I recently claimed that the audyssey mics were inconsistent
and was called out by you know who. I linked to a thread
similar to this and now this supports my claim even more.
Good timing. Thanks.

 

Good work, Luke.

More reasons why I don't care for hamfisted REQ applications or absolute priority taken to frequency response. Can't even count on the mic(s) to be consistent.

 

Hmmph. That said ....

...why does your frequency response suck so much?

 

Given how cheap the mics are, that's a pretty decent spread. I don't see how this is an issue though since these are +/-2dB devices being used to correct the +/-15dB errors caused by speaker and room interactions. Does it really matter that the receiver will think reference is at 85dB when actually it's 82dB and that >6khz is off by 3-4dB? I'm not putting forth the argument that it's inaudible, just that it's not significant compared to the benefit gained with the low frequency corrections.

 

There's a lot of discussion about "this receiver sounds so much better" even when switching from a Denon with XT32 to a Marantz with XT32, when in reality they shouldn't sound any different.

For example, if you were about to hear two different Marantz processors in the same HT, and one was from 2010 and the other was the latest and greatest $4,000 flagship, you'd probably expect the newest model with better internal hardware in every way to at least sound a little better correct?

Case in point:

http://www.avsforum.com/forum/90-re...s/2670641-marantz-av7703-vs-av7702-vs-av7005-2ch-stereo-critical-listening.html

You'd be wrong

With the same frequency response, they're indistinguishable, so what's wish all the reports/reviews saying they are distinguishable? I think these $5 mics are most, if not all, of the reason.
@zorax2 calibrated both his X4200 and 8802 (both use XT32) with the same mic and bingo, no audible difference.

 

interesting results luke.


any idea if the levels in the avr's are calibrated to the various mics or is the theory that these should all be 'hot swappable' between avr's?

 

All Audyssey mics released from 2008 to present are swapable between any Audyssey AVR/AVP regardless of Audyssey version being used.

 

I found a surprising delta in a couple mics I measured a couple years back too. It does matter jdsmoothie, and 12b4a. If you have one mic consistently set the treble 4-5dB hotter for instance (or the vocal range, or the bass - - or any combination of EQ significantly different) then that is noticeable. When I ran Audyssey multi EQ XT32 on my ONKYO 5508 I never liked it. When I swapped out to my Denon 4520ci with the same speakers I liked Audyssey multi EQ XT32. When my Denon went back in the shop one time I used the Denon mic to calibrate the Onkyo and realized there was the EQ'ed sound I liked. I tried the Onkyo mic on the Denon, upon it's return, and didn't like the calibrated sound. I switched back and forth a few times and realized my personal sound preference always followed one mic, and not the particular AVR vendor. These differing mics were the same part number.

My Onkyo mic EQ'd for 5dB more treble above 10khz. That's a huge swing in the overall sound presentation and explains why I had always prior thought the Onkyo Audyssey calibrations were too bright. (hurt my ears bright - especially with dynamic EQ engaged which of course makes it even brighter at lower listening levels)

That story is here:

http://www.avsforum.com/forum/90-re...rocessors/1764065-brief-comparison-between-two-audyssey-multi-eq-xt32-mics.html

Meanwhile you read professional reviews about how much better one particular model AVR's Audyssey implementation is over another. (A major reason I bought the Denon 4520ci was based on a review like that). And we the consumers can easily fall prey to mis-information. It is a mic lottery if you will. Armed with this type of knowledge -- we know better!

Luke keep it up. I love it.

I hope we can get more of these measurements and tests from other members with multiple mics. It is knowledge!

 

Apologies for the Necro reply to this thread but this "might" explain why my new Denon sounds better than my old Marantz. Yes they are a few years apart but even so there shouldn't be "that much" difference between a Marantz SRxx13 reciever and a Denon x3800. However, the main difference is Atmos cohesion so I'm not sure how the mic makes a difference there.

 

Whelp, I guess there's a market for golden ACM1HB mics now.

 

Perhaps this is why I think my 7702 has a more fuller sound than the 4520?? Great job guys. Wish I could have been there.

Typed on a little keyboard, excuse any grammatical errors.

 

Some people just don't have their priorities straight...

 

Anyone that thinks that those frequency responses won't sound significantly different must not tinker around with manual EQ very much.

 

A couple more tech points..
1. Another crucial spec is output level vs frequency
2. The inexpensive, target condensor mike element works reasonably well out to 13-14KHz, but beyond this its response decreases steeply... Thus if one is using high resolution loudspeakers with efficient tweeters such as horn or small dome, so then processor DSP must then create the higher frequency response...


Just my $0.02...

 

@SteveH - Is Denon/Marantz aware of the variability in quality of this mic that is used for Audyssey calibrations? It's a shame they use a relatively inaccurate mic (with this much variability) in particular with the high end Marantz 8802 and 7703 as well as the 6000 and 7000 series Denons. Are there any plans that you are aware of to make available a more accurate calibration microphone?

 

@zorax2,

That's the Audyssey pro kit (which isn't supported on the newest Denon x300 series???)

