MiniDSP

[Camverton]

@Camverton I've just spent a bit of time trying to see if I could work out what's going on better in terms of the ETC changes you've seen. I can't see anything like 15 dB reduction you've seen but I can see some peaks reduced by up to a couple of dB, and a smaller number actually increased in amplitude with DL on. I'm sure this is all tied up with the changes to the impulse response shape, how surfaces reflect different frequencies, and how the contributions of multiple reflections all sum up. It's complicated! Definitely something that's made me think but it still remains a physical impossibility for DL to reduce boundary reflections.

Just to check actually, your 15 dB change is for a single speaker measured on it's own I assume? Rather than two playing together?

Cross posted, thanks for your reply. Yes, this is of course a measurement of just one speaker not a stereo pair.

Just to be clear the mathematics involved doubtless require a far better brain than mine and it is indeed complicated so I speak from what I have observed (and hear) rather than knowledge of how it works. Do bear in mind that the 15 dB reduction is when using an omnidirectional speaker which is likely to be more affected by dirac.

Current setup in my main, arguably too lively, room is dirac up to 99 Hz with SHD handling crossover at 40Hz from SVS subs to German Physiks omnis positioned over 2 metres from the front wall and slightly less than 1 metre from the side walls. The speakers are positioned to give the most realistic illusion of having a performer in front of me, although that is not the best position for even bass, hence the subs and use of dirac. For small scale ensembles or tightly focussed soloists I have a slight preference for using dirac. For large scale orchestral music the reverse. It’s all about the illusion, and wonderful to have a choice - cheaper than changing components!

 

[Camverton]

For clarity I'm not questioning the ETC changes happen but refute the conclusion that the changes demonstrate that DL reduces boundary reflections. This has nothing to do with the DL algorithms as it is physically impossible for them to alter loudspeaker directivity. The complication of looking at an ETC is it sums up what happens at every frequency all at once, which is why the factors I described can change what is seen but without there being any change in direct vs reflected signal amplitudes at each frequency.

Ultimately it makes no practical difference whether I'm right here of course. It's whether someone enjoys the sound of the result that does, no matter how it is arrived at.

One thing’s for sure you’re quicker at posting than me! As for your last paragraph - amen to that, and listen to some wonderful music is exactly what I am about to do now!

 

[MartinC]

Do bear in mind that the 15 dB reduction is when using an omnidirectional speaker which is likely to be more affected by dirac.

That could certainly be a reason why what you can see is rather different to my experience in terms of ETC changes, but won't change whether DL can affect boundary reflections or not. It's the interpretation of the ETC change where we differ, not that they occur.

 

[MartinC]

Somewhere above I made the point that the frequency 'curtains' in Dirac Live that can be dragged to define the correction frequency range do so for both the amplitude and phase correction components. I've just done some testing to prove this point and have actually come across a recent bug but first I'll share what I'm sure is the intended behaviour. This is currently valid for the older style linear target curve, and the new style after the 'convert to points' option has been selected (I'll cover the bug in a subsequent post).

Below are graphs showing first the amplitude and second the phase response of example Dirac Live filters (not the in-room response)*. These show what adjustments are being made by Dirac Live. Each graph show a full frequency range example (10 Hz to 24 kHz) and one where to frequency curtains have been dragged in to 21 Hz and 150 Hz. The amplitude response adjustment is clearly limited to the selected to frequency range. The phase response is slightly affected by very small time differences that the time shift resolution in REW prevents me from removing, but if we focus on the detailed frequency-dependent peaks and troughs this is clearly limited in the same way.




*I measured the filters by using a USB loopback to REW, with my SHD acting as a 'microphone' rather than this being from an acoustic measurement.

 

[MartinC]

Here's the bug I've found... With Dirac Live version 3.3.0 there was a new default target curve introduced that uses low and high shelf filter shapes. From the extremely limited changelog notes I believe the intention is to broadly match the existing in-room response and so to not drastically change the existing tonal balance. What I've spotted this morning though is that if this is used and the frequency range is limited then there are amplitude response changes outside the defined range. Below is a comparison like my previous example but using the new target curve. With the amplitude response example its clear that a +5.5 dB low shelf and - 2.0 dB high shelf filter is being applied beyond where the frequency limit curtains have been set. If the option to 'Convert Shelves to Control Points' is selected then this problem is fixed and the behaviour is as my in my previous post. The behaviour I'm showing is apparent within Dirac Live itself, where it's clear that the indicated result shows changes beyond the defined frequency range in the same way.




I'll contact Dirac about this as this doesn't look like the intended behaviour to me.

Edit: here a screenshot from the Dirac Live application demonstrating how it also shows changes outside the set frequency range.

 

[Camverton]

Just to confirm that I’ve observed the same behaviour. With the right hand curtain dragged to 105 Hz the treble region is reduced by a couple of dB when using auto but not when manually adjusting the target curve to look much the same as the auto target up to the right curtain position.

As it happens I normally use manual adjustment of the target curve but as I was only correcting up to 105 Hz I thought I would try auto. Bit surprised to find the effect on the treble and then remembered Martin mentioning something about it.

 

[eddie-baby]

What are the thoughts on DIRAC anyway, can't you just tune with REW manually to a target curve and then to taste by ear?

Why do you need someone else's software?

 

[MartinC]

Just to confirm that I’ve observed the same behaviour. With the right hand curtain dragged to 105 Hz the treble region is reduced by a couple of dB when using auto but not when manually adjusting the target curve to look much the same as the auto target up to the right curtain position.

As it happens I normally use manual adjustment of the target curve but as I was only correcting up to 105 Hz I thought I would try auto. Bit surprised to find the effect on the treble and then remembered Martin mentioning something about it.

For info. I have reported this to Dirac and when requested I provided them with a couple of example data sets showing this but have heard nothing more. My bet is they're working on a fix that will be rolled out.

For info. there have been two new Dirac updates since my post above but I've not tried them.

 

[MartinC]

What are the thoughts on DIRAC anyway, can't you just tune with REW manually to a target curve and then to taste by ear?

Why do you need someone else's software?

Dirac Live is 'smarter' than what can be done manually in REW in terms of how it improves the phase response, although how audible this measurable improvement may be I'm unconvinced of. Have gone from doing things manually in REW for me the single biggest change in going to DL was that it makes it easier to generate corrections based on a weighted (I think) average response over the measurement volume, and I do now think it's best to base EQ on spatial averages rather than single point measurements as a result.

For context I don't personally like the effect of Dirac Live run over the whole frequency range and only use if for the low bass.

Edit: I should have stressed that Dirac Live is fundamentally more user-friendly that REW for the inexperienced/those who don't want to spend lots of time learning how to do new things well.

 

[eddie-baby]

Dirac Live is 'smarter' than what can be done manually in REW in terms of how it improves the phase response, although how audible this measurable improvement may be I'm unconvinced of. Have gone from doing things manually in REW for me the single biggest change in going to DL was that it makes it easier to generate corrections based on a weighted (I think) average response over the measurement volume, and I do now think it's best to base EQ on spatial averages rather than single point measurements as a result.

For context I don't personally like the effect of Dirac Live run over the whole frequency range and only use if for the low bass.

Edit: I should have stressed that Dirac Live is fundamentally more user-friendly that REW for the inexperienced/those who don't want to spend lots of time learning how to do new things well.

That's interesting Martin. As you've highlighted here 'if' DIRAC does things perfectly to your environment its still down to personal preference generally that wins the show.

I can kind of understand what it's about though and I do definitely think it has a future (along with many others of which are possibly not yet with us and will be a lot less costly).

I do think DSP's and the way they are developing are the future of tailoring hi-fi to individual needs though. And if you are using a multiple speaker set up and trying to integrate subs as well they are a no brainer, if the tech and software compliment each other.

I only this week tried yet another digital DSP ('not a miniDSP brand') and the overall effect of integrating mid bass's and subs was staggering but it had other issues which caused me to bin the project. I am kind of waiting (hoping) for the Flex HT to be the answer and to be released as I think it may tick most of the boxes I need. Hopefully the price wont be silly and I can finally get a DSP system done to my liking. I really don't think I'll be paying the extra for the DIRAC though, no interest at the moment anyway. Think its costly and doesn't guaranty anything! Maybe they should consider a trial period, then if you really like it you pay for it!

 

[Tony_J]

If you run Dirac on a PC you can get 2 weeks free trial. Feed USB out into your miniDSP - job done. If you like the result it is cheaper to do the miniDSP upgrade than it is to buy the PC license.

 

[Camverton]

What are the thoughts on DIRAC anyway, can't you just tune with REW manually to a target curve and then to taste by ear?

Why do you need someone else's software?

Yes you can just tune with REW manually and, with experience, get a good result, although one still needs to feed the results from REW into software or hardware. My current setup has a preset which does just that but in practice I use mostly the Dirac Live corrected presets. To my ears the greatest advantage of using Dirac is that it tidies up the impulse response which leads to more precise imaging. The effect of this varies with not only the speakers but also, to some degree, the recording.

At first sight Dirac Live appears to do it all for you but there are choices to be made, both in measurement and target curve manipulation, that will affect the final result. I suppose it’s a case of if at first you don’t succeed try again and possibly make things worse before eventually making the sound a lot better. It must be said that when it comes to sound quality one man’s better is another man’s worse!

 

[MartinC]

That's interesting Martin. As you've highlighted here 'if' DIRAC does things perfectly to your environment its still down to personal preference generally that wins the show.

There is isn't really any such thing as 'perfect' for Dirac to achieve, and as @Camverton pointed out it is very flexible in terms of target curve and the defined frequency range of affect to be tunable to taste. If it were only possible to run Dirac Live over the whole frequency range then I would still be doing things manually instead.

I can kind of understand what it's about though and I do definitely think it has a future (along with many others of which are possibly not yet with us and will be a lot less costly).

I do think DSP's and the way they are developing are the future of tailoring hi-fi to individual needs though. And if you are using a multiple speaker set up and trying to integrate subs as well they are a no brainer, if the tech and software compliment each other.

Even without subs I think the key appeal is in sorting out big bass response problems. Without DSP I'd be using smaller main speakers that didn't excite a horrible 35 Hz mode in my room, and many audiophiles have historically been forced to make this trade off.

I only this week tried yet another digital DSP ('not a miniDSP brand') and the overall effect of integrating mid bass's and subs was staggering but it had other issues which caused me to bin the project. I am kind of waiting (hoping) for the Flex HT to be the answer and to be released as I think it may tick most of the boxes I need. Hopefully the price wont be silly and I can finally get a DSP system done to my liking. I really don't think I'll be paying the extra for the DIRAC though, no interest at the moment anyway. Think its costly and doesn't guaranty anything! Maybe they should consider a trial period, then if you really like it you pay for it!

I should expand a little on the potential advantage of Dirac Live when run over the full frequency range as I probably undersold it. First from a technical perspective it is 'wrong' to take a single point measurement in REW and then apply filters that make the response fit a chosen target curve over the full frequency range. The reason is that our hearing handles reflected sound differently to what arrives directly from the speakers whereas a microphone doesn't, so what we see on a graph in REW does not fully represent what we hear. In the low bass region though the graphs do represent what we hear and so applying EQ here is much more obviously 'correct'. A key distinction with Dirac Live is that it at least tries to be smarter in terms of a full frequency range correction. This is due to its multiple measurement point approach and algorithm I think distinguishes the first arriving sound to at least some degree.

Where miniDSPs have an advantage over most rivals is subwoofer integration and flexibility. On the latter I'll just highlight that whilst I don't run Dirac Live over the full frequency range my miniDSP SHD does allow me to apply manual EQ on top of the Dirac Live filters, so I can for example let Dirac Live handle the bass response but then apply manual EQ to apply a treble lift in more of a tone-control like fashion which is VERY different to forcing an in-room response to fit a target curve with a treble lift.

 

[eddie-baby]

There is isn't really any such thing as 'perfect' for Dirac to achieve, and as @Camverton pointed out it is very flexible in terms of target curve and the defined frequency range of affect to be tunable to taste. If it were only possible to run Dirac Live over the whole frequency range then I would still be doing things manually instead.



