Knowledge of basic theory; components, drivers, acoustics, circuit topologies, radiator types etc. is valuable for sure. Close to mandatory in practice. Simulation of some parts could also be valuable or mandatory if information is needed to buy drivers compatible with selected radiator types, or environment is impossible for LF measurements due to special radiator type when also near field does not help. First question: should we format that theory package to look designing of quite complete speaker?
I could also design something without measurements by theory and datasheets or by traced curves or by measurements in IEC baffle or by listeing only, but never done and never will for myself and especially for customers.
I remember one thread on local forum with some drama when a first timer started to design his first and last diy speaker. Selected method was to calculate (=guess) few values for each XO component and create final design by swapping components and evaluating by listening. My head almost exploded from that, but he made it and left the building for good. Another question: should we write formal instructions specially for group which does not have (or not guaranteed) personal development plan and interest with diy speakers other than single shack?
An addition, why to list methods and tools without any comments or comparison what are features, strengths and weaknesses/limits/faults. That happens in both instruction and generic threads almost daily. Why to put first...second timer to situation where he/she has to test and compare with zero background info something he has not yet experience, knowledge and perspective. Second timers with the same history make recommendations, and the result is "international diy forum".
Some of you have probably noticed that my plan was not to introduce VituixCAD and recommended measurement methods on any English forum due to "status quo" and required time and nerves to cause significant reaction or a change. Eventually diypass started this, but it would help if I wouldn't have to do close to everything related to VCAD alone. That could include also criticism of status quo and traditional "ANSI tools and methods".
Sorry about OT - again.
You knew this day was coming, it was only a matter of time .. 
Request:
- Measurement of 3 axis enclosure with free & fixed positioning of microphone (Virtual 3D Anechoic Environment).
It's "just" an extension of the diffraction tool, but wouldn't that be something. Imagine the possibility of in advance be able to work on a virtual cabinet (free form or geometry) together with XO response adjustment (live XO adjustment) and Qa + Ql + Vent adj options.
Request:
- Measurement of 3 axis enclosure with free & fixed positioning of microphone (Virtual 3D Anechoic Environment).
It's "just" an extension of the diffraction tool, but wouldn't that be something. Imagine the possibility of in advance be able to work on a virtual cabinet (free form or geometry) together with XO response adjustment (live XO adjustment) and Qa + Ql + Vent adj options.
Not directly anything.
Maybe long vacation from diyaudio, htguide and VituixCAD, or skipping of November which is quite depressing time here close to arctic circle.
Or rebooting diy audio forums to utopia(/nightmare) where VituixCAD is only known crossover simulator and no Umik-1 or OnmiMic v2 to avoid waking up to Groundhog Day repeating the same crap. Not necessarily VCAD, but some kind of culture change where users stop recommending tools and methods to others by own laziness, limits, low motivation or investments, and start to think what is probably the best - most suitable, versatile, efficient/fast and growing by needs for the others (who is asking help). This needs much more than just quickly said "I use The Edge" or "The Edge exists".
I've been pondering what speaker project to tackle next. Some ideas are either for a more traditional 2 way or then a 2.5way or a lower cost 3way.
I would be willing to do "a whole project write up" including the measurements and Xover design with VCAD etc.
If Kimmo you would be willing we could do it so that you'd first validate what I come up with (process wise) offline and then once done we'd publish a process with the example write up as well as all the files (as you have for the two sample projects now).
This might help the community forward... but as one project does not cover the many different types of speakers people come up with, it might not still be enough.
I would be willing to do "a whole project write up" including the measurements and Xover design with VCAD etc.
If Kimmo you would be willing we could do it so that you'd first validate what I come up with (process wise) offline and then once done we'd publish a process with the example write up as well as all the files (as you have for the two sample projects now).
This might help the community forward... but as one project does not cover the many different types of speakers people come up with, it might not still be enough.
Hmm, how can one create a Poll in the thread. Would be interesting to know what type of design would be the one that interests most as a 'design walk through'.
It is possible somehow as this thread has it for example...
It is possible somehow as this thread has it for example...
But my personal problem is that I have some kind of mission to save as many as possible from poor tools and methods (imo or objectively) what few others recommend or just mention as a possibility.
I'm also painfully aware that there are many designers (also experienced) who actually have voluntarily selected to use poor tools and methods. Some message template to copy-paste from or written public instructions helps to avoid continuous repeating for beginners in infinite queue.
Maybe those could include some disclaimer: "this is not recommended crap method, but continue if you are sure and willing suffer consequences like a man..."
I have not purchased standard and not aware whether it has some recommendations for measurement distance, number of mic orbits or measuring on spherical surface all around the speaker with equally spaced measurement points.
Anyway, power response is mean of squared pressures on spherical surface around DUT as a function of f i.e. average intensity (f). If sound pressures are measured along one...two...multiple orbits - typically in horizontal and vertical planes knowing/assuming that speaker has certain main axis direction and vertical center line is not tilted left/right, each intensity (squared pressure) value is weighted to scale data from single...dual...multiple planes to spherical surface. Each weighting factor is slice area of sphere what that individual pressure measurement covers. Factors are quite close to sine(off-axis angle), but typically calculated with Archimedes' Hat-Box Theorem to improve accuracy, though result is never exact due to limited amount of measurement points.
