Is it smart to use MiniDSP 2x4HD for DSP/FIR/convolution in an Atmos/Auro3D setup?

[baron-bob]

Well, the computer voice is wrong

Both Dirac and Audyssey use mixed IIR and FIR filters. Granted, Audyssey filters are more limited. Still, both can and will do time domain corrections.

Here are some pointers for better correction with Audyssey:

Configuring Audyssey – The Right Way

Find out how to get the best out of you Audyssey receiver in this article by configuring Audyssey the Right Way!

simplehomecinema.com

And:

Audyssey – Simple Home Cinema

Posts about Audyssey written by Roland

simplehomecinema.com

I already got the Secrets of Audyssey PDF in the past and I think I got the best out of it.

But I do not understand your point in general.

Is automated calibration convenient and the best solution, if one do not know how to do manual calibration or does not have the interest to do it? Absolutely! It is not wrong to use it but I want to get more out of my system.

Do you think automated calibration is better or equal to manual calibration at the moment? If yes, we have to agree to disagree on that I guess.

 

[voodooless]

Do you think automated calibration is better or equal to manual calibration at the moment? If yes, we have to agree to disagree on that I guess.

That is a difficult question... It depends on so many things, not the least on the knowledge of the person wielding the tools

Something like Dirac ART will be very hard to do "by hand", and also involves mixing various bits of channels into other channels. You'll need equipment that can actually handle that. As for normal corrections in the time and frequency domain: I think one can do those manually just fine with a bit of knowledge and fiddling. results will be equal to or possibly better than Dirac, mainly because you may have more than the 1 or 2K taps that Dirac uses.

 

[baron-bob]

That is a difficult question... It depends on so many things, not the least on the knowledge of the person wielding the tools

No point in arguing with that. Of course one (I) has to learn a lot. But one reason for me to become a member here is to get the information I need to do it right. My plan is to post my process in detail at some point so that anyone can tell me what I'm doing wrong or were I could do better. I will try to learn what others know and willing to share. At the end there is perhaps a synthesized version of the knowledge in form of a step by step guide with detailed explanation for everyone.

Something like Dirac ART will be very hard to do "by hand", and also involves mixing various bits of channels into other channels. You'll need equipment that can actually handle that.

ART is I think a very sophisticated technology. I can not imagine to reproduce that by my self at any point in the future. I totally lack the mathematical, electronic and physical knowledge. I will be happy to grasp the concepts behind normal corrections. But I'm curios if others will reproduce it somehow and how they gonna do it. I would also like to experience ART some time in a well calibrated system. I'm also sure that I can get better results with MSO when I had the time to dig deeper into it and wondering how strong the difference will be. What I saw on measurements of ART so far looked very impressive but it is hard for me translate that into the possible experience.

As for normal corrections in the time and frequency domain: I think one can do those manually just fine with a bit of knowledge and fiddling.

That is the impression I also got so far.

Results will be equal to or possibly better than Dirac, mainly because you may have more than the 1 or 2K taps that Dirac uses.

If I can get as good as Dirac I would be glad. My system already sounds really good in a way that I can not complain about it. What drives me is that I know it can still be optimized within reason. It doesn't have to be perfect but I want to take all relevant aspects of a HC into account and try to do them sufficiently good. I already explained why Dirac is not an option for me right now but if I already had it or could get it within my budget I would take it of course. Until then I'm fine with manual correction I'm sure. Also the DSP I'm probably going to buy can always be sold ; )

If you can recommend some sources that cover explanations and how-to's of manual room correction I would be glad if you share them here. You seem to have come around a lot on the forum judging by your profile.

 

[voodooless]

There are quite a few free tools to help you:

GitHub - erdewit/HiFiScan: Optimize the audio quality of your loudspeakers

Optimize the audio quality of your loudspeakers. Contribute to erdewit/HiFiScan development by creating an account on GitHub.

github.com

GitHub - greenm01/porc: Python Open Room Correction (PORC)

Python Open Room Correction (PORC). Contribute to greenm01/porc development by creating an account on GitHub.

github.com

rePhase FIR Tool

miniDSP is a leading manufacturer of Digital Audio Signal Processors for the HomeTheater, Hifi, headphone and Automotive market. Join our large community of Audiophiles, Engineers and DIYers using our innovative products.

www.minidsp.com

DRC: Digital Room Correction

DRC: Digital Room Correction filter generator

drc-fir.sourceforge.net

Using REW's new inversion feature for room correction

I am posting this with the kind permission of @Serkan who drew my attention to this brand new method for room correction In this video you will see how to try this with an early access version of REW, available here: https://www.avnirvana.com/threads/v5-20-10-early-access.10787/ I have tried...

www.audiosciencereview.com

 

[OCA]

MiniDSP 2x4 HD practical FIR tap limit per channel is 2042 taps in a dual channel set up, you need to allocate a min of 6 taps in each unused channel. The delay introduced by that convolution filter (96kHz IEEE bin) will be 1 / 96,000 * 2042 / 2 = 10.64ms

You can use these FIR slots for speaker phase correction although what you can achieve in the low bass region is quite limited with only so many taps but you can also use up to 8 biquads/channel in addition to add first or second degree allpass filters.

 

[voodooless]

MiniDSP 2x4 HD practical FIR tap limit per channel is 2042 taps in a dual channel set up, you need to allocate a min of 6 taps in each unused channel. The delay introduced by that convolution filter (96kHz IEEE bin) will be 1 / 96,000 * 2042 / 2 = 10.64ms

You can use these FIR slots for speaker phase correction although what you can achieve in the low bass region is quite limited with only so many taps but you can also use up to 8 biquads/channel in addition to add first or second degree allpass filters.

Practically you should be able to do quite a bit with 2k taps. The delay dependance the location of the impulse response. Since you don’t want too much pre-ringing, the actual delay can be much less than 10ms. By using the delay function of the normal DSP configuration, you can probably save a bit on phase correction latitude as well. Remember that something like Audyssey also only works with about 1k taps.

 

[OCA]

Audyssey also only works with about 1k taps

Audyssey doesn't correct for phase and uses a frequency warping technique to deal with bass filtering which helps its FIR filters act like IIR filters in the low frequency region.

 

[voodooless]

Audyssey doesn't correct for phase and uses a frequency warping technique to deal with bass filtering which helps its FIR filters act like IIR filters in the low frequency region.

It does the correct phase. Obviously, it doesn't have that much resolution in bass frequencies, but generally, the result is still quite good. It also uses IIR filters in a mix. For what you get with the limited number of taps, it's really not bad.