My first REW measurements(KEF R3 META)

[staticV3]

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?

You can use a miniDSP 2x4 HD/DDRC-24/Flex to get bit-perfect audio transmission and still apply room correction and sub integration.

 

[Keith_W]

I wouldn't worry about bit perfect playback if I were you. What the room does to your audio is far more important and it is what needs to be corrected.

The number 1 cause of unsatisfactory DSP correction is improper corrections made by the user. Subwoofers need bass management. In fact all systems need some form of bass management. Bass management is more than cutting peaks, you also need to align phase, time align, and much more.

At this point I encourage you to persist with DSP, but you will need to choose a DSP program. Some of them are "one button DSP" where a software algorithm decides on a correction for you. Other software have minimal automation, or no automation at all. This gives you finer control, but with a steep learning curve (and also the potential to make many mistakes along the way). I am skeptical of how accurate software automation can be compared to the manual approach. I have seen some good results, and some bad results. REW + RePhase are examples of software with no automation. If you go beyond its normal measurement functions and start considering using it for correction, you will quickly hit a brick wall and you will need a lot of guidance and reading. But trust me, it's worth it. You learn so much. You need to decide whether you want to go the easy route or take a more manual approach.

To get you started, here is a list of common software packages used by many members of ASR with a very brief description. Remember: more automation = easier, but also more dependent on software algorithm and less flexible. Less automation = steeper learning curve, but more user control.

- Dirac: a lot of automation, minimal user controls. Expensive.
- Focus Fidelity: new software, less automation than Dirac, user can over-ride algorithm. Inexpensive.
- Audiolense: more complete than Focus Fidelity, more advanced functions. Has some automation. User can over-ride algorithm. Inexpensive.
- Acourate: very minimal automation, more manual approach, very powerful. Extremely flexible. Inexpensive.
- REW + Rephase: zero automation, very manual, surprisingly powerful. Extremely flexible. Free.
- REW + MSO: some automation (software does the analysis for you) but the rest is manual. For bass only. MSO has a learning curve. Free.
- Audioweaver, Matlab, Octave: forget it unless you are an engineer or a maths nerd.

You could also use Room Shaper. This is a VST (meaning it is installed in your signal path) and it is almost plug and play. Corrects bass up to mid frequencies only.

Do some research, download a trial, and have a play.

 

[kolestonin]

You can use a miniDSP 2x4 HD/DDRC-24/Flex to get bit-perfect audio transmission and still apply room correction and sub integration.

I gave a look and I would say that DDRC24 and Flex are my best options given they come with a Dirac module.

And let's suppose that Flex would be my choice.
My current chain is pc—> DAC —> preamp —>power amp

Do I have the options to?

a) keep all my current gear and insert the Flex in the chain? If yes, where in the chain?
b) replace my current DAC with Flex and use it in DAC mode with my current preamp?
c) completely remove my current DAC and preamp, and use Flex as DAC and preamp(volume controller)?

I need to say, that this last scenario is the most interesting.
Not only cause I am a big fan of simplicity, but also cause I will be able to sell my current equipment and cover the cost of the Flex. So I can give the money I was intended to spend for buing Flex in order to buy a Dirac license.
You need to know I use my Topping pre90 purely as a volume controller and not as a source selector cause I only have one source. But my understanding is that a big benefit of using a preamp as volume controller is it's analogue volume control. Cause a device with digital volume control require full volume in order to get the best out of it. And below maximum volume it's ability to expose detail is reduced cause SINAD and dynamic range will be reduced. If that's the case with c scenario, I will need to stick with a or b cause I am not willing to sacrifice absolutely nothing in terms of sound of quality no matter the simplicity or money gain.

I wouldn't worry about bit perfect playback if I were you. What the room does to your audio is far more important and it is what needs to be corrected.

The number 1 cause of unsatisfactory DSP correction is improper corrections made by the user. Subwoofers need bass management. In fact all systems need some form of bass management. Bass management is more than cutting peaks, you also need to align phase, time align, and much more.

At this point I encourage you to persist with DSP, but you will need to choose a DSP program. Some of them are "one button DSP" where a software algorithm decides on a correction for you. Other software have minimal automation, or no automation at all. This gives you finer control, but with a steep learning curve (and also the potential to make many mistakes along the way). I am skeptical of how accurate software automation can be compared to the manual approach. I have seen some good results, and some bad results. REW + RePhase are examples of software with no automation. If you go beyond its normal measurement functions and start considering using it for correction, you will quickly hit a brick wall and you will need a lot of guidance and reading. But trust me, it's worth it. You learn so much. You need to decide whether you want to go the easy route or take a more manual approach.

To get you started, here is a list of common software packages used by many members of ASR with a very brief description. Remember: more automation = easier, but also more dependent on software algorithm and less flexible. Less automation = steeper learning curve, but more user control.

- Dirac: a lot of automation, minimal user controls. Expensive.
- Focus Fidelity: new software, less automation than Dirac, user can over-ride algorithm. Inexpensive.
- Audiolense: more complete than Focus Fidelity, more advanced functions. Has some automation. User can over-ride algorithm. Inexpensive.
- Acourate: very minimal automation, more manual approach, very powerful. Extremely flexible. Inexpensive.
- REW + Rephase: zero automation, very manual, surprisingly powerful. Extremely flexible. Free.
- REW + MSO: some automation (software does the analysis for you) but the rest is manual. For bass only. MSO has a learning curve. Free.
- Audioweaver, Matlab, Octave: forget it unless you are an engineer or a maths nerd.

You could also use Room Shaper. This is a VST (meaning it is installed in your signal path) and it is almost plug and play. Corrects bass up to mid frequencies only.

Do some research, download a trial, and have a play.

Can't thank you enough for that!
Gathering all possible eq solutions with a brief explanation on each one of them is almost sticky material.
Very helpful.

 

[staticV3]

keep all my current gear and insert the Flex in the chain? If yes, where in the chain?

PC->Flex Digital->DAC->Preamp->Power Amp

But really, PC->Flex->Power Amp would make a whole lot more sense.

Use the Flex as DSP, DAC, and Preamp.

But my understanding is that a big benefit of using a preamp as volume controller is it's analogue volume control. Cause a device with digital volume control require full volume in order to get the best out of it. And below maximum volume it's ability to expose detail is reduced cause SINAD and dynamic range will be reduced. If that's the case with c scenario, I will need to stick with a or b cause I am not willing to sacrifice absolutely nothing in terms of sound of quality no matter the simplicity or money gain.

That's a flawed understanding of digital vs analog volume control.

Unless you can actually hear the DAC's noise floor through your speakers at your MLP, which is highly unlikely with modern electronics, nothing is lost compared to an analog preamp in terms of detail retrieval and dynamic range.

 

[dogmamann]

Finally found the time to take some REW measurements.

I made 3 measurements in the main listening position.
Red=Left speaker only
Green=Right speaker
Blue=Both speakers

Sharing only the screenshots.
mdat files are available and can be uploaded if you think is necessary.

I just wasn't to be sure my REW settings are correct before moving forward, taking more measurements, sharing them on line etc.

That's why I also added 2 screenshots with my soundcard and measure settings.

Any comment, help, suggestion will be highly appreciated.

That’s a wobbly response right there, which makes everyone wonder why the speakers costs so much, and why would anyone prefer this over r11.

 

[staticV3]

That’s a wobbly response right there, which makes everyone wonder why the speakers costs so much

Not everyone. Just those that don't understand the difference between anechoic and in-room response

 

[kolestonin]

That’s a wobbly response right there, which makes everyone wonder why the speakers costs so much, and why would anyone prefer this over r11.

I think it's my (difficult)room to blame and not the speakers.

But your clearly negative comment on my curves is very interesting to me.
As I understand you see something that is undoubtedly bad and I would be interested to know what is this and where I should focus my efforts to improve it.
This the purpose of this thread after all as I am not capable to read and interpret my measurements.
Like a patient showing his blood tests to a doctor and asking for the best cure

 

[Keith_W]

I think it's my (difficult)room to blame and not the speakers.

But your clearly negative comment on my curves is very interesting to me.
As I understand you see something that is undoubtedly bad and I would be interested to know what is this and where I should focus my efforts to improve it.
This the purpose of this thread after all as I am not capable to read and interpret my measurements.
Like a patient showing his blood tests to a doctor and asking for the best cure

Ignore him. You are right, it is your room that is causing the issue and not the speakers. That wobbly curve is normal. And you might want to read what Floyd Toole said about wobbly curves - link. In particular, pay attention to this: "Don't worry about little ripples. When I see exceptionally smooth high-resolution room curves I strongly suspect that something wrong has been done. The measurement microphone is no substitute for two ears and a human brain."

You can easily prove this to yourself. Place your mic where your left ear would be and do a sweep. Then move it to where the other ear would be and repeat. Apply 1/24 smoothing and overlay the two curves. What you will see: bass remains almost the same. But the wobbles are vastly different at high freqs. Fortunately this is not what you hear, it has to do with Equivalent Rectangular Bandwidth (don't bother learning about it in detail, just be aware that it exists). A sweep from an omnidirectional microphone in a single point in space does not reflect reality. You have to be aware of its limitations and what information you can derive from it. You can derive phase and timing information from a single sweep and information about bass. But it should not be used to correct treble freqs at high resolution.

