I have wondered about not being able to do phase adjustments on the averages of log sweeps or RTA measures. So I get why you are doing what you are doing … but do you think it’s ok to take the phase adjustments from the log sweep and mix them in with/adjust the eq filters from the RTA measurements? Don’t phase and FR adjustments work hand in hand? If one is moved then it affects the other? And so you need to create phase adjustments using the same baseline measurements?
As I said earlier, I don’t know much about the theories behind phase, but I think what rePhase does is to fix the decrease of correct phase when doing filter modifications, and the filter modifications themselves does not contain any phase information. So using the filters from RTA measurement should work, provided you use them on a measurement with phase information (hence the need for extra sweeps in step 16).
But this was an experiment, and I would not be surprised if my way of doing it in step 16 is less than optimal. But it did produce a quality improvement.
Btw, now its night here in Sweden and I have played music on low volume for the last hours. The phase corrections is even more noticeable on low volume for some reason (unlike before, the clarity and details are still there even at low volume).
2 Likes
Thank you for this amazing guide! I really appreciate that you have continued to update it and valuable insights and improvements. I will hopefully receive my UMIK-1 mic today and look forward to following this guide over the weekend.
One part that is not clear to me – how are you outputting the audio test tones from REW through the the rest of the chain? I would like the play the sound files through my full chain (Roon > mR > Brooklyn > amp > speakers – and all associated cables) so that the corrections account for any colorations imposed by my setup. Can you add a tip in your guide on how best to do that? Also, would it be best to disable the Roon DSP (updamling, headroom management, volume leveling) for each zone measured before running the measurements?
Thanks again!
1 Like
I have my DAC connected directly to my computer, and then to speakers, so I haven’t testes how to properly do the whole chain. One option is to save the “Pink PN” sound in REW to a wav file, convert it to a flac and then copy the flac to your music library. Then you can play the sound from Roon, and you should at least remove up-sampling from Roon (I would remove volume leveling and headroom management as well just to be sure).
This way gets more complicated if you want to do step 16 as well (but not impossible), since you need to save the sweeps, and then tell REW to start measurement when it hears an acoustic signal.
But simplest should be to use a long USB cable from your computer, and connect it directly to the DAC. Maybe use a laptop, that you can have in your lap while doing the measurements, and then just select the DAC as output in REW.
Let us know how it goes with your room correction. 
EDIT: laptop in lap is probably a bad idea, due to the influence this will have on your microphone, instead place the laptop behind you on the floor, or something similar. Or in another room, and have a friend for help to start/stop measurements.
1 Like
I’m a bit lost on the instructions, not because there’s a problem with them, but with me never having done ANY Room Correction before.
What if you have a 2.1 set-up over the standard 2.0 set-up, as that’s how the instructions read to me. Does the addition of a Sub make the instructions wrong for my set-up (2 front mains and a sub).
Thanks
John
Im using a 2.1 system myself, so it will work fine. If you feel insecure how to do it all, leave step 16 until later (you can do it anytime later as long as you save the measurements in REW).
In fact, I would guess 2.1 systems in general benefits even more from room correction, since room correction makes the crossover and volume settings on the sub less important (almost redundant).
@koen
So last night I followed the updated instructions from @Magnus and I’m even happier with the results. I tried step #16 but in essence skipped it as I don’t have rePhase (Windows program, no Mac version) As @AndersVinberg pointed out room correction is more than just correcting bass and treble but its the only tangible things I can describe for my listening environment. I know its doing more but I lack the right phrases to describe what I’m now hearing. Presence is what comes to mind. Cliche I know but on a good recording its like the musicians are in front of me. I think its going to take some time to re-adjust to my system which is a great thing. Guess I was used to hearing harshness in the highs and muddled bass. Now I can add more volume w/out it being overwhelming for my small room.
Has room correction made a difference? In my situation hell yes! Music now sounds more musical if that makes sense. What I’m hearing sounds more 3 dimensional in my space with no negative effect on the rhythm, timing, and pace which I love. Thanks again to @Magnus for this guide. Its what I was looking for. I can really tell the difference between a good recording and a bad one 
now
2 Likes
Thanks for sharing your findings, Mark! They absolutely do make sense. I’ve just ordered my UMIK-1 
If you do my guide except step 16 than what you are doing is basically adjusting volume for different frequencies. As far as I know, no time/phase fixes is done (that’s step 16). But fixing a flat frequency is a big step, and definitely leads to more precise sounds, better defined bass, better stereo imaging, etc.
The guide is also fixing frequencies for each speaker individually, so difference in speaker, or room related differences due to speaker positions, will also be fixed.
But I am also more and more suspect of my step 16, how much it does (if anything). I think I can hear a difference with and without step 16, but if so its very very small and might be a slight difference in volume, or just wishful thinking
I have sent an email to the one who developed rePhase, and asked for input to my step 16, but so far no reply.
