Moode audio - dsp Update!!

[tripitaka]

Update!!

I really was dreading trying to add a convolution file to my new RPi ...but it seems easy as Pi after all.

I simply emailed my impulse files (as WAV) to my phone, then from my phone opened the Moode IP address -> configure -> CamillaDsp then downloaded the two files (L/R) from within the Camilla Convolution File panel so that they became available for selection within the Quick Convolution Filter panel, which I then activated in the General dropdown menu then pressed Save. That has got to be least painful thing I have ever done with a Raspberry Pi.  

Is that really all there is to it? If I known about this, I would've been onto it ages ago.

[End Update]

 

Hi, I've fired up Moode Audio (free software) over RPi4/HifiBerryDigipro2.2 and I am impressed with Moode

The backstory is that I decided to try and build on earlier efforts tinkering with miniDSP convolution, by embarking on a new and possibly quixotic quest to find if convolution can be done (for free) on a rapsberry.  Yes, I know I need my head examined

 

I had to sadly give up my beloved Rune Audio (not Roon) to move over to new hardware, since Rune updates are now rather hard to follow (the result of some sort of ownership dispute a few years back).

 

Anyway, I had assumed Volumio would be the answer but after trying it I have nothing good to say about it (sounded bad and has no functions of real value without paying a hundred bucks a year). 

 

Moode was my next adventure.

Moode avoids having an app at all, by letting you access it over your phone browser - works beautifully (for Android you need to bind the Mac address of your device in order to find it, of course, and then pin the link to your home screen).  I am now a convert and reckon avoiding an app is actually the winning formula, even though I specifically wanted an app originally (which was why I wanted Volumino).

 

Anyway, Moode accesses my 4Terabyte hard drive beautifully, while I also merrily stream Tidal over Bubbleupnp, everything has worked seamlessly so far.

 

And... Moode has Camilladsp functionality, which will be my next project.

 

Usual caveats, I am not an audio expert, so everything I have written may be wrong but this forum knows by now that I am just a punter with an open mind, so if any other punters find value from my stumblings then I am happy.

 

Edit: oh I certainly should have mentioned that I always use an external reclocking device in conjunction with the above, placed before my DAC - the sound is very ordinary otherwise (though maybe a good enough DAC might still be able to compensate, IDK)

 

 

 

 

 

Edited September 7, 2022 by tripitaka

[Guest]

6 minutes ago, tripitaka said:

The backstory is that I decided to try and build on earlier success tinkering with miniDSP convolution experiments, by embarking on a new and possibly quixotic quest to find if convolution can be done (for free) on a rapsberry.  Yes, I know I need my head examined

. . .

And... Moode has Camilladsp functionality, which will be my next project.

I am running convolution on CamillaDSP, so it can done.  It is easy to set up and the documentation is excellent, a nice change from other free DSP software. The hard part is, of course, generating the convolution file.

 

From what I have read (not using Pi), Moode have done a good job of integrating CamillaDSP. 

 

Have fun...

[tripitaka]

5 minutes ago, Snoopy8 said:

The hard part is, of course, generating the convolution file.

Thanks Snoop, no surprise you are already ahead of me to this space

 

I have an earlier thread ("Adventures in phase") which, for all it's shortcomings, I think does a decent job of explaining how to generate a convolution file

 

Edited September 5, 2022 by tripitaka

[Muscat]

I was a long time Moode Audio user and only just switched to Volumio over the weekend.  Reason for the switch was that I couldn't get Spotify to work with Moode.  I must say, Volumio setup is a lot more seamless than Moode, and Spotify works!  However, I did get the impression that the Volumio audio quality is a little subpar compared t what I remember of Moode.  Interesting that you mention the same.

[tripitaka]

1 minute ago, Muscat said:

I was a long time Moode Audio user and only just switched to Volumio over the weekend.  Reason for the switch was that I couldn't get Spotify to work with Moode.  I must say, Volumio setup is a lot more seamless than Moode, and Spotify works!  However, I did get the impression that the Volumio audio quality is a little subpar compared t what I remember of Moode.  Interesting that you mention the same.

