VituixCAD v2

[kimmosto]

3.0.1.4 (2026-06-07)

Main

• Audibility warning and Audibility alarm lines added to GD & Phase graph.

Impulse

• Step and RMS curve smoothing reduced.
• Single IR file export filename changed.

[technodanvan]

Just a note kimmo, the 'download' link in your signature points to VituixCAD2.

[kimmosto]

3.0.1.5 (2026-07-28)

Main

• GD audibility warning and alarm traces in GD & Phase graph modified. See user manual for more information.
• Normalization of group delay traces to cycles removed because in practice audibility threshold is quite constant in milliseconds (not in cycles).
• Value of the selected (bold) curve at the mouse cursor is displayed as interpolated if there are only a few data points.

[draki]

I was reading this white paper on the SEAS Storeglimt project.

What caught my attention was the Z-offset value attributed to the drivers on the V-CAD crossover schematics.

Looking at the enclosure, it has a straight front, all the drivers are on the same Z-plane.
Knowing what I know about V-CAD so far, their Z offsets should all be zero..?

What am I missing here? I presume the SEAS team knows more that I do....

[Reet]

Seas basic design has not surpassed the skillset of a beginner DIYer. The response data appears to be minimum phase, data is only measured to 60 deg anyway, so CTA-2034 is already lost.

[kimmosto]

I guess it's simple on-axis design with measured phase, and timing of the drivers has been biased manually or automatically; IR peak start to 0 s or so... Relative acoustic distances are adjusted with Z.
Seas has anechoic so actual total to on-axis can be compared to simulation, and Z mm and component values can be adjusted if responses with reversed mid don't match. It's 3-way with dome mid so the risk of high directivity error is quite low.
Many DIYers capture more data, but don't have anechoic to verify anything below midrange. Everyone does what they can and want to do. Some can't, and some others don't want 😀

[kimmosto]

3.0.1.6 (2026-08-14)

Main

• Position of splitter bar between Drivers/Crossover/Room tabs and chart area is saved to user.config also when the main window was maximized.
• Target curves of SPL, CTA-2034, GD&Phase and Filter charts are hidden when Optimizer is closed with X button.
• The graphs initially show the target curves used by the Optimizer.
  Exception: Phase target is visualized with ref. axis and listening window target, even though the phase is not optimized.

I'm not entirely sure if the change in the logic for hiding/showing target curves is a significant improvement. Hiding has caused extra work after optimization - especially if the charts were published. I also received some feedback from target hiding, but my interpretation might not be correct.
One option would have been to add new button or menu command for hiding all target curves.

[Reet]

3.0.1.6 (2026-08-14)

Main

• Target curves of SPL, CTA-2034, GD&Phase and Filter charts are hidden when Optimizer is closed with X button.

I'm not a big fan of this change. I prefer the manual control over which traces are shown. Perhaps a toggle in the options to allow auto show/hide of targets or not would satisfy everyone?

[kimmosto]

I prefer the manual control over which traces are shown.

Targets stay visible and can be hidden via context menus as earlier if you close Optimizer with Hide button. New logic still shows and hides the targets according selected option in the Optimizer. I publish charts quite often so hiding target curves every time after closing the Optimizer has started to glitch.

This is probably not the final solution so new ideas are welcome. For example, both SPL and CTA-2034 targets are now shown when 'Preference rating' is optimized even if the slopes are not optimized (SL_ON, SL_LW, SL_PIR and SL_SP are unchecked). Target curves are not used it the slopes are not optimized.

[Reet]

Hide button is a bit odd UI interaction, but it works for my purpose. Thanks.

[kimmosto]

Hide button is a bit odd UI interaction, but it works for my purpose. Thanks.

I might have forgotten original purpose of Hide button. Perhaps it was for very small or low resolution displays and low CPU performance; to monitor optimizing and curves without keeping window open above chart area. Optimizer stays in memory and can run with hidden window. Closing with X stops and clears optimizer from memory. So it's kinda vintage feature.

[kimmosto]

This is done, but not yet packed and published:
Hide button is replaced with Close button (ends optimizing and closes the window) and 'Hide targets...' checkbox added to hide or leave curves visible w. What do you think?