What I would suggest is to use 1/3 smoothing and look at the high freqs. 1/3 smoothing better reflects what we hear. If you see broad wobbles in the FR, you might want to fix it because it will affect the tone - or as Toole says, "broadband, low Q, tone controls".

 

[dogmamann]

Ignore him. You are right, it is your room that is causing the issue and not the speakers. That wobbly curve is normal. And you might want to read what Floyd Toole said about wobbly curves - link. In particular, pay attention to this: "Don't worry about little ripples. When I see exceptionally smooth high-resolution room curves I strongly suspect that something wrong has been done. The measurement microphone is no substitute for two ears and a human brain."

You can easily prove this to yourself. Place your mic where your left ear would be and do a sweep. Then move it to where the other ear would be and repeat. Apply 1/24 smoothing and overlay the two curves. What you will see: bass remains almost the same. But the wobbles are vastly different at high freqs. Fortunately this is not what you hear, it has to do with Equivalent Rectangular Bandwidth (don't bother learning about it in detail, just be aware that it exists). A sweep from an omnidirectional microphone in a single point in space does not reflect reality. You have to be aware of its limitations and what information you can derive from it. You can derive phase and timing information from a single sweep and information about bass. But it should not be used to correct treble freqs at high resolution.

What I would suggest is to use 1/3 smoothing and look at the high freqs. 1/3 smoothing better reflects what we hear. If you see broad wobbles in the FR, you might want to fix it because it will affect the tone - or as Toole says, "broadband, low Q, tone controls".

What’s the point in then buying the reference 3 instead of r11 if both are going to have ripples in a real room?

 

[staticV3]

What’s the point in then buying the reference 3 instead of r11 if both are going to have ripples in a real room?

The R11 Meta has deeper bass extension and less THD than the R3 Meta.

The R3 Meta costs less and is more compact.

 

[dogmamann]

The R11 Meta has deeper bass extension and less THD than the R3 Meta.

The R3 Meta costs less and is more compact.

I was asking about Reference 3 and R11. Not R3

 

[staticV3]

I was asking about Reference 3 and R11. Not R3

Ah my bad.
Compared to the R11 Meta, the Reference 3 Meta has better controlled vertical directivity and less THD.

 

[dogmamann]

Ah my bad.
Compared to the R11 Meta, the Reference 3 Meta has better controlled vertical directivity and less THD.

Measurements says r11 has lower distortion. Please check it. Can you post the vertical directivity comparison here?

 

[er|κzvio1in]

Finally found the time to take some REW measurements.

I made 3 measurements in the main listening position.
Red=Left speaker only
Green=Right speaker
Blue=Both speakers

Sharing only the screenshots.
mdat files are available and can be uploaded if you think is necessary.

I just wasn't to be sure my REW settings are correct before moving forward, taking more measurements, sharing them on line etc.

That's why I also added 2 screenshots with my soundcard and measure settings.

Any comment, help, suggestion will be highly appreciated.

With the SVS app: apply PEQ to reduce the enourmous 50 hz resonance, that alone will give an immediate boost in sound quality.

 

[staticV3]

Measurements says r11 has lower distortion.

According to KEF's specs, the Reference 3 Meta has less THD:


Graphed:

Can you post the vertical directivity comparison here?

Below is the vertical directivity.

R11 Meta:


Reference 3 Meta:


The Reference doesn't have that lobe at 1kHz.

 

[kolestonin]

With the SVS app: apply PEQ to reduce the enourmous 50 hz resonance, that alone will give an immediate boost in sound quality.

It was the 1st thing I tried but it seems I was not so thorough.
Thanks to your post I gave it another try but this time with the volume cranked up and going through my full list of reference tracks with each frequency slot(3 in total) of the application.
I used attached measurement and shot my 3 bullets to 50hz, 79hz and 101hz(also attached)

I left only one sub in the chain, found the optimal setting for 1st PEQ profile, and then passed to the 2nd, and finally to the 3rd one.
When all profiles were set I connected the 2nd sub and replicated the settings in the app.
Went again through my speaker testing tracks and surprisingly(cause instant change with 2subs simultaneously is not possible) found out where I needed to go back again with only 1 sub and fine tune 2nd and 3d profile.
I dropped a bit the db's in two last profiles so to eliminate a suspicion of boominess.
And finally came to a result proving that this SVS app PEQ feature work really good.
I could say it does wonders but I want to be moderate and live a bit with my 'new sound' before making such statements.
What I can say as a generic comment is that now my sound reminds me a lot the sound of a pair of JBL LRS 305's I had for some years in a different setup on the narrow area of same living room at an almost near-field listening scenario.
And I think my brain will need some time to adapt again to more accuracy. I will also definitely go back and repeat the procedure to eventually further fine tune the 50hz setting, and of course re-tune(if necessary) the other two frequencies based on a possible new 50hz setting. This 50hz filter had the biggest impact and maybe impressed me. So I did not test it so thoroughly.
I would also like to repeat the measurements with all filters on and off.

