A deep dive presentation on the fundamentals of “proper” Digital Room Correction (DRC). Includes hands-on DSP FIR Filter Designer demos using Acourate and Audiolense.
Having participated in many audio forum discussions, having watched online videos on Digital Room Correction (or DRC), and having reviewed over a dozen DRC products over the past 11 years, I have come to two conclusions. One is that there is considerable misunderstanding about DRC, how it works and even what problems DRC is trying to solve. And, just as important, understanding what is possible using the SOTA of DRC. I hope you find the content educational and practical.
If you are referring to Hang Loose Convolver, the software can be downloaded for free and one can configure the software for Roon to verify it is working. See HLC Operations Guide page 33 for both Mac or Windows configuration.
Once working, purchase a license key from my site and start comparing using your own filters. If after 14 days you don’t like it, for whatever reason, I am happy to refund the full purchase amount.
@Mitch_Barnett - Great video, thank you. To apply DRC to my Vandersteen 5A speakers, should I adjust their built-in EQ first and then measure them? Or should I plan to not use the built-in EQ at all and just rely on DRC?
Thanks @Nick Re: V5A speakers - nice! It is one half dozen or the other as DRC is going to correct towards the minimum phase target response regardless if the built-in eq is engaged or not. Unless the speakers have been already tuned with the built-in eq according to the manual, I would let DRC do it’s job as it is a much higher resolution solution. I.e. thinking of the FIR filter as a graphic eq, there are 32,768 eq sliders. Then there is the excess phase correction provided by the DRC.
Long ago, I was a Tact RCS user and despite putting in a lot of time/energy, I never got great results. My Tact RCS is now in the back of a closet. (I think I recall your name from the old Yahoo! Tact group.)
I’ve been wanting to give DRC another try for a while. Since I’ll need to re-setup my speaker’s EQ when the crossovers come back from repair, this might be the time to do it. I’m debating between using Dirac 3 and Accurate. The ease of use of Dirac 3 is appealing, but the ability of Accurate to output Roon filters is also handy. Plus, I remember Uli’s work from the Tact days, and respect him a lot. With Dirac, I’d probably buy a miniDSP SHD Studio, so it would cost a bit more. OTOH, the Accurate UI looks a bit overwhelming, so I’m not sure which way to go yet.
Hi @Nick if you are asking me in the context of the thread title, then there are some technical considerations.
re: minDSP if we are talking FIR filtering, then it is a technical limitation of the DSP chips used in the h/w devices. It is the number of FIR filter taps versus FIR low frequency resolution and interval spacing as described in this post on the miniDSP site: https://www.minidsp.com/support/forum/opendrc-series-support/11444-looks-like-opendrc-has-some-new-competition?start=6 The chart is a bit old, but I have not seen anything new… The number of taps affects both the low frequency limit of the filter and its frequency resolution. The formulas are in the article. While there is no industry standard, the examples in the video and results achieved are using 65,536 tap FIR filters or 131,072 taps. However, 4096 taps is definitely better than nothing
Dirac did not make the list because of a handful of technical issues. It is not possible to explain each one in a forum comment. The main technical issue is Dirac’s use of IIR filters at low frequencies. IIR filters cannot address the non-minimum phase behaviour at low frequencies that exists in virtually every room. This requires excess phase correction, which means a FIR filter. While Dirac does use FIR filtering, it is not at low frequencies. One can verify this by using a FIR filter designer to open up a Dirac correction filter, extract the excess phase response and observe no excess phase correction at low frequencies.
In most cases, any DSP/DRC system should improve the sound quality. It is a matter of to what degree considering the thread title.
As far as ease of use, for two channel work, Acourate is similarly Wizard driven (i.e. using Room Macros) as Dirac as shown in the video. Macro 0 is optional and Macro 6 Inter-Channel-Phase-Alignment (ICPA) is new.
@Mitch_Barnett Thanks for a great insightful video. I have a somewhat similar question to Nicks. I’m using Genelec GLM software with a couple of 8341 and a 7350 sub. Do you have any insights into the GLM software compared to Accurate or Audiolense. GLM is very easy to use and I like the relative easy sub integration. I store the GLM corrections in the speakers. Would I potentially be better off skipping GLM and only use Accurate or Audiolense filters directly in Roon?
Hi @TreeZorro thanks for your comments. Actually, it was this review of GLM software on ASR that was the catalyst for me to put this video together.
If you look at the REW before and after measurements in the link, GLM is virtually doing nothing. Put simply, PEQ’s are not Digital Room Correction. Plus the notes I made for @Nick on Dirac similarly apply to GLM as well.
Yes, if one wants SOTA in DRC, Acourate, Audiolense and Focus Fidelity Designer are at the top of the heap. Compare the correction made by GLM in the ASR thread to Audiolense being used to biamp JBL M2’s and digital XO to 4 separate subs. There is no comparison.
Yes, the FIR filters designed with the DSP software mentioned above will run in Roon’s convolver no problem.
Thank you very much for the explanation. I didn’t realize that GLM was just PEQ filters. Very interesting read about GLM on the ASR thread.
I will have to look into one of those DRC programs and figure out a way to add convolution filters to other sources than Roon while still maintaining easy couch volume remoteability.
Hi @Mark_M unfortunately, there are no DRC software on the Mac that meet the requirements. However, these Windows DRC programs can run under Parallels or Bootcamp as an option. Some folks purchase a miniPC for $250 and use that to take the measurements, design and generate the FIR correction filters. The resulting FIR filters are platform independent and only require a convolver (like in Roon) to host the filters.
I guess Dirac was trying to avoid pre-ringing by keeping all FIR in HF region. In fact I think what makes your correction method SOTA is avoiding any kind of ringing and pre-echo despite full inversion of excess phase (as in time reversal).
Fast switching of convolution files is a must to properly compare corrections and I have seen what is possible with your Hang Loose convolver (literally instant switching). Roon convolver takes a few seconds at best and ruins the experience. I have asked Roon to update their DSP convolver engines as a new feature request a good while ago but there has not been any response.
It could be a good idea for Roon to purchase Hang Loose as a license and implement it. We pay a lot of money for this service and clearly deserve the best convolvers.
@Mitch_Barnett I heard https://www.focusfidelity.com/ is coming in a MacOS version soon - this will make many Mac users happy as I think the convenience and performance of this tool is really outstanding.
Another improvement which is coming is „clock drift correction“. I wasn’t actually aware but having read into this and it‘s very interesting. How would you judge its impact / possibilities of further improvement?
No, unfortunately Roon keeps ignoring my request on that matter. I guess they are heavily involved with that ARC thing these days.
Although you’ve asked the question to Mitch, here’s my two cents as per REW on the matter: With USB mics (like the Umik) and especially with longer sweeps, the clock drift was quite significant. To the extent that I was having to use the analog Audyssey mic to adjust speaker timings accurately Recently though, John Mulcahy added a clock adjustment tool and process to REW and the problem seems to be solved:
@Markus_Hubner David’s Focus Fidelity Designer is SOTA DSP and will be very welcome on the Mac. I know David is also working on clock drift compensation in Impala. Audiolense and Acourate already have clock drift built in.
Btw, it is not just clock drift correction with USB mics, but some DAC’s, especially “audiophile” multichannel DAC’s, with no ADC, have a hard time getting repeatable measurements where the time of flight measurement for each driver is consistent. If one is thinking about a MCH system, or biamp/triamp, or multiple subs, a pro DAC with built in mic pre/ADC with an analog mic will save a lot of frustration.
@OCA thanks for your kind words re: Hang Loose Convolver. Implementing instant switching (and automatic gain level matching) of FIR filters is a bit of work First one needs to start with a 0ms convolver, which Roon’s is not. Then how to deal with switching between a minimum phase filter (i.e. no delay) and a linear phase filter with variable excess phase (i.e. delay) in each filter to be compared. Now add multichannel to the equation, going beyond 8 channels, and provide instantaneous, glitch free switching…