VituixCAD

Well into wishes and suggestions not bugs with this one, so feel free to ignore. :D

I just had a look at the impulse response section for the first time - yes I had not looked at it before because I assumed that it was part of a basic built in MLSS measurement system and didn't realise it's purpose was actually to display the impulse response from the simulation instead - quite a handy tool. :)

(Yes it helps to read documentation too, but I find that trying to use a program without reading all the documentation first helps to uncover usability issues, and in this regard vituixcad is so good that documentation almost isn't needed)

I did run into a usability issue when trying to use the "axial response of driver" mode to view the impulse response of one driver though.

In this mode you must click on the driver in the schematic to choose which driver to show the impulse response for. I was then interested to see what the effect on the impulse response was if I deliberately de-tuned some of my notch filters, however I soon realised that the moment I clicked on a component to adjust the value the impulse response window no longer knows which driver is selected (because no driver is selected) and drops back to total axial SPL mode.

So any attempt to vary a component value to witness the result on the impulse response becomes extremely tedious especially without an overlay facility as every change keeps putting it back to total axial SPL mode, so by the time you change it back again its very hard to tell if a small change made was worse or better.

What would make sense to me is if there was a drop down box in the Source response group of the impulse response window that lets you select the driver directly there rather than having to click on a driver back in the schematic. Not only does this keep all the settings pertaining to the impulse response in the same window and make it quick and easy to select between drivers without swapping windows, it would make it possible to then go and start adjusting component values quickly and see the results without constantly deselecting the driver.

Just a suggestion. :)
 
Last edited:
^Original purpose of impulse response for each way was to produce data for convolver plugin or other active FIR engine. That has not changed with version 2. IR or step response of single driver (from SPL or filter magnitude) is quite indirect information for human and therefore difficult to adjust with crossover components.

I had (and still have) opportunity to add driver combo box to Impulse response window when driver list was added to Optimizer window. I did not add because driver selection from network is easy enough for IR file export. Separate driver list is much more important with Optimizer.

IR and step overlays would be easy to add and have some value with total SPL. I think that will is the next step.
 
^Changes are perfect, but I don't believe in 0.8/f. That threshold could be okay with earphones but speaker with that much GD is probably too weak already, though I can't prove it here and now. I have some threshold to make such statement to program if personal excess GD target is max. 0.3/f.
 
Last edited:
Hi Kimmosto,

Any chance we can see a baffle diffraction calculator that accounts for if a driver is not flush mounted (or rear mounted)? In my case where the driver is rear mounted, the cutout on the baffle causes significant diffraction that can be seen in the measurements. I would love to be able to see this in simulation.

Thanks!
 
bcodemz,
What is nominal size/diameter (") of the driver and cone material and thickness of the baffle? What is the lowest frequency where significant difference between flush/front and rear mount is visible? Have you tested rounding or bevels?

Diffraction model (DED) includes first order diffraction from cabinet edges with ideal piston radiator(s). In addition to baffle thickness and possible rounding/bevels, significance of rear/flush/front mount depends on directivity of the radiator. Directivity of other than piston and non-ideal cones is not supported in the simulation.
I suspect that it's not worth to add rear/flush/front mount alone without radiator's directivity. That would also be at least a bit out of scope because primary purposes of Diffraction tool are:
a) To convert near field responses to far field (including directivity to full space) up to splicing frequency in Merger tool. Everything above that should be measured as far field to get reliable data for crossover simulation.
b) To find good location for mid and HF drivers.
 
See this video if you have wondered what is "Datasheet import" in Updated field of driver table in Enclosure tool. That is my private tool for updating and adding new drivers to setup package and online database. Local database can be updated with online database, and the latest database is installed to local from setup.exe if user\Documents\VituixCAD\Enclosure\VituixCAD_Drivers.txt has been deleted or renamed.

The best way to get new drivers to online database is just asking if decent and uniform pdf datasheets are available. You can also copy new rows to text file, attach to e-mail and send to me. Soloing with own local is not so productive and recommended with freeware.
 
Just another Moderator
Joined 2003
Paid Member
importance of good measurements

The datasheet import is pretty cool kimmo! You really have got everything in this program!!

I did a new round of measurements on Sat. 0 through 180 deg in 10deg increments for each of the three speakers in my MTM. I used REW in dual channel mode. The results are much better!

The only problem was I changed my measurement setup a bit and got a dip in the tweeter repsonse around 9Khz that wasn't there before, but I'll just ignore that.

The power response and polar map look so much better than the last lot of non equally spaced measurements I did that only went to 90deg. I should have good data now for doing the crossover redesign.

Attached is the sim of the current crossover with the new measurements.

Tony.
 

