Roon & Home Audio Fidelity (Room Correction / convolution filter creation)


I think if you do I search on ROON and/or CA you will find explanations of how to proceed. I bought a simple $75 mic and downloaded the free REW software and after a bit of effort was able to do it. Once, I had the hang of it it took 15-20 min max. I found the following text on the CA forum:

As an introduction to REW software, there is a good and simple tutorial available here

Once you have set the level (by using left loudspeaker for instance) you’ll have to keep it for all the different sweep measurements.

From a practical point of view I suggest to follow this procedure:

  • set the microphone at the sweet spot position
  • set the level using the left loudpseaker
  • perform the sweep measurement for this position & left loudspeaker and rename it L1
  • perform measurement for this position but right loudspeaker this time, rename it R1
  • move microphone to the next position & perform the measurements for right then left loudspeaker, name it R2 & L2
  • repeat the procedure for the others different positions (I recommend sweet spot + 4 others as a minimum)
  • save the bundle of measurements as a .mdat fil

Mind you, it sounds easier then it is if you have never done it. But, if I can figure it out you can too. There are also lots of much smarter folks then me who have tried to describe how to do it on the different forums.



Thanks @DJD, there actually is a great post @Magnus started:

It is even more elaborate than the link you posted (thanks for that). On the basis of that post, I purches a UMIK-1 and tried all the steps, to find my filter horrible. Which led me to the conclusion that I must have done the sweeps wrong, or something. So I put it aside for the time being, but must admit my curiosity is coming back when reading all these positive reviews about Thierry’s filters.


I say give it a go! There are lots of opinions out there on subjects like do you leave the furniture in the room, do you put the mic at the speakers or straight up etc…and I haven’t a clue what is correct.

A couple points from my experience:

For my measurements I cleared the furniture, set the mic at tweeter height (I built a little stand out of wood), made a grid on the floor that had 9 spots (center spot being the sweet spot), each spot (of tape) was about 12" apart, pointed the mic to the center of each speaker, started at the first spot in the first row, then moved to the second spot in the first row, etc.

In REW you choose the left speaker - take a measurement - save it - then choose the right speaker - take a measurement - save it - then pick up the stand and move to the next spot, etc…

Good luck if you give it a go,


(Mark Allen) #148

I followed @Magnus guide for taking good measurements. Using this diagram as a guide.

You may want to consider a good microphone stand it helped me tremendously.


For those of you using the HAF cross-talk correction filters, how do you find the sound when you move away from the sweet spot?

(Mark Allen) #150

I’m curious as to others response to your question, but in general I rarely move too far from the sweet-spot as that’s where I critically listen to music. That being said I have moved away plenty of times from my sweet-spot and its not too bad at all, frankly I haven’t noticed. I’m sure this is why HAF recommends multiple measurements.


When I move 4-5 feet away (to the left or right) from the listening position the boomy base starts to re-assert itself.


Guys, a quick update: Thierry from HAF has been using me as a guinea pig (I happily volunteered😉). He provided me with some software (not REW), making it easier — at least for me — to do measurements. Also, as already has been said many times here, his support is great.

We were almost in the process of finalising measurements, working towards creating a filter, but we both agreed it would be wise to wait until the large christmas tree leaves the small chamber. So the good news for non-techies like me: Thierry is working on a process to provide us with a smooth and easy process, and he’s currently perfecting it.

As soon as I have some test files or a filter and done some listening, I will report back! Can’t wait for that christmas tree to leave the room already!:wink:

(Mark Allen) #153

Wow this sounds awesome! I’d be more than happy to do another set of measurements using Thierry’s beta tool for measurements. I’m currently using his “head” x-talk filters(beta) and can’t imagine not listening without them. I’ve never been so enveloped in music. It’s so addicting. I wonder if he needs another “volunteer”? :slight_smile:


It does sound awesome, doesn’t it? I suggest you just ask him, I think he will be happy with the help. He’s not ready yet and needs to prepare a quick guide and all, perhaps he can use another guinea pig.

In fact, I’ve just advised him to join this forum, as there’s a lot of interest in HAF and a lot of Roon users with questions. Don’t know if he has the capacity and/or time to do this on top of his work, but I think it would actually save him some time, so both “sides” would benefit.

Either way, meanwhile, my challenge is to think of a valid reason to declare our christmas tree a danger for our safety and get rid of it asap!:laughing:

(Mark Allen) #155

Email sent. I’m more than happy to help if I can. I do think an exact step-by-step guide from HAF would help in HAF adoption rates. I had familiarity with REW and taking measurements so it was easy for me, but like most people I’d love to send HAF the best measurements possible.


Great! I must confess it’s the same for me. I have been intrigued for a while, but didn’t understand anything about it. @Magnus did an excellent effort with his thread and detailed REW instruction, but it was still too difficult for me.

Finally, with Thierry’s step by step assistance, I am confident I will get there. Perhaps not “there” as in the equivalent of a 25K treated room, but a noticeable and affordable inprovement for me. (Please note I still have to hear the first results, but I’m confident the results will be worth the effort.:smiley:)

(Mark Allen) #157

Well you are in for a treat. On first listen I thought something was wrong as I’ve never heard my system sound so good. I thought the bass had gone missing then I realized how it was masking all the detail from the other frequencies. On a good recording you’ll be amazed at the Soundstage as well. I could go on, but its best if you experience it yourself


Hey Mark, I’m very curious. The thing is, my Heresy’s are known for their rather light (but tight and punchy) bass. I’m not a basshead, but I hope bass doesn’t vanish completely. Either way, I think I’ll need to take some time to get used to the sound of the filters once they’re here. I’m particularly looking forward to a more natural sound/balance and the “disappearing” of my speakers. Can’t wait!

(Mark Allen) #159

koen, the bass won’t disappear. For me it just became more natural. For example when listening to Jazz the bass sounds like a real bass instrument. Lots of bass notes that were masked are now easy to pick out. It hard to explain but I think you just have to listen for yourself. I’m not a basshead either. I didn’t lose any bass it’s just clearer and more accurate/natural if this makes sense. Perhaps others can explain it better than I


Thanks Mark, your explanation is clear enough for me :wink: Also, if the results would be anything near to what you describe, that would definitely be worth the money for me!

(Rémi) #161

Now under Beta testing : binaural recording of room response (requires a special headset with microphones at ear level). This goes into the territory of the €€€€/€€€€€ BACCH solutions, probably at a fraction of the cost. Good to see Thierry is still improving the technology.


If I didn’t know you’re enthusiastic about Thierry’s filters, I would think this is a joke. Are you serious? If so, is the headset required for measurements only, or also for the applied filters to experience its effect?

(Rémi) #163

No, not a joke :slight_smile: Don’t expect Head Tracking though, and BACCH does it. But this is a much more expensive and software/hardware intensive solution, prices start at 5k$. Not the same target market

NB: the headset is only needed once for measurements. Again, I haven’t tried it yet, and I am unsure that Thierry will actually propose it if he believes the performance doesn’t really have business potential…

(Mark Allen) #164

This could be interesting. I’m looking forward to his “magic” in the New Year.