Hi,
I just tried to download the newest version of the by me very appreciated software on a new computer of mine, but it failed. It says that the page doesn´t exist. I also did see that there are some issues with the Google drive etc. Does anyone, of maybe the creator itself, know what is going on? Any solution in sight?
Best regards
//H
@Sharkythefrog
File reanalyzed with Virus Total, no problem, G have their opaque reasons.
https://www.virustotal.com/gui/file...fb347797eb84c2f6911ef65884e5faff766f7/details
File reanalyzed with Virus Total, no problem, G have their opaque reasons.
https://www.virustotal.com/gui/file...fb347797eb84c2f6911ef65884e5faff766f7/details
I am on it again after some year(s) filled with other stuff.
Some questions to the community:
1) Has anyne done or found component lists with price (and type)?
2) Has anyone done any XO:s and speakers based on room measurements? (Asking as I am about to build a speaker a little bit similar to the Larsen 8 or 9 - well actually not as mine will have a MLTL as a bass box 50x50x60 cm BxDxH and an upper part looking like the Carlsson OA-61 which is known as "omni directional" since the 1970s with lots of the sound bouncing from the back wall and the roof. Tricky with just free air and close measurements)
Some questions to the community:
1) Has anyne done or found component lists with price (and type)?
2) Has anyone done any XO:s and speakers based on room measurements? (Asking as I am about to build a speaker a little bit similar to the Larsen 8 or 9 - well actually not as mine will have a MLTL as a bass box 50x50x60 cm BxDxH and an upper part looking like the Carlsson OA-61 which is known as "omni directional" since the 1970s with lots of the sound bouncing from the back wall and the roof. Tricky with just free air and close measurements)
Last edited:
Perhaps you should take those anechoic measurements and compare with in room power sets for various locations. I assume your goal is finding a location for them which requires little EQ (other than overall smooth adjustments), since this would indicate a balanced location.
Hi @XMechanik - nice to see your post in the 3-way thread. Any update from Google? You know the exe could probably fit as an attachment on this thread.
Still no feedback from Google Drive "trust and safety team".
I'll try to recompile a 64bit version of the software and remove the functionality of new version notification (which is useless considering how often new versions are released). I suspect this functionality may cause false alarms by some heuristic scanners. If this helps XMachina will be available for download again.
I'll try to recompile a 64bit version of the software and remove the functionality of new version notification (which is useless considering how often new versions are released). I suspect this functionality may cause false alarms by some heuristic scanners. If this helps XMachina will be available for download again.
I started using XMachina a few hours ago. It is really intuitive and I got the first designs pretty soon.
My aim is to have a passive crossover for a two way MEH with relatively low part count and EQ the results with a DSP. Some of the results look really promising.
I followed this thread on and off - can I somehow force the program to suppress the peaks above cancellation notch so that it influences the HF less? Would it work to do the LF part as one way separately?
The reason why I finally tried XMachina is because I have like zero experience with designing passive crossovers.
My aim is to have a passive crossover for a two way MEH with relatively low part count and EQ the results with a DSP. Some of the results look really promising.
I followed this thread on and off - can I somehow force the program to suppress the peaks above cancellation notch so that it influences the HF less? Would it work to do the LF part as one way separately?
The reason why I finally tried XMachina is because I have like zero experience with designing passive crossovers.
Hi Allen,
These type of speakers rely very much on wall/ceiling resonance and have the drivers placed all kinds of directions. So measure them in free space will give a very different result than how they will perform under their correct position. Google Larsen 4 (or 8 or 9), or OA58.
Yes, I understand. I figured your goal is to have the direct and reflected have a similar trend, which means smooth DI?
Do you mean something like this? Did you try extending upper frequency range above 20kHz?
I need to get rid of the encircled part:
Just in case, the measurements are attached. Unfortunately, they are in dBV and not dBSPL, but from using these in the program, I do not see this is a problem when I set the target at the same levels. I have already tried one of the circuits in XSim and it seems it will work pretty well.
I will try to extend the range above 20 kHz.
I have another question - the generated crossovers without common ground tend to have a little less components in total - is that correct? Is there any reason not to use them? I have not seen this before, it is not very common as far as I can tell.
Just in case, the measurements are attached. Unfortunately, they are in dBV and not dBSPL, but from using these in the program, I do not see this is a problem when I set the target at the same levels. I have already tried one of the circuits in XSim and it seems it will work pretty well.
I will try to extend the range above 20 kHz.
I have another question - the generated crossovers without common ground tend to have a little less components in total - is that correct? Is there any reason not to use them? I have not seen this before, it is not very common as far as I can tell.
I don't see any problem in attenuating the > 1 kHz band of the LF-way (or maybe I'm not getting something). I loaded the characteristics, smoothed them out a bit first as they looked a bit noisy. Here is the result: the 2 kHz region LF ripple is 35 dB lower than the system response, which seems sufficient, but if you want more, use a steeper low-pass slope.
The circuits are not simple but still implementable(RLC 3/6/7). The characteristics are quite uneven and several components are needed to equalise them. For example, here is a comparison of the HF characteristic and the filter characteristics that was designed to compensate it.
Attached design file.
Let me know if this doesn't solve the problem, maybe I didn't get something.
The circuits are not simple but still implementable(RLC 3/6/7). The characteristics are quite uneven and several components are needed to equalise them. For example, here is a comparison of the HF characteristic and the filter characteristics that was designed to compensate it.
Attached design file.
Let me know if this doesn't solve the problem, maybe I didn't get something.
Thank you very much for this! Yes, the HF driver just measures this way according to the datasheet and the bump in the LF is from the bandpass chamber - I might actually try to remove the cone volume filler to see if the bump goes away - I would expect that maybe worse problems could arise since this peak is relatively well behaved compared to what I measured in the previous build.
I will have a detailed look at this later, I really appreciate your input on this!
Edit: By the way, the smoothing can be done in the program itself? I must have missed that. I used 3 points per octave for the task to reduce the crossover complexity.
I will have a detailed look at this later, I really appreciate your input on this!
Edit: By the way, the smoothing can be done in the program itself? I must have missed that. I used 3 points per octave for the task to reduce the crossover complexity.
There is actually an option for averaging in the frd context menu. But it only works well at magnitude and there is a bug in phase processing, so it's better not to use it.