• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Is the LXmini by Siegfried Linkwitz the answer for home audio?

Burning Sounds

Senior Member
Joined
Dec 22, 2016
Messages
422
Likes
594
Location
Co. Durham, UK
That said, and based on system photographs I've seen, and a few of those I've listened to, a number of owners don't have their speakers located optimally. However, the speakers are very portable and can be moved from a storage position to a listening position very easily when the wife is gone.......which is the procedure followed by some owners. :)

Dave.

Absolutely agree. Mine used to be in my music room where I could place them optimally. However, they didn't get used so much because the LX521s are there also. As I mentioned in my earlier post they are not in an ideal position in the living room, but work much better than I expected for everyday TV use and casual music listening. My wife primarily uses this system for watching French and Spanish movies so dialogue is the most important thing not crash, bang, wallop explosions etc. For serious music listening it is easy to pull them further out and turn them to the right so they are firing down the room.

If you are using a PC and JRiver as your source it is easy to build the required DSP in JRiver eliminating another box and cables. Just be aware that because of peculiarities in the way JRiver defines Q for high and low pass filters and shelving filters you can't simply drop the MiniDSP parameters in and get the best out of the LXMinis. They will sound OK and the adjustments are relatively small, but it is not as SL intended.
 
OP
A

abdo123

Major Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
3,733
Likes
3,262
Location
Brussels, Belgium
If you are using a PC and JRiver as your source it is easy to build the required DSP in JRiver eliminating another box and cables. Just be aware that because of peculiarities in the way JRiver defines Q for high and low pass filters and shelving filters you can't simply drop the MiniDSP parameters in and get the best out of the LXMinis. They will sound OK and the adjustments are relatively small, but it is not as SL intended.

What about Equalizer APO?
 

Burning Sounds

Senior Member
Joined
Dec 22, 2016
Messages
422
Likes
594
Location
Co. Durham, UK
What about Equalizer APO?

Sorry, I've never used it so don't know how it defines Q. But it is important to know that it is not always straightforward to move X-over/EQ parameters from one piece of software (and some hardware units, too) to another. The only way to know for sure is to check the transfer function - it can be done using REW and an inexpensive ADC/DAC.
 

mdsimon2

Addicted to Fun and Learning
Joined
Oct 20, 2020
Messages
513
Likes
556
Location
Detroit, MI
Sorry, I've never used it so don't know how it defines Q. But it is important to know that it is not always straightforward to move X-over/EQ parameters from one piece of software (and some hardware units, too) to another. The only way to know for sure is to check the transfer function - it can be done using REW and an inexpensive ADC/DAC.

I have written a very crude Octave/Matlab routine that will allow you to generate .frd files based on a combination of filters. The original intent of the routine was to generate biquad coefficients for a miniDSP but it can be used to model any combination of IIR filters.

From my brief testing the Q implementations in the routine match what is used in the miniDSP. I can definitely confirm this is the case for the LXmini transfer functions based on comparing the modeled .frd to actual measured transfer function from a miniDSP 2X4HD.

It might be helpful for those that want to have an understanding of what the LXmini filter transfer function looks like without purchasing a miniDSP and measuring the transfer function. You can then use your DSP platform of choice and compare that transfer function to the miniDSP transfer function.

I've attached the routine to this post. I believe it should work in Matlab but I have only used it in Octave so far. You will need to have the signal and control packages installed and loaded in Octave for this to work. I am not a programmer or a DSP expert and I should warn you in advance that the code is crude, not well commented and has minimal error checking.

The routine will output two .frd files which can easily be imported in to REW. One is called "out.frd" and represents the ideal analog transfer function for the filters you have modeled. The other is called "outdig.frd" and represents the DSP transfer function including high frequency warping effects (more prevalent at 48 kHz sample rate than 96 kHz).

I've also attached a few screenshots of the routine in action (not using the LXmini transfer function as I do not want to share this information with those that have no purchased plans).

https://www.gnu.org/software/octave/
https://octave.sourceforge.io/signal/index.html
https://octave.sourceforge.io/control/

Michael
 

Attachments

  • Screen Shot 2021-01-27 at 1.08.36 PM.png
    Screen Shot 2021-01-27 at 1.08.36 PM.png
    289.5 KB · Views: 60
  • Screen Shot 2021-01-27 at 1.08.46 PM.png
    Screen Shot 2021-01-27 at 1.08.46 PM.png
    97.8 KB · Views: 74
  • Screen Shot 2021-01-27 at 1.08.59 PM.png
    Screen Shot 2021-01-27 at 1.08.59 PM.png
    425.5 KB · Views: 64
  • biquad.m.zip
    5.6 KB · Views: 27
D

Deleted member 2944

Guest
Very good!
I installed Octave on my Manjaro installation and it works perfectly.

