MMM approach and a new calibration app (magic beans)

[Newman]

The spinorama give the direct sound information. This Toole says listening tests reveal to be the single most important thing to get right above the transition zone. So how can anything be 'far more important' than that?

OK, let's take an example. Suppose you bought a perfect speaker. The designer used an anechoic chamber, placed a mic 1m from the speaker, and tuned it until it is perfectly flat.

Perfectly flat. Got it.

You take it home and put it in your room. You place your mic 1m away and you see a perfect flat response above the transition zone, below this it's a dog's breakfast. You move your mic 2m away. The frequency response starts to fall. Re-measure at 3m, 4m, etc. You will see that as you get further away, the falling frequency response becomes more pronounced.

Sure, but that’s not the direct sound, that’s the summed sound. The direct sound is still flat, except for a tiny attenuation due to air absorption, which is dependent on humidity and I think safe to ignore for domestic room sizes.

Thus the tonality of the speaker depends on listening distance. Sit too far away, and you will get bass heavy sound with a pronounced downward treble tilt, regardless of what the predicted in-room response says. You don't have to take my word for it. Get your mic out and go check it for yourself. Why bother with the predicted in-room response when you can easily check your actual room response?

The primary determinant of tonality is the direct sound. The reflected component is a secondary determinant, sure, but you don’t want to worsen the primary determinant. Out of the frying pan into the fire.

The fact is, once loudspeakers leave the sanitized test conditions where they were designed and into the great unwashed public like you and me, we do all sorts of awful things to speakers. Just look at some of the systems posted on ASR. You will see coffee tables in front of speakers, bookshelf speakers on computer tables, bookshelf speakers in the corner of a kitchen, dipole speakers pushed up against the front wall, and so on. So never mind what the Spinorama says. Spinoramas are artificial, what matters is what you measure.

Like I say, direct sound is king, and none of your examples above affect the direct sound. If your ‘great unwashed public’ insist on doing ‘all sorts of awful things’, then they are creating compromises that, I assure you, are more likely to be harmed than helped with in-room EQ.

I tried to find a reference for what I am about to say, and gave up. But in a nutshell, Toole asserts that the evidence is that humans unconsciously detect the direct sound separately from the indirect sound, and prioritize it. Hence the anechoic data is still relevant to in-room listening. This makes it inappropriate for us to measure summed sound, and think that it determines what we perceive, for example tonality.

Toole himself says that "broad tone controls" can be applied to the upper frequencies. Granted, the Magic Beans app does not do that, its aim is to analyse and remove the "room transfer function". But still - we need to realise that it's not only freqs below the transition zone that are affected by choice of speaker placement and MLP, but also the upper frequencies.

Toole suggests tone controls in order to deal with less-than-great mastering. He does not recommend them as a way to fix room sonics or speaker FR. Our best options above the transition frequency are to buy speakers with a flat and smooth anechoic FR (ie direct sound) in that range and don’t muck around with them, or, second best option, buy speakers that are amenable to EQ and EQ their direct sound only…to be flat and smooth. NOT to measure their summed FR in the room and deal with that as a tonality issue.

cheers

 

[Newman]

@Rednexala a corrected NF response is also available from spinorama, and more accurately.

How would this work, for instance in the case of @joentell's A.T. screen example?

One could check with the manufacturer for the attenuation data.
Or reference independent measurements,
https://www.avsforum.com/threads/up...reen-materials-reviewed-and-measured.3206678/
Or make a reasoned deduction of the attenuation from research,
https://screenexcellence.com/downloads/ISVR_screens_evaluation.pdf

Then apply that in conjunction with spinorama. More reliable and less misleading, IMHO.

 

[joentell]

The whole Toole idea is we can hear through the room above transition,just like we hear an instrument playing in a room and we don't feel the need to "correct"' it.
And it makes sense.