Attachments

  • MTM_vituix2_Six-pack.png
    MTM_vituix2_Six-pack.png
    395.1 KB · Views: 280
I was thinking about the UI issue of no longer having easily toggled "ways" as Version 1 did, which leads to having to break and reconnect circuit lines to emulate ways.

Instead of adding multiple generators, perhaps a cleaner and easier to implement approach would be to keep the single generator and simply add "switch" elements to the circuit designer which can be toggled on and off with a row of checkboxes like the old ways checkboxes, located somewhere on the crossover tab outside of the scrolling circuit area.

To emulate ways the user then just connects the switch elements in the appropriate locations to isolate each crossover section from the generator when the switch is open.

This can sort of be done now by abusing the open and short context menu options of a resistor, however this is a bit ugly as it means a resistor that does not really exist will show up in the parts list, and it requires right click manipulation on the scrolling circuit area rather than a simple checkbox outside of the circuit area.

Further, the switches and associated checkboxes could have a more general purpose use beyond just convenient activation of ways, such as switching in or out two different versions of a crossover or different sections of a crossover for comparative purposes, emulating a "cut/boost" switch in an L-pad and so on.

If a single pole changeover version of the switch is also available it would make it even more flexible while using a minimum number of switches.

Just a thought. :)
 
Instead of adding multiple generators, perhaps a cleaner and easier to implement approach would be to keep the single generator and simply add "switch" elements

Very early test versions had switch in component menu but I didn't leave it because 'Active gain/delay buffer' or any other series component such as resistor (0.001 Ohms) enables on/off-toggling of "Ways" with Open/Short selection.
The main issue is still impedance response and output power curves for each "Way". That would need some linking association between selected crossover node and curve. Many spice simulators have probe symbols for current and voltage. VituixCAD could also have P/U/I/Z probes to link nodes to graphs.
 
Rev 2.0.3.10 (2018-07-10)

* Shelving frequency at middle gain checkbox "finally" added to Options window. If checked, nominal frequency of generic active shelving filters (not OPA) is at middle gain in dB. If not checked, shelving frequency is nominal frequency of the low/high-pass part.
Check with dbx DriveRack, Hypex AS/PSC/DSP/DLCP, miniDSP (Basic), Xilica XA/XP. Uncheck with Behringer DCX, FourAudio PPA.
Setting is saved in project file (vxp).

* User manual in German updated yesterday. It's compatible with rev. 2.0.3.9.
 
At least Band pass type 1 would be doable quite fast because it has single port which can be replaced with passive. Other types are more difficult due to limited UI space with current layout.
Yes, I was mainly speaking about the type 1.
Higher types aren't very cost effective anymore anyway.
(although maybe still interesting)
 
Last edited:
Rev 2.0.3.10 (2018-07-10)

* Shelving frequency at middle gain checkbox "finally" added to Options window. If checked, nominal frequency of generic active shelving filters (not OPA) is at middle gain in dB. If not checked, shelving frequency is nominal frequency of the low/high-pass part.
Check with dbx DriveRack, Hypex AS/PSC/DSP/DLCP, miniDSP (Basic), Xilica XA/XP. Uncheck with Behringer DCX, FourAudio PPA.
Setting is saved in project file (vxp).

* User manual in German updated yesterday. It's compatible with rev. 2.0.3.9.
thanks for Your work
Q function is now missing in new version for 2nd order circuit blocks in LIB
And it would be nice if you added first order circuit blocks in LIB
thanks again
 
Q function is now missing in new version for 2nd order
first order circuit blocks

Both versions have 2nd order Bessel, Butterworth and Linkwitz-Riley. In practice 2nd order with variable Q is the same as adjusting expected/estimated load impedance (Z parameter), though adjustable Q factor could be more logical for users. Anyway, I will add "Adjustable Q" option into Passive 2nd order LP & HP.

1st order LP & HP are doable for sure :D
 
As an extensive user of VituixCAD I must confess I don't like new graphical filters interface. It is not convenient to frequently deleting and adding new connecting lines inside of the circuit while altering count and placement of the active filters blocks connected in series.

Wouldn't be better to create kind of 'intelligent' wires from the generator to the first filter block and from last filter block to the driver? Such an intelligent wire should always stay connected on its both ends regardless of movements of the blocks, driver and generator...

Or another idea - adding checkbox to the particular (first or last) filter block: 'always connected to source' and 'always connected do load'.
Or assume that first block from the left stays always conneted to source, last from the right stays always connected to load and then forget hassling with these stupid wires. :)

I also vote for command 'replace filter block'. Selected filter can be quickly changed for example from LT to shelving etc.

I also don't quite understand how to conveniently switch over 'ways' in this new GUI. Old interface was so much better for active multi-driver designs in terms of rapid on/off/invert/time/level checks.
 
Last edited: