Dirac/Acourate/etc for dummies

[ForgeusAudio]

I have a project I'm working on to replace an aging Marantz Pre/pro and HTPC. My current plan is to build a new HTPC that takes over both functions.
Current setup:
Kodi HTPC --hdmi--> Marantz AV7005 Pre/Pro --balanced audio--> Outlaw Audio 7500 Amp/Custom Rythmik subs (2)
Potential setup:
Kodi HTPC --USB--> Okto DSP8 Pro --balanced audio --> Outlaw Audio 7500 Amp/Custom Rythmik subs (2) [later the amp will be replaced with a custom Hypex NCXXXMP setup]

The Kodi HTPC I'm using now is Linux based and I'd like to stay with that as the interface is really good and the OS allows a true appliance feel. The primary use is a home theater with a 5.2 setup. I do listen to stereo audio too, but primary use is home theater.

Looking around I found Dirac and Acourate that can do the DSP/auto EQ function, however both seem to be Windows only. I don't mind having a Windows system to do the initial measurement but would like to stay away from Windows for the HTPC.

Is this possible? Can I do the measurement on one system and then do playback from another? I see there are convolvers for Linux, but just because it is called a convolver that doesn't mean it would work with the files generated by Dirac or Acourate. Logically, there must be a linux based solution as I just don't see MiniDsp using a Windows OS in their Dirac based systems.

 

[Sonnie]

You can use Dirac in Windows or Mac in the miniDSP. I'm not familiar with Linux. I think you would need something that produces the filters for you that will load into convolver. Have you looked at Audiolense (support forum here at AV NIRVANA)?

 

[ForgeusAudio]

As best I can tell, Audiolense doesn't do auto EQ and their convolver is Windows.

I'd like to use a solution that does the auto room EQ. Doing it manually or semi manually isn't of much interest.

 

[Sonnie]

Accurate Sound may can be of assistance if you don't find a viable solution for auto-EQ. Mitch can produce the filters for you and will likely do a better job than the auto-EQ does.

Digital Room Calibration Services, Convolvers, Headphone Filtersets

We help deliver digital room calibration for studio quality sound for any room in your house worldwide. Headphone filtersets and the Hang Loose Convolver.

accuratesound.ca

 

[Focus Fidelity]

The suggestion to contact Mitch is a good one, he might know more about a Linux solution. Our software generates filters as *.wav files so are not proprietary and can be imported by the convolvers in Roon, JRiver, Foobar2000.

 

[BobStern]

Except for Dirac, all software mentioned in this thread can export their correction filters in a standard format that is compatible with almost any Linux program or audio plugin that performs convolution. Dirac's filters are proprietary and cannot be exported for use on Linux.

The easiest option of all is a 109€ service called Home Audio Fidelity that guides you on making the measurements and then generates the correction filters for you.

If you want to design your own filters, you can use Room EQ Wizard (REW), Focus Fidelity, Audiolense or Acourate.

REW is free and runs on Mac and Windows, but it's not as easy to use as Audiolense. (Its forum is here on AVNirvana.)

Among the four commercial Windows programs, Audiolense is the easiest to use and has the unique advantage of a free, unlimited time demo mode that lets you make measurements and use all its filter design features. The main limitation of the free demo is that it cannot export filters, but it can play the first 90 seconds of any music track so you can test the effect of its filters.

Based on my perusal of the user guide, Focus Fidelity seems as easy as Audiolense for filter design, but it requires you to use another measurement program.

If you look at an unsmoothed frequency response measurement, you'll be astounded at how ragged it is. Each of these programs generates a smoothed version of the frequency response measurement and then inverts that to produce its correction filter. A big difference among the programs that strangely is not discussed much is that they use markedly different smoothing algorithms. Another difference is that Focus Fidelity uses measurements at multiple positions to reduce correction of cancellations that vary over small distances.

 

[Focus Fidelity]

Except for Dirac, all software mentioned in this thread can export their correction filters in a standard format that is compatible with almost any Linux program or audio plugin that performs convolution. Dirac's filters are proprietary and cannot be exported for use on Linux.

The easiest option of all is a 109€ service called Home Audio Fidelity that guides you on making the measurements and then generates the correction filters for you.

If you want to design your own filters, you can use Room EQ Wizard (REW), Focus Fidelity, Audiolense or Acourate.

REW is free and runs on Mac and Windows, but it's not as easy to use as Audiolense. (Its forum is here on AVNirvana.)

Among the four commercial Windows programs, Audiolense is the easiest to use and has the unique advantage of a free, unlimited time demo mode that lets you make measurements and use all its filter design features. The main limitation of the free demo is that it cannot export filters, but it can play the first 90 seconds of any music track so you can test the effect of its filters.

Based on my perusal of the user guide, Focus Fidelity seems as easy as Audiolense for filter design, but it requires you to use another measurement program.

If you look at an unsmoothed frequency response measurement, you'll be astounded at how ragged it is. Each of these programs generates a smoothed version of the frequency response measurement and then inverts that to produce its correction filter. A big difference among the programs that strangely is not discussed much is that they use markedly different smoothing algorithms. Another difference is that Focus Fidelity uses measurements at multiple positions to reduce correction of cancellations that vary over small distances.

Hi Bob,
I don't want to divert from the original topic of this thread too much but I can elaborate on the multiple measurements and smoothing as implemented by Focus Fidelity. Multiple measurements are used to build a model of the speakers and listening room. This is much more than just an averaging of multiple measurements, it enables the identification of resonant modes and non-minimum phase behavior. The modeling process also means that the conventional approach of smoothing the magnitude response in the frequency domain is not required. The model does produce a smoothed response but is done in such a way that phase and time domain properties are preserved.
Kind Regards,
David