Redefine your budget room EQ 'flat' target curve to Harman's pro curve

JackD201

WBF Founding Member
Apr 20, 2010
12,308
1,425
1,820
Manila, Philippines
A couple of your articles helped me a lot when I was setting up my home theater. Seems as good a time as any to thank you :)
 

CherylJosie

Member
Apr 18, 2015
51
0
6
So among all these replies I am still not seeing any comment on my suggestion to smooth the speaker response above the transition and extrapolate it back below the transition but avoid changing the overall spectral balance i.e. don't tilt the target curve off the slope that the speaker finds on its own when placed into the listening environment.

I was thinking that was potentially an evolutionary step toward determining a house curve by introducing the concept that a correctly chosen speaker in a well treated room already does a fairly good job under its own steam because that is the way it was designed to work in the first place.

That thread here:

http://www.hometheatershack.com/forums/rew-forum/96-house-curve-what-why-you-need-how-do.html

The problem is that an ideal house curve varies from one room to the next. The rule of thumb is, the smaller the room, the steeper the slope needs to be.

completely validates my suggestion IMO.

The rationalization I use to explain what so many EQ-hating people report is that humans are very forgiving of aberrations that do not affect psychoacoustic perception.

While the power response is directly audible to a microphone with slow sine sweeps, power response is sort of non-sequitor to humans. What we care about more is being able to locate a sound source in space (imaging) with good intelligibility and that is strongly dependent on our ability to distinguish between direct and ambient sound using precedence effect.

On-axis speaker response is generally tuned flat and that tuning is psychoacoustically preserved by precedence effect in a small room despite power response being thrown off with ambient reflections, modes, cabin gain etc. so the motivation, the 'problem I am trying to solve' in Amir's parlance, is not letting the room EQ corrupt that on-axis response with a target curve that alters the spectral balance of the sound power, disrupts the perception of the precedence effect, and destroys the imaging.

I was sort of wondering why no one is marketing an EQ that detects the native 'house curve' of the speaker with no EQ on it and just smooths that native response rather than imposing a target upon it.

Regarding the phantom power, does Onkyo microphone really require it? Seems to work in my sound card...

I am working on the concept of the attenuator. As a rough guess I figure anywhere between +4dB and -10dB line level graphic EQ out needs attenuation back to about -60dB or -70dB dynamic mic level so I figure about 6-7 decades or 2^6=64:1 2^7=128:1 about 100:1 resistive divider or maybe 10K:0.1K should do it? Not that familiar with audio.
 

NorthStar

Member
Feb 8, 2011
24,305
1,323
435
Vancouver Island, B.C. Canada


Thanks. I do subscribe to a number of the Forums and get daily notifications of the threads, but don’t find much that I’m qualified to comment on.
You guys are way over my head. :)

Regards,
Wayne A. Pflughaupt

On what topic? ;)
 
Last edited:

Wayne A. Pflughaupt

Well-Known Member
Oct 15, 2012
100
38
268


So among all these replies I am still not seeing any comment on my suggestion to smooth the speaker response above the transition and extrapolate it back below the transition but avoid changing the overall spectral balance i.e. don't tilt the target curve off the slope that the speaker finds on its own when placed into the listening environment.
I only speak plain English, but it you’re talking about EQing the main channel speakers, that was only vaguely clear in your original post. Basically I have no problem with it. Based on the frequency response graphs I see in the magazine reviews, most speakers could benefit with some EQ. The problem is that most people don’t know how to do it right, and / or use cheap equalizers to do it, and then when they screw it up they go on the internet ranting that EQ doesn’t work, causes as many problems as it solves, etc.


Regarding the phantom power, does Onkyo microphone really require it? Seems to work in my sound card...
If you’re talking about something that came with a AVR, that is not a measurement mic. For starters, you have no custom calibration file for it, so you don’t know where its response ends and the room’s begins. Those mics only work for AVRs because the calibration is built into the receiver itself.

First the cheap equalizer, now the cheap mic. I hope this doesn’t sound too rude, but if you’re serious about all this, you need to “up your game” and spring for the right (read “good”) equipment.


