VituixCAD v2

[technodanvan]

Kimmo,

I observe ASR maybe weekly (at best) and find the reviews there to provide fairly valuable data. That said, in my (albeit very limited) opinion I have to agree with you. Opinions run strong at ASR and I'm not particularly confident they are varied enough to really provide valuable information. Frankly, they do seem afraid of total data, something with which they are provided gobs of (objectively) and in a very limited manner (subjectively). I admire what Amir & Co. is trying to do but the approach does indeed seem limiting...and common users are not helping to improve that.

This is from someone who appreciates hard data, and the processes to gather it, and believes anything heard can be measured in some way, shape, or form.

I'm still barely a novice in using your software but I believe, based other reviews, mastery of what you have created will help ensure mastery of the subject matter. Someday I'll be there, hopefully.

I guess what I'm saying is, continue what you're doing and what you believe in. For what it's worth, I think you are on the right path.

[Quza]

Do you have any recommended reads after 'Master Handbook of Acoustics' and 'Sound Reproduction'?
Because that's the level I think ASR is on. Where to go from there?

[kimmosto]

Rev. 2.0.75.4 (2021-07-28)

Main, Preference rating
* Added On-axis slope target text box to allow optimizing ON slope other than 0.
* Added Listening window slope target text box to allow optimizing LW slope other than -0.3.

Main, Drivers tab
* Added Color column to Drivers table. Colors are compatible with legacy version 1.1. Modification of color, width and dash style may be allowed later.

[kimmosto]

Do you have any recommended reads after 'Master Handbook of Acoustics' and 'Sound Reproduction'?
Because that's the level I think ASR is on. Where to go from there?

ASR is just a platform, and individual member posting there could be anything between rookie and chief designer of significant (speaker) manufacturer or university professor. Measurements for reviews cover few sections from CTA-2034-A. It's not DIY or audio R&D forum so correct direction is out if speaker areas on this forum or diyaudio are more suitable. Books you listed are quite thick and extensive, but not so deep and adequate for designing. Some other sources such as papers at AES e-library and Acoustics Sound fields and Transducers by Beranek&Mellow are too deep in my opinion. Dickason's cookbook is practical for basic DIY.

[kimmosto]

Modification of color, width and dash style may be allowed later.

Rev. 2.0.75.5 (2021-07-30)

Main
* Added color palette button to the right side of driver list in Drivers tab for adjusting trace color, width, dash style and visibility of each driver model. Separate settings for magnitude and phase responses.
* 'Hide traces' command removed from context menu of crossover schematic. Visibility of driver traces are controlled via Drivers tab.
* Added 'Custom trace properties' checkbox to Options window. If checked, properties of chart curves are read from user config when program starts.
* Trace properties of Power&DI chart are not saved to project file (vxp) anymore.

[kimmosto]

'Hide traces' command restored to context menu of crossover schematic.

[Reet]

Thanks! The ability to save the trace options was a much needed addition, it was real tedious to change trace options from defaults every time the program opens.

Just something to think about for future development, but a "dark mode" option would be a great addition IMO. charts and schematics on a black background.

[Reet]

Playing around with it a bit, I think the option to save the trace options should simply be an "always on" feature, but then the program needs a "reset to default" button on the trace config windows.

[kimmosto]

The ability to save the trace options was a much needed addition, it was real tedious to change trace options from defaults every time the program opens.

I have never needed that. It's just an opinion related to personal history with other tools or possibly color blindness. Opinions may also be limited and "wrong" e.g. if never simulated line arrays. Tweaking with individual colors need more work than simulation itself. Just waste of time. So default colors and widths are still recommended.

Just something to think about for future development, but a "dark mode" option would be a great addition IMO. charts and schematics on a black background.

Program supports high contrast themes with dark background. E.g. dark traces on default white background are system colors which turn automatically to light when background is dark enough.

[kimmosto]

I think the option to save the trace options should simply be an "always on" feature, but then the program needs a "reset to default" button on the trace config windows.

I know. Problem (=work) at the moment is that default colors coded in the program are set when each form is loaded. Duplicate palettes are needed in order to switch to default colors with single button in Options.

[Reet]

Program supports high contrast themes with dark background. E.g. dark traces on default white background are system colors which turn automatically to light when background is dark enough.

Interesting, so I tested and see that black backgrounds do occur for "high contast" mode, but that mode is a bit extreme for normal use for me. What I am referring to is the "dark mode" setting in Windows shown here, which unfortunately does nothing in VituixCAD. I prefer this dark mode as more of a personal preference, while the high contrast mode is more geared towards accessibility features for people with vision problems.




Of course it doesn't have to be tied to the windows options. A manual selection in the options for background colour would be perfectly okay for me.

[Quza]

Thanks for your book recommendations!

I have never needed that. It's just an opinion related to personal history with other tools or possibly color blindness. Opinions may also be limited and "wrong" e.g. if never simulated line arrays. Tweaking with individual colors need more work than simulation itself. Just waste of time. So default colors and widths are still recommended.

While VituixCAD probably isn't made for this purpose, playing around with crossover settings for speaker arrays is very intuitive. When I need good cancellation to the sides, but not the back, I increase the line width of 70-110° angles and play around. Thanks for the 'custom trace properties' feature!
/edit: oh, nevermind, this is for the drivers only. Whatever, it only takes 10 seconds to change

[kimmosto]

"dark mode" setting in Windows shown here, which unfortunately does nothing in VituixCAD. I prefer this dark mode as more of a personal preference.

'Dark app mode' switch is quite new feature, and not widely supported in applications. Basically it's just a bit in Win32 registry for "legacy" platforms such as .NET Framework 4 of VituixCAD so (some) programmers are on their own. I've already done something simple: background of charts and XO schematic turn to dark and some colors change when 'Dark app mode' is switched on and program is restarted.

[kimmosto]

oh, nevermind, this is for the drivers only. Whatever, it only takes 10 seconds to change

It's for almost every chart in the program. Few charts such as Impulse response and Calculate T/S parameters have fixed properties. Convert IR to FR has possibility to save session with trace properties and overlays as a file so it's the most featured. Anyway, restoring default properties is moving from Options window to View menu in the main program.

[kimmosto]

Rev. 2.0.75.7 (2021-08-01)

Main
* Added View->Default colors command. Restores default colors, widths, dash styles and visibilites of chart traces in the Main program, Enclosure, Merger, Calculator and Diffraction tools.
* 'Custom trace properties' checkbox removed from Options window.
* Added simple support for 'Dark app mode'. Background of charts and XO schematic turn to dark when 'Dark app mode' is on and program is restarted.

[Reet]

'Dark app mode' switch is quite new feature, and not widely supported in applications. Basically it's just a bit in Win32 registry for "legacy" platforms such as .NET Framework 4 of VituixCAD so (some) programmers are on their own. I've already done something simple: background of charts and XO schematic turn to dark and some colors change when 'Dark app mode' is switched on and program is restarted.

Thanks! This is a good change for me, my only recommendation would be to also change the default "blue" trace colour to cyan or green for better visibility on the black background.

[kimmosto]

Dark app mode is just a mess. Few problems and colors fixed to revision 2.0.75.8 (2021-08-02).

[kimmosto]

Question: Normal phase, minimum phase and excess phase responses are now in SPL chart. Is GD & Phase chart better place for all of them? It would be more logical, but traditionally total on-axis SPL and phase have been in the same graph.
How about duplicating normal phase and moving excess and minimum phase curves to GD & Phase chart? I don't like to duplicate all three.

[kimmosto]

Rev. 2.0.75.9 (2021-08-03)

Main
* Added 'Dark mode' checkbox to Display group in Options window.
* Detection of High contrast and Light/Dark app mode removed. Toggle Dark mode in Options if needed to get suitable colors for selected theme.
* Added Dark mode support to Top and Side images in Room tab.

Also wires and components in crossover schematic are now lighter while dark mode.

[tktran]

Hi Kimmo,

You have a user guide for REW, ARTA and SoundEasy for taking measurements of drivers.

Do you have a quick start guide for multichannel sound-cards for active equalisation?

[kimmosto]

Rev. 2.0.75.10 (2021-08-03)

Traces window
* Added '1 px' button. Changes line width of selected/all rows to 1 px.
* Added '2 px' button. Changes line width to 2 px.
* Added 'Solid' button. Changes dash style to Solid.
* Added 'Dash' button. Changes dash style to Dash.
* Added 'Phase Dash' button. Changes dash style of phase responses to Dash.

[kimmosto]

Do you have a quick start guide for multichannel sound-cards for active equalisation?

No I don't.

[Reet]

Manual selection of "dark mode" in the options is perfectly okay with me. I do agree however with the user on DIY Audio that suggested the grid colour is a bit distracting in dark mode. I would recommend the grid colour on the schematic to be a more subdued grey - pick the grey shade used for the chart graticule divisions (#6d6d6d).

[kimmosto]

...the grid colour is a bit distracting in dark mode. I would recommend the grid colour on the schematic to be a more subdued grey - pick the grey shade used for the chart graticule divisions (#6d6d6d).

Same answer to here:

I don't keep grid visible because it is useless and disturbing imo so grid color was not checked and fixed in earlier revisions and builds.

Grid color is DimGray while dark mode and Gray while light mode in the latest build 2.0.75.10 (2021-08-04).

[Reet]

Same answer to here:

I don't keep grid visible because it is useless and disturbing imo so grid color was not checked and fixed in earlier revisions and builds.

Grid color is DimGray while dark mode and Gray while light mode in the latest build 2.0.75.10 (2021-08-04).

I'm okay with it as it is in 2.0.75.10. Thanks for taking the time to make dark mode happen, I'm a fan, and the 6pack exports look real nice too.

[kimmosto]

Rev. 2.0.75.11 (2021-08-04)

Main
* Color of crossover grid changed to more subdued. Display, bitmap export and SVG export to use the same color.
* Dash style of overlays changed to DashDot to distinguish from phase responses set with 'Phase Dash' button in Traces window.

[Reet]

Something appears a bit odd with the the traces in this latest version. I have 1 project that isn't displaying the woofer trace even though it is checked in the drivers tab, I have to manually check it in the SPL, phase, and filter charts. I'm not sure why this is, it is a 3-way project and I am able to load a different 3-way project without issues. Checking to display the traces and then saving the project is not resolving this issue. This project has 2 crossover variants, and only R1 is not displaying the woofer trace. In fact all I have to do is select variant R2, then go back to R1 and the traces are turned off again. I've attached the project files here for your reference.

Northern Pikes 03 Aug 2021.zip

Something else I noticed while troubleshooting this, is that if I rename a driver, it isn't being renamed in all crossover variants, only the variant that is active is being updated. It's a bit unusual to change the driver name after multiple crossover variants are created, generally the driver names are worked out at the start of the project, but this is something that should be corrected regardless.

[kimmosto]

isn't displaying the woofer trace even though it is checked in the drivers tab,

Traces of individual driver instance are hidden. Right click above driver symbol and uncheck 'Hide traces'.
Show column in 'Driver traces' window controls all instances of driver model but does not override setting of individual driver instances. This is confusing for sure because show/hide traces status for each driver instance is not visible in crossover schematic.

..if I rename a driver, it isn't being renamed in all crossover variants

No surprise. Variant feature is new compared to driver renaming. I'll try to improve that.

[kimmosto]

I'll try to improve that.

Done to rev 2.0.75.12

[Reet]

Traces of individual driver instance are hidden. Right click above driver symbol and uncheck 'Hide traces'.
Show column in 'Driver traces' window controls all instances of driver model but does not override setting of individual driver instances. This is confusing for sure because show/hide traces status for each driver instance is not visible in crossover schematic.

Ah, I recall the introduction of the hidden "hide traces" option, and must have selected it for the woofer in this project to test it's functionality. The confusion here is that it's in conflict with the "show" column for the driver. I can hide the traces by right clicking on the driver, but the "show" checkbox for the driver remained checked. I can go to the traces menu on the SPL chart and show the driver, but the "hide traces" option remains selected so when I move to a different crossover variant and back the "hide traces" option is taking priority and unchecking the SPL trace for the driver. Rather confusing behaviour.

The solution here is that the "hide traces" should not be an independent option, but simply a convenient toggle that unchecks the "show" column for the driver. There would then be no confusion when "show" is checked for the driver as well as "hide traces" as this situation could not exist. This of course would turn off the driver traces for all crossover variants, but I am okay with that.

[Reet]

Done to rev 2.0.75.12

Perfect

[kimmosto]

The solution here is that the "hide traces" should not be an independent option, but simply a convenient toggle that unchecks the "show" column for the driver.

That would not work because scope of Show column and Hide traces is not the same. Show column is for driver model i.e. all driver instances in crossover using the same response data. 'Hide traces' in context menu is able to hide/show traces of individual driver instances. For example Jack has line array with single driver model having 12 driver instances in crossover. Hide traces in context menu enables hiding 1...12 of 12 driver instances. This may be helpful e.g. if crossover has three different groups (4 drivers each) for power tapering, and designer wants to look traces of single group at a time.
'Driver traces' window was introduced few revisions ago. Before that hiding of all driver instances required selecting all and checking of 'Hide traces' in context menu. One possible modification is to prevent using Show column, but status of show/hide traces is still difficult to realize as long as driver symbols on XO schematic don't have any visualisation for it.

[Reet]

status of show/hide traces is still difficult to realize as long as driver symbols on XO schematic don't have any visualisation for it.

Ok, I understand now. Is it much work to place an icon on the schematic, a little eyeball or something next to the driver when the "hide traces" option is selected for it? There really needs to be some indication that the feature is enabled besides in the driver context menu. There is visual indication for all the Open, Short, Invert, Mute selection, just need one more

[kimmosto]

Updating OSIM buttons cause performance problem so I'm not big fan of them. For example H or H~ text inside driver symbol wouldn't be difficult at all.

[kimmosto]

2.0.75.13 (2021-08-06)

Main
* Show and Ovl columns hidden in Driver traces window.
* Added H character into driver symbol in XO schematic if traces of driver instance are hidden.
* Added H button below crossover schematic to show/hide chart traces of selected driver instances.

More versatile option is available without overlapping.

[kimmosto]

2.0.75.14 (2021-08-13)

Enclosure
* Added optional acceleration traces to Excursion chart. Show Acceleration checkbox to context menu.

[haraldo]

hi guys, this looks insanely great, but... one very basic question, I don't see where I add new driver types
Could you help me there please?
Missing the Peerless STW-350F-188PR01-04

[Reet]

hi guys, this looks insanely great, but... one very basic question, I don't see where I add new driver types
Could you help me there please?
Missing the Peerless STW-350F-188PR01-04

Look for the green + button in the enclosure tool.

The help file is here:

https://kimmosaunisto.net/Software/VituixCAD/VituixCAD_help_20.html

[haraldo]

Look for the green + button in the enclosure tool.

The help file is here:
https://kimmosaunisto.net/Software/V...D_help_20.html

Thx, I expected it somewhere else, but found

[kimmosto]

2.0.76.0 (2021-08-18 )

Merger
* Scale, Delay and Invert controls are removed from High frequency part. These are available for Low frequency part only to simplify user interface.
* Search (binocular) button removed from HF part.
* Added BS (baffle step) checkbox column to Low frequency response list. Checked with cones/ports located in front baffle creating baffle step to on-axis direction.
* Added 'Polar pattern' list box to allow forcing textbook directivity for near field LF responses if directivity of far field HF responses is too unreliable at LF due to window function, and directivity at LF is known textbook pattern.
* Added 'below Hz' text box for adjusting highest frequency using textbook polar pattern. Directivity is blended within textbook range and transition frequency.

Note that old Merger projects (vxm) give wrong result due to missing BS column in LF response list and possible scaling of HF part. Open old Merger project, check correct BS cell(s) and adjust LF scaling if needed until result is okay. Save vxm.

[kimmosto]

New build with yesterday's revision number.

2.0.76.1 (2021-08-19)

Merger
* Fixed value update of Diam and Area columns in LF response list.

Options
* 'Klippel NFS' renamed to 'Swap planes'.