Good point, Volumino was effortless to set up. That is something positive I can say after all

[Guest]

8 minutes ago, tripitaka said:

Thanks Snoop, no surprise you are already ahead of me to this space

 

I have an earlier thread ("Adventures in phase") which, for all it's shortcomings, I think does a decent job of explaining how to generate a convolution file

Good that you are using REW & RePhase. I prefer the results to DRC (which is much easier to use).

[Grizaudio]

48 minutes ago, Snoopy8 said:

I am running convolution on CamillaDSP, so it can done.  It is easy to set up and the documentation is excellent, a nice change from other free DSP software. The hard part is, of course, generating the convolution file.

 

From what I have read (not using Pi), Moode have done a good job of integrating CamillaDSP. 

 

Have fun...

I found the Camilla dsp tutorial for RPI4 not so straight forward. A pre configured image would be far easier to work with, with a simple DAC selection table etc. 

 

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

 

ATM I use Roon for convolution, but would appreciate a software like Camilla that would allow my Motu and Pi4 combo to use the Dacs spdif/toslink inputs, whilst maintaining convolution crossover. 
 

atm I’m limited to Roon only. 

Edited September 5, 2022 by Grizaudio

[Grizaudio]

52 minutes ago, Snoopy8 said:

I am running convolution on CamillaDSP, so it can done.  It is easy to set up and the documentation is excellent, a nice change from other free DSP software. The hard part is, of course, generating the convolution file.

 

From what I have read (not using Pi), Moode have done a good job of integrating CamillaDSP. 

 

Have fun...

will Moode allow additional DAC inputs like spdif/toslink etx to feed through the convolution filters? From something like a Motu mk5 lite 

[Guest]

@Grizaudio, convolution and configuration with Roon is much, much simpler. And you have worked it out for your multi channel setup. Not sure why you want to even consider CamillaDSP at all ?????

 

I know enough of CamillaDSP to get it working for my needs.  All I did was answer OP's question on whether convolution can be done.  Sorry, but you should ask someone else who knows a lot more...

[Grizaudio]

2 minutes ago, Snoopy8 said:

@Grizaudio, convolution and configuration with Roon is much, much simpler. And you have worked it out for your multi channel setup. Not sure why you want to even consider CamillaDSP at all ?????

 

I know enough of CamillaDSP to get it working for my needs.  All I did was answer OP's question on whether convolution can be done.  Sorry, but you should ask someone else who knows a lot more...

Hi Snoopy,

 

I’m only considering Camilla due to its ability to allow the use of spdif inputs on my dac. 
 

atm I’m limited to Roon as the source. 
 

I believe audiolense will do this, if Camilla can’t. 

[Guest]

1 minute ago, Grizaudio said:

Hi Snoopy,

 

I’m only considering Camilla due to its ability to allow the use of spdif inputs on my dac. 
 

atm I’m limited to Roon as the source. 
 

I believe audiolense will do this, if Camilla can’t. 

Comparing the capability of AudioLense to CamillaDSP is an exercise in futility.

If you know AudioLense can do the job, end of story...

 

And I do not have the CamillaDSP knowledge to help you, so will bow out of this thread.

[tripitaka]

On 05/09/2022 at 8:31 PM, Grizaudio said:

I found the Camilla dsp tutorial for RPI4 not so straight forward. A pre configured image would be far easier to work with, with a simple DAC selection table etc. 

 

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

 

I am possibly the least IT-savvy bloke on this whole forum, but I just added Convolution files to Moode, at least I think I did

Update added to first posting.

[PKay]

On 5/9/2022 at 8:01 PM, Muscat said:

I was a long time Moode Audio user and only just switched to Volumio over the weekend.  Reason for the switch was that I couldn't get Spotify to work with Moode.  I must say, Volumio setup is a lot more seamless than Moode, and Spotify works!  However, I did get the impression that the Volumio audio quality is a little subpar compared t what I remember of Moode.  Interesting that you mention the same.

I’m new to Volumio and have just got LMS up and running. A very different sound presentation. I’m going to do some comparisons over the weekend. 

[RoHo]

On 07/09/2022 at 9:23 PM, tripitaka said:

 

I am possibly the least IT-savvy bloke on this whole forum, but I just added Convolution files to Moode, at least I think I did

Update added to first posting.

I'm interested to know if you think the whole process was worthwhile.   Is it room eq you are applying in the convolution?

I currently use MiniDSP for subwoofer XO and room eq but have also been playing with the same Pi4-Moode system.  For me ATM I purely use it as a way to play digital files - stuff I can't buy on CD- but it's also been a good learning experience.  I must say my eyes still glaze over when I try to read lines of terminal  code but I still have got it up and running 

I have toyed with using Camilla to apply the room eq but are there major advantages?

Edited September 14, 2022 by RoHo

[frednork]

@tripitaka if you need help with moode I have some experience in using convolutions within it

[tripitaka]

46 minutes ago, RoHo said:

I'm interested to know if you think the whole process was worthwhile.   Is it room eq you are applying in the convolution?

I currently use MiniDSP for subwoofer XO and room eq but have also been playing with the same Pi4-Moode system.  For me ATM I purely use it as a way to play digital files - stuff I can't buy on CD- but it's also been a good learning experience.  I must say my eyes still glaze over when I try to read lines of terminal  code but I still have got it up and running 

I have toyed with using Camilla to apply the room eq but are there major advantages?

 

Thanks for asking as I had been intending to report back on this.

I am coming to the view that there is a trade-off. I had been trying to flatten the frequency output of my speaker (as measured nearfield).

 

Pluses:

Moode Audio is a great renderer and a keeper for me. What I have come to love is that there is no app to get 'crossed wires' with Bubbleupnp.

CamillaDSP is unbelievably simple to use and the convolution has definitely improved timbral accuracy.

 

Minuses:

I am coming to understand that using a general microprocessor, as opposed to something that is optimized for DSP suffers not just with really dreadful inefficiency but also in error management, such that the noise floor increases.

In any case, I feel that the increased timbral accuracy has come at a cost in transparency. This would likely be less of an issue with a worse system than mine and more of an issue with a better system. So horses for courses.

 

Sadly I am going to disconnect my DSP for the time being, until I can think of a best-of-both-worlds approach which hopefully doesn't involve me operating a PC to play tunes or paying subscriptions

 

[frednork]

14 minutes ago, tripitaka said:

 

Thanks for asking as I had been intending to report back on this.

I am coming to the view that there is a trade-off. I had been trying to flatten the frequency output of my speaker (as measured nearfield).

 

Pluses:

Moode Audio is a great renderer and a keeper for me. What I have come to love is that there is no app to get 'crossed wires' with Bubbleupnp.

CamillaDSP is unbelievably simple to use and the convolution has definitely improved timbral accuracy.

 

Minuses:

I am coming to understand that using a general microprocessor, as opposed to something that is optimized for DSP suffers not just with really dreadful inefficiency but also in error management, such that the noise floor increases.

In any case, I feel that the increased timbral accuracy has come at a cost in transparency. This would likely be less of an issue with a worse system than mine and more of an issue with a better system. So horses for courses.

 

Sadly I am going to disconnect my DSP for the time being, until I can think of a best-of-both-worlds approach which hopefully doesn't involve me operating a PC to play tunes or paying subscriptions

 

Thanks for your comments above.  Am interested in what you think may be the cause of the problem you perceive.

 

Sorry I didnt see above how you have implemented moode dsp within your system and what is the normal config (w/o moode), so a brief outline of that would be helpful as well.

[MrBurns84]

1 hour ago, tripitaka said:

 

Minuses:

I am coming to understand that using a general microprocessor, as opposed to something that is optimized for DSP suffers not just with really dreadful inefficiency but also in error management, such that the noise floor increases.

