This product looks promising:
Hi-end DSP based multi-channel integrated Preamp/Crossover/DAC project
You will need a PC with convolver (JRiver or similar) to host and generate Acourate xovers. Plus of course multi-channel soundcard/DAC.
Acourate suite (for xovers) can not be used with hardware like miniDSP.
However, the good news is that MiniDSP works fine with Vituix-generated filters. Either IIR or FIR (not all MiniDSP units have FIR capabilities). So in this case you don't need Acourate - Vituix has it all covered. Just select the according DSP platform to match the hardware. The number of tabs for FIR filters is limited though, compared to a vastly bigger capacity of a PC-hosted environment.
My idea was to use VituixCAD for my filter creation and then export that to an external crossover. So perhaps as you say, all I need is MiniDSP.
The DAC is a different story and I keep that separate from the crossover.
It is all doable, but there is a caveat regarding MiniDSP hardware, AFAIK:
- Nano Digi will give you digital o/p for external DACsm but it is IIR only;
- 2x4 HD and DRC-DA8 have both IIR and FIR but come with integrated DAC - analog o/p only - unless you know how to access the pre-DAC digital signal directly from the board to use it with your external DACs (I don't/ can't);
- DRC-DI has both IIR and FIR (and fairly large number of taps), outputs digital signal, but is 2-channel only. You will need multiple units for multi-way.
- Nano Digi will give you digital o/p for external DACsm but it is IIR only;
- 2x4 HD and DRC-DA8 have both IIR and FIR but come with integrated DAC - analog o/p only - unless you know how to access the pre-DAC digital signal directly from the board to use it with your external DACs (I don't/ can't);
- DRC-DI has both IIR and FIR (and fairly large number of taps), outputs digital signal, but is 2-channel only. You will need multiple units for multi-way.
Choose HPF, LPF, Shelf or peak and select FIR type filter from Shape dropdown box
Sorry for that.About the instructions given in measurements preparations xxx, I didn't found what to do with ref time in REW. Have you got any advices ?
Hello kimmosto !
As usual ! Thanks you for this tool and for all your efforts put into this !
I didn't find with the 'search' function of this thread one information, so I asking ou directly
It's about the "Measurement Preparations with REW" document, precisely, the "Export far field responses of tweeter" chapter.
I understand that you advice to select axial response “ hor 0”, adjusting time window, and to "apply windows to all"
That's mean to me that the hor 0 ref time will be used on each other measurements. (Current version of REW allow to keep the ref time of indiciduals measurements).
Do I understand well ? Phase is for me an hard topic !
In the same way, as the medium and woofer are shorter, I don't get if I must use the woofer driver (for example) hor 0 ref time for its other measurements or the tweeter hor 0 ref time ?
See you !
The same value for all measurements of all measured drivers if you measure them at 1 meter. The whole point of this mess is to measure also timing differences so reference time and distance from mic to mechanical rotation center of driver must be constant.
I have a question about this. Is it the case that each radial set must have identical timing? So, a set of measurements done at 0 deg might have time set at 2.9ms, while a set done at 60 deg has timing of 2.85ms (these are arbitrary figures). Would that be OK? This was my experience:
Before I started my speaker measurements, I made an assessment of what the shortest time of flight would be from the closest driver over the rotational range of measurements. I made the measurements starting at -90 degrees but, as I progressed past 0 degrees, I noticed that the signal was getting cut off due to my time delay being too long. I had made a mistake in my assessment of minimum time of flight. The problem arose due to me having chosen the centre of the baffle as my pivot axis, combined with the fact that the midrange and tweeter were offset. I had overlooked that on the positive range the drivers would be swinging closer to the microphone. So my remedy was to start my further measurements at +90 deg, where drivers would be closest to the mic. So all measurements from +10 deg to +90 deg were done with the revised time setting and those from 0 deg to -90 deg were done with the original time of flight.
This is no less than perfect. Since version 2.0.5.0 (2018-10-02) there has been Convert IR to FR tool in VituixCAD. That has also Left window for sound earlier than Reference time. After clicking Far button, Left window is set to Rectangular 1.5 ms = 516 mm @344 m/s i.e. edge noise of subwoofer with cone diameter of 1032 mm will be captured inside time window assuming that acoustic center of tweeter is not further that 516 mm from the edge. In practice you can have horn tweeter, but problems are not expected with 18" woofer.
If you suspect that very early noise of subwoofer is cut out, check with 80 deg off-axis measurement (before exporting woofer's frds) is Left window of 1.5 ms enough. You can increase it up to 2.9 ms when measuring at 1 meter. That equals to subwoofer hitting the mic while measuring 90 deg off-axis
REW has also Left window so no problem.Exporting pir files with ARTA or mls files with CLIO need different approach because those programs don't have Left window. PreDelay is needed with ARTA, and time shift with CLIO. My instructions do not include exporting with those programs anymore.
Last edited:
Few basic rules for linking XO of driver in the main program and driver+enclosure system in Enclosure tool:
* Accurate extended impedance model of driver is mandatory in Enclosure tool if crossover in the main program is passive or semi-passive. Default impedance model with Re+Le only damages transfer function of XO because it's usually very far from fact.
* Number of drivers in Enclosure tool must be one (1) because main program feeds transfer function of single driver to radiator simulation in Enclosure tool. Box and possible vent should be scaled for single driver. Drivers having different filter in the main program should be investigated separately.
* Impedance response loaded from Enclosure tool back to main program is not saved to project file. It's just temporary trick in memory -> no need to be afraid that project is damaged. Original impedance response in zma file loaded to Drivers tab can be restored by changing impedance scale to 1.05 and back to 1.0. Until something is changed in Enclosure tool.
* Link between crossover in the main program and enclosure tool is contrived. I hope that users could design speakers without it.
Zip also enclosure project if you send something.