 

Good test! FWIW, Chris K., the creator of Audyssey, was always very upfront that the mics have a potential error factor of +/- 2db. That could mean a 4db swing between two random mics. In practice, most people seem to have found that the differences are less than that. I keep a back-up mic on hand, and interchange the two with pretty consistent results. But, there is a certain amount of chance involved in that. If I had a mic that consistently produced a 5 or 6db difference at higher frequencies, I would suspect a bad mic, because the error factor should be consistent across the full frequency spectrum as it was in this test.

Incidentally, for best results in testing, and EQing, something like this would be better, as there would be a lot of potential comb filtering from both the board, or the sofa back, from close proximity to the omnidirectional mic. https://www.amazon.com/gp/product/B00BQOFG6W/ref=ox_sc_sfl_title_13?ie=UTF8&psc=1&smid=ATVPDKIKX0DER

 

Comb filtering or sound reflections off the board in this test is immaterial. All the mics were treated the same. If any undesired reflections happened to one it happened to all. The meat is this ---there was a 4dB swing in treble on one of three mics in this test and a 5 dB swing in treble on my test. Out of five Audyssey mics tested that doesn't bode well for the broader audience and it may be that hitting the outer tolerance limits is not unusual. By the way, without comparing these to a calibrated mic there is no way to know which is accurate --- or if any actually are.

 

I get that and if you noticed, I didn't challenge the test methodology or results. I said it was a good test. I merely noted that it was better to use a mic stand. I think you did miss another point that I made, however. If there were a 5db swing in treble frequencies with one of your mics, then you almost certainly had a bad mic. They do degrade over time, and any component can go bad at any time, anyway. If one mic showed a 5db difference across the full frequency range, that would be outside normal parameters, but still understandable, I guess.

But, if a mic showed only that difference in a specific portion of the frequency range, then it was a bad mic. As you observed from the tests, all of the mics remained consistently either high or low in SPL, compared to each other. And, all had essentially the same frequency response from the standpoint of shape.

It would indeed be better if there were even smaller tolerances among the standard Audyssey mics, but even calibrated SPL meters costing much more have an error factor of about +/- 1.5db. As a practical matter, as long as the mic's response is fairly uniform across the full frequency spectrum, there wouldn't be a lot of difference between calibrations, in the same way that there wouldn't be a lot of difference in calibrating a system with a 75db test tone, and a 77db (or 73db) test tone. So, the +/- 2db shouldn't really change much with respect to the overall EQ, from a curve standpoint, although it might change your MV setting by that amount.

Other than the occasional bad mic, or need for a microprocessor reset, I suspect that people who find one AVR or pre-pro sounding so much better than another may be experiencing slight differences in Audyssey mics, or of their own implementations of Audyssey, or they may simply be hearing things that they want to hear with the upgrades. We have heard of that happening from time-to-time with other audio components.

Regards,
Mike

 

Here are the receivers/processors that these three mics (courtesy of @zorax2 )were supplied with, along with the Audyssey version, mic model, and date on the microphone:


1. Denon X2200 (MultEQ XT), mic model ACM1HB (Date = 09-17-2016 (this is a recent refurb))
2. Denon X4200 (MultEQ XT32), mic model ACM1HB (Date = 08-16-2013)
3. Marantz AV8802 (MultEQ XT32), mic model ACM1HB (Date = 11-05-2014)


To start, we wanted to make sure all the mics were measured from the exact same position, but they're so light and flimsy they wouldn't sit on the back of the couch with just their own weight...

We then secured each mic to a metal clip for a tripod (which we didn't have) and then secured the metal clip to a piece of plywood:

I plugged these into the mic input on my HTPC and we set the input level to 40.


First up, let's compare the X2200 and X4200 mics. They're both the same model #, but the receivers are XT vs. XT32 so who knows:

Wow, sensitivity at 1khz is identical, and the biggest deviations are at ~200hz and ~9khz but less than 1db off at both of those places. Not bad!

Next, let's add in the 8802 mic. Again, same model #:

Ummm...what on earth is this?!? To level match at 1khz with the other two mics it has to be lowered by 2.8db and then it looks like this:

Even when level matched, around 9khz it's over 3db lower than the X2200 mic and over 4db lower than the X4200 mic!

Both the X4200 and 8802 use the same version of Audyssey, but with the massive variation between both mics you'd get very different results depending on which mic you used.

With this much variation from mic to mic, even though they're the same model # and provided with receivers that use the same version of Audyssey, I think it's easy to see how users are getting a different experience with different receivers/processors.

Just for fun, I'll throw in an even older mic:

4. Marantz AV7005 (MultEQ XT), mic model #ACM1H (Date = sometime around 2010/2011)

All level matched at 1khz:

The test setup is flawed...
When measuring a transducer be it a loudspeaker or microphone, the device being measured must be totally isolated from any object or construction that may or can/will create reflections that bias/corrupt the root results.
Therefore...
Either an anechoic chamber or free-air setup is required.

Just my $0.02...

 

This was for a "relative" comparison rather than measuring the "absolute" frequency response of the mics. All of the mics were tested in identical (though a perhaps flawed position on the board), nevertheless, there were some significant response differences. These differences would also be apparent in an anechoic chamber or free-air setup.

 

Here is a comment from another thread regarding the Audyssey microphones:

Lets kleer the air about the Audyssey EQ microphones..
1. The brand needs to procure this through Audyssey
2. Its factory FOB cost is

 

I would like to see how consistent is each microphone. If tested in quadruplicate with identical results it would really support finding a measurement Mic that fits your listening preferences. +\- 2 dB is pretty large range for a calibration microphone. That is the window I am shooting for from my speakers at the listening position, I would expect the microphone of a $2k product to exceed this window.

You would think the flagship models would use a better performing microphone than the low end models. Marantz could learn from Grado, in that they grade identical phono cartridges on measurements, the tighter to spec the more they sell for even though they all cost the same to make (although if Grado only sold the best performing you would have greater cost due to manufacturing losses).

Interesting thread in my view!

 

All the more reason to feel free to tweak the results of any calibration.

In my case it was crossovers for the mains, increasing volume level on the center and decreasing volume level on the rears.

Even if I had the capability to do manual EQ of frequency ranges, like I can with my DD15 sub, I couldn't be bothered to be perfectly honest.

 

All the more reason to feel free to tweak the results of any calibration.

In my case it was crossovers for the mains, increasing volume level on the center and decreasing volume level on the rears.

Even if I had the capability to do manual EQ of frequency ranges, like I can with my DD15 sub, I couldn't be bothered to be perfectly honest.

DEQ strikes again!

 

From the Denon website:

Is the supplied Denon (Audyssey) microphone calibrated?

Yes, the supplied Denon (Audyssey) microphone is calibrated to a 1/4" industry standard measurement microphone. The correction is applied to the measurements as they are being taken. It is important to use ONLY the microphone that is supplied with the Denon receiver as the calibration curve built into the receiver is specific to that microphone. It is also critical to point the microphone upwards and to place it at ear level. Any other microphone will have different characteristics and will not produce the correct results. Each model's microphone has different calibration curves and will not be interchangeable between models.

 

I must of missed something...
Can someone pls explain/define for me what 1/4" industry standard measurement microphone is..

Thnx..

Just my $0.02...

 

That article paragraph is talking about only the same model number. To my knowledge all current Audyssey equipped AVRs use the same mic.


https://audyssey.zendesk.com/hc/en-us/articles/212345843-Calibration-Mic
https://audyssey.zendesk.com/hc/en-us/articles/212347783

 

Something else I noticed is that I couldn't find any serial number or any other unique identifier on the Audyssey mics. There was a sticker with a date on it, but I'd imagine there are hundreds of mics with the same date.

To me that's a pretty good indication that they're not maintaining a database of sensitivity + curve data for each and every mic.

 

I am absolutely certain that they are not, from any number of sources, including the creator of Audyssey. The ACM1HB mics are interchangeable. As your test showed, although there was some variation in SPL, the frequency responses of the microphones you tested were remarkably similar. That could only happen with interchangeable mics, all calibrated in the same way, and operating within their specified tolerances of +/- 2db.

And, to reiterate for the benefit of the poster who was concerned about the size of that variation. The variation in calibrating an HT system to 73db versus 75db, or 77db, is essentially irrelevant as long as all channels are calibrated equally at the MLP, so that they produce the same SPL at the MLP; and as long as all control points within channels are set consistently. And, properly operating Audyssey microphones are quite reliable in those respects.

Regards,
Mike

 

That is what I took from your post above, but I took away some different opinions based on the testing.

For starters, this is only a test of 3 different mics, and yet we saw deviations greater than +/- 2db. The 8802 mic was almost 3db more sensitive, and when level matched at 1khz, it had a deviation greater than 4db over 9khz.

The way I see it, that's +/- 4db with an extremely limited test of only 3 mics. The more mics you test the more deviation you'll see.

If you were to run Audyssey with the 8802 mic vs. the other two on the same receiver, "reference" will be ~3db lower in volume, but the high end will subjectively sound "brighter" because of the 4db of HF roll-off that Audyssey will correct/raise with EQ.

The 3db at reference isn't a huge deal, but the 4db difference on the HF side, in my opinion, is.

I think that's was contributing to "I moved from Denon with XT32 to a Marantz with XT32 and the difference was life changing" type of experiences.

 

Has anyone ever posted calibration data for these Audyssey mics?

 

I had mine calibrated by CSL. See attached for PDF report.

 

Hmmm....
No mention of the situation when a major mass production lot of out-of-spec Audyssey microphones were shipped out...
The respcetive AVR brands won't confirm this but it did happen..

Just my $0.02...

 

That time-frame wasn't in the last couple years no?

 

At that time Audyssey officially declared, that they would (free of charge) exchange those microphones against the current model. But this has already happened several years ago... (around turn of the decade ?)

 

Photobucket hosed all the images in the first post. I just re-uploaded all of them to imgur and fixed the links.