I mostly agree with this. But, we are dealing with speakers that might not accurately reproduce that instrument, hence why I prefer to call it speaker response correction. Even a good measuring speaker behind an AT screen will need some compensation compared to the same model speaker that isn't behind one.

 

[joentell]

One could check with the manufacturer for the attenuation data.
Or reference independent measurements,
https://www.avsforum.com/threads/up...reen-materials-reviewed-and-measured.3206678/
Or make a reasoned deduction of the attenuation from research,
https://screenexcellence.com/downloads/ISVR_screens_evaluation.pdf

Then apply that in conjunction with spinorama. More reliable and less misleading, IMHO.

Those are good ways in theory, but I've found that AT screens can possibly affect the response in more ways than just HF attenuation. I have my center channel behind an AT screen, but also close to a rear wall and I find that it seems to slightly affect the dispersion pattern. My best guess is that certain frequencies are reflecting off the back of the screen towards the rear wall, and eventually coming out in a wider pattern. I'm not certain, but it seems that way.

In any case, if I can take a direct measurement and correct for it accounting for the performance of that speaker in the actual location, I prefer to do that.

We haven't even discussed production variation which is more common than some people think.

 

[Descartes]

Since he won't tell you, i'll tell you In a nutshell, the Magic Beans app restores the tonality of the speaker by subtracting the effect of the room. The idea of it is quite brilliant IMHO.

I would want to try it before forking out that kind of $$$

 

[joentell]

I would want to try it before forking out that kind of $$$

You can test the method using measurements in REW. The app takes more things into account than what's possible to calculate using REW's trace arithmetic, but you can get an idea. The point of the app was to simplify things when you're doing this for 16 speakers for a HT system for example. There's a lot of room for error when you're doing a bunch of speakers and it takes a lot of time doing it manually.

 

[GXAlan]

I would want to try it before forking out that kind of $$$

The challenge, is that if you try it, you also don’t need to buy it because you cannot erase the knowledge you had from the measurement.

The challenge, is that if you don’t have an opportunity to try it, you don’t know if you are paying a few hundred dollars for a 5% improvement or a 25% improvement in sound…. Is it making +/- 0.2 dB differences from the default curves or +/- 5 dB?

I was initially very excited about magic beans but then it was a bit pricey as an uncertainty and then I had the luck of getting a second-hand Trinnov at a price too good to pass which is a bit unfair since that was way more expensive than Magic Beans, and maybe Magic Beans with the HTP-1 would have worked as good or better than default Trinnov!

 

[Edthomp]

The challenge, is that if you try it, you also don’t need to buy it because you cannot erase the knowledge you had from the measurement.

The challenge, is that if you don’t have an opportunity to try it, you don’t know if you are paying a few hundred dollars for a 5% improvement or a 25% improvement in sound…. Is it making +/- 0.2 dB differences from the default curves or +/- 5 dB?

I was initially very excited about magic beans but then it was a bit pricey as an uncertainty and then I had the luck of getting a second-hand Trinnov at a price too good to pass which is a bit unfair since that was way more expensive than Magic Beans, and maybe Magic Beans with the HTP-1 would have worked as good or better than default Trinnov!

I am using a Trinnov Alititude 16. And I am curious about the target curve. Meaning, the Trinnov does its acoustic correction to a target curve for those frequence ranges not excluded in the excursion curve (whether that be flat, Trinnov Light, or some custom Harmon-like curve, etc.). But I think the driver of the Magic Bean software is that a generic target curve may not work in your room/speaker/MLP combination due to directivity characteristics of the speakers and room impacts...The Trinnov doesn't take near field measurements, nor does it request speakers make/model, so it has zero near field information about the capabilities of the speakers.

If I am understanding correctly the technique of the MB software, it provides a target curve that your room/speaker/MLP combination is capable of, warts and all, as some things like bad directivity of a speakers cannot be corrected.