I am working on the concept of the attenuator. As a rough guess I figure anywhere between +4dB and -10dB line level graphic EQ out needs attenuation back to about -60dB or -70dB dynamic mic level so I figure about 6-7 decades or 2^6=64:1 2^7=128:1 about 100:1 resistive divider or maybe 10K:0.1K should do it?
If I understand where you’re coming from here, the equalizer isn’t going to give a massive boost to the signal. It will go out at the same level it came in, +/- any overall changes in gain brought by the filters boosted or cut.

Regards,
Wayne A. Pflughaupt
 

CherylJosie

Member
Apr 18, 2015
51
0
6
[POST]204230[/POST]

I looked over that post and frankly it confuses me.

I just started watching the video but I read this paper:

Robust Loudspeaker Equalization Based on Position-Independent Excess Phase Modeling

http://202.114.89.42/resource/pdf/1545.pdf

They show a couple of zeros that they are interested in EQing out of the response in fig 1 (they settle on 135Hz zero).

Then they explain that they are adding this correction in the phase rather than the amplitude.

They speak of keeping the pre-ringing down to an acceptable level. Fig. 4 Impulse Response Maximum Level Envelopes is supposed to illustrate what, the magnitude of the pre-ringing? What is a Maximum Level Envelope (google reveals nothing). The only thing I could find that looks promising is this:

https://upload.wikimedia.org/wikipedia/commons/b/b3/EnvelopeAnim.gif
https://en.wikipedia.org/wiki/Envelope_%28mathematics%29

so I am assuming that what they are presenting in that graphic is a family of curves derived off impulse responses but no function is given to define the family of impulse responses. I fail to understand why impulse responses would appear to converge to a non-zero level at infinite time though... they should all converge to zero.:confused:

Fig. 5 shows that the step response is faster and squarer but what is an 'average energy step response'? Fig. 6 shows something called Schroeder Decay with apparently significant improvement with a full spectrum signal but nearly no improvement below 300Hz when they were equalizing a zero at 135Hz and this result is counterintuitive.

Something is flying way over my head. Does anyone at WBF understand this paper?
 

amirm

Banned
Apr 2, 2010
15,813
37
0
Seattle, WA
I looked over that post and frankly it confuses me.

I just started watching the video but I read this paper:

Robust Loudspeaker Equalization Based on Position-Independent Excess Phase Modeling

http://202.114.89.42/resource/pdf/1545.pdf

They show a couple of zeros that they are interested in EQing out of the response in fig 1 (they settle on 135Hz zero).

Then they explain that they are adding this correction in the phase rather than the amplitude.
If you are able to read and understand that paper, you don't need the rest of us :). That paper is way, way into the weeds for the topic at hand. It is written for someone designing an Auto-EQ, not as a method to teach people what to do. Specifically, they are trying to solve the problem of the room having different response with respect to seating/mic location. If you listen alone and in the same sweet spot, then you don't need to worry about that. If you do want to create a wider listening area, the solution to that is to measure multiple locations and average them. While there are a lot of papers attempting to show better methods, none are fool proof and simple "spatial" averaging works as well.

"Phase" correction does come into play here but its use is confusing in that we are just talking about delay. Adding delay in the sub for example can help optimize the response in the room relative to the mains or multiple subs. See this article of mine: http://www.madronadigital.com/Library/Computer Optimization of Acoustics.html

They speak of keeping the pre-ringing down to an acceptable level. Fig. 4 Impulse Response Maximum Level Envelopes is supposed to illustrate what, the magnitude of the pre-ringing? What is a Maximum Level Envelope (google reveals nothing). The only thing I could find that looks promising is this:

https://upload.wikimedia.org/wikipedia/commons/b/b3/EnvelopeAnim.gif
https://en.wikipedia.org/wiki/Envelope_%28mathematics%29

so I am assuming that what they are presenting in that graphic is a family of curves derived off impulse responses but no function is given to define the family of impulse responses. I fail to understand why impulse responses would appear to converge to a non-zero level at infinite time though... they should all converge to zero.:confused:

Fig. 5 shows that the step response is faster and squarer but what is an 'average energy step response'? Fig. 6 shows something called Schroeder Decay with apparently significant improvement with a full spectrum signal but nearly no improvement below 300Hz when they were equalizing a zero at 135Hz and this result is counterintuitive.

Something is flying way over my head. Does anyone at WBF understand this paper?
I do :). And again, my advice is to not go there. This is not the type of paper you want to read. It is like trying to figure out how to drive a car by reading a manual on how the valves are designed in your engine.
 

CherylJosie

Member
Apr 18, 2015
51
0
6


I only speak plain English, but it you’re talking about EQing the main channel speakers, that was only vaguely clear in your original post. Basically I have no problem with it. Based on the frequency response graphs I see in the magazine reviews, most speakers could benefit with some EQ. The problem is that most people don’t know how to do it right, and / or use cheap equalizers to do it, and then when they screw it up they go on the internet ranting that EQ doesn’t work, causes as many problems as it solves, etc.


Yes I agree the way I posted the concept was confusing.

I discussed two options. One was using external EQ to tweak the microphone response and tilt the Audyssey target toward a Harman target. The second was measuring the actual spectrum of the speaker in the room and using a mathematically smoothed (linear fit) version of that response above the Schroeder frequency to tilt the EQ target. I also discussed merging those two approaches into a combined target, and I discussed using the smoothed native response per speaker rather than per system as the target. Confusing, yes.

If you’re talking about something that came with a AVR, that is not a measurement mic. For starters, you have no custom calibration file for it, so you don’t know where its response ends and the room’s begins. Those mics only work for AVRs because the calibration is built into the receiver itself.

When attempting to use an EQ on the mic to tilt the target curve of the auto correction algorithm, I have to use the mic that came with the receiver. The calibration file is in the receiver's firmware.

For REW I would have to buy a microphone. Not there yet but I got recommendation on the UMIK. Amirm suggested that I try using the receiver mic to get familiar with REW. I need to install updated OS first because my sound card is not working with the current version. Somehow the driver broke on the prior release. Hazards of using open source is hardware support is sporadic.

All is in process.

First the cheap equalizer, now the cheap mic. I hope this doesn’t sound too rude, but if you’re serious about all this, you need to “up your game” and spring for the right (read “good”) equipment.

I already sank over $6K into this system. Funding is not immediately available for such expenditure.

I have considered doing a GoFundMe on this EQ tweak project but at this point I am content to just make do with what I can scrounge. Sort of the whole point of re-defining the budget EQ target in the first place is saving money. Frugality comes with the territory. If I cannot do it without spending gobs of money on extra equipment that sort of defeats the whole purpose.

I am hoping to get it working first and measure it later. Same with the room treatments I installed. They are hung on light curtain rods and easily removed so while I make permanent seams instead of pins I expect I will be taking measurements too. Proving the improvements is not the priority. Achieving them is, so that is what I am focusing on.

If I understand where you’re coming from here, the equalizer isn’t going to give a massive boost to the signal. It will go out at the same level it came in, +/- any overall changes in gain brought by the filters boosted or cut.

Regards,
Wayne A. Pflughaupt

Correct, any signal from the mic would come out at mic level and that is a problem with the gain mismatch of mic and EQ. The s/n on the EQ is not large enough to overcome the gain mismatch and the impedance likely has some mismatch also. I need to use a pre-amp with the receiver's mic to boost it to line level for the EQ and then attenuate it back to mic level for the receiver input.

Suspect that maybe you are not getting the big picture. Unsurprising since I have not posted any. Not literate with Linux graphics tools, but as I get into this I might draw up a schematic.

Analog EQ Plan is:

Audyssey mic -> preamp (mixing board) -> analog EQ -> 100:1 attenuator -> receiver

Set all bands to unity gain and adjust the in/out gains of the preamp/EQ until I can duplicate the receiver's 75dB subwoofer SPL in the auto cal routine. Now gains are all matched.

Adjust the EQ to the inverse of the Harman curve with unity gain at 600Hz.

Run auto cal. Cal result resembles Harman curve rather than Audyssey 'flat' curve.

Sound card (DSP) Plan is:

Audyssey mic -> sound card w/VST etc. EQ -> 100:1 attenuator -> receiver

etc...

Until I get a mic and learn to use REW I will have to tune the EQ by ear. I already tuned the absorption by ear. Seemed to help.
 

CherylJosie

Member
Apr 18, 2015
51
0
6
Those are phenomenal articles Wayne and must read for anyone interested in this topic. They are just as valuable today as they were then for our audience here which does not use AVRs or Audyssey.

We agree on the recommendation to simply boost the sub output as a cheap/easy/free approximation to a sloping down curve for people who do use Audyssey.

I agree that those articles are a landmark.

I have tried that sub boost and not liked the boomy/muddy result. The vocals still sound nasally also. I basically discounted that approach.

I did not try combining sub boost with the other methods I posted. Maybe the flat curve with tone controls and sub tweak is a possible approach but it still seems counterintuitive.

The B&K curve would seem to counter-indicate any subwoofer boost but rather seems to indicate treble cut. In that respect the Audyssey 'reference' or 'movie' curve is more accurate approximation and sounds more accurate to me also than just subwoofer boost. Problem is Audyssey reference treble attenuation is neither smooth nor broadband. It does not span enough of the frequency range. Neither does subwoofer boost. Also, shelving countour is not present in the B&K or Harman targets at any frequency, but more shelves in the treble (not the sub-bass) might help with approximation to B&K curve.

I am still wondering about my supposition that the precedence effect counter-indicates any tweak of the general slope to the power response. That video at:

http://twit.tv/show/home-theater-geeks/164

seems to indicate that the head-related transfer function and brain processing of the precedence effect allows us to hear the generally flat character of the on-axis frequency response of the speaker through the room despite the treble attenuation in the power response picked up by a measurement microphone.

Also, the power response and the on-axis frequency response converge as the wavelength increases and the radiation pattern becomes omnidirectional. This explains the flattish nature of the bass portion of the B&K curve and would tend to indicate that the lower frequency portion of the Harman target may be incorrect anyway.

Presumably then the subwoofer EQ should be flat too since it likely exists entirely below Schroeder.

Since the power response is going to be placement-dependent (how much reflected energy is reaching a given listening position?), that means that any fixed target EQ curve is going to be wrong. The target has to be modeled on each individual speaker in its given location or the 'room curve' will distort the wideband frequency balance of the speakers.

This was one of my objections to the method used in the Harman paper. They seemed to cherry-pick a curve to one speaker in one room and then wrote it up as if it were a panacea. Surprise, surprise when the slope of their curve very much resembles a straight-line fit to their measured un-EQd B&W power response in that room above the Schroeder transition frequency.

Still waiting for someone in the know to jump on this concept and give it a thumbs-up or a thumbs-down.

I understand that common practice is to use a single target throughout the room. My contention is that the target should be flat on-axis response native to a quality speaker and that response is only indirectly measureable in the actual non-flat power response of the speaker above the Schroeder transition frequency with that particular speaker in that particular placement. Preserving that flat on-axis response above the Schroeder transition frequency is the goal I was referring to.

Taking it one step farther, it seems that maybe a linear fit is not correct at all. Maybe it needs a second-order fit to reproduce that B&K curve smoothly through the Schroeder transition frequency. Maybe developing that second-order fit off the unequalized in-room power response and replicating that smooth inverted bowl shape in the B&K curve is the better approach.

This would more closely resemble the approach used in dedicated sub-only EQ that is popular among pro audiophiles -- flatten the omnidirectional bass response, and let the speakers do their own thing above the Schroeder frequency. I propose using the measured in-room response of each speaker to approximate that same approach for all frequencies below the Schroeder transition and a 2nd order fit to the native in-room power response would accomplish that while smoothing out any profound medium/narrow-band frequency response deviations from that curve.

One more time. Comments?:)

Or did my muddy language and lack of graphic illustrations confuse everyone again?

I might be able to post some diagrams eventually. Gotta run. Let me know what you think.
 

NorthStar

Member
Feb 8, 2011
24,305
1,323
435
Vancouver Island, B.C. Canada
Hi Cheryl,

I always enjoy reading your fascinating sound exploration on a high technical level. I would love to see/hear Dr. Floyd E. Toole himself in person to reply to your posts and questions.
Only true experts of sound propagation and room's acoustics could really answer your highly technical quests, I believe. ...With years of scientific sound exploration behind them.

What is the true best/perfect house curve? ...The target curve that we shall all abide by? ...Be it "totallement horizontale ou penchante".
...Before, and after. I just don't know; I go with what has been written by the certified scientific sound reproduction masters.
Here; Amir, Wayne, Nyal, Sean, Michael, ...and several others.

