Hello,
I have a general question about the accuracy of VituixCAD. I have been using it for a month to simulate different type of speakers, mainly two-ways and three-ways. I exploited all the important tools such as crossover, diffraction and enclosure simulations. I chose drivers from Dayton so I didn't need to create my own FRD and ZMA files. My question is: how reliable is it as a simulation tool? Are there any specific areas where it might produce errors? Would you build a speaker only according to the data from the simulation, without measuring the drivers on the actual baffle/enclosure?
Thanks!
I have a general question about the accuracy of VituixCAD. I have been using it for a month to simulate different type of speakers, mainly two-ways and three-ways. I exploited all the important tools such as crossover, diffraction and enclosure simulations. I chose drivers from Dayton so I didn't need to create my own FRD and ZMA files. My question is: how reliable is it as a simulation tool? Are there any specific areas where it might produce errors? Would you build a speaker only according to the data from the simulation, without measuring the drivers on the actual baffle/enclosure?
Thanks!
My personal answer to that is sometimes.
Because sometimes its really near simulation in reality mesaurements, and sometimes not so near.
Don´t know why because i do mesaurements-file FRD and ZMA the same way.
My "badest" time was with a 4 way, one 20 ohm one 16 ohm and two 8 ohm drivers.
And i use both VituaXcad and Xsim.
There are enough posts that show sim vs reality. Nor did i say it is not reliable.
Have you studied the vituixcad documentation including help ?
It actually gives a quite good indication, but how accurate do you want to be?
How close is it simulating your specific cabinet shape?
How accurate is the driver data privided by the manufacturer?
How do you know it is accurate enough?
Have you studied the vituixcad documentation including help ?
It actually gives a quite good indication, but how accurate do you want to be?
How close is it simulating your specific cabinet shape?
How accurate is the driver data privided by the manufacturer?
How do you know it is accurate enough?
I expect most of the variation will be in the drivers. Drivers often have 20% error in the Fs and other parameters due to variation in the spider stiffness. Of course this also changes with break in of the driver and eventually aging. Drivers have been found to shift parameters back close to the new state when left sitting for just a few months. There will be internal box standing wave modes and panel resonances as well that I don't think Vituixcad models. The great thing about DIY speaker design is that it's very complex if you take into account every variable. So you never run out of things to do. I really recommend prototyping with digital crossovers as you can quickly and easily handle all sorts of things that would be prohibitively expensive or impossible with passive crossovers.
Hi, VituixCAD is reliable and accurate in sense it's algorithms do math. Question is only how close the math used is to reality and I bet it is close enough otherwise nobody would use it. I bet kimmosto has made sure that it is best tool available.
Question of accuracy is mostly from the data you input, if you load bad data in the output is accurate to the data but not to reality. Also, since it is quite flexible program suite it is possible to do mistakes operating it, which can lead to inaccurate/false results.
For example the diffraction tool is simplified, only front edge effects not the whole box, no intricate details like other cones on baffle, no second order effects etc. it's something one can use to draft things for example. Add bafflestep to nearfield measurements which is something people without anechoic chamber can utilize, and so on.
It is very good tool, and accurate, you just have to know what you feed into it, and what you do with it, to get results you can trust. You could use bad/approximate data as well, experimental data, anything, if you know what you are doing and magnitude of error the data yields and can live with it.
In general, enclosure and diffraction tools and manufacturer data can be used to help with design process, but there is going to be some error. You must build speaker and measure it well, process the data well, set it up into the program well, design crossover and implementit as designed. Now if you measure the whole thing it should match pretty close to the simulated response.
All the tools and features can be used during the process, but you'd always need to measure the thing at some point, likely many times. Basically all you need to do is to understand what you are doing and VituixCAD will help you do it 🙂 Start from measurement manual, also try think why the measurement manual steps are what they are, it should help you build intuition / understanding what's this stuff all about, help you forward.
Question of accuracy is mostly from the data you input, if you load bad data in the output is accurate to the data but not to reality. Also, since it is quite flexible program suite it is possible to do mistakes operating it, which can lead to inaccurate/false results.
For example the diffraction tool is simplified, only front edge effects not the whole box, no intricate details like other cones on baffle, no second order effects etc. it's something one can use to draft things for example. Add bafflestep to nearfield measurements which is something people without anechoic chamber can utilize, and so on.
It is very good tool, and accurate, you just have to know what you feed into it, and what you do with it, to get results you can trust. You could use bad/approximate data as well, experimental data, anything, if you know what you are doing and magnitude of error the data yields and can live with it.
In general, enclosure and diffraction tools and manufacturer data can be used to help with design process, but there is going to be some error. You must build speaker and measure it well, process the data well, set it up into the program well, design crossover and implementit as designed. Now if you measure the whole thing it should match pretty close to the simulated response.
All the tools and features can be used during the process, but you'd always need to measure the thing at some point, likely many times. Basically all you need to do is to understand what you are doing and VituixCAD will help you do it 🙂 Start from measurement manual, also try think why the measurement manual steps are what they are, it should help you build intuition / understanding what's this stuff all about, help you forward.
Last edited:
When I complete a speaker project, I always measure the complete system. If my final measured results do not match the VituixCad simulation within +/- 0.5 dB, I consider that an unacceptable error and I go looking for the cause... which is usually a simulation error on my part, but sometimes it is a sloppy measurement (also my fault).
The diffraction tool models the baffle shape and edge, the driver spacing and positioning. It does not model the depth of the cabinet or the shape of the back of the cabinet. That is OK, because these effects are of secondary importance. The baffle diffraction tool has enough resolution to predict the sound power response, the directivity index, and the early reflections response. That is good enough for my purposes...
The diffraction tool models the baffle shape and edge, the driver spacing and positioning. It does not model the depth of the cabinet or the shape of the back of the cabinet. That is OK, because these effects are of secondary importance. The baffle diffraction tool has enough resolution to predict the sound power response, the directivity index, and the early reflections response. That is good enough for my purposes...