 

[joentell]

I am using a Trinnov Alititude 16. And I am curious about the target curve. Meaning, the Trinnov does its acoustic correction to a target curve for those frequence ranges not excluded in the excursion curve (whether that be flat, Trinnov Light, or some custom Harmon-like curve, etc.). But I think the driver of the Magic Bean software is that a generic target curve may not work in your room/speaker/MLP combination due to directivity characteristics of the speakers and room impacts...The Trinnov doesn't take near field measurements, nor does it request speakers make/model, so it has zero near field information about the capabilities of the speakers.

If I am understanding correctly the technique of the MB software, it provides a target curve that your room/speaker/MLP combination is capable of, warts and all, as some things like bad directivity of a speakers cannot be corrected.

We have some people with Trinnov's in the MB Discord group. You can ask in there and ask what their experience has been.

 

[joentell]

The challenge, is that if you try it, you also don’t need to buy it because you cannot erase the knowledge you had from the measurement.

The challenge, is that if you don’t have an opportunity to try it, you don’t know if you are paying a few hundred dollars for a 5% improvement or a 25% improvement in sound…. Is it making +/- 0.2 dB differences from the default curves or +/- 5 dB?

I was initially very excited about magic beans but then it was a bit pricey as an uncertainty and then I had the luck of getting a second-hand Trinnov at a price too good to pass which is a bit unfair since that was way more expensive than Magic Beans, and maybe Magic Beans with the HTP-1 would have worked as good or better than default Trinnov!

Anyone with enough experience to do this manually in REW probably doesn't need the app. I've been open about the method so people who want to do it for free can do it without the app. The app is made for simplicity and time savings when calibrating multiple speakers.

 

[GXAlan]

Anyone with enough experience to do this manually in REW probably doesn't need the app. I've been open about the method so people who want to do it for free can do it without the app. The app is made for simplicity and time savings when calibrating multiple speakers.

I am not your target customer anymore, but I wonder if there was a way your app could take a nearfield measurement and then a MLP measurement and IN THE DEMO mode, have an estimate of whether you think the difference is small, medium, or high?

On the one hand, you might lose potential customers if they see that your tool estimates a small difference between default target and magic beans target (the room transfer curve is similar to the Harman/M2 curve). On the other hand, if the difference is big, it can make that prediction and then convince more people to buy the app.

People know your approach is a reasonable one at this point. Doing it in REW is challenging, which is where your app really is valuable.

But it’s novel/different enough that instead of FOMO, you have people wondering if they are going from 120 dB SINAD to 240 dB SINAD or going from 20 dB to 40 dB

 

[joentell]

I am not your target customer anymore, but I wonder if there was a way your app could take a nearfield measurement and then a MLP measurement and IN THE DEMO mode, have an estimate of whether you think the difference is small, medium, or high?

On the one hand, you might lose potential customers if they see that your tool estimates a small difference between default target and magic beans target (the room transfer curve is similar to the Harman/M2 curve). On the other hand, if the difference is big, it can make that prediction and then convince more people to buy the app.

People know your approach is a reasonable one at this point. Doing it in REW is challenging, which is where your app really is valuable.

But it’s novel/different enough that instead of FOMO, you have people wondering if they are going from 120 dB SINAD to 240 dB SINAD or going from 20 dB to 40 dB

I like the idea. I would have to be able to quantify what low, medium, or high difference would be. That's kind of tough, but doable. Maybe we can show deviation from Harman, whatever downward slope Dirac uses, and Audyssey's default Reference target. Maybe we can assign a letter grade based on the total dB variation above and below the transition region. Something to think about. Don't be surprised if we implement something like that.

Another idea I had was just to look at directivity info for their speakers so we could say whether their speakers are good candidates for EQ. I like your idea for a demo more. Maybe we can do both.