Anyway, doing my guide but leaving out step 16 gets you very close to what I got in Dirac, I thought earlier that step 16 was the big change from earlier, but it turns out the measuring pattern is the big quality gainer compared to before. I can hear some very small differences between Dirac and the guide without step 16, but its so small and hard to tell which one is actually better. Maybe someone with better ears and better audio equipment than me can tell.
2 Likes
Magnus
Thanks for the reply, and good to know you also use a 2.1 set-up 
So do you just leave the Sub alone when doing the measurements, only the front 2 speakers are measured? Not sure what to do with measurements for the Sub, as at present my Devialet is set-up with the Main Speakers
to Low Pass 2nd Order at 80Hz for the Sub (I’ve a Paradigm Sub 1).
just ordered my UMIK-1, so when it arrives next week can have a better look at your guide (looks gobbly gook to me at present, but just need to take one step at a time and should get it happening).
Great work with the guide 
, and if it wasn’t for this I don’t think I would have taken the leap.
Cheers
John
I measure left/right channel with the sub connected and playing. Doing separate measurements of speakers and sub is possible, but then care has to be taken about overlapping frequencies interacting and it gets complicated (to complicated for me )
Something interesting you will notice, is that the frequency played mainly by the sub will still be different for left and right speaker measurement, even though the same microphone measuring pattern is used and the channels output the same signal to the subwoofer. I would guess this is related to how sound from different sources interacts, but can also be related to timing/phase problems between sub and monitor.
Generally speaking, I want to measure where I sit, and what my ear hears, this is about room correction after all, and not fine tuning individual speakers and/or sub. You can do that also, for example to try and fix phase problems between drivers or between monitors and sub, but that’s another issue and nowhere near as crucial (in my opinion) as getting the frequencies right in your room.
Magnus, thanks again mate
Yes, makes sense as the sound reaching my ears is the important point, and not what’s coming specifically out of each individual speaker. I just wasn’t sure if the sub being sent the signals for 80Hz and below by itself would make a difference, but off course it’s only the sound reaching my ears that matters…
Can’t wait to get my UMIK-1 next week to give it a go
I have followed Magnus’ steps (only in REW) and generated WAV files which I imported into HQPlayer playing thru ROON. It sounds good, but, if I turn it up I’ll start to get the sound from each speaker cutting out. It’s almost as if the sound on one side goes quiet and the other louder - then it alternates to the other speaker. Back and forth it goes. It’s fine if I turn it back down. And I’m not talking about wicked loud for the problem to occur. Is this ‘clipping’?
Any thoughts?
This is a great thread.
Don’t think its clipping, what happens if you bypass HQPlayer and use the .wav files directly from Roon?
Just tried disabling the convolution in HQP and using filters in ROON - while still using HQP for up-sampling - same result. Also eliminated HQP altogether and set up filters in ROON - same result. Now just tried no convolution in Roon but still up-sampling - same problem.
I’m going go back and check all connections - seems like something odd is going on.
Updated step 6 and 16 (the rePhase stuff I had before turned out to be less useful). Step 6 now contains a more reliable movement pattern.
2 Likes
First, many thanks to @Magnus for taking the time and trouble to document this. It’s brought me to the point where I want to have a go - I’ve also invested in a UMIK-1, so I’d better take the plunge!
One thing though, as I understand it, the guide uses a setup whereby the PC/Laptop soundcard is used to produce the pink noise for the measurements, and I’ll need to connect the output of the soundcard into my hifi pre-amp.
The alternative presumably is to use REW to generate a pink noise WAV file and put this into Roon, and then use Roon to play this back, using REW to measure the sound as being reproduced by my actual audio setup.
Is this actually a better way of doing the measurements (using Roon), or are any differences negligible in the grand scheme of things?
Hi @Magnus, one question while waiting for my UMIK-1. I’ve checked that both REW and the UMIK-1 are compatible with Macs, but I overlooked one thing: the USB part about the USB mic.
My Mac is three meters away from my listening position, how do I measure? How did you? (I could use a Windows laptop, but don’t know if I can easily import the files into Roon on my Mac.)
[quote=“Geoff_Coupe, post:57, topic:23800, full:true”]
Is this actually a better way of doing the measurements (using Roon), or are any differences negligible in the grand scheme of things?
[/quote]I would suggest to save the Pink PN to wav which can be done from Generator in REW, convert it to flac with for example Audacity and play it from Roon. Just make sure you remove all DSP in Roon before playing and measuring it. This way you will measure the full chain, as used when you normally play music.
Another option is to temporarily connect the DAC digitally to the computer (USB is probably best), but depending on hardware and how you have setup Roon that might not work. Playing Pink PN directly from soundcard seems less optimal, but it might not matter.
koen: if UMIK-1 is compatible with Mac then its just to plug it in with the USB, you might need an USB extension though. I have neber used a Mac myself though, so someone else might be better to help you with this issue.
1 Like
Thanks Magnus, I’ll wait until the mic arrives and then see what I need. Mac shouldn’t be a problem.