The Intellectual People Podcast - Mitch Barnett of Accurate Sound

[jtwrace]

Mitch Barnett of Accurate Sound tells us about his journey to become a leading expert in Digital Sound Processing (DSP). Accurate Sound is a remote service using Audiolense, Acourate and Room Eq Wizard (REW).

 

[Sonnie]

Very interesting story thus far (haven't finished yet)... and glad to put a mug and voice with the emails we've been swapping.

I've only read one of his articles thus far, and I thought it was really good.

 

[JStewart]

Starting at about 40 min Mitch’s observations on full range vs partial range correction were interesting in that this always seems to be topic of debate.
Thanks for asking about that @jtwrace

I first spotted Mitch on ASR about a year ago. Sometime after I read his articles on room correction products. Then I bought and read his book. Conclusion? If I want to pursue it I'm calling Mitch. It's not that I can't figure out how the tools work, its how to get the best result from using them. I think that's going to take some experience and art.

This is Mitch's (MItch Barnett, Mitchco, Accurate Sound) website for anyone not already familiar:

Digital Room Calibration Services, Convolvers, Headphone Filtersets

We help deliver digital room calibration for studio quality sound for any room in your house worldwide. Headphone filtersets and the Hang Loose Convolver.

accuratesound.ca

 

[Sonnie]

You will see some ads for them on our site... combined with Audiolense.

Yeah... reading Floyd Toole's book, he does NOT recommend equalizing anything over 500Hz. This is probably one of the few things I don't agree with him on for me personally. I have listened to both short term and long term (to allow for what Toole calls adaptive hearing), and equalized just sounds better to me long term over a broader range of music. Still yet... I plan to do some more comparisons after I finish rearranging the absorption and diffusion in my room.

 

[AJ Soundfield]

Starting at about 40 min Mitch’s observations on full range vs partial range correction were interesting in that this always seems to be topic of debate.

Not among those with >high school level Physics education and/or reliance on actual scientific evidence.
For the rest, it does make for an Enticing marketing story
https://www.aes.org/e-lib/browse.cfm?elib=17839
A source of unending entertainment perhaps?

cheers,

 

[AJ Soundfield]

Yeah... reading Floyd Toole's book, he does NOT recommend equalizing anything over 500Hz.

Not quite. Not recommended with neutral loudspeakers or as "permanent" filter. i.e., program could call for some EQ "fix.
Those almost always missed nuances are actually critical.

cheers

 

[Sonnie]

Here's a test that was done, albeit it does not include Dirac Live and is rather old, perhaps before the current systems improved... AND it was performed for multiple seating locations, which I agree, may not make that much difference in some cases. This was also full range testing, not just 500Hz and below... although ARC in the Anthem back then did only correct to 500Hz, so one of them was only 500Hz.

The Subjective and Objective Evaluation of Room Correction Products

In a recent article, I discussed audio’s circle of confusion that exists within the audio industry due to the lack of performance standards...

seanolive.blogspot.com

For me... I only correct the main listening position, which I believe does much more for the serious music listener than trying to correct for multiple locations, which ultimately degrades what the single measurement position can do.

 

[AJ Soundfield]

,,,with a rather non-neutral speaker.
And not vs a manual, anechoic EQ's version.
And read the comments section carefully, especially Seans.
And...I think I posted that link about a 1000x

 

[AJ Soundfield]

Btw, I wish when I was in school I could turn in 6 different answer sheets for a Math test and tell the teacher that despite all 6 being different, the answers were "accurate".

 

[Sonnie]

The largest measured differences among the different room corrections occur below 100 Hz and around 2 kHz where the loudspeaker had a significant hole in its sound power response. The room corrections that were able to fill in this sound power dip received higher preference and spectral balance ratings.

Conclusions

There are significant differences in the subjective and objective performance of current commercial room correction products as illustrated in these listening test results. When done properly, room correction can lead to significant improvements in the overall quality of sound reproduction. However, not all room correction products are equal, and two of the tested products produced results that were no better, or much worse, than the unequalized loudspeaker. Room correction preferences are strongly correlated to their perceived spectral balance and related attributes (coloration, full/thin, bright/dull). The most preferred room corrections produced the smoothest, most extended in-room responses measured around the primary listening seat.

 

[Sonnie]

Next GTG we have at my house, we'll have to do some A/B testing, with the listener not knowing which presets are no-EQ, EQ to 500Hz and full EQ.

We'll be sure to use some neutral speakers.

 

[AJ Soundfield]

The most preferred room corrections produced the smoothest, most extended in-room responses measured around the primary listening seat.

