Are you talking about main program or Enclosure tool or something else?
Main program saves visible-status of curves which are not so mandatory. For example excess and minimum phase and excess GD. Visibility can be controlled with Show... commands in context menu (right click). "Mandatory" curves i.e. traces required for designing speaker/crossover are initially visible with default width and color. Visible-status will not be saved.
Enclosure tool selects visible curves by selected radiator type so there's not much sense and possibilities to filter curves by user's feelings and opinions. SPL curve has quite many traces but you can hide extras.
Last edited:
Why not. I've designed three speakers with Purifi W04 and passive radiators
Note that measurement and design instructions are very important with multi-driver projects.P.S. 4x W04 work okay with 6x passives so vented would be more rational.
Measure full range with with MLS or white noise, or measure in dual-channel mode with 1st order high pass filter with program having cross-correlation with frequency response of reference channel. For example ARTA is perfect for that. Some programs with full cross-corralation have measurement signal with high-pass slope.
In practice you can measure domes, compression drivers, magnetostatic planars / AMTs with full range sine sweep. Just reduce voltage to e.g. 1 Vrms and do not measure too far.
True ribbons need some protection. One common option is to design fixed passive protection with 2nd order high pass and include that to measurements and design.
Anyway, VituixCAD has features to merge measured HF and LF with manual extrapolation and minimum phase extraction. Calculator and Merger tools are needed. Not very simple and fast but doable if no other way. This method tries to capture also directivity at LF.
One more important thing is that truncated measurement range does not produce accurate measured phase because impulse response is also cropped. So measurement with REW from 700 Hz is not necessarily very accurate at LF ~700-1.5k.
Why not. I've designed three speakers with Purifi W04 and passive radiators
P.S. 4x W04 work okay with 6x passives so vented would be more rational.
Did you design the drivers into one enclosure with one big compartment, or did you design them to be each in separate chambers?
Are there benefits of one of those designs over another?
Main program. SPL, Power & DI, GD & Phase charts. I adjusted the scale, line thickness, which graphs to display, worked, pressed the save button, and closed the program. After 2 hours I started it, and I want to continue what I was doing. I have to adjust these parameters again. And this happens many times a day. It would be nice to keep these settings.
^As already mentioned, width of traces will not be saved to user settings or project file. No use to repeat this request without better arguments than "it would be nice". Other items are explained also in user manual:
- Manual SPL max is saved to project file.
- Initial Y max/span of other than SPL chart is set in Options windows and saved to user settings (not to project file).
- Power & DI and unnormalized Directivity chart follow Y max of SPL chart.
- Manual SPL max is saved to project file.
- Initial Y max/span of other than SPL chart is set in Options windows and saved to user settings (not to project file).
- Power & DI and unnormalized Directivity chart follow Y max of SPL chart.
My projects had single W04 and two passive radiators.
Shared large volume as bass reflex would be the most economical for multiple drivers. For example 40 liters for 4xW04 and single 105x350mm vent.
Bringing Phase into equidistance
Hi,
It seems like this should be doable but I have not seen it done so perhaps it is just me thinking off track. When merging signals, you align the phase by bringing everything into minimum phase. But the farfield phase is an accurate measured phase for that distance. Is there a way to adjust the nearfield phase so that it aligns with the farfield phase at the merger point and adjusts the entire nearfield phase measurement as if it had been measured at the same farfield point? This would provide a better power response calculation and off axis graphics in the nearfield range (I would think).
Hi,
It seems like this should be doable but I have not seen it done so perhaps it is just me thinking off track. When merging signals, you align the phase by bringing everything into minimum phase. But the farfield phase is an accurate measured phase for that distance. Is there a way to adjust the nearfield phase so that it aligns with the farfield phase at the merger point and adjusts the entire nearfield phase measurement as if it had been measured at the same farfield point? This would provide a better power response calculation and off axis graphics in the nearfield range (I would think).
Ok. Another.
Main program. Drivers tab. At the bottom there is a minimum phase checkbox. And on the SPL graph, to can turn on the minimum phase. What is the correct way to see the minimum phase graph? Mark a checkbox for each driver, and look at the normal phase graph on the SPL graph (which will be the result of adding the graphs of the minimum phases of individual drivers)? Or not to mark this checkbox, and on the SPL chart, by right-clicking, mark the display of the minimum phase (which will be the minimum phase from the resulting chart of individual drivers)?
Main program. Drivers tab. At the bottom there is a minimum phase checkbox. And on the SPL graph, to can turn on the minimum phase. What is the correct way to see the minimum phase graph? Mark a checkbox for each driver, and look at the normal phase graph on the SPL graph (which will be the result of adding the graphs of the minimum phases of individual drivers)? Or not to mark this checkbox, and on the SPL chart, by right-clicking, mark the display of the minimum phase (which will be the minimum phase from the resulting chart of individual drivers)?
What??? Merger tool uses measured phase by default and aligns phase responses at transition frequency by adjusting delay of LF sum. Exactly what you're suggesting.
You can force HF responses to minimum phase by checking Minimum phase, and add excess group delay of original measured HF response (at 1/2 octs above transition f) by checking with GD of HF. But this is just optional in case measurement data is not captured properly: bad gear, setup, software, settings or other screw up.