[Reet]

* Added 'Polar pattern' list box to allow forcing textbook directivity for near field LF responses if directivity of far field HF responses is too unreliable at LF due to window function, and directivity at LF is known textbook pattern.

I am wondering if you can expand a bit on the use cases for this feature. I compared the result of a small driver where I had good off-axis data, and the result of "from HF part" was somewhere in between omni and cardioid response. Should I expect good near and far measurements to track one of these textbook functions?

In one scenario I have measured a 10" woofer that is in a tower speaker but at floor level. As a result the floor interaction is an important aspect of the driver response. It took a fair bit of effort to get what I would consider a good response on a single axis, that I simply used the diffraction tool to generate the off-axis data. Would using this textbook polar pattern function provide a greater degree of accuracy in this case?

[kimmosto]

ARTA is very sensitive to (drifting) D.C. offset and crosstalk in dual channel mode. Directivity information could be quite random - almost unacceptable at LF:


Also different window functions produce different result and error due to DC offset by LF content multiplied with window function:
Cosine window


Hanning window


Tukey 0.50 window


All have some error, but Hanning could capture really good directivity information without forcing to any textbook pattern. That is case dependent so no guarantee. Merger tool has been using directivity information of far field measurements also for near field LF measurements assuming/hoping that there's no (drifting) D.C. offset or crosstalk and window function is smooth producing logical directivity information at LF. But we may use other than Hanning, Nuttall etc. to avoid error in amplitude response at LF, or soundcard could be too bad for ARTA. So forcing to textbook pattern at LF is valuable if designer "knows" that radiator is e.g. omni up to 50 Hz. No need to simulate directivity with Diffraction or select certain window function. Just force pattern at LF and blend with measured directivity up to merging frequency. This is cheap but very fast trick which saves quite much work compared to more accurate simulations.

This is the first (bad) measurement with ARTA forced to omni below 50 Hz which is presumably very close to fact in that case (narrow vented enclosure with 6.5" driver). Merging at 400 Hz.


Gradient radiators are more complex because directivity and polarity to certain off-axis angle are usually not constant, and actual directivity of cardioidish is usually very difficult to "know" without actual measurements in large space. Cardioidish patterns are there but they may not be very valuable.

Forcing to textbook pattern is okay also for cones and ports very close to floor. That case could be quite compatible with 'No baffle loss' options because floor reflection is able to compensate baffle loss. Another option is to simulate on-axis with floor reflection with Diffraction tool and load baffle effect response to Merger.

[kimmosto]

Users have been asking do they need measurement data in vertical plane, negative angles, smaller than 10 deg steps etc. ASR Directiva project has measurement data for some conclusions, applicable to circular radiators in horizontally symmetrical small box.
Solid lines = both planes and negative angles, step 5 deg, totally 143 responses / driver.
Dash-dot lines = positive angles in horizontal plane only, step 10 deg, totally 19 responses / driver.

[Reet]

ARTA is very sensitive to (drifting) D.C. offset and crosstalk in dual channel mode. Directivity information could be quite random - almost unacceptable at LF:

Thanks for the detail. If I understand you correctly, when measuring off-axis data, The far field measurements should converge at the same point as frequency decreases towards 0Hz, assuming a truly omnidirectional response. I am using a Steinberg UR22mkii for measurement which is a fairly competent device, so I have little concern of DC or cross-talk in the measured data.

For my own comparison, I took one recent driver I measured, and Loaded the far-field measurements, normalized like your plot shows. Unfortunately the low frequency response slope is far from the ideal in your example above.



I am very happy with the crossover that resulted from these measurements, but wanted to compare the results to the "omni" textbook function to "fix" the low frequency slope. Below is a power response comparison of the merged responses from the text book function versus the measured "from HF part". The dashed lines are the measured data, solid is omni textbook.

There is some change to the power response below 300Hz. This driver is a small 120mm midbass, so probably not the best example for low frequency results.I used this driver as a midrange in a 3-way, and in this use case, the difference was absolutely zero in the complete design power response.

Of course, drivers in a cabinet aren't completely omni even at 20Hz, so maybe the cardioid text book function would be a more fitting comparison here.