Room-tuning software

Gracie is her name and she is my shadow. Unless someone else has food then the mooch in her takes over.

She sits on my recliner with me and the loud music or movies do not bother her in the least. She sits and sleeps right through them. Large sudden loud things in movies get her once in a while. But not much else. Dog barks in movies and she pops her head up and looks for them. Too funny.

3 Likes

OMG, that is SO cute!! We’re down to 2 dogs at present - had 4 a few years ago, also have two birds - a cockatiel and an African Gray parrot. A few years ago, my wife’s Double Yellow-Head parrot passed away. She had her for 30 years. They could sing opera duets together, like nothing you’ve ever seen. One of our dogs right now, a Chihuahua/Jack Russel mix, is the only dog I’ve ever had who is totally fooled by animals on TV - growls and barks at them as though they’re in the room. Our other dog looks at her as if to ask, ā€œWTF is wrong with you?ā€

We had four until about 3 months ago. She was 17. That was a bit much. 3 is a bit much.

Hey Brian, I want to thank you directly for recommending Focus Fidelity. In response to today’s Paul’s Post, I placed this comment:

This whole series of Posts has coincided beautifully with my first experiments using DSP room correction software. About two weeks ago I followed the advice of someone in the PS Audio forum’s new Rooms & Treatments category(something I suggeted to Paul McGowan to add and surprisingly he thought it was a good idea), bought a measurement microphone ($100) and software for measurement (free) and filter-generation ($250) from Focus Fidelity. There are certainly MUCH more expensive options for this, but FF is in my budget and a huge shout-out to David there who answered my newbie questions and recommended a fantastic book to help explain the whole procedure.

The general philosophy behind this is that your room - whether it’s a bedroom, family room, outhouse, or dedicated isolated listening room like mine, has unique acoustic characteristics that WILL change how your system delivers music to your ears. Sounds like EQ, right? But it’s a lot more than EQ. It makes many of the assumptions Paul made when he said in this Post, ā€œSome designs simply can’t reproduce spatial information well—either because their off-axis response is uneven, or because they smear phase and time alignment.ā€ Just like speakers, rooms affect these characteristics. What’s more than EQ is this statement from the intro to the book: ā€œthe goal is to accurately reproduce music as close as possible to the content on the recording.ā€ ( Barnett, Mitch. Accurate Sound Reproduction Using DSP, p. 6). So it’s not ā€œcoloringā€ sound as much as doing what it can to reproduce what the engineers and producers heard when putting the recording together. The author is singing my song because he has great experience on BOTH the recording and playback sides.

So last night with great anticipation and a bit of nervousness, I plugged the digital filters into Roon… and… I was amazed. From the get-go, I noticed that I lost no detail, but the detail was more homogenous with the overall sound. The center-channel features - vocals, instrumental - are still there, but better mixed, less ā€œedgy,ā€ a bit softer. It’s funny, but in another Post I was saying how my ears had to get used to that center channel being so pronounced when I first started listening in my room a few years ago, but now I think they might have to adjust to NOT softening the center channel so much. The soundstage is every bit as wide, but central to this Post, DEEPER and more 3-dimensional. And reading the suggested book, I see how it has a lot to do with time accuracy. Not just in the center, but all across the stage, things were immediately deeper. As for definition, I need to listen further, but nothing has been lost.

So the reason I replied to you [pmj in the Post] is to use your list:
1 phantom center.vocalist - BETTER
2. the singer is in the room and the speakers disappear - BETTER
4. the soundstage is 3D and is beyond the width and depth of the speakers - BETTER
5. 4. I am transported to the venue - BETTER, except there were no glasses clinking.

Again, thanks for the recommendation. More experimenting to come, but I think this might be the best single improvement I’ve seen since building the room.

Glad to hear it worked it. I found the same things when I did it the first time. I had someone else do it for a fee and they used REW. I looked into and i was like yuck..free, but yuck. Found Focus Fidelity and for $250 figured I would give it a try. The directions looked much easier. And they were.

Doing frequency, timing, and phase correction convolution files in REW is tedious, but possible. I’ve done it, and it’s tedious! :weary_face: I wonder how the Focus Fidelity convolutions compare?

There are many adjustments you can make in there, some of which I take the default. Some I change. I do not do anything above 5khz for instance. Some people don’t do more than 1k. I found it sounded better doing up to 5k. I found out that’s what Anthem ARC does too. Made me feel better about my findings. There were a ton of internet know it all’s that insisted do not go higher than 1k.

This time around I used the quick measure tool in it as I moved speakers and seat around to find best results. Then did the 9 measurements. Getting the subs using high level inputs really takes a leap forward when you room tune.

I end up using the FLAT response from FF. They do have other curves they can apply. But I use flat and then add a PEQ for pushing the bass up in Roon. That lets me adjust the PEQ on the fly for different material. Some you want a bit more bass to get the feel of the drums, others it seems too much. Using Roon PEQ makes that easy to change on ipad.

That’s interesting what you say about not doing anything over 1 or 5k (glad I’m not only using my 78 rpm records because they top out not far from there depending on when they were recorded). When I emailed with David at Focus Fidelity, he didn’t recommend using ā€œflat,ā€ because most people find it a bit on the brittle side. But, I argued with him, my goal is to hear what the engineers, mixers, etc. heard, so why would I want to emphasize anything? He pointed out something I’d never thought about - which IS covered in the book he recommended - don’t forget that the engineers, mixers, etc. were also in a room that affects the sound. Studios, not only in the recording rooms, but also for monitoring and mixing, go to a lot of trouble to set up characteristics based on knowledge of acoustics, human habits, etc., so THEY are also using DSP to tune their environments. He recommends using filters that emulate these control / mixing rooms. Focus Fidelity only has 6 options other than flat, so I chose the Harman one to begin with, but I’ll experiment with others. Here is a screen capture of how it went with my room and measurements. Basically the translucent graphs are from the measurements and the solid lines are what the Harman filter is doing.


David at FF said that lower hump is probably a node in my room due to its small size.

I tired the harmon and went back to flat. I do not remember why..lol

This is the way. Using a cutoff around the upper limit of human vocals (5K is a little beyond that, but to each his own) is a smart way to do it.

Yep. I have something similar, although not as humpy. Am I seeing this right, though? The convolution filter is raising that ~60Hz notch by about 7-8dB? That’s seems like an awful lot. Everything I’ve read suggests that if any upward boost correction is done at low frequencies, it shouldn’t be more than 2-3dB at most. Beyond that and the recommendation seems to be to use appropriately placed sub(s) to cancel the notch.

I believe its 45ish to 42ish… so only 3.

1 Like

Ah! Missed the little dots above the solid red. Thanks.

software can get confusing with all the lines. I just leave on the before and after so I can see what it averages out to for fixing.

You are correct it does not do much for the dips, as you know it cannot make stuff up… peaks are taken care up unless really big.

Takes about a minute with Bacch ORC.
Put a microphone in each ear, two 10 second tone sweeps, 25 seconds of processing. Done.

4 Likes

By far the easiest room set up there is! Even I can do it! It’s like having a genius (Edgar) come into my room and say this is how it’s supposed to sound!! It’s sounds great but you already know this.

3 Likes

Takes about a minute with Bacch ORC.

Yeah, but at my salary it would take a lot longer to afford it :face_holding_back_tears:

2 Likes

And I’m getting ready to try the flat - no idea why. lol. I’m also experiencing a bit of a shift to the left of my center channel. And it’s a touch more spread out. I’m a bit torn because spreading out the center channel is probably more realistic, but I’ve grown to love it to be pronounced, especially for vocals. I’m going to live with it for a little while to see if my ears adjust. And I tried moving my left speaker in about an inch because I found that it was not exactly the same distance from its wall as the right speaker. So I guess I’m going to need to re-do my measurements, which is probably a good idea anyway. To deal with that center-shift, I did something I’ve never done - adjusted my pre-amp balance control, but my brother told me I could just set up a Procedural filter in Roon to bump one side up a couple dB. He also said I could use a Procedural filter to compensate for the 6 - 8 dB reduction caused by the Convolution filter so I could more easily compare the filter being off vs. on.

In other words, I ain’t done sperimenting yet.

Yeah shift left seems like you should be able to see in your magnitude results.

If I had to guess it seems like its a bit higher on the right except in the 300-600 range or so. Maybe a re measure is called for and see what happens. You can swap between the two filters live (windows only)

What I’ve read, which makes sense, is boosting a room related suck out can lead to driver distortion. So it’s best to PEQ peaks and fill suck outs with subs.