MiniDSP in a Naim system

I use a miniDSP nano AVR DL. It's an integral part of my home cinema system, with my main Naim stereo connected via the unity gain function on the 552 preamp, from an Anthem processor. A great bit of kit!

That's what I need to hear,thanks! I think ill buy one as it's in my nature to tinker and the reviews are generally positive. 

Just spend a bit of time setting it up WRC. Note that you also need their microphone, but it's relatively cheap & very good quality.

I took a different approach:

I used a calibrated microphone to accurately determine the frequency response of my system and room, adjusted the sub to achieve the best integration.  I then measured the frequency response of my system and room again.  This response I inverted and used as a calibration for a filter in Audacity.  I used this to edit a copy of the source files on my NAS for playback through my ND5 XS.

Net result:  Good integration of the sub and reduction of the bass boom of the room without risking signal degradation by adding additional real-time ADC and DAC hardware to the playback chain.

Hi WRC, such a device will degrade the sound to some extent.. But it may be a case of lesser evils if you have acoustic issues in your room. Just remember that device in such a config is acting as an ADC and DAC .. and so probably needs to be about twice the cost of your source DAC to have equivalent transparency. Probably better to try and do in the digital domain first to reduce the lossy conversions. Huge's suggestion  is a good DIY solution with reduced processing artefacts if you don't mind doubling up on your media.. could be sizeable and time consuming though.

The Nano AVR DL's purely digital and incorporates Dirac room correction software. The latter in my experience is excellent and in my system has only positive effects. You can apply very gentle and targeted correction, but it does take a bit of experimentation to get the best from it.

Yes, the adc to dac is not ideal but I'm looking at this as a last resort. My room is not the most acoustically friendly and my listening position is smack bang centre, also not ideal because I'm seated in a semi-suck-out zone. 

I've struggled to integrate my two REL T7 Subwoofers, literally days of wasted time,  but I know for certain that these are very fast and musical subs because one of my friends had one perfectly integrated with his hifi and it's quite magical. The only solution I can see is some proper dsp even though I'm wary of degrading the signal I've spent so much time (and money) acquiring! 

I think custom dsp is the next big thing in hifi. Just look at the latest sonos developments and there's already a strong precedent.  Makes perfect sense because, as we know, the end-users' room is the biggest unknown for every hifi manufacturer. 

Perhaps, but as a Sonos user I was totally underwhelmed by the recent their DSP foray.. though was quite fun wandering around waving an iPad up and down..

i must admit, I still feel it's better to to try and get the room to be comfortably correct rather than deliberately distort the sound... Given the source material is ultimately resolution and band limited, these limitations to my way of thinking could well be exagerated with a downstream DSP unit.

DEQX make great DSP gear, and you can still get the basic Premate instead of the DSP/preamp models. Some high quality speaker makers use DEQX in their designs.

A friend of a forum friend uses A MiniDSP in his vinyl playing system and swears by it. I'm rather reluctant to do the analogue-digital-analogue thing. Which is a bit silly because how can I comment without trying it?

I looked at the minidsp stuff at CES a few weeks ago. Their software is.... shall we say... somewhat falling short in the transparancy about what it is doing and what it is telling yo... they get the software from the developers Dirac, and I had some somewhat stern words with Dirac's rep there in the Dirac room. 

OK, to tell you of my concerns. I have no problem with the basic idea. I like the mulitple measurement idea, and the pretty diagram of the sofa.

It then shows you the averaged response as measured.

It then calculates the corrected response, and shows you a corrected line.

You assume, because why wouldnt you... that the corrected line is what will happen in room. It isnt. It is the synthetic calculated correction. Dirac dont make this clear at all to the use. Dirac *should* then go back and do *another* set of in-room measurements and then tune in from that, iteratively. But they dont. You never actually as a corrected in-room response on their software. This is verging on the deceptive. Dirac accept it might be a little "simplistic"

Also, there is no indication whatsoever in the software about where the errors will occur especially at low frequency due to room boundary issues. Frankly, anything below (say) 100Hz is guesswork anyway. And their software should indicate this. But it doesnt.

I like the idea. But the implementation as demonstrated to me by miniDSP and Dirac themselves left me with significant concerns.

Jon - interesting insight - thanks

S

Jon,

Very good observations.  I think you are right on.  While it is tempting to "correct" to a "flat" response at some location, or even averaged over many locations, I expect that would abuse the music signal with too much correction.  A challenge is to arrive at a minimal correction which "does no harm."  A much greater challenge us to create software which accomplishes that for any arbitrary room/speaker combination, in the hands of any arbitrary user.  Perhaps Dirac have done some clever things (I have not used their system, and I am not specifically criticising them), but I would expect that in most cases the best results would be obtained by an informed user (or dealer/consultant?) who uses the system interactively.

I am using "room correction" with 3 or 4 parametric equalizer "audio units" on my Mac, from Audirvana+.  The result is very compelling.  The additional digital computation has insignificant effect on sound quality to my ears, but the sound of the bass in my room is more convincing;  instrumental and vocal timber are more convincing due to the blend of upper bass and lower midrange.  I arrived at this by  extended listening to a wide variety of music, and mildly guided by measurements.

My opinion is that room/loudspeaker correction has enormous potential for improving music replay in the home.  I am guessing we have some way to go before a "turn key" system will work well in any situation.

WRC, it would seem to me that with a good sounding Naim system to incorporate digital room correction (filtering), it should be applied to a digital source before conversion to analogue.  Of course, that precludes application to vinyl source.  For analogue source, as has been stated, you are dependent upon the quality of conversion in the ADC and DAC.  So now you are faced with trade offs, and let your ears be the guide.  I think there are many, even on this forum, who are willing to sacrifice some of "sound quality" of digital conversion in exchange for the "sound quality" improvements which can result from well applied room correction.

Charlie

Jon,

Thank you.  Despite thoroughly reading the Dirac marketing material (and some of their technical documentation) I was unaware of those limitations, although I had noted that they glossed over the detail in those areas.

Charlie,

As with you I apply parametric correction in the digital domain before the data get to the DAC.  In my case I have generated the 'inverse room function' and applied that to the data files using Audacity.  I also did not attempt to achieve a fully flat response, instead limiting the filters so that the amplitude of the resultant peaks in the bass are of just slightly higher magnitude than the extent of the peaks and troughs in the mid-range response.

As with yourself I have measured the resultant response 'in room' and then fine tuned the correction by measurement and by ear (the theoretical correction was fairly close but needed a bit of adjustment in one frequency band).  The final result is a distinct improvement as now no part of the bass response overwhelms any other part of the signal.  Even the mid-range sounds clearer, and readability is improved; and I didn't actually expect either of those advantages.

Huge,

yes, I am aware of your solution.  I have been reluctant to apply the filters to my files.  But I may do that, as my approach precludes UPnP streaming.  We are both applying filters to the signals sent to all speakers.  I think results could be better if we could filter every speaker independently (though that would require a great deal more effort).

We will see significant improvements in this area I think.  And perhaps those of us who can coax good sound out of available tools should hang up a shingle - or at least ccoontinue to share our experiences.

Charlie

CharlieP posted:

Jon,

Very good observations.  I think you are right on.  While it is tempting to "correct" to a "flat" response at some location, or even averaged over many locations, I expect that would abuse the music signal with too much correction.  A challenge is to arrive at a minimal correction which "does no harm."  A much greater challenge us to create software which accomplishes that for any arbitrary room/speaker combination, in the hands of any arbitrary user.  Perhaps Dirac have done some clever things (I have not used their system, and I am not specifically criticising them), but I would expect that in most cases the best results would be obtained by an informed user (or dealer/consultant?) who uses the system interactively.

Charlie

 

Despite what the Minidsp company might say, I don't think anyone who's used Dirac would be content with its automatic correction. The beauty of the software is its adaptabilty and the ability to apply subtle correction over a very narrow frequency range. I use Dirac in both my stereo systems via Amarra Symphony, and with my 7.1 surround system with the Minidsp avr Nano hdmi box. The challenge when trying to optimise multichannels is considerable but the results are really excellent.

 It'd be good if the other Dirac users on this forum could relate their experiences with the software. Amarra offer a free trial period for those who fancy a play.

I agree with Jon that the post-Dirac graphs could be clearer in labelling them as an estimated correction response. A lot of people have moaned about this on various forums. However, to be fair to Dirac and MiniDSP, those who have compared the estimated correction with an actual weighted average response across the multiple Mic positions using REW or equivalent, then the consensus seems to be pretty good agreement.  

And I have to say when using the MiniDSP website to set up REW using a calibrated U-MIK usb microphone on a Mac, the online instructions were incredibly useful.

If you think the information on the MiniDSP little black boxes is lacking, then have a look at the technical documentation and instruction manuals for the new Arcam AVRs with Dirac eq on board! 

jon honeyball posted:

I looked at the minidsp stuff at CES a few weeks ago. Their software is.... shall we say... somewhat falling short in the transparancy about what it is doing and what it is telling yo... they get the software from the developers Dirac, and I had some somewhat stern words with Dirac's rep there in the Dirac room. 

Jon ... just so no one is confused, I think its fair to point out the DIRAC and miniDSP are independent of each other.

Some miniDSP devices use software other than DIRAC, and DIRAC room correction is used on devices other than miniSP.

Sorry for the delay, unfortunately there were unavoidable reasons.

However, I've just been looking at the possibility of using a miniDSP 2x4 to apply my room correction filters using only in the analogue feed to the sub, i.e. leaving the signal from source through the amp to the main speakers unaffected.  I think by doing this I may be able to get the best of both worlds:  
1  The main part of the sound quality of the system (upper bass, mid and HF) will be left completely unaffected by the lower quality ADC and DAC in the DSP box,
2  The DSP box only affects the lower bass frequencies where the ear is a lot less discriminating.
3  Yet at the same time it can still markedly reduce the influence of the major room modes as these only affect frequencies that are solely handled by the sub.
Since I'm only correcting for the resonance of the main room modes (and then only partially, aiming for similar peaks and troughs to the rest of the frequency range rather than a totally flat response), I shouldn't get too many side effects.

I was also concerned that the ADC / buffer / DSP / buffer / DAC propagation could cause significant delay, but reading the literature it seems that the minimum group delay through the box appears to be less than 40μs, so not significant in a sub channel.

For want of the cost two short RCA cables and a miniDSP2x4 (I already have a UMIK-1), I'd be silly not to try it.

Interesting stuff Huge, and a very sound tactic. Keep us posted!

MiniDSP 2x4 box and extra cables have been ordered:

Watch this space: " ".

(That space won't actually change , but I will post my results in this thread when I've installed the DSP.)

The miniDSP and cables have arrived and I'm in the process of optimising the system.  First results are very promising.  The practice seems to correspond with my hypothesis, which is quite encouraging.  At the moment I would say the miniDSP control of the subwoofer solution has a small but significant advantage over applying room correction to the music files.  It's not "night and day", but I feel it has a clear advantage 75% of the time, and only very rarely does it seem to work less well (this only happens with a very few specific pieces of music).  It's notable that the differences are dependent on the piece of music, but are consistent for each piece.

When I've got it fully optimised I'll publish the tools, methods, results and limitations in a new thread (and link in from this one) as effectively I'll be publishing a cookbook solution that others could use.

That's encouraging.  I look forward to reading your write up!

I've published my results here:
https://forums.naimaudio.com/to...on-using-a-subwoofer