A guide how to do room correction and use it in Roon

Thanks again, @Magnus! My listening position is a bit right from the middle, can I ignore this or do I have to measure differently? I could measure from the middle, but then when I sit down and play music, I guess the results aren’t optimal. It will be hard to change my listening position in our room though. What’s the best alternative in your opinion?

1 Like

can you not export the Dirac filters into Roon to give you access above 96/24?

.sjb

Nope, they’re proprietary, I’m afraid. But even if I could – I’m not terribly impressed by anything over 96. My Mojo goes to 11 (well, 768), but still I don’t particularly care for upsampling.

Anyway – we’re drifting off (sorry 'bout that) – I think @Magnus is on a path that is interesting to a lot of Roon users.

2 Likes

I wish you could, but I doubt Dirac would allow a free trial period then :slight_smile:

I am currently experimenting with using the same measure points as Dirac does, and using sweep instead. I would guess those measure points are carefully selected. Lets see how that turns out.

You can use this guide anyway, but you probably wan’t to use different targets for left/right speakers instead of as I suggest using the same (step 11 and 13). Just press “Set target level” or manually adjust for both speakers in step 11.

Then if you like you can always add +dB to one speaker in Roon to get sound more centered.

2 Likes

I have no experienced using network player, but for measurements maybe it would be easiest to use the DAC directly connected to computer.

I just tried moving the microphone in a “Dirac” pattern, which made a small improvement. Basically, imagine you have a 20x20x20 cm box on your shoulder, and move the microphone to one corner, wait 10 RTA averages, then the next corner and so on until all 8 corners are visited on one side, then move over to other shoulder and repeat. In total 160 averages for each speaker.

Updated the guide with this (only step 6 was changed).

5 Likes

I can’t wait to try this update later tonight. The results so far have been a huge improvement for my listening situation.

1 Like

That’s it, @Mark_Allen! I’m in! :wink: I’ll order a USB mic and am very curious what this will bring :smile:

1 Like

@koen

Its pretty amazing I’ve got to admit. Guess I’ll be re-listening to a bunch of albums in the next few days! This guide is extremely helpful(along with a few other threads about room correction) especially if you’ve muddled through the various online videos and tutorials scattered on the web. They are helpful to a degree just so you have a basic understanding but as @Magnus points out REW isn’t the most user-friendly piece of software. Keep reading and sharing that’s what the community is all about and thank the developers for incorporating DSP functions to Roon.

2 Likes

Mark, would you mind describing in your own words (without tech jargon) exactly what improvements you experience? Interested to read a bit more about your findings, as I’m completely new to room correction :slight_smile:

2 Likes

@koen
Sure, I’d be more than happy to, but I’d like to try the update @Magnus has made to his/her guide first. I’m not very good at describing my feelings (that’s what my last girlfriend says) but i’ll try lol. What i’m hearing right now is less /no harshness in the highs and a hell of a lot more accurate bass. More presence I suppose.

2 Likes

I tried a trick to fix phases in the output, by using REW and rePhase in combination, and the result closed the gap between what I had and Dirac. I can not tell them apart, if anything the REW/rePhase room correction sounds slightly better. But then I have spent like 100 hours on REW and about 1 hour on Dirac, so I guess its not quite fair to make that comparison.

But its a very tricky step to make, and I am not very familiar with rePhase. I tried to describe how I did in step 16 (I replaced the old advanced step which isn’t very useful now when we make measurements on both sides of head with both speakers). But if you get it right, it sound pretty amazing, and gives that complete sound that Dirac gives.

3 Likes

Thanks for your efforts @Magnus! I have tried using REW to emulate the EQ side of Dirac - I took measurements in the same 9 positions as Dirac suggests and did an average. Struggled to get a good set of EQ filters out of it though. Didn’t try RTA so looking forward to trying this out. And I didn’t try any phase/time domain stuff via rePhase though it was on my to-do list. I will embark on this using your process as a new baseline and see what happens.

@magnus has done fantastically helpful work, I applaud it.
I did a much smaller summary of how to use Acourate, @rovinggecko added a lot to it.
Just one observation to the concern of @koen and others: good room correction does much more than tweaking bass and treble, because they address phase, which means time. The audio signals we work with have two components, we usually talk about amplitude and phase, we could equivalently talk about x and y. When you only look at amplitude, which is what bass and treble controls talk about and simplistic equalizers, you are missing half of the signal. It’s like saying, how do I get from New York to Boston, and the answer is, drive 200 miles! Which direction? Never mind, just drive 200 miles! 200 miles is the distance, the amplitude, the length of the line, but we need the direction, the phase.

Phase is not easy to think about because by itself we can’t hear it, but phase shifts among the various frequencies mess things up. Various parts of the signal arrive at different times which causes confusion. Imagine you hook up a subwoofer with a perfect crossover but you put it at the other end of a football field. The bass arrives 1/3 of a second after the rest, that would be bad. That’s a gigantic phase shift. Silly, but phase errors matter even when more realistic.

That’s what a good room correction system does.

I’m no expert on this, but I am concerned with just using the parametric equalizers, because just moving those sliders does not address phase. I have tried that, in several systems, and have been dissatisfied.

But a good room correction system deals with the whole story. I was not happy with my experience with REW, but I may not have done it right. (Didn’t have @magnus guide!) I eventually figured out Acourate and it is great. This is not a critique of any systems, just a comment on why RC can be so good.

It really obliterates all the focus on details in DACs and power supplies and cables. The room causes huge distortions, compared to those other issues which are real but smaller. It’s like listening to bit-perfect signals through a cardboard tube.

4 Likes

I’d like to thank all of you for your invaluable contributions to my listening experiences. Thanks to this forum – and thus also thanks to the Roon team, because without them there wouldn’t even be a forum – I have learnt SO much more about music and listening than “just” some CD tips (also invaluable, and my most visited thread: “what are you listening to now?”). Roon and my co-Roon users continue to expand my musical horizons in so many ways and I just want to say: thanks for that! :blush:

5 Likes

[quote=“AndersVinberg, post:35, topic:23800, full:true”]
But a good room correction system deals with the whole story. I was not happy with my experience with REW, but I may not have done it right. (Didn’t have @magnus guide!) I eventually figured out Acourate and it is great. This is not a critique of any systems, just a comment on why RC can be so good.

It really obliterates all the focus on details in DACs and power supplies and cables. The room causes huge distortions, compared to those other issues which are real but smaller. It’s like listening to bit-perfect signals through a cardboard tube.
[/quote]I added the phase stuff today, and it did bring a final touch to the sound, giving a greater clarity and a more complete sound. However, for me, considering the room I sit in, the basic stuff gave a much greater improvement (look at my before and after screenshot at the end of the guide).

I am not 100% sure I did the phase calculations in rePhase correctly though, it definitely improved the sound but I am not familiar with rePhase or phase calculations. Maybe someone more familiar with rePhase could check the screenshot in step 16 and see if I did it optimally :slight_smile:

Great tutorial @Magnus
I am curious why did you choose to us the PinkRN to do the sweep rather than the LogSweep that REW suggests to use? My understanding is that the logSweep provide extra information that is useful to calculate delays, sound reflection and so on
Thanks
Marco

You are correct about the extra information that a sweep provides, that is why I use that in my advanced step 16, since I need phase information for rePhase. However, for normal frequency modification its not needed (and its removed when you do averaging of several sweeps).

And using the RTA has some advantages, it automatically averages the response on a good level, and you can move the microphone so you avoid small pockets with lots of + or - level of higher frequencies. Having said that, you can do 16 sweeps instead of using the RTA at step 6 and get a very similar result.

For example, if I do a sweep at my head position, I get a +7 dB level area at around 1500 Hz, but if I move the microphone 10 cm I get - there instead. Using continuously RTA and moving the microphone gives a more reliable measurement.

1 Like

got it!
Thanks I will give it a go on the weekend

cheers

Marco