[Reet]

Sounds good to me 😊

[kimmosto]

3.0.1.7 (2026-08-20)

Main, Optimizer

• Hide button replaced with Close button.
• Added 'Hide targets when closed' checkbox.

Main, Copy biquad coeffs...

• Eliminated useless error message when coefficients were exported before project was saved.

[Reet]

Hi Kimmo,
I am designing a large system that uses a pair of the same driver in series. For this type of system, VituixCAD shows the actual SPL and filter response that each individual driver sees, so the response is 6dB lower than what the total output is. Ideally in this scenario I would like to see a plot of the total output of the pair of driver in series, this will make it easier to determine where the specific crossover frequency is, as well for exporting IR response. Currently I need to boost the level by 6dB to represent the correct output level. It would be great if there were an option for this.

Another addition you might consider is a directivity plot in the diffraction tool. I understand that the simulation of directivity is limited in accuracy, but I think it can provide a bit more insight into baffle shape/size choices and the impact of large vs small roundover.

[kimmosto]

...plot of the total output of the pair of driver in series

Workaround is to mute others, save on-axis to overlay and then unmute. Needs redoing after changes. For example, 4-way with two side woofers:

Challenge is that certain driver name or index in driver list does not necessarily create a group or way. At least in theory the same driver data can be used for different ways/bands, though 0.5 ways and power tapering solutions are more typical cases where the same driver data is distributed with different filters into a single band.

[kimmosto]

directivity plot in the diffraction tool.

I use this workaround. Export off-axis data to temp directory with default name and feed to driver. Fast to redo after a change in diffraction settings.

[Reet]

Workaround is to mute others, save on-axis to overlay and then unmute. Needs redoing after changes. For example, 4-way with two side woofers:

Challenge is that certain driver name or index in driver list does not necessarily create a group or way. At least in theory the same driver data can be used for different ways/bands, though 0.5 ways and power tapering solutions are more typical cases where the same driver data is distributed with different filters into a single band.

Thank you, this is helpful. The design I'm working on is a WWMTMWW, the woofers are 8 ohm drivers in series/parallel arrangement, and the midrange is a pair of 4 ohm drivers wired in series. I am running some filter transfer function simulations using IR export and convolution via APO EQ, I just have to remember to add 6dB of gain to the IR export for a single driver.

For this design I have attempted the 0.5 way filter for the outer most woofer, however I didn't obtain a better result than just running all 4 drivers in parallel with a lower cost filter.

[kimmosto]

For this design I have attempted the 0.5 way filter for the outer most woofer, however I didn't obtain a better result than just running all 4 drivers in parallel with a lower cost filter.

Sounds normal. W-W distance is usually shorter than half wavelength at XO so 0.5-way filtering can be even harmful. M-M pair could have some 0.5 way features for upper mid-range if mid drivers are large causing too much vertical directivity right below upper XO. It also turns main axis slightly upwards to support listening while standing. Not much, but flexibility might improve depending on listening habits.

[Reet]

In my case, currently I have a crossover point around 500Hz, and W-W distance of 620mm for the inner pair of woofers, so the distance is longer than half wavelength, closer to 1x. Perhaps I should lower the crossover frequency, but am trying to balance the ideal frequency ranges for distortion/SPL performance of the drivers, asking for low frequency output from the midrange drivers will increase IMD through the midrange which I would like to avoid. Vertical directivity looks a bit of a "laser beam" at the moment, however horizontal, power & DI are looking good to me.

[kimmosto]

...W-W distance of 620mm for the inner pair of woofers, so the distance is longer than half wavelength, closer to 1x.

That is okay if target is to avoid early vertical reflections. Vertical beam will be quite narrow at some frequencies.
My target is usually to attenuate primarily the lowest floor dip at ca. 300 Hz with WW-WW pair.

This is also WWMTMWW but with cardioid woofers. Woofer responses are simulated ideal super-cardioid. Not exactly comparable with your projects but might give some idea what lower XO would do.

[Reet]

Thank you Kinmo, I appreciate your insight.