I'm here just like you; to learn the ultimate truth about what's best. :b
 

Fitzcaraldo215

New Member
Nov 3, 2014
394
2
0
So among all these replies I am still not seeing any comment on my suggestion to smooth the speaker response above the transition and extrapolate it back below the transition but avoid changing the overall spectral balance i.e. don't tilt the target curve off the slope that the speaker finds on its own when placed into the listening environment.

I was thinking that was potentially an evolutionary step toward determining a house curve by introducing the concept that a correctly chosen speaker in a well treated room already does a fairly good job under its own steam because that is the way it was designed to work in the first place.

That thread here:

http://www.hometheatershack.com/forums/rew-forum/96-house-curve-what-why-you-need-how-do.html



completely validates my suggestion IMO.

The rationalization I use to explain what so many EQ-hating people report is that humans are very forgiving of aberrations that do not affect psychoacoustic perception.

While the power response is directly audible to a microphone with slow sine sweeps, power response is sort of non-sequitor to humans. What we care about more is being able to locate a sound source in space (imaging) with good intelligibility and that is strongly dependent on our ability to distinguish between direct and ambient sound using precedence effect.

On-axis speaker response is generally tuned flat and that tuning is psychoacoustically preserved by precedence effect in a small room despite power response being thrown off with ambient reflections, modes, cabin gain etc. so the motivation, the 'problem I am trying to solve' in Amir's parlance, is not letting the room EQ corrupt that on-axis response with a target curve that alters the spectral balance of the sound power, disrupts the perception of the precedence effect, and destroys the imaging.

I was sort of wondering why no one is marketing an EQ that detects the native 'house curve' of the speaker with no EQ on it and just smooths that native response rather than imposing a target upon it.

Regarding the phantom power, does Onkyo microphone really require it? Seems to work in my sound card...

I am working on the concept of the attenuator. As a rough guess I figure anywhere between +4dB and -10dB line level graphic EQ out needs attenuation back to about -60dB or -70dB dynamic mic level so I figure about 6-7 decades or 2^6=64:1 2^7=128:1 about 100:1 resistive divider or maybe 10K:0.1K should do it? Not that familiar with audio.

Cheryl - not sure if they are helpful, but here are several points.

I used Audyssey XT and XT/32 for years with the Pro kit. It is not the greatest. Dirac Live, which I use now, is better in many ways and it sounds much better. But, on the other hand, Audyssey sounded way better than no EQ with music as well as movies in all the rooms in which I tried it. This may be at odds with the Sean Olive paper on room EQ systems, but a number of friends and I were unanimous in that view, primarily with music. That is not to say there are not better Room EQ systems, but when I started with Audyssey 8 years ago, there were not too many.

There may be many quibbles with the Audyssey target curve. I find its only really major flaw to be the incorporation of the BBC Dip at 2.5 kHz or so, which I was able to quite simply remove via the Pro kit. In other ways, the target curve is generally similar to many other default target curves, with some usually smallish differences. Maybe there was not enough room gain type bass lift, but that did not bother me.

Contrary to your interpretation, I find the standard Audyssey curve with BBC dip eliminated was closer to other systems' Room EQ target curves than it is to the cinema X-curve. The Audyssey flat target curve, mistakenly called the "music curve", was also far less preferable. I do not know of any small room target curve that is flat in the upper frequencies, nor have I seen any notable commentators who advocate that. Why it is even there or why it is called a "music curve" does nothing but cause confusion. I see no use for it.

I also found no use for Audyssey Dynamic EQ or Dynamic Volume. They are not really calibrated for music, my primary interest. Superficially, they seem like good ideas. I just could not get them to sound right, even with movies.

A consideration in attempting to trick Audyssey via your sound card workaround is the mike's own calibration curve. The stock Audyssey mike has a batch mike calibration curve built into the AVR. The Audyssey Pro mike has its individual calibration curve in a uniquely numbered file which is loaded on the PC with the Pro software, which refers to it.

I am still somewhat confused about your approach, its feasibility and whether it is worthwhile. Personally, I have not had much temptation or success with target curve manipulation. I generally always found I was happiest with the standard Audyssey (BBC dip removed ) or Dirac curves. I also think Dirac's sonic superiority is much less a function of how its target curve differs from Audyssey (BBC dip removed) than about other technical aspects of these two EQ tools.