...none of which were compared to manual intervention, just each other.
You have REW + mic, so far more than Auto EQ for Dummies kit that comes with AVR for the masses....
Keep reading the Comments section below the article, focusing on Seans responses.

 

[AJ Soundfield]

Next GTG we have at my house, we'll have to do some A/B testing, with the listener not knowing which presets are no-EQ, EQ to 500Hz and full EQ.

We'll be sure to use some neutral speakers.

You also need clean anechoic/spinorama type speaker measurements to set the base line. Forget your MLs unless NRC type data exists.
Quasi-anechoic at 1m with a good turntable would allow for point source types, presuming a known design axis.
No need to reinvent the wheel IMO, but always good fun.

cheers

 

[Sonnie]

I think that study is too old to be meaningful for what we have today.

Not to say there aren't some good points in it, but we need a more modern day comparison.

 

[Sonnie]

You also need clean anechoic/spinorama type speaker measurements to set the base line. Forget your MLs unless NRC type data exists.
Quasi-anechoic at 1m with a good turntable would allow for point source types.
No need to reinvent the wheel IMO, but always good fun.

cheers

I've got some other speakers that are not MartinLogan, but we could still compare both to see which is preferred.

Bottom line, which sounds better in my room with what we have to work with. Not everyone will have a neutral speaker, not everyone will have access to data. Most will actually be in the same "real world" scenario that exist in my room.

 

[AJ Soundfield]

I think that study is too old to be meaningful for what we have today.

Not at all. It's just always misinterpreted. Context lost without reading the comments.

Not to say there aren't some good points in it, but we need a more modern day comparison.

There is very good reason for Magic EQ peddlers to not want that to happen. It would have to be someone independent, perhaps a McGill U type study. Didn't turn out well for magic MQA...

cheers

 

[AJ Soundfield]

Not everyone will have a neutral speaker, not everyone will have access to data. Most will actually be in the same "real world" scenario that exist in my room.

Yep, hence the proliferation of auto-Magic EQ on nearly every AVR etc.
Vs zero tends to win out.

cheers

 

[Sonnie]

Not at all. It's just always misinterpreted. Context lost without reading the comments.


There is very good reason for Magic EQ peddlers to not want that to happen. It would have to be someone independent, perhaps a McGill U type study. Didn't turn out well for magic MQA...

cheers

I don't think you can compare these systems of 12 years ago with what we have today. Well... you can, and then you see how vastly different the results are. But having REW, a much better mic, and better systems gives us a lot more flexibility than we had 12 years ago when all of this was in it's infancy.

 

[AJ Soundfield]

I don't think you can compare these systems of 12 years ago with what we have today. Well... you can, and then you see how vastly different the results are.

Where can we see that???

But having REW, a much better mic, and better systems gives us a lot more flexibility than we had 12 years ago when all of this was in it's infancy.

DSP, microphones and most importantly, perceptual science have been around a bit longer than 12 years.
REW + mic = Manual intervention.

 

[Sonnie]

Where can we see that???

You'd have to look back at results from years ago... REW forum back in 2009, etc.

DSP, microphones and most importantly, perceptual science have been around a bit longer than 12 years.

And DSP has improved in 12 years, would you not agree.

REW + mic = Manual intervention.

Exactly... and there is manual intervention with Dirac Live and Audiolense and with what Mitch does as well.

 

[AJ Soundfield]

You'd have to look back at results from years ago... REW forum back in 2009, etc.

Ah, so not in the form of a controlled, perceptual aka "ears" variety.

And DSP has improved in 12 years, would you not agree.

Yes and enticing marketing. DSP speaker controllers came out 20 yrs ago fyi.
Btw, note the dates for what is possible with knowledge rather than enticing magic programs.

Exactly... and there is manual intervention with Dirac Live and Audiolense and with what Mitch does as well.

Not the same thing whatsoever, when completely free from anechoic data.
A great story though.

cheers

 

[Sonnie]

Ah, so not in the form of a controlled, perceptual aka "ears" variety.

There's apparently only been one controlled study, and we linked it. But I don't think we need controlled studies to necessarily prove something has progressed and is better. Maybe some do... strictly a personal preference IMO.

Yes and enticing marketing. DSP speaker controllers came out 20 yrs ago fyi.
Btw, note the dates for what is possible with knowledge rather than enticing magic programs.

You brought up DSP, not me. I simply insinuated that DSP has improved over the years and asked if you would not agree.

Did you mean to say REW + mic + anechoic data = Manual intervention?

