LS50 meta's plus sub vs More expensive speakers

[BrokenEnglishGuy]

Aye, the LS50 is not a high-volume speaker. Certainly can see an argument for pushing the crossover higher if aiming for significantly louder volumes, to keep that sub-200Hz distortion down. But then it was probably the wrong choice of speaker - loud volume needs bass drivers, Uni-Q or not. The distortion is from the size of the driver, not the Uni-Q-ness.

I just don't see any real evidence of any of KEF's line taking any unconventional decisions about crossovers in multi-way speakers that might be related to Uni-Q. If anything they use *fewer* crossovers than other manufacturers - they don't have any current 4-way designs at all.

Anyway - this is relevant to the OP. Towers will be higher-sensitivity and go louder than LS50+Sub. By going for the LS50 you're kind of resigning yourself to lower volumes; sub+HPF will boost that somewhat, but still not up to tower-type volume levels. But I think maybe this isn't an issue?

They have their kef muon 199.990 xD


Adding more volume and another woofer increase the sensitivity and lower distortion/increase sub bass
This is from kef, free fild or large room without bass boost

 

[aarons915]

That seems like over-solving it. Sure, the LS50 performance tails off below 200Hz, but crossing up that high is overkill and just causes a different set of problems. Crossing as low as 80Hz already gives dramatically reduced excursion compared to the way it behaves unfiltered. Lots of people seem extremely happy with their LS50s totally unfiltered, where the midbass driver will be slamming in and out, so to suggest it needs to be crossed as high as 200Hz seems like hyperbole.

Agreed, we all know the LS50 aren't the tool for the job if someone wants to listen at 100+db but thankfully most people listen in the 75-85 db range. I usually look at the THD plots to choose a crossover frequency with 80 being the lowest I would ever go. With the LS50 Meta most of us can use the 86db distortion plot since this is closer to our listening levels, you can kind of visualize that even a 100Hz 2nd order high pass will keep the meta under 1% THD at these levels. For the 96 db range I would bump that up to 120Hz-150Hz but much more care needs to be taken to ensure no subwoofer localization occurs. Also to others who think this is too high, remember we're talking 2nd order high pass and 4th order low pass so the acoustic crossover in-room ends up being 10-20Hz lower and should be ok.

 

[cany89]

Apologies if I've missed that someone else has already mentioned this but the Hegel H120 has pre-amp outputs and so there is no need for a high-level connection here.

NP, I already replied to that. I forgot that Hegel got line outs from pre.

 

[Tangband]

That seems like over-solving it. Sure, the LS50 performance tails off below 200Hz, but crossing up that high is overkill and just causes a different set of problems. Crossing as low as 80Hz already gives dramatically reduced excursion compared to the way it behaves unfiltered. Lots of people seem extremely happy with their LS50s totally unfiltered, where the midbass driver will be slamming in and out, so to suggest it needs to be crossed as high as 200Hz seems like hyperbole.

Anyway, another random thought - is a passive high-level HPF for the speakers an option? Seems like it could be the answer, but I can't see anything readily available off-the-shelf. I guess filter results depend on speaker impedance too much, so it's not a viable generic product?

If there's no theoretical problem, building your own passive HPF - with advice from someone in the DIY speaker area for component value+selection based on the measured LS50 Meta impedance curves - might be a fun little project. And potentially the cheapest option, even if you have to buy a soldering iron.

Well, it depends how steep the crossover frequency is. Crossing HP at 80 Hz 6 dB/oct with passive components, ( a simple 100-150 uF capacitor in series with the + terminal will do fine ) doesnt releave the bassdriver in ls50 meta from big excursions. But it can be better integrated with a subwoofer if doing so.

A dsp in series with the signal must be transparent enough to not destroy the qualitys of Kef ls50 meta. With a dsp you can do HP crossover at 100 Hz with 48 dB/oct, which is a big benefit if one wants to reduce cone-excursion.
But the signal must go trough an AD/DA extra, making it less transparent.
I dont think a cheap mini-dsp is good enough.

Making the kefs really good can surely be done by bypassing the passive crossover inside the loudspeaker, and making two terminals extra on the back of the loudspeaker, making the whole loudspeaker active with a dsp crossover and two power amps.
Then one can use a 48 dB/oct filter for HP at 100 Hz , and enjoy the benefit of actively drive the ls50 with a crossover at 2,1 kHz for tweeter and midrange. If the dsp crossover is good enough ( meaning it has to have digital inputs ) then the sound will be clearer than any passive options.

And such a speaker, optimized with dsp in the crossover between tweeter and midrange and with two subwoofers in stereo, would probably beat the Kef r7 soundwise.

The NHT xd system is made in this way. Measurements has been done by stereophile.
https://www.stereophile.com/content/nht-xd-active-loudspeaker-system-measurements

 

[Ultrasonic]

But the signal must go trough an AD/DA extra, making it less transparent.
I dont think a cheap mini-dsp is good enough.

Just on this point, there is no additional AD step if a miniDSP is fed with a digital signal.

I did use a 2x4 HD with the power and and speakers in my signature, and a BK Monlith subwoofer. The very slight compromise from the DSP/DAC quality was more than offset by the benefit of being able to properly integrate a subwoofer and apply EQ to manage room modes. The SHD I have now is slighlty better but it costs a lot more. Obviously the 2x4 HD isn't a streamer too, and the lack of a display makes it far less pleasant to use as a preamp.

 

[Tangband]

Just on this point, there is no additional AD step if a miniDSP is fed with a digital signal.

I did use a 2x4 HD with the power and and speakers in my signature, and a BK Monlith subwoofer. The very slight compromise from the DSP/DAC quality was more than offset by the benefit of being able properly integrate a subwoofer and apply EQ to manage room modes. The SHD I have now is slighlty better but it costs a lot more. Obviously the 2x4 HD isn't a streamer too, and the lack of a display makes it far less pleasant to use as a preamp.

Yes, the mini dsp option with digital input is probably better than the cheaper mini dsp option with only analog inputs.

Im not very fond of combining dsp with analog crossover filters . You can very easy end up with the drawbacks of both techniques. Its much better soundwise to let the dsp also do the crossover between the tweeter and midbass in ls50.
That way, you can throw away the passive sounddegrading components inside the ls50.
But this option demands DIY on the loudspeaker, a dsp crossover, a good microphone and measurement programs. No garanties to succed.

 

[Ultrasonic]

Yes, the mini dsp option with digital input is probably better than the cheaper mini dsp option with only analog inputs.

Oh if you were talking about the 2x4 rather than the 2x4 HD I'd definitely avoid this. Going by everything I've read this really isn't a good quality unit at all. The 2x4 HD is MUCH better. I've quite often come across people who have been totally put off miniDSPs by an early experience of the original 2x4 but this really doesn't seem to be representative of their subsequent products.

 

[phoenixdogfan]

Agreed, we all know the LS50 aren't the tool for the job if someone wants to listen at 100+db but thankfully most people listen in the 75-85 db range. I usually look at the THD plots to choose a crossover frequency with 80 being the lowest I would ever go. With the LS50 Meta most of us can use the 86db distortion plot since this is closer to our listening levels, you can kind of visualize that even a 100Hz 2nd order high pass will keep the meta under 1% THD at these levels. For the 96 db range I would bump that up to 120Hz-150Hz but much more care needs to be taken to ensure no subwoofer localization occurs. Also to others who think this is too high, remember we're talking 2nd order high pass and 4th order low pass so the acoustic crossover in-room ends up being 10-20Hz lower and should be ok.

View attachment 149278

Right now I'm experimenting. I've created an entirely 4th order crossover at 100 hz using Dephonica with phase compensation, and a second crossover with 4th order low pass and 2nd order high pass. Just did that last night and created full range Dirac Live corrections for each crossover profile. Will decide which one I like better, though my cursory listen last night had me leaning toward the one with 4th order slopes on both legs. And at 100 hz there was no localization issue. I might also do a third crossover at 120 hz with 4th order low and 2nd order high pass, just to see how I like it.

 

[phoenixdogfan]

Yes, the mini dsp option with digital input is probably better than the cheaper mini dsp option with only analog inputs.

Im not very fond of combining dsp with analog crossover filters . You can very easy end up with the drawbacks of both techniques. Its much better soundwise to let the dsp also do the crossover between the tweeter and midbass in ls50.
That way, you can throw away the passive sounddegrading components inside the ls50.
But this option demands DIY on the loudspeaker, a dsp crossover, a good microphone and measurement programs. No garanties to succed.

