Jump to content

LMS (with Daphile) to DAC via USB - BruteFIR equalizer


Recommended Posts

Just wondering if anyone has played with the USB (to DAC) based audio settings in LMS (Daphile player) changing the equalizer parameters through BruteFIR

When I click on advanced audio and choose sampling via BruteFir, I can load a right/left channel file.

I wanted to apply my eq settings but can't find any documentation how the file should look like and the format it accepts.

If anyone has done so, can you please share yours. Thanks

heaps

 

 

Link to comment
Share on other sites



  • 3 months later...
  • 3 weeks later...
  • 2 years later...
4 hours ago, Rrobot said:

I know this thread is a bit stale now but id also love to know how others are going with daphile and brutefir? 🙂

 

Slightly different, but here is a tutorial for LMS and Camilla DSP. 

https://www.audiosciencereview.com/forum/index.php?threads/rpi4-camilladsp-tutorial.29656/

Worked for me when I tried it. 

 

 

  • Like 1
Link to comment
Share on other sites



52 minutes ago, Grizaudio said:

 

Slightly different, but here is a tutorial for LMS and Camilla DSP

 

 

Excellent, thanks! 👍👍

 

I do actually run LMS on my system so maybe it's actually better to implement the filters upstream on the server? I had read that someone had done a, plugin for CamillaDSP in LMS. My daphile PC is a very gutless atom so it may struggle with FIR.  All my DSP efforts so far have largely been with the Behringer ultradrive. 

Edited by Rrobot
  • Like 1
Link to comment
Share on other sites

35 minutes ago, Rrobot said:

Excellent, thanks! 👍👍

 

I do actually run LMS on my system so maybe it's actually better to implement the filters upstream on the server? I had read that someone had done a, plugin for CamillaDSP in LMS. My daphile PC is a very gutless atom so it may struggle with FIR.  All my DSP efforts so far have largely been with the Behringer ultradrive. 


Camilla dsp will happily run on the Pi4 with the LMS endpoint (this is how it achieves loop back via Linux). 

 

I used it successfully to run my 96k 96000 taps FIR crossovers (2way) 
 

It’s worth noting this guide doesn’t allow for sampling rate changes. The Camilla setup requires a fixed sampling rate. 
 

It is possible to implement sample changes and FIR for each SR, but it becomes more complicated. 
 

I just use Roon now, which does it all easily. 
 

Edited by Grizaudio
  • Like 1
Link to comment
Share on other sites

Ill give it a go! I'm kindof at the point where i might decide to do my active adventures totally in software although the cost of multi channel dacs is still a significant impediment which starts to bring it all up closer to the cost of minidsp or hypex. 

 

Slightly off topic but It'd be nice to have more consumer hifi targeted hardware dsp alternatives using the AD or TI or even ALLDSP chips  to be honest...the t.racks and omnitronic units from Germany look pretty good for the money but the postage to Australia kills it a bit. 

Edited by Rrobot
  • Like 1
Link to comment
Share on other sites

10 minutes ago, Grizaudio said:

Check out my profile for how I do it if you are interested. 

 

 

Thanks! I will have a look 🙂

I'm just really at reasonably  basic levels with REW still...measuring, looking at the various plots of the  drivers, setting crossovers, PEQ filters and level matching in the DCX2496. Ive read the minimum phase articles in the REW documentation heaps of times but must admit phase  still baffles me slightly. More reading required! 

  • Like 1
Link to comment
Share on other sites



Ok, so I've made a little progress. Have measured one of my speakers  from about 1m as an exercise and generated some PEQ filters in REW, exported via xml to Rephase and applied some minimum phase paragraphic EQs and generated a wav file. I think my Atom N2600 processor may be too insipid to process the filters however as when I apply the bruteFIR files in Daphile the audio just glitches when trying to play.  Can I fiddle with the impulse settings much or do I need a better platform ?

 

image.png.595ced314709f275b60af9dcd23fab0a.png

Edited by Rrobot
Link to comment
Share on other sites

2 hours ago, Rrobot said:

Ok, so I've made a little progress. Have measured one of my speakers  from about 1m as an exercise and generated some PEQ filters in REW, exported via xml to Rephase and applied some minimum phase paragraphic EQs and generated a wav file. I think my Atom N2600 processor may be too insipid to process the filters however as when I apply the bruteFIR files in Daphile the audio just glitches when trying to play.  Can I fiddle with the impulse settings much or do I need a better platform ?

 

image.png.595ced314709f275b60af9dcd23fab0a.png

 

Tap count is moderate. 
My Pi4 would easily process 64000-96000 taps for my 2way. Your atom only has 1.6ghz and 2cores to work with.
 

 

Edited by Grizaudio
  • Thanks 1
Link to comment
Share on other sites

Thanks for confirming I wasn't pushing things with the taps...

 

Actually id forgotten the specs on this old fanless PC.. It's a single core 😝 N270 atom so very ancient! Could be time to crack out the NUC I reckon! I quite liked the low power, silent nature of the old jetway board but it's really too old I think. 

Edited by Rrobot
  • Like 1
Link to comment
Share on other sites

47 minutes ago, Rrobot said:

Thanks for confirming I wasn't pushing things with the taps...

 

Actually id forgotten the specs on this old fanless PC.. It's a single core 😝 N270 atom so very ancient! Could be time to crack out the NUC I reckon! I quite liked the low power, silent nature of the old jetway board but it's really too old I think. 

 

I've you are going to push 192,000k and 1:1 tap ratios, a NUC would be smart.

The Pi is starting to get a heavy load running LMS endpoint, and implementing 192k / 2way FIR filters. 
 

There are so many small fan-less options these days:

https://www.hardkernel.com/shop/odroid-h3-plus/

 

  • Like 1
Link to comment
Share on other sites

There are some great boards out there! 🙂 

 

Ok, so installed Daphile on the NUC and the convolutions work now...thanks for the encouragement to persist! 👍👍 Whether I've improved the sound is debateable though. lol I need to go deep diving on REW measurement, minimum phase, excess phase and so on. I'm still on the journey of understanding I must concede. 

 

One thing I notice with Daphile though is it seems to keep the DACs discreet...as in a separate LMS player effectively for each DAC. I'm unclear how or if I could run a 2 way active crossover using BruteFIR in Daphile unless I synchronise the two players/DACs in LMS into one ?

Edited by Rrobot
  • Like 1
Link to comment
Share on other sites



Noob alert! Well, I'm finding this all very baffling to be perfectly honest. I though I'd successfully loaded a convolution filter into Daphile. I'm using the same convolution for left and right but that shouldn't matter? It accepts the .wav convolution file in both left and right channels 

image.png.33fff21490ce74858afad9dee3e6d82b.png

...yet when I re-test response in REW using a sweep exported from REW and played locally from the Daphile machine, i get exactly the same measured amplitude curve as before any EQ was applied, even if I apply extreme EQ dips in rephase to test if it's working. Suggests to me the convolution is doing nothing in Daphile. Unfortunately there's not much guidance around about Daphile and Brutefir. Below is the Rephase amplitute and phase graph after some paragraphic Gain EQ and some filters linearization for crossover points plus me hacking about with Paragraphic phase EQ. 

image.thumb.png.2d78e0d1673e5273c88dfaed1f00d525.png

image.thumb.png.4fafd43df4f1804e954eb9fc7894499c.png

Is there something obvious I'm doing wrong? 🤦‍♂️ 

 

 

Edited by Rrobot
Link to comment
Share on other sites

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...
To Top