Yes... I owned a Yamaha DSP-100U back in 1990. And that processor is nowhere near a comparison to the miniDSP SHD processor I have now. Hence how they have improved, and I would say significantly.

Not the same thing whatsoever, when completely free from anechoic data.
A great story though.

I'm not saying it's the same thing, however, Dirac Live does me no good if I don't have REW + mic and that manual intervention. Not all REW + mic = manual intervention (that you brought up and mentioned) is with anechoic data. Rarely is there any anechoic data available for most home users. That does not make the comparisons unimportant or inconsequential to the vast majority of people living in the real world. Sure, if we are closed minded and believe things can only be achieved in one way based on one study or one man (or maybe even a couple of men), then it may be insignificant to those people.

 

[AJ Soundfield]

There's apparently only been one controlled study, and we linked it.

Right, of "Room correction vs each other vs none". Not the droids we're looking for.
I've linked another with the preference scores. They are hardly surprising given what all other tests of speaker/listener preferences show.
Of course the excuses aren't surprising either, too old, too many seats etc, etc.
Never that the onus is squarely on the magic EQ peddlers, no one else...and the evidence thus far is zero.

I don't think we need controlled studies to necessarily prove something has progressed and is better. Maybe some do... strictly a personal preference IMO.

Science/I do. Nothing personal.

Did you mean to say REW + mic + anechoic data = Manual intervention?

Nope, REW + mic means zero need for auto/magic enticing marketing story EQ,
It means you can measure both speaker and speaker/room. That distinction is critical.
With gating and out in room away from objects, a 1m REW measurement will give sufficiently accurate quasi-anechoic data for the typical point source-ish speaker > 500hz +/-. Unfortunately not for a large panel like your ML.

I'm not saying it's the same thing, however, Dirac Live does me no good if I don't have REW + mic and that manual intervention. Not all REW + mic = manual intervention (that you brought up and mentioned) is with anechoic data. Rarely is there any anechoic data available for most home users.

You're missing the point, see above. None of the magic systems, that I am aware of tasks the user with first obtaining both speaker and speaker/room, before applying filters.
Worse, they have "target" curves, which the user can then "tweak". That is an exercise in pure subjectivity. Naturally, the audiophile is then happy with the results. There is zero relation there to an actual controlled "ears" only study for actual preference. Where results can be amusingly different.
YMMV.

cheers

 

[Sonnie]

I fully get your point, but the fact still remains that most people don't have that anechoic data and likely can't get it, nor is it available for every speaker, nor can everyone accommodate a way to measure it. In these cases, as least we can measure the speaker in the room and see what it looks like, and see what's happening with what we will actually be attempting to correct. Then we can use Dirac Live (or whatever we choose) and possibly make it sound better... and that is the ultimate goal.

While having the anechoic data may help us get closer to accurate, it is not going to make us hear the difference between corrected (whether it's accurate or not) and uncorrected response any better, it's just data. If it's my mind playing tricks on me that Dirac or Accurate/Audiolense correction sounds better than uncorrected, I'm gonna let my mind play tricks on me every time I listen to music, because to my mind it sounds significantly better corrected.

If we do a comparison, and no one knows what preset is corrected and which one is not, and we play both, then the very least we can determine is if I minds are actually playing tricks on us.

The target curve most users are using is that of Harman... which is anything but subjective.

Whether controlled or uncontrolled, results are going to be amusingly different. When we did one of our speaker evaluations, not everyone preferred the Dirac correction to the uncorrected on every single speaker.

Either way... I don't think anyone is gonna go wrong using Dirac or Accurate/Audiolense, especially up to 500Hz. As Toole points out, his major concern is what these systems do above 500Hz, not below. I say compare what you have to work with and go with what you like. And there are plenty of guys out there that don't like the corrected versions, so for those guys, correcting it didn't make their minds cause them to think it was automatically better.

 

[AJ Soundfield]

I fully get your point, but the fact still remains that most people..

Sigh, Groundhog Day/ Einsteins Insanity
https://www.avnirvana.com/threads/avr-for-2-ch-and-ht.8204/page-2#post-62260

In these cases, as least we can measure the speaker in the room and see what it looks like, and see what's happening with what we will actually be attempting to correct.

NO. This is what the 99% seemingly can't grasp. The "in room" at LP is what you/the mic sees, NOT what you hear (>500hz). Please keep reading Tooles book. Its the classic error. Thus the "corrections" are for what you "see", NOT what your 2 ears/brain hear. I've shown the preference scores before for this type error. Makes no sense to continue, the Einstein thing again.

cheers