I played around with a minidsp years ago and it took ages to get a house curve that was a livable compromise...i could never perfect it based on the amount of time i had away from the family.  Over time i just gave up and lived with my room acoustics, which to be honest was not bad at all.

[RoHo]

1 hour ago, tripitaka said:

 

 

 

Sadly I am going to disconnect my DSP for the time being, until I can think of a best-of-both-worlds approach which hopefully doesn't involve me operating a PC to play tunes or paying subscriptions

 

You've just written a product description for MiniDSP  

 Seriously, the new Flex series is very tempting: upgraded DACs and nicer "human interfaces". And add DIRAC for auto EQ if you wish.

I think that what you were attempting with the convolution filter (using a global DSP filter to adjust a speakers NF response)  is problematic and possibly added to the less than perfect outcome.

  I would be really interested in how it all worked if instead you were  EQing the rooms bass response.

[tripitaka]

23 minutes ago, RoHo said:

You've just written a product description for MiniDSP  

 Seriously, the new Flex series is very tempting: upgraded DACs and nicer "human interfaces". And add DIRAC for auto EQ if you wish.

I think that what you were attempting with the convolution filter (using a global DSP filter to adjust a speakers NF response)  is problematic and possibly added to the less than perfect outcome.

  I would be really interested in how it all worked if instead you were  EQing the rooms bass response.

I do also have a miniDSP - the OpenDRC which is claimed to be optimised for FIR filtering.  But, being greedy, I wanted to improve on the 6K tap limit for that device.  I avoided the Flex simply because I wanted complete visibility of what was going on inside (I'm funny like that).

I will take stock for a while as I'm in no great hurry to do anything.

[RoHo]

8 minutes ago, tripitaka said:

I do also have a miniDSP - the OpenDRC which is claimed to be optimised for FIR filtering.  But, being greedy, I wanted to improve on the 6K tap limit for that device.  I avoided the Flex simply because I wanted complete visibility of what was going on inside (I'm funny like that).

I will take stock for a while as I'm in no great hurry to do anything.

What about the Flex do you find more "opaque" than the older models?  I think the DSP software is the same as in my 2x4HD.

[tripitaka]

2 hours ago, RoHo said:

What about the Flex do you find more "opaque" than the older models?  I think the DSP software is the same as in my 2x4HD.

I thought I would enjoy making my own correction impulses from scratch so that I could 'see how the sausage was made'.

Not sure I really did enjoy it but anyway

[tripitaka]

4 hours ago, RoHo said:

You've just written a product description for MiniDSP  

 Seriously, the new Flex series is very tempting: upgraded DACs and nicer "human interfaces". And add DIRAC for auto EQ if you wish.

I think that what you were attempting with the convolution filter (using a global DSP filter to adjust a speakers NF response)  is problematic and possibly added to the less than perfect outcome.

  I would be really interested in how it all worked if instead you were  EQing the rooms bass response.

 

Oh yeah I see what you mean, what I was really doing was smoothing the NF output above 100hz.

Below that I was cutting the output to shreds in favour of dual subs - using listening position measurements as a rough guide.

 

I never really finished TBH, I do things in fits and starts

[tripitaka]

Signing off :  My view remains that CamillaDSP (via Moode) is a fabulous, user-friendly choice for zero dollars, but I have returned to using my MiniDSP OpenDRC.

 

In fact I have returned back to my old RPi3, which I had rigged to provide both SPDIF and I2S outputs and so now I'm using the I2S path as the 'bypass channel', with the x4 MiniDSP DSP curves selectable over the SPDIF output. I feel like this is as close as I can get to the 'best of both worlds' outcome that I was originally seeking from onboard CamillaDSP processing using RPi4.  I'm also quite glad I discovered Moode which I'm still using (albeit without its brilliant Camilla interface).

 

Edited September 29, 2022 by tripitaka