I'd be very loath to go DIY on my LS 50s, especially the Metas. Good way to void a warranty, mess up resale value AND mess up the balanced sound of a brilliant speaker. Kef already provides a way to go with electronic active crossovers. It's called the Wireless II.

 

[Ultrasonic]

I'd be very loath to go DIY on my LS 50s, especially the Metas. Good way to void a warranty, mess up resale value AND mess up the balanced sound of a brilliant speaker. Kef already provides a way to go with electronic active crossovers. It's called the Wireless II.

Using a miniDSP isn't DIY or anything that would void a warranty . In the way suggested in this thread anyway!

Edit: I realise I slightly missed what you were replying to there. I don't see an issue with use a miniDSP to provide a crossover between main speakers and sub whilst maintaining whatever crossover is in the speakers (passive or active).

 

[abdo123]

Oh if you were talking about the 2x4 rather than the 2x4 HD I'd definitely avoid this. Going by everything I've read this really isn't a good quality unit at all. The 2x4 HD is MUCH better. I've quite often come across people who have been totally put off miniDSPs by an early experience of the original 2x4 but this really doesn't seem to be representative of their subsequent products.

they both use the same DAC chip btw. the only different thing is the chip doing the digital signal processing and the extra inputs.

both of them are okay imo.

 

[Ultrasonic]

they both use the same DAC chip btw. the only different thing is the chip doing the digital signal processing and the extra inputs.

both of them are okay imo.

I'd suggest the quality of the DSP chip is pretty fundamental though. Seriously, I'm a miniDSP fan but would strongly discourage anyone from buying 2x4 for the sort of application were discussing.

 

[abdo123]

I'd suggest the quality of the DSP chip is pretty fundamental though. Seriously, I'm a miniDSP fan but would strongly discourage anyone from buying 2x4 for the sort of application were discussing.

Sigfreid Linkwitz had no problem using the 2x4 to design his final speaker the LXmini. Since he literally took part in designing one of the crossover filters you would typically use in a miniDSP I would think they're pretty okay.

both the 2*4 and the 2*4HD have a SINAD of ~98 dB.

I think some people had amplifiers that couldn't be driven to full output with 0.9V. so the loss of 3-4 dB made the sound less dynamic for them.

the 2*4 is really meant for subwoofers or amplifiers with adjustable gain. this way the lower than typical output is not an issue.

 

[Ultrasonic]

the 2*4 is really meant for subwoofers or amplifiers with adjustable gain. this way the lower than typical output is not an issue.

And use solely to process a subwoofer signal may be OK but I genuinely have seen lots of people dissatisfied with the performance of the 2x4 used as a sub/main crossover (which is what we're talking about here). My guess is that the SINAD figure you're quoting was for pure DAC operation and so doesn't tell us anything about DSP performance.

 

[kukocz]

Not sure about 2x4 but 2x4 HD and DDRC-24 have the same hardware and you can use RPi with camilladsp to xspture output and redirect to another DAC. Then you stay in digital domain.

Changing settings for 4 channels only in capture section doesn't work:

Jul 22 19:37:21.162 ERRO Mixer 'minidsp2DAC' has wrong number of input channels. Expected 4, found 2., module: camilladsp

Then I figured out that declaration of 4 channels is also needed in mixer (but without mapping):

devices:
  samplerate: 48000
  chunksize: 8192
  enable_rate_adjust: true
  queuelimit: 4
  capture:
    type: Alsa
    channels: 4
    device: "hw:CARD=DDRC24"
    format: S32LE
  playback:
    type: Alsa
    channels: 2
    device: "hw:CARD=LAQXD1"
    format: S32LE
mixers:
  minidsp2DAC:
    channels:
      in: 4
      out: 2
    mapping:
      - dest: 0
        sources:
          - channel: 0
            gain: 0
            inverted: false
      - dest: 1
        sources:
          - channel: 1
            gain: 0
            inverted: false
pipeline:
  - type: Mixer
    name: minidsp2DAC

Conclusion is that you need always read 4 channels from DDRC-24 in hw mode (but why I can read only 2 channels in plughw mode? )

Everything is working now but I experiencing another issue - audio glitches in random period of time. Chunksize doesn't matter so I need to think what else can be tuned

 

[phoenixdogfan]

Using a miniDSP isn't DIY or anything that would void a warranty . In the way suggested in this thread anyway!