Few examples of weighting factors with constant angle step visible here: post #1590.
With older VCAD versions number of different factor sets was close to infinite because angle step was not necessarily constant. Now simulator calculates with constant angle step due to response interpolation.
VituixCAD shows response of total pressure 10log(4pi) + 10log(weighted average of squared pressures). That is also convenient to separate power from axial response, listening window average and directivity index curves.
Last edited:
I like to do this one small step at a time. I see what you want to do is very kind to others, but not as kind to yourself.
VC is stable. It has features that are working. Buglist must be low. Now you can do what you think is best for you.
Thanks for trying with the standard.
I got impression that Pallas has it so maybe he could quote possible specs for count, directions and distance of measurement points.
Did you mean ANSI/CTA-2034A "Standard Method of Measurement for In-Home Loudspeakers." or something else? That exists, but I don't have it. Someone offered me spinorama factors for VCAD. Power & DI curves of front sector would be possible and easy to add, but difference to half space (or +/-60 deg) power & DI might be quite small.
I'm gona be a big baby now... my latest entry #1622 got ignored.. "Maybe, if we are all quite, he will go away" ... LMAO.
On another note. I will design and build my first loudspeaker next year. It will "only" be a 2-way design to keep things simple.
From my experience listening to the PMC Twenty5.22/24, I liked the Peerless HDS 830990 and the SEAS 27TFF seemed to do a decent job.
I am not completely sure I wana proceed with the 27TFF after listening to Accuton C25-6-158 and two other tweeters I will try is Viawave SRT-7 and Bliesma T25A-6.
I might end up trying all of does tweeters in order to be able to experience them properly and have some references.
I will also try ported (bass reflex), aperiodic and passive radiator as companions to 830990.
The software to do it all in will be VCAD. I am sure there are members who are curious,
so I will create a thread and keep it as an open source project. I will "advertise" the thread here but not clutter this thread. It will be an exciting project since it will contain so many aspects.
Driver combo, cabinet style and XO variations. Besides this, I need a proper way to measure things. CLIO 12 is interesting but not cheap, so I will explore the options available.
On another note. I will design and build my first loudspeaker next year. It will "only" be a 2-way design to keep things simple.
From my experience listening to the PMC Twenty5.22/24, I liked the Peerless HDS 830990 and the SEAS 27TFF seemed to do a decent job.
I am not completely sure I wana proceed with the 27TFF after listening to Accuton C25-6-158 and two other tweeters I will try is Viawave SRT-7 and Bliesma T25A-6.
I might end up trying all of does tweeters in order to be able to experience them properly and have some references.
I will also try ported (bass reflex), aperiodic and passive radiator as companions to 830990.
The software to do it all in will be VCAD. I am sure there are members who are curious,
so I will create a thread and keep it as an open source project. I will "advertise" the thread here but not clutter this thread. It will be an exciting project since it will contain so many aspects.
Driver combo, cabinet style and XO variations. Besides this, I need a proper way to measure things. CLIO 12 is interesting but not cheap, so I will explore the options available.
A full project write-up would be appreciated by many.
English is not the primary language of many interested in Vituix cad, therefore searching for answer is not easy. Then you find youtube video where the person is not using it properly, or at least not to it's full potential.
Please try to document not just what you do, but why you do it this way.
Please do not
Could you clarify what you meant. I was too confused (in addition to stressed & tired) to answer anything because message was mix of "measurement of 3 axis enclosure", diffraction simulation, XO adjustment and Qa + Ql + Vent adj options. Especially what is "measurement of 3 axis enclosure" if this is finally just simulation?
That standard (and other CTA standards) can be downloaded free of charge from the CTA site, though a sign-up is required.
LEAP has its EnclosureShop which would manage an imported 3D file to perform advance Diffraction Analysis. You can design the enclosure in a CAD software and import that shape which then becomes the basis for the diffraction analyze.Please do not
You can do 90-0-90 degree horizontal and Vertical measurements of the cabinet.
That got me thinking; What if VCAD also could do that, plus allow the user to work on the crossover at the same time, with the option of adjusting Qa + Ql and Vent or Port size and position
So that is the basis for my post and request
It's ridiculously easy to design enclosure without severe diffraction problems, and impossible to design simulator able to predict directivity of real life radiators with selections and parameter values user have to find and enter so that whole process would be significantly faster and easier and as accurate as measurement of actual prototype or final speaker.
Simulating of almost everything in loudspeakers except electrical crossover circuit assumed as linear has the same philosophical problem: simulation is waste of time if designer cannot provide parameter values giving decent accuracy compared to simple measurements.
Bad diffraction is rarely a design variable you could combine with any crossover design so that the result would be as good as designing good enclosure shape in the beginning. Diffraction is directivity feature so effects have to locate conveniently at possible XO freq range to have some changes to fix with XO.
I'm just a simple country boy who cannot cope any sci-fi simulations. In addition, too high integration of driver, enclosure, diffraction, crossover and room simulation is something I don't appreciate due to inaccuracies and more difficult and slow starting of a project, and therefore this program won't be like e.g. LspCAD 6.
Last edited: