My first REW measurements(KEF R3 META)

[kolestonin]

I applied above suggested settings and performed 3 new measurements.
Left, Right, and L+R, at the main listening position.
Both subwoofers were on.

I would like to have some feedback or any conclusions on my curves.
We can tell there is a peak at 50hz?
How can I try to tame it? Just as a first try.
Can you guide me on how to apply some adjustments?

My SVS app has a parametric eq function and it could be a quick and easy first try for me.
Is the logic behind the attached screenshot correct?

The App gives me 3 presets which can be applied all together and the frequencies range is from 20 to 200hz.

 

[OCA]

Taming those peaks is as easy as the click of a button with REW, it's a bit harder to align the subs between each other and correctly cross them over with the speakers. You have to take individual measurements for each sub and speaker first. If you think you are done with it already, then use REW's auto EQ. If you install the new beta, now it has even automated shelving filters. Just determine a common target and target level for both speakers and "Match response to target" from 20Hz to 200Hz (with no smoothing on the speaker response, no boosting filters) and you will get 90% of the way to optimal correction.

 

[Keith_W]

May I ask what you are using as a source? If you are using a computer as a source, sending audio to a DAC, you have much more powerful capability to tame that peak and perform other corrections. PEQ's are imprecise by comparison, and should only be used if you have no other choice. It is a quick and dirty solution which sort of works, but @OCA's suggestion is much better. Otherwise, the Q in your PEQ which is centred around 50Hz is much too wide, I would narrow the Q so that it more closely matches the peak that you observe. And I would use two of them - one at 50Hz, the other at 92Hz.

 

[kolestonin]

I am doing some experimentation while familiarizing my self with REW.
I made again 3 measurements with attached parametric EQ settings.
I don't see much of a difference. Not sure if I am doing it all wrong or if it is simply not efficient in my room.

I want to try the auto eq rew feature as suggested but I will have to to make a bit of a research as I have no clues on how to proceed.

ps: the right speaker is the one close to the open door. If we look on these and the previous measurements, can we say they are what we would expect to see based on speakers placement?

 

[Keith_W]

Your new set of measurements looks the same as the previous set that you posted yesterday. You did not tell us what you are trying to measure? The effect of the PEQ?

That open door will affect the sound depending on the frequency band. Depending on the size of the door, it may increase the apparent size of the room and shift the pattern of standing waves in lower frequencies. For the upper frequencies, it will change the pattern of reflections depending on the radiation pattern of your speaker and how far the speaker is from the door. You may not see this in a FR sweep, you would have to look at the ETC (energy time curve) or spectrogram for that.

Rather than trying to predict which frequencies will be affected, it is easier to do a sweep comparing open door vs. closed door. FYI I have made dozens of sweeps in my own listening room, comparing the effect of open/closed doors and windows, open/closed curtains, furniture placement, and so on.

On a side note, I am not sure if I like your door. I would be looking at it closely to make sure it does not resonate. Subwoofers pressurize/depressurize the room and can cause doors and windows, which seal the room to the outside, to move. I fixed the annoying rattles in my listening room which is full of doors and windows by installing rubber gaskets in the jambs.

 

[kolestonin]

Your new set of measurements looks the same as the previous set that you posted yesterday. You did not tell us what you are trying to measure? The effect of the PEQ?

This is indeed what I did.
I applied these PEQ settings on both of my subs via the SVS app and repeated the same exact measurements so to find out if they make any difference.
But as you say there is no difference with my previous measurements.
So this leads me to think that or I am doing something wrong or that the PEQ SVS feature will not work in my room/case and I shall not loose more time with it.

That open door will affect the sound depending on the frequency band. Depending on the size of the door, it may increase the apparent size of the room and shift the pattern of standing waves in lower frequencies. For the upper frequencies, it will change the pattern of reflections depending on the radiation pattern of your speaker and how far the speaker is from the door. You may not see this in a FR sweep, you would have to look at the ETC (energy time curve) or spectrogram for that.
Rather than trying to predict which frequencies will be affected, it is easier to do a sweep comparing open door vs. closed door. FYI I have made dozens of sweeps in my own listening room, comparing the effect of open/closed doors and windows, open/closed curtains, furniture placement, and so on.

I get what you say about ETC and spectrogram but the problem is I will not be able to interpret this data.
And this is what happens with REW each time I try to dive into it.
The guides, posts, suggestions are unlimited and, if you don't have some background or knowledge of sound science, you find yourself lost in the options and ending up in a discouraging constant doubt about your choices.
In plus, for the majority of guides the existence of devices like minidsp or streamers with dirac modules and similar is a prerequisite. While I don't have one in my chain(topping e30>topping pre90>arthur 3409/N2 power amp) and this makes it even more complicated.
Can I add a minidsp in my current chain? What will be the benefits?
What could be my other options given above limitations?

On a side note, I am not sure if I like your door. I would be looking at it closely to make sure it does not resonate. Subwoofers pressurize/depressurize the room and can cause doors and windows, which seal the room to the outside, to move. I fixed the annoying rattles in my listening room which is full of doors and windows by installing rubber gaskets in the jambs.

I hate this door(s) more than you.
I even completely removed it when I firstly placed my speakers on their current position, but it had to be put back as I faced a house revolution minutes after the removal.
It has these terrible glass fully resonating tiles. Which are hold in place by a metal frame. In a door with 54 years on its back.
I can remember them rattling with smaller, but differently placed, speakers. They do not give an identifiable vibration or sound with my current setup though. But they are still a very negative influence in a room that, while non treated, is furnished with 'sound friendly' objects.
I will inspect more closely as you suggest.
Really thank you for the rubber gaskets suggestion. I think I will add them in both doors.

 

[Keith_W]

This is indeed what I did.
I applied these PEQ settings on both of my subs via the SVS app and repeated the same exact measurements so to find out if they make any difference.
But as you say there is no difference with my previous measurements.
So this leads me to think that or I am doing something wrong or that the PEQ SVS feature will not work in my room/case and I shall not loose more time with it.

I agree that the measurements show that your DSP is not applying your PEQ. I don't know why, might be a software setting.

I get what you say about ETC and spectrogram but the problem is I will not be able to interpret this data.
And this is what happens with REW each time I try to dive into it.
The guides, posts, suggestions are unlimited and, if you don't have some background or knowledge of sound science, you find yourself lost in the options and ending up in a discouraging constant doubt about your choices.

That is what ASR is for, we are here to help. Feel free to post whatever graphs and .MDAT's you have trouble interpreting. Subwoofer integration is not easy, there are multiple threads on ASR of people struggling with their subs.

In plus, for the majority of guides the existence of devices like minidsp or streamers with dirac modules and similar is a prerequisite. While I don't have one in my chain(topping e30>topping pre90>arthur 3409/N2 power amp) and this makes it even more complicated.
Can I add a minidsp in my current chain? What will be the benefits?
What could be my other options given above limitations?

Of course you can add a MiniDSP to your current chain. Broadly speaking, you have two options when it comes to DSP: hardware based, or software based. Hardware based (e.g. MiniDSP, DEQX) have the advantage in that they are one box integrated solutions with software included. They have other niceties like built-in DAC's with the correct connectors, remote control, and a large user base with their own support forum. However, I do not like the MiniDSP and these reasons may not apply to you. It is limited in processing power, and it uses mixed phase filters. Mixed phase has some advantages, it is less computationally intensive and it has low latency. However, it can not correct bass to any high degree of resolution.

The ultimate solution would be a software solution similar to what @OCA recommended - i.e. use your computer to perform DSP. However, going down this route is more complex and has a steep learning curve. To see what you are in for, take a look at OCA's Youtube channel and see if you can understand what he says in his videos. I have an above average understanding of DSP, and I still struggle to understand half the things he says. You can take your pick from free software (REW/RePhase) to paid (Acourate, Audiolense, Dirac, Focus Fidelity). These software packages differ in how much control they allow you to have, and ease of use for beginners. In general, software which gives you more control, i.e. forces you to look at graphs and make your own decisions, is better and more powerful. Examples are REW and Acourate. I do not suggest beginners take this route, unless you are exceptionally persistent and prepared to make DSP your new hobby. You are better off going with something like Audiolense or Focus Fidelity which have more automation and is easier for beginners.

For software based correction, you will need: (1) a PC, (2) an interface card with as many DAC channels as you need like the RME Babyface (4 channels) or the Motu Ultralite Mk5 (8 channels), (3) 48V phantom power microphone and stand, (4) the software.

The advantage of going down either route is much better integration of your sub with your mains. Subs need to be adjusted in amplitude, phase, and time to match your mains. It is simply not possible to do this with any precision without DSP. I believe that anybody who owns a sub needs DSP, otherwise they are not getting the most from their sub. But that's just my opinion.

I hate this door(s) more than you.
I even completely removed it when I firstly placed my speakers on their current position, but it had to be put back as I faced a house revolution minutes after the removal.
It has these terrible glass fully resonating tiles. Which are hold in place by a metal frame. In a door with 54 years on its back.
I can remember them rattling with smaller, but differently placed, speakers. They do not give an identifiable vibration or sound with my current setup though. But they are still a very negative influence in a room that, while non treated, is furnished with 'sound friendly' objects.
I will inspect more closely as you suggest.
Really thank you for the rubber gaskets suggestion. I think I will add them in both doors.

It's not only the doors that can rattle in the jambs, the glass panes can rattle as well. If it's the glass panes you might need silicone sealant. Run a bead of sealant between the glass and the frame. It took me months to fix the issues in my listening room - windows, doors, furniture, picture frames, light fittings, you name it. I played test tones and inspected everything for vibration.

 

[kolestonin]

That was very informative and I will investigate all mentioned options.

For the moment I would need some more basic advice as during the weekend I will have the time to make some more measurements. Take a nice bunch of measurements around the listening position and check the average, or eventually try the moving microphone method.

So I would like to know what is the apropriate way to do it.

Measure the left speaker, then the right one, and combine them for every position?
Measure left, right and pair, and combine them for every position?
Measure only pair for each position?

 

[Keith_W]

Depends on what you are trying to achieve. FYI I take dozens of measurements, depending on what question I want to answer. Some of the crazier things I have done include generating a set of filters for listening in my dining room while I have dinner.

If you want to see if there is much variability at different listening positions, take L/R sweeps at different positions. I sometimes take mono measurements (both speakers playing together) because you don't listen to an individual speaker, you listen to them both - and sometimes you might see something in a mono measurement that you won't see in a single speaker. You have to ask yourself what you want to see before you even pull out the mic.

 

[kolestonin]

I had the time to make some thorough measurements.
1 at the main listening + 16 more around the main listening position.
A total of 17 pair(L+R) measurements which I averaged. I hope one day I will be in grade to create some filters based on this average and import them to EQ Apo.

The result looks really good to my untrained and clueless eyes.

But I would like to have comments on what people see or what they don't see there.
Conclusions, EQ suggestions or generic advices based on these measurements.

What I share is:
All SPL Average
All SPL Average+Main listening position(measurement 1)
Waterfall+Main listening position(measurement 1)

 

[PeterNL]

I had the time to make some thorough measurements.
1 at the main listening + 16 more around the main listening position.
A total of 17 pair(L+R) measurements which I averaged. I hope one day I will be in grade to create some filters based on this average and import them to EQ Apo.

The result looks really good to my untrained and clueless eyes.

But I would like to have comments on what people see or what they don't see there.
Conclusions, EQ suggestions or generic advices based on these measurements.

What I share is:
All SPL Average
All SPL Average+Main listening position(measurement 1)
Waterfall+Main listening position(measurement 1)

Please repost all graphs with SPL scale 50-100dB SPL, NOT -230 160dB otherwise is not readable (everything will look perfectly flat).
Did You take measurements with Acoustic Timing Reference ? (You should)

 

[Keith_W]

But I would like to have comments on what people see or what they don't see there.
Conclusions, EQ suggestions or generic advices based on these measurements.

What I share is:
All SPL Average
All SPL Average+Main listening position(measurement 1)
Waterfall+Main listening position(measurement 1)

What you should be doing is asking yourself a question, and then taking the measurement to answer the question.

What you are doing at the moment is taking a bunch of measurements and asking ASR what they see. You will not get very helpful replies this way.

Some examples of questions you could ask yourself:

- "I wonder what sound my wife hears when she sits next to me. Let's see what the measurements look like over there". Ans: one L/R sweep at MLP, another L/R sweep at the other listening position.
- "What is the effect of the door when it is open vs. when it is closed?". Ans: one L/R sweep at MLP with door open, another L/R sweep with the door closed.
- "I wonder if my speakers are optimally placed and whether they should be pulled out / toed in / etc. more". Ans: multiple L/R sweeps at MLP with the speaker repositioned between sweeps.
- "I wonder if my speakers are distorting / exhibiting volume compression". Ans: calibrate your mic to an SPL meter, then place your mic 1m on axis from the speaker. Run repeated sweeps at 85dB, 95dB, and 105dB.
- "I wonder if the MMM method gives me more reliable information than multi-point single sweeps". Ans: try it yourself.

When you eventually get your MiniDSP, there is even more tinkering you can do.

In all cases, clearly label your sweeps. There is a world of information and understanding you can gain by asking yourself questions and doing measurements to answer them.

 

[kolestonin]

Please repost all graphs with SPL scale 50-100dB SPL, NOT -230 160dB otherwise is not readable (everything will look perfectly flat).
Did You take measurements with Acoustic Timing Reference ? (You should)

Yes, measurements have been taken with Acoustic Timing Reference.

I am reposting the graphs with correct(I hope) db scale.
If still not readable, please guide me on how to apply the 50-100db spl cale.

 

[PeterNL]

Yes, measurements have been taken with Acoustic Timing Reference.

I am reposting the graphs with correct(I hope) db scale.
If still not readable, please guide me on how to apply the 50-100db spl cale.

Click "Apply Settings" because it is still not 20-20kHz and 50-100dB scale, also waterfall.
Also use VAR smoothing if You can and try Psychoacoustic smoothing (it will help you to read graphs and frequency response how you hear it).

 

[kolestonin]

What you should be doing is asking yourself a question, and then taking the measurement to answer the question.

What you are doing at the moment is taking a bunch of measurements and asking ASR what they see. You will not get very helpful replies this way.

Some examples of questions you could ask yourself:

- "I wonder what sound my wife hears when she sits next to me. Let's see what the measurements look like over there". Ans: one L/R sweep at MLP, another L/R sweep at the other listening position.
- "What is the effect of the door when it is open vs. when it is closed?". Ans: one L/R sweep at MLP with door open, another L/R sweep with the door closed.
- "I wonder if my speakers are optimally placed and whether they should be pulled out / toed in / etc. more". Ans: multiple L/R sweeps at MLP with the speaker repositioned between sweeps.
- "I wonder if my speakers are distorting / exhibiting volume compression". Ans: calibrate your mic to an SPL meter, then place your mic 1m on axis from the speaker. Run repeated sweeps at 85dB, 95dB, and 105dB.
- "I wonder if the MMM method gives me more reliable information than multi-point single sweeps". Ans: try it yourself.

When you eventually get your MiniDSP, there is even more tinkering you can do.

In all cases, clearly label your sweeps. There is a world of information and understanding you can gain by asking yourself questions and doing measurements to answer them.

I got your logic since your previous posts
And took the measurements with that in mind.
I took measurements with curtains pulled and in place, doors open, doors closed, doors covered with blankets.
Carefully labeled and saved in different labeled folders.

I am in the process of analyzing this info.
And I will surely need to take some more measurements.
And do some more reading, and watch some more videos etc.
But this will take time.

A little 'boost', help, opinions while I am in the process of checking the data will not harm.
Especially given my limited knowledge on the matter is obvious.

 

[Keith_W]

Overlay your graphs and choose the appropriate vertical scale to compare them. It will make it easier to see the difference.

 

[kolestonin]

Click "Apply Settings" because it is still not 20-20kHz and 50-100dB scale, also waterfall.
Also use VAR smoothing if You can and try Psychoacoustic smoothing (it will help you to read graphs and frequency response how you hear it).

I am giving it another try.
50-10db scale+var smoothing(it does not seem to be applied on average curve but it does not change no matter what I try)

 

[kolestonin]

I need to say I am a bit excited as I managed to create filters with rew, import and apply them to eq apo.
So now I know every step of the process and I just need to become confident with it. And of course continue reading and watching videos so to overcome the struggling with the unlimited options, measuring methodologies etc.
And I am excited cause I am entering into eq world and I realize the possibilities ahead are a lot and promising.

I also tried a quick manual setting of filters and q/slopes based on my measurement.
So to check if it will make any positive difference. And to be honest I think it did. I need to give it some more listening obviously. My first impression is that it removes a bit of 'boominess' but also makes the bass a bit 'thin' but not necessarily 'sterilized'. I need to fine tune and experiment so to find the sweet spot maybe. Or try to rise the gain of my subs as well as different crossovers.

But I would also like to have your quick feedback on my quick experimental try. Just to check if I read the measurement correctly, and the logic behind my approach is correct. or any filter suggestions/fine tuning from your side.

Feel free

edit: I now realized I can import and make appear in Peace the filters I created with REW. I was not using Peace. Just importing the filters directly to eq apo. So I had no clues when I was setting the filters manually(first screenshot). I am adding the screenshot of my REW created filters as well. Which I think sounds even better. But a bit low in volume. I don't know yet.

 

[staticV3]

I need to say I am a bit excited as I managed to create filters with rew, import and apply them to eq apo.
So now I know every step of the process and I just need to become confident with it. And of course continue reading and watching videos so to overcome the struggling with the unlimited options, measuring methodologies etc.
And I am excited cause I am entering into eq world and I realize the possibilities ahead are a lot and promising.

I also tried a quick manual setting of filters and q/slopes based on my measurement.
So to check if it will make any positive difference. And to be honest I think it did. I need to give it some more listening obviously. My first impression is that it removes a bit of 'boominess' but also makes the bass a bit 'thin' but not necessarily 'sterilized'. I need to fine tune and experiment so to find the sweet spot maybe. Or try to rise the gain of my subs as well as different crossovers.

But I would also like to have your quick feedback on my quick experimental try. Just to check if I read the measurement correctly, and the logic behind my approach is correct. or any filter suggestions/fine tuning from your side.

Feel free

Nice work!

A few notes:
You should apply an appropriate preamp so that the total filter gain never exceeds 0dB. Otherwise you can get nasty digital clipping.

Peace has a feature where it'll automatically decrease the preamp every time it detects clipping, until it no longer occurs. You may have to enable that first.

In Peace, you have to change a setting in the preferences to be able to enter filter gains with decimal places. By default, it'll always round to 0.5 or 1.0dB steps (can't remember which).

I honestly recommend you use the default Equalizer APO GUI instead of Peace.
One massive advantage is that you can apply your preamp and PEQ filters separately and toggle each on/off separately.
That allows you to turn just your EQ on/off without affecting the overall volume.

In Peace, if you have the appropriate preamp applied (as you should!), then the green toggle button in the top right corner will always turn both the EQ and Preamp on/off, which means that you can never actually A/B
just the EQ alone.

To import a PEQ Preset.txt into EQApo's default GUI, use the "Include->Include configuration file" block.

The end result should look kind of like this:


With the default GUI, you can also experiment with more advanced convolution filters.

Here's an awesome website that'll create such convolution files for you: https://www.loudspeakers.audio/en/

 

[kolestonin]

My enthusiasm of having discovered what eq can do backtracked a bit if not disappeared completely.
This happened cause I did some A/B comparisons of Qobuz web player with Qobuz desktop application.
And even with my best eq filters the difference in favor of the app is very big.
So big, that it made me quit my attempts to make my eq better cause I realized that there is no way I will even get close to the sound of the app no matter what I do. It is literally impossible.
And as 90% of my listening is via the app, I abandoned my EQ APO experiments and started reading how I can eq wasapi exclusive mode.
To find out that if you eq wasapi exclusive mode, you literally undermine all it's benefits
So I now focus on finding a way to better integrate my two subs as my last chance to get some more of my gear.
But then again, I started thinking that if I want to integrate my subs I will need to to use a software, which will again undermine the bit perfect playback. Is this true? If I want to use wasapi exclusive mode I will need to forget any kind of intervention?