Edit: I realise I slightly missed what you were replying to there. I don't see an issue with use a miniDSP to provide a crossover between main speakers and sub whilst maintaining whatever crossover is in the speakers (passive or active).

My impression was you were advocating surgery on the LS 50 to substitute an electronic crossover of the passive one in the speakers themselves, so you could have an all inclusive electronic crossover for sub, mid, and tweeter.

My mistake, sorry.

Edit:

Actually, it wasn't your comment I was initially commenting upon (using miniDSP with LS 50's for Xover, Parametric EQ, and DSP is a fine idea). Tangband, however, did recommend surgery on the LS 50's. So, yes, that is indeed IMHO a bad idea, and the initial comment stands.

 

[Ron Texas]

Speaking from experience as I have a pair of LS50's (original) and two Rythmik L12 subs:
Adding a sub to my speakers (and low end performance of the Meta's is the same) makes a big difference. Visible cone excursions disappear with a crossover as low as 90 hz. At one time I had a single Martin Logan sub. I had bass localization problems, possibly from crossing over above 80 hz or some weird room problem. Two subs are much better than one. Beyond that this discussion goes off in a lot of directions about subs in general.

The MiniDSP 2x4 HD is a nice solution as it is the only inexpensive device out there which will do both the high pass and delay on the mains. Plus, it gives system wide eq for sources other than a computer or streamer. SINAD is around 90, not 99 as someone mentioned. This kind of device also has a limited number of taps per filter. Some perfectionists will stick their nose up at this device. Why should I care since I'm using a Crown XLS 1502 with a SINAD of 78. How good is good enough the rest of you can discuss on your own time.

The Crown has a built in 4th order adjustable high pas filter, and some other DSP tricks. It provides just over 1 ms of delay for the mains. The L12's subs have low latency so positioned right next to the mains the timing is good. SVS subs have considerably more latency from their DSP making delay of the mains a more difficult task.

Someone mentioned using a 200 hz crossover. That's ok if you have a sub designed for it like the dual 8" sub Rythmik makes. Typical 12" subs just aren't fast enough to go past 120 hz. Also, a high crossover point means running the subs stereo and getting everything right just gets harder. I tried it going as high as 175 hz. Around 100 works better.

I haven't heard the Metas, but my takeaway from some video reviews is there isn't much difference between a Meta and originals which have been eq'ed to sound like a Meta.

To answer the OP's question, I would prefer a system which can cut it without subs. Integration isn't as easy as some think and there are a lot more things to go wrong. People make a big deal about flat to 20 hz, but other than a pipe organ instruments don't go lower than 28 hz. For whatever reason, when I hear live music bass is somewhat attenuated. Then again, gut punch bass is fun...

 

[Senior NEET Engineer]

Agreed, we all know the LS50 aren't the tool for the job if someone wants to listen at 100+db but thankfully most people listen in the 75-85 db range. I usually look at the THD plots to choose a crossover frequency with 80 being the lowest I would ever go. With the LS50 Meta most of us can use the 86db distortion plot since this is closer to our listening levels, you can kind of visualize that even a 100Hz 2nd order high pass will keep the meta under 1% THD at these levels. For the 96 db range I would bump that up to 120Hz-150Hz but much more care needs to be taken to ensure no subwoofer localization occurs. Also to others who think this is too high, remember we're talking 2nd order high pass and 4th order low pass so the acoustic crossover in-room ends up being 10-20Hz lower and should be ok.

View attachment 149278

Does distortion not matter for peak SPL? Even with EDM music I can measure 10+ dB higher peaks. Standard for movies is 20 dB of headroom.

 

[Haint]

How do they justify the price tag on this? It's $3600.00 in Canada.

The company pays most of the big Youtuber commentators to be their go-to recommendation, gotta make that marketing budget up somewhere.

 

[phoenixdogfan]

The company pays most of the big Youtuber commentators to be their go-to recommendation, gotta make that marketing budget up somewhere.

They say that, I say miniDSP SHD AND VTV Purifi Eval 1 Eigentact, for $1200 less, better performance, 2 x 4 digital crossover, 10 band 4 channel parametric eq, and Dirac Live 3.

Oh, and ignore the paid You Tube shills.