Dave.
 

Ron Texas

Major Contributor
Joined
Jun 10, 2018
Messages
4,133
Likes
5,407
Location
A mysterious place with no name.
Sorry, I've never used it so don't know how it defines Q. But it is important to know that it is not always straightforward to move X-over/EQ parameters from one piece of software (and some hardware units, too) to another. The only way to know for sure is to check the transfer function - it can be done using REW and an inexpensive ADC/DAC.

Be especially careful with JRiver in their Wiki they say Q for shelf filters is mislabeled, is really S. They give you a way to recalculate it, but a Q of .7 is about S=1 and Q of 1 is about S=2.
 

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,021
Likes
5,788
Interesting, I am honestly a bit skeptical of room correction but I understand this perspective. Certainly anything that helps you differentiate between room response and speaker response is helpful.

Michael

It's worked well for me so far :). The 3 speakers I own that Amir has measured(M105, 308p, 8030c) have all gotten better because of it.
 

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,021
Likes
5,788
I think his design entertains the concept of wide directivity up until the point of where are hearing is most sensitive to spatial cues (1k to 5k Hz). From there, your speakers has to have narrow directivity or your stereo image will inevitably collapse as i mentioned in part 1.


Logically, this actually makes more sense to me than going omni all the way up(ie Morrison or MBL), as most things in life become more directional as the frequency increases.
 

ooheadsoo

Member
Joined
Aug 14, 2020
Messages
52
Likes
36
It's worked well for me so far :). The 3 speakers I own that Amir has measured(M105, 308p, 8030c) have all gotten better because of it.
It's hard to live without room correction for bass, once you've lived with it for a while. I suppose some rooms may be better behaved than others, but I greatly prefer my systems with the bass resonances notched flat.
 

ooheadsoo

Member
Joined
Aug 14, 2020
Messages
52
Likes
36
Logically, this actually makes more sense to me than going omni all the way up(ie Morrison or MBL), as most things in life become more directional as the frequency increases.
You can play with this by varying the level of stuffing in the full range driver mounting/diffraction tube. I thought the sound was too bright without the specified stuffing.
 

Burning Sounds

Senior Member
Joined
Dec 22, 2016
Messages
422
Likes
594
Location
Co. Durham, UK
Be especially careful with JRiver in their Wiki they say Q for shelf filters is mislabeled, is really S. They give you a way to recalculate it, but a Q of .7 is about S=1 and Q of 1 is about S=2.

Yes, Yes and No.

Yes, be careful with JRiver's DSP PEQ - it is not straighforward. Yes, S=1 is 0.707, but No, S=2 is 1.414. And this is only true for high and low pass filters. However you won't see S being used anywhere on the high and low pass dialogue box in JRiver's PEQ. What you will see is Bandwidth (Q) - there is no mention of S. So if you enter 1 in Bandwidth (Q) it is actually 0.707 (actually 0.71 as JRiver can only do 2 decimal places). And to create Linkwitz-Riley filters you will need to cascade two Butterworth filters. It's all a bit messy really, especially as there is no graphic visualisation of what you have just input.

The calculator on JRiver's Wiki that @Ron Texas is referring to is only for shelving filters. Calculating the Q for those is necessary because of the way Q changes in JRiver's PEQ depending on the gain of the shelving filter. I've never found the suggested calculation resource of Math Papa to work well for me, but either WolframAlpha or Cymath work fine and are easy to use.

Anyway I think we've probably got a bit OT here as the OP asked about the LXMini for home audio which to me means keeping it simple and why I've reverted to using the MiniDSP2x4. In my living room the system is turned on and off with one switch. In my music room I use a PC source and I find there are a lot of advantages to doing the DSP in JRiver.
 

Ron Texas

Major Contributor
Joined
Jun 10, 2018
Messages
4,133
Likes
5,407
Location
A mysterious place with no name.
Yes, Yes and No.