I'm just imagining someone measuring their speakers, and the app saying, your speaker SUCKS! LOL. Kidding of course. There is some truth to that though. Some people's speakers are just bad and no EQ can really fix it. It's just that the typical automatic methods can make the response even worse, so less bad is still an improvement. There are times when I'll do a remote calibration for someone, not having spinorama data for some older speakers, and the conclusion is that they should probably get better speakers. Even these inexpensive Polk XT20's look decent from a directivity standpoint. https://www.spinorama.org/speakers/Polk Audio XT20/ErinsAudioCorner/index_eac.html
Much better than some other "high-end" speakers I've measured.

 

[ShadowBoxer]

Someone please check me on cost:

MB-std ($249) + MultiEQ-X ($200)

Correct? (with no demo)

 

[joentell]

Someone please check me on cost:

MB-std ($249) + MultiEQ-X ($200)

Correct? (with no demo)

That sounds right assuming you already have a calibrated mic.

I've heard a few mentions of demos. Are there other DSP correction software that provide free demos? I've tried a lot of them and I don't recall seeing any.

 

[Reverend Slim]

That sounds right assuming you already have a calibrated mic.

I've heard a few mentions of demos. Are there other DSP correction software that provide free demos? I've tried a lot of them and I don't recall seeing any.

I could see maybe a version that lets you go through the process and see a coarse version of the curve at the end, but without letting you do the export at all. You could have it default to that when there's no login, perhaps. Anything more would be problematic.

 

[Rednaxela]

Someone please check me on cost:

MB-std ($249) + MultiEQ-X ($200)

Correct? (with no demo)

You might be able to get away with using Audyssey One. See this post.

Didn’t extensively research this, so I may be overlooking something. If not, you’d only need the $20 Audyssey MultEQ Editor app instead of MultEQ-X.

 

[GXAlan]

That sounds right assuming you already have a calibrated mic.

I've heard a few mentions of demos. Are there other DSP correction software that provide free demos? I've tried a lot of them and I don't recall seeing any.

Dirac AVR products have a 30 day money back guarantee and the PC/Mac version have a 14 day trial, but it’s easier to control licensing.

Audyssey’s iOS app is a lot cheaper to gamble on.

Every other AV product is bundled?

It might be interesting to see if you can talk Emotiva into licensing magic beans with their BasX line which just has some PEQ for room correction or even Monoprice with the HTP-1. They said that the “freebie” moved a lot of a Thighmaster’s in the 80’s and it would help popularize your approach.

I'm just imagining someone measuring their speakers, and the app saying, your speaker SUCKS! LOL.

It’s how you phrase it.

It’s not that the speaker sucks. It’s that the Magic Beans cannot predict the effect of the target curve in your room or predict the effect of EQ.

There’s a bit of “your speakers and room are so good that the default Harman target match what we would tell you what to do” as well as scenarios like your acoustically transparent screen and your speakers aren’t great for EQ

 

[Buckeye]

I finally got around to recalibrating using Magic beans (taking my time and actually paying attention to the directions). The calibration is fairly quick if you understand the process. There are videos in the program that explain the process. The great thing is that I didn’t have to run Dirac again! I just loaded the previous project and imported the Magic Beans target curve for each speaker.

My initial opinions are that the sound seems clearer and I hear the background effects much better. Does the difference blow my sock off, no. Was it worth the price for me, yes. I don’t have time to learn REW or understand all the stuff that goes into it. I just want something that I can run a simple program and get a good result.

Now I just have to figure out how to convince Joe N Tell to add multi sub calibration to this application and my system will be perfect!

 

[holbob]

That sounds right assuming you already have a calibrated mic.

I've heard a few mentions of demos. Are there other DSP correction software that provide free demos? I've tried a lot of them and I don't recall seeing any.

I definitely tried a demo of dirac through my pc before adding it to my old 2X4 HD.

 

[ShadowBoxer]

Are there other DSP correction software that provide free demos?

MathAudio.

Test tone beeps in demo mode.

As others have mentioned, $450 is a lot to gamble on an unknown result.

However, downloading a static exported result to the AVR is different than real-time processing like MathAudio. So, it's a bit of oranges & tangerines.