Minimum phase of what? Single driver or whole speaker?
Minimum phase checkbox in Drivers tab manipulates phase response of frequency responses loaded into Drivers tab. That feature is added for users who cannot measure timing & phase as recommended. Something is screwed up in response data. For example phase response is totally missing or timing is normalized at peak of impulse responses which ruins the data. Normally that checkbox should be unchecked.
Minimum phase curve of SPL graph is minimum phase response of whole construction to reference angle. It could be target for phase response when designing speaker with minimum phase total SPL. Another purpose could be target when adjusting delay offset of all drivers to set excess group delay to zero at HF. This is mostly visual, but standard for plotting responses of construction.
One more time: Read user manual and don't expect that I repeat everything here forever.
Sorry Kimmo. That makes a lot more sense; I think I knew that but had stepped away for a while and am just returning after several months. Wonderful. Thanks.
Is there away to do the following in Vituix:
So I've designed an enclosure for my subwoofer using WinISD. But due to space constraints, I have to use a less than ideal Qtc. Now, the SPL response (in WinISD) shows that it's boosted at around 50Hz range for +4dB. Fortunately, my subwoofer manufacturer's frequency response shows a massive -5dB dip at 50Hz. So I thought, that's perfect. My enclosure should compensate my driver's dip.
But now, how do I apply the effects of my enclosure to my driver's frequency response graph in VituixCAD when I'm designing a crossover?
Thank you.
So I've designed an enclosure for my subwoofer using WinISD. But due to space constraints, I have to use a less than ideal Qtc. Now, the SPL response (in WinISD) shows that it's boosted at around 50Hz range for +4dB. Fortunately, my subwoofer manufacturer's frequency response shows a massive -5dB dip at 50Hz. So I thought, that's perfect. My enclosure should compensate my driver's dip.
But now, how do I apply the effects of my enclosure to my driver's frequency response graph in VituixCAD when I'm designing a crossover?
Thank you.
^I guess the same question is also in my e-mail inbox...
Basic problem is that neither half space simulation by T/S parameters alone nor manufacturer's measurement in datasheet are not reliable data for crossover design. Typically manufactures do not inform measurement conditions or lab is too small and reflecting to produce accurate response at LF or response at LF is hidden or simulated. Useless by default.
LF response simulated with Enclosure tool could be more accurate, but requires that T/S parameters produce accurate simulation. We can't know that without own measurements.
So simply follow measurement and design instructions linked below. Measure near field, simulate baffle effect response with Diffraction tool, merge to time-windowed far field measurements with Merger, load to Drivers tab for crossover design.
Basic problem is that neither half space simulation by T/S parameters alone nor manufacturer's measurement in datasheet are not reliable data for crossover design. Typically manufactures do not inform measurement conditions or lab is too small and reflecting to produce accurate response at LF or response at LF is hidden or simulated. Useless by default.
LF response simulated with Enclosure tool could be more accurate, but requires that T/S parameters produce accurate simulation. We can't know that without own measurements.
So simply follow measurement and design instructions linked below. Measure near field, simulate baffle effect response with Diffraction tool, merge to time-windowed far field measurements with Merger, load to Drivers tab for crossover design.
Thank you for your reply! I will follow your advice and re-read the manual and the instructions you linked. One question, when you say 'measure near field', you mean I have to measure the T/S parameters myself, right? In what condition should I measure it, i.e. do I put the drivers in the enclosure, or lay it on a table, etc?
I definitely have to dabble more with Vituix, this is a very powerful program! Thank you for your effort.
^No. Near field response is measured at 5-7 mm from the center of dust cap. Low supply voltage (-20...-30 dB compared to far field measurements) to avoid clipping of microphone and mechanical hitting.
Nothing is needed from driver manufacturer to produce measurement data for crossover design. Sd or Dd is needed for scaling in Merger tool and Diffraction simulation, but Dd can be measured with tape or ruler.
Nothing is needed from driver manufacturer to produce measurement data for crossover design. Sd or Dd is needed for scaling in Merger tool and Diffraction simulation, but Dd can be measured with tape or ruler.
Warning! Do not use B&O ICEpower or compatible amplifiers having balanced/bridged output and high D.C. voltage
in both speaker terminals. Speaker minus terminal should be in the same potential with 0 V terminal of line input.
Is this just for Impedance measurements for dual measurements as well?
in both speaker terminals. Speaker minus terminal should be in the same potential with 0 V terminal of line input.
Is this just for Impedance measurements for dual measurements as well?
Class-D amp is a poor choise for measurements. In Canada for example this would be a good alternative
LM3886 Done Right: Best DIY implementation of the LM3886 80W gainclone – Neurochrome
Class-D amps do filter the RF they create in output, but generally not enough for this purpose and there is lots of it remaining to "confuse" the AD converter in your soundcard. In best case you get an almost accurate frequency response, but I would not trust the THD measurements for sure.
LM3886 Done Right: Best DIY implementation of the LM3886 80W gainclone – Neurochrome
Class-D amps do filter the RF they create in output, but generally not enough for this purpose and there is lots of it remaining to "confuse" the AD converter in your soundcard. In best case you get an almost accurate frequency response, but I would not trust the THD measurements for sure.