Yes, be careful with JRiver's DSP PEQ - it is not straighforward. Yes, S=1 is 0.707, but No, S=2 is 1.414. And this is only true for high and low pass filters. However you won't see S being used anywhere on the high and low pass dialogue box in JRiver's PEQ. What you will see is Bandwidth (Q) - there is no mention of S. So if you enter 1 in Bandwidth (Q) it is actually 0.707 (actually 0.71 as JRiver can only do 2 decimal places). And to create Linkwitz-Riley filters you will need to cascade two Butterworth filters. It's all a bit messy really, especially as there is no graphic visualisation of what you have just input.

I used the calculator link in the JRiver wiki to get the values for S that I posted. Your value for S=2 of 1.414 seems intuitively correct. What could have possibly went wrong?
 

Burning Sounds

Senior Member
Joined
Dec 22, 2016
Messages
422
Likes
594
Location
Co. Durham, UK
I used the calculator link in the JRiver wiki to get the values for S that I posted. Your value for S=2 of 1.414 seems intuitively correct. What could have possibly went wrong?

Probably nothing wrong - when you use the calculator for shelving filters you get a different value for S depending on the gain of the filter. Enter a Q of 1 and a gain of 3 you will get 1.97. Then change the gain to 18 and rerun the calculator. You should now get 1.46.

The fixed values I mentioned (S=1 is actually 0.707 and S=2 is actually 1.414) is just in reference to high and low pass filters only. Peaking filters are fine and behave as expected.
 

Phorize

Addicted to Fun and Learning
Forum Donor
Joined
Apr 26, 2019
Messages
938
Likes
1,095
Location
U.K
ASR can be an odd place sometimes.

For those insisting on a NFS test, what do you think that this test will show that is not already known from SL’s outdoor measurement and significant explanation of the design on his website?

Michael
Agree. Obviously the klippel is an excellent development that widens access to rigorous speaker measurement, but no one should assume that speaker designers lived the dark ages before its inception, or that a klippel is the only way to objectively measure a speaker.
 

dshreter

Active Member
Joined
Dec 31, 2019
Messages
215
Likes
219
Would an ASR test of these make them more viable for your consideration??
These speakers were not designed with any sort of CEA 2034 objective in mind. You fellas understand that, right?

I live really close and considered taking my spare LXmini pair to Amir for NFS testing, but based on the speaker reviews from Amir I've seen so far, I don't see any point in it. These speakers are outside his understanding of what a speaker radiation pattern objective might be.

There are hundreds of LXmini's (kits and commercial) that have been constructed and are in the field providing enjoyment to their owners. A number of them have been augmented by the inexpensive subwoofer setup I designed for them.

Objective testing is fine and dandy, but the large number of folks enjoying the sound of these speakers is a far better (ultimate) indicator of the design success Mr. Linkwitz achieved here.

Dave.

Comprehensive measurement would be interesting for interpretation rather than critique. I agree boiling it down to a preference score would be superfluous.
 

dshreter

Active Member
Joined
Dec 31, 2019
Messages
215
Likes
219
ASR can be an odd place sometimes.

For those insisting on a NFS test, what do you think that this test will show that is not already known from SL’s outdoor measurement and significant explanation of the design on his website?

Michael
It will show measurements made with a common basis. We spend time here and have learned to interpret the data and visualizations as they are presented here, so getting it on the klippel would add to that body of knowledge.

There shouldn’t be any astounding revelation, but almost all of the speakers measured by Amir also have measurements existing somewhere else. As well, I’m assuming people are interested in Amir’s interpretation of the data and subjective listening feedback.
 

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,021
Likes
5,788
This is my use case as well – did you ever proceed?

No. I'd really love to, but I'm not sure I'm up for the challenge. I'd probably be more comfortable trying to build the Minis first. That seems like something that would be closer to my ability level.

Still, these are super intriguing as end game speakers in some respects.
 

F-Bomb

New Member
Joined
Feb 11, 2021
Messages
2
Likes
0
I'm curious, how do those with this speaker like it when playing decidedly not acoustic music?
Do you perhaps mean live music? From what I have read, all Linkwitz speakers are designed to recreate the feel of a live performance, e.g. that of an orchestra. But most records nowadays are studio records (e.g. for rock music) instead of live records.

Are LXmini owners satisfied with these speakers for studio records?
 
Top Bottom