Even without subs I think the key appeal is in sorting out big bass response problems. Without DSP I'd be using smaller main speakers that didn't excite a horrible 35 Hz mode in my room, and many audiophiles have historically been forced to make this trade off.



I should expand a little on the potential advantage of Dirac Live when run over the full frequency range as I probably undersold it. First from a technical perspective it is 'wrong' to take a single point measurement in REW and then apply filters that make the response fit a chosen target curve over the full frequency range. The reason is that our hearing handles reflected sound differently to what arrives directly from the speakers whereas a microphone doesn't, so what we see on a graph in REW does not fully represent what we hear. In the low bass region though the graphs do represent what we hear and so applying EQ here is much more obviously 'correct'. A key distinction with Dirac Live is that it at least tries to be smarter in terms of a full frequency range correction. This is due to its multiple measurement point approach and algorithm I think distinguishes the first arriving sound to at least some degree.

Where miniDSPs have an advantage over most rivals is subwoofer integration and flexibility. On the latter I'll just highlight that whilst I don't run Dirac Live over the full frequency range my miniDSP SHD does allow me to apply manual EQ on top of the Dirac Live filters, so I can for example let Dirac Live handle the bass response but then apply manual EQ to apply a treble lift in more of a tone-control like fashion which is VERY different to forcing an in-room response to fit a target curve with a treble lift.

Interesting Martin. We are certainly getting into another dimension with hi-fi now, pretty soon I can see arguments not over DACs or Cable but who's got the best frequency/sound algorithm

I don't know what's worse amps with no tone controls or an infinite amount of adjustment I can see why people like to keep thing simple anyway.

 

[bencat]

Eddie in my case the argument is over which is the better clock and this improves ( I think not tried it yet with DIRAC) all digital files by some margin .

 

[eddie-baby]

Eddie in my case the argument is over which is the better clock and this improves ( I think not tried it yet with DIRAC) all digital files by some margin .

So your still not in the DIRAC camp yet but happy with what DSP can do is it?

 

[bencat]

No Eddie I have been in the DIRAC camp for over six years now starting with the MiniDSP DDRC 24 two way and then moving on to the stand alone DIRAC unit DDRC 22D of which I have two but one is currently not being used . I love what DIRAC live does right across the board . The external clock revelation i have had works with any digital chain that has an input for an external clock . As far as I can tel you it is the use of a more accurate and stable clock as the master timing clock that makes any digital file just sound better . I am currently using it is my main system which has no DIRAC unit in the chain . However as all things DIRAC are digital I think that adding this to one of my other systems will give them the huge improvements I have heard with my current system . I am going to do this for the two way active system I am showing at the Wam Show but I am resisting giving this a try out and just doing it when i arrive and see what happens . Unlike MartinC I have always used DIRAC right across the spectrum and like what it does from the highs to the lows . The good thing about it though is is has such versatility that you can dial in what you want just the treble or the bass or the mid range or right across all of them . I am also impressed that it sounds better now than when I first tried it and they keep working on it .

 

[Kubs]

I'm thinking of signing up to Qobuz ... How would I play this using the minidsp studio as a streamer? There is no volumio plugin and nor can I cast or use qobuz connect (if there is such a thing) .....

Advice / steer grately received. Thanks

 

[MartinC]

I'm thinking of signing up to Qobuz ... How would I play this using the minidsp studio as a streamer? There is no volumio plugin and not can I cast or use qobuz connect (if there is such a thing) .....

Advice / steer grately received. Thanks

The version of Volumio on the SHD series includes Qobuz support without an additional plugin, which is why there isn't one. I use Qobuz on mine. I think it's under My Music where you enter your Qobuz login details and then the icon for it will appear.

 

[Kubs]

Cool. Thanks. I will check it out when I get home. I'm assuming the UI will be the same basic volumino interface?

Would be great to control qobuz from the app on the phone as the experience is more enjoyable. I assume there needs to be a connect type arrangement to link to the shd. ....?