In all cases I am satisfied as this was a first easy win that gives me the drive to continue playing and furtherly improving.
A Flex for subs time alignment seems the second easy win to achieve down the road.

 

[er|κzvio1in]

It was the 1st thing I tried but it seems I was not so thorough.
Thanks to your post I gave it another try but this time with the volume cranked up and going through my full list of reference tracks with each frequency slot(3 in total) of the application.
I used attached measurement and shot my 3 bullets to 50hz, 79hz and 101hz(also attached)

I left only one sub in the chain, found the optimal setting for 1st PEQ profile, and then passed to the 2nd, and finally to the 3rd one.
When all profiles were set I connected the 2nd sub and replicated the settings in the app.
Went again through my speaker testing tracks and surprisingly(cause instant change with 2subs simultaneously is not possible) found out where I needed to go back again with only 1 sub and fine tune 2nd and 3d profile.
I dropped a bit the db's in two last profiles so to eliminate a suspicion of boominess.
And finally came to a result proving that this SVS app PEQ feature work really good.
I could say it does wonders but I want to be moderate and live a bit with my 'new sound' before making such statements.
What I can say as a generic comment is that now my sound reminds me a lot the sound of a pair of JBL LRS 305's I had for some years in a different setup on the narrow area of same living room at an almost near-field listening scenario.
And I think my brain will need some time to adapt again to more accuracy. I will also definitely go back and repeat the procedure to eventually further fine tune the 50hz setting, and of course re-tune(if necessary) the other two frequencies based on a possible new 50hz setting. This 50hz filter had the biggest impact and maybe impressed me. So I did not test it so thoroughly.
I would also like to repeat the measurements with all filters on and off.

In all cases I am satisfied as this was a first easy win that gives me the drive to continue playing and furtherly improving.
A Flex for subs time alignment seems the second easy win to achieve down the road.

ypou should be carefull with those boosts and you need to keep measuring.

 

[kolestonin]

ypou should be carefull with those boosts and you need to keep measuring.

I have no idea what I was listening and why I was finding these extreme boosts to make a positive change. By more listening I realized the sound was way pronounced, low frequencies exaggerated, annoying.
Removed all PEQ and just left the 50hz setting, but instead of -10db I set it at -0,6db and at the end I am sure I will completely remove it. As I started from 10db and by trial and error(listening) I ended up at 0,6db. So if I continue testing some more(even now that I write these lines I catch myself to be a bit annoyed by some cymbals hitting in the background), I will definitely remove it.
And that's the story of my life with all ways of EQ, PEQ, sound correction in general. It starts with excitement and ends at the start point(no EQ). The fact I cannot measure before and after, so to have a concrete(like an absolute number) proof that my sound is improved, is feeding the doubt that my intervention 'broke' something and ends up with frustration and quit of the tries. It is the frustration of putting time, effort, energy to something and have a big zero as a result. Repetitively and each single time I decided to try my luck with it. And you will say that it is not true that I can't measure, as I already measured and I can re-measure again as I have all the equipment. But I don't trust my measuring methodology, my REW and mic settings, and of course I cannot properly read the measurements and come to useful conclusions. And who says that this is simple, he just has to throw himself on youtube as a noob and give a look to these 10000 different tutorials, with 20000 different ways of setting everything up and 30000 different set ups.

 

[kolestonin]

I still made 0 tries and progress on the EQ front, but I have a (not relevant) update as I found a good deal and ordered a pair of Monitor Audio Silver 500 7G so to try them and replace my R3 Meta if I like them more.
I like these Meta's so much that I find really difficult to imagine how can a speaker sound better. But I had these towers under my radar since a long temp and when I read a new series is released I took my chance to contact some dealers, found an excellent deal and decided to give them a try.

Is there a way to instantly A B them with my current equipment without the need of adding new gear? I don't think it's possible but I have nothing to loose by asking.
Windows 10 pc>Topping E30 DAC>Topping Pre90 Preamp>boXem Arthur 3409/N2(nc252mp) Poweramp
(+2 subs hooked on my preamp)

I would also be interested to perform some comparison measurements, but I am a bit reluctant given the frustration REW settings are giving me each time I try, and the fact I do not enjoy the procedure. I will also need to re-install the software as I recently formatted my pc. Maybe with some guidance from your side on which settings to change after installation so to setup REW correctly I will give it a try.

 

[staticV3]

Is there a way to instantly A B them with my current equipment without the need of adding new gear?

No.

Maybe with some guidance from your side on which settings to change after installation so to setup REW correctly I will give it a try.

Use the Java EXCL input/output type in the soundcard settings.

The rest depends on what specifically you're trying to do.