal volume is 1 Vav or 1/8 watt. Again my home amp maxes out at 140 w both channels driven. I have a 800 w amp for performances in halls.
If I was going to spend $14000 on a speaker system, I would rather spend $18000 on an actual musical instrument. I/ve been eyeing a bassoon for $18000.I played one with skill ending with graduation in 1968. I have never owned one.
As for diy, my pair were stolen 9/20. They had pawn shop value. I bought drivers to replicate them in a case so ugly it would have no pawn shop value. Since after a 15 month search I found used ones at 2/3 what I paid originally, that project is waiting on replacement of the power saws, clamps, tools etc the burglar also stole.
Attachments
Last edited:
Don't get me wrong, i'm not critisizing your choice and or preference.
I was just pointing to the fact you use horn loaded /bass reflex two ways designed for PA initially, with steeper slope filters.
The size and requirement ( in power and number of drivers) for a sealed loudspeaker able to put out 30hz with Qtc of 0.6 at 115db spl @1m will be different from yours, definitely.
Yes there is no more evidence other than what he said in this interview, but.. it's not the only one he did and if you search a bit you'll find other things from him or users, etc,etc..
Anyway, i'm not going to argue about this, i've no gain in this game ( i'm not trying to sell anything to anyone and i'm not trying to promote a 'guru' or anything like that) and my own experience of the ones i heard will not relate to home use as they were used in a pro treated room ( and it wasn't SC-VI anyway...).
That said the whole approach of the guy was a milestone along vertical aligned loudspeakers design, like D'Appolito was before him and to my eyes Dave Smith Snell's late effort ( which put a bridge to line array, shading, etc,etc,..).
Next step in this evolution is Horbach-Keele to me which synthesize all this: Fir filters ( no smearing of transients and steep slope) and a way to maintain vertical directivity under control ( and not only a 'wide' pattern one), does it with sealed/closed box and you'll be as close as possible a Magnus could have been. Probably even 'better' in some ways.
That said they'll need an answer to the 'waveguide problem' this approach have.
I was just pointing to the fact you use horn loaded /bass reflex two ways designed for PA initially, with steeper slope filters.
The size and requirement ( in power and number of drivers) for a sealed loudspeaker able to put out 30hz with Qtc of 0.6 at 115db spl @1m will be different from yours, definitely.
Yes there is no more evidence other than what he said in this interview, but.. it's not the only one he did and if you search a bit you'll find other things from him or users, etc,etc..
Anyway, i'm not going to argue about this, i've no gain in this game ( i'm not trying to sell anything to anyone and i'm not trying to promote a 'guru' or anything like that) and my own experience of the ones i heard will not relate to home use as they were used in a pro treated room ( and it wasn't SC-VI anyway...).
That said the whole approach of the guy was a milestone along vertical aligned loudspeakers design, like D'Appolito was before him and to my eyes Dave Smith Snell's late effort ( which put a bridge to line array, shading, etc,etc,..).
Next step in this evolution is Horbach-Keele to me which synthesize all this: Fir filters ( no smearing of transients and steep slope) and a way to maintain vertical directivity under control ( and not only a 'wide' pattern one), does it with sealed/closed box and you'll be as close as possible a Magnus could have been. Probably even 'better' in some ways.
That said they'll need an answer to the 'waveguide problem' this approach have.
Last edited:
@indianajo , don't get me wrong, I currently have a Diatone DS-90. These speakers have a 16" woofer. And my speakers at any, even ultra-low volume, play very well (with a few nuances). I really like the low frequencies that these 16" woofers produce.
But the mids and highs don't sound very clean to my ears. I had a Focal Opal 615 speaker system, which played magically in the mids and highs. This is exactly what I want: to have such low frequencies as the Diatone DS-90, and the mids and highs the same or better than the Focal Opal 615.
About Dunlavy... I read an interview with Mr. Dunlavy. Even the reporter who interviewed him was surprised that his speaker system plays exactly as intended by the sound engineer, if infrasound is recorded, it will play infrasound. 🙂
But the mids and highs don't sound very clean to my ears. I had a Focal Opal 615 speaker system, which played magically in the mids and highs. This is exactly what I want: to have such low frequencies as the Diatone DS-90, and the mids and highs the same or better than the Focal Opal 615.
About Dunlavy... I read an interview with Mr. Dunlavy. Even the reporter who interviewed him was surprised that his speaker system plays exactly as intended by the sound engineer, if infrasound is recorded, it will play infrasound. 🙂
My experience of Dunlavy's is they are clinical sounding ( in the good way, revealing). But this wasn't without it's own set of issues: sweetspot wasn't very large ( and i hate that 'head in a vise' effect) despite an ( heavily) acoustically treated room.
I had similar issue with the SCM110a i lived with in the last control room i've spent lot of time in.
Both share the 'very wide spread' directivity. For the ATC i know from where the issue came from ( the mixing desk was 4m wide, 1,2m depth... iow ( in other words) a very nice reflector inducing comb filtering... and as i listened to them with and without desk in same room i know what the 'reflector' brought with it).
For the Dunlavy i think the issue was related to the fact they are so accurate that some issues usually 'hidden' are revealed in their plain glory ( phantom image is an illusion and if you are not located where you should... there is less illusion!) and they were designed for a limited listening spot since the start as i understood.
This is were i find Dave Smith comments about directivity interesting. A chance E. Geddes developed an approach to Constant Directivity with same radiated sound everywhere ( and W.Parham discovered a waveguide which had the same quality).
Indianajo, do you know which Meyer's loudspeaker it is? About your loudspeaker they seems to be a 'poor man' 's Tad Tsm2. I understand why you like them for acoustic source. You can't beat high efficiency when it comes to dynamics imho.
Theft. What a nuisance. I'm with you and i hope you'll have at least all your tools back at one point or another in close future.
I had similar issue with the SCM110a i lived with in the last control room i've spent lot of time in.
Both share the 'very wide spread' directivity. For the ATC i know from where the issue came from ( the mixing desk was 4m wide, 1,2m depth... iow ( in other words) a very nice reflector inducing comb filtering... and as i listened to them with and without desk in same room i know what the 'reflector' brought with it).
For the Dunlavy i think the issue was related to the fact they are so accurate that some issues usually 'hidden' are revealed in their plain glory ( phantom image is an illusion and if you are not located where you should... there is less illusion!) and they were designed for a limited listening spot since the start as i understood.
This is were i find Dave Smith comments about directivity interesting. A chance E. Geddes developed an approach to Constant Directivity with same radiated sound everywhere ( and W.Parham discovered a waveguide which had the same quality).
Indianajo, do you know which Meyer's loudspeaker it is? About your loudspeaker they seems to be a 'poor man' 's Tad Tsm2. I understand why you like them for acoustic source. You can't beat high efficiency when it comes to dynamics imho.
Theft. What a nuisance. I'm with you and i hope you'll have at least all your tools back at one point or another in close future.
Last edited:
Theft? I'm sorry, I do not write and speak English very well, but I sincerely sympathize with you. It's strange that things like loudspeakers are still being stolen...
By the way, somewhere I was not attentive, I didn’t notice, did you write what speakers are you using now?