Just yesterday, I did help a friend create a set of modified Dirac target curves for his comparative listening. Dirac is much easier to use than Audyssey Pro for this or other purposes. He wanted that, but I am skeptical that the modified target curves will be worthwhile.

Note that Dirac allows quick, on the fly selection of up to 4 different filter sets which can be just different target curves. I think the ability to quickly audition alternative target curves is vitally important in making any sonic assessment. I am not aware of any other tools that allow this.

My own upgrade to Dirac was not just about seeking a better EQ tool, but that result came as a byproduct of a shift in my system architecture, which Dirac helped make possible. I evolved as follows:
(a.) disc player via HDMI to prepro with Audyssey
(b.) PC optical disc playback and rips in JRiver via HDMI from PC to prepro with Audyssey
(c.) PC optical disc playback and rips in JRiver with Dirac via HDMI from PC to prepro in Direct Mode (no Audyssey)
(d.) PC optical disc playback and rips in JRiver with Dirac via asynch USB from PC to Exasound e28 DAC

I have never been happier with my sound than I am now, thanks to Dirac and the Exasound. My Integra 80.2 prepro and Oppo player are now out of the system. (My TV cable box is also, but that is another story.) The friend I mentioned still uses (c.) above with a Marantz 8801 prepro. He was also quite happy with the sonic upgrade from the Audyssey Pro calibration I had done for him several years ago to Dirac Live for about the last 6 months.


I have rambled on. My advice is forget about your complex scheme to work around Audyssey and bite the bullet. Use the PC as your player for rips and discs via its optical drive, and get Dirac on the PC using my (c.) alternative above with your AVR. If you still wish to play around with target curves, Dirac will let you do that quite easily.

PS - I should have mentioned this, but the bulk of my listening is to hi rez multichannel recordings of classical music. I also go to good number of live classical concerts, about 2/month on average.
 

Wayne A. Pflughaupt

Well-Known Member
Oct 15, 2012
100
38
268


I agree that those articles are a landmark.

I have tried that sub boost and not liked the boomy/muddy result. The vocals still sound nasally also. I basically discounted that approach.
I’m going to hazard a guess that you have not equalized your sub? If so, that’s the problem, not the “approach.” If the vocals sound “nasally,” I’m going to hazard a guess that you really mean they sound “thin” and lacking warmth? If so, that’s an indication that your speakers may be too small for the room.

What speakers are you using, and what is the size of your room (cu. ft.)?


For REW I would have to buy a microphone.
I already sank over $6K into this system. Funding is not immediately available for such expenditure.
The mic with calibration file will cost less than $100. The Yamaha EQ I mentioned can be ebay’d for usually less than $150.


Analog EQ Plan is:

Audyssey mic -> preamp (mixing board) -> analog EQ -> 100:1 attenuator -> receiver

Set all bands to unity gain and adjust the in/out gains of the preamp/EQ until I can duplicate the receiver's 75dB subwoofer SPL in the auto cal routine. Now gains are all matched.

Adjust the EQ to the inverse of the Harman curve with unity gain at 600Hz.

Run auto cal. Cal result resembles Harman curve rather than Audyssey 'flat' curve.
Okay I think I finally get it. You’re ultimately running the signal into the receiver's calibration mic input? I don’t think you mentioned that specifically, so I figured you were using one of the regular inputs. Hence my comments that no post-EQ gain adjustment is necessary.

An easy way to decrease the output after the EQ is to simply use a second pro-audio mixer. Plug the EQ output into one of the channels, then reduce the main output to where you need it.

And the whole idea is fool Audyssey into auto-calibrating a Harman curve? Problem is, it’s never a good idea to use a “pat” house curve. As fully discussed and explored in my house curve article, every room is different. The slope you need will depend on the size of the room and your distance from the speakers.

Honestly, you could easily accomplish the general Harman slope with the receiver's tone controls, after Audyssey has auto-calibrated flat response. A bit up with the bass control, a bit down with the treble control would do it.

Regards,
Wayne A. Pflughaupt
 

CherylJosie

Member
Apr 18, 2015
51
0
6
Hey veryone thanks for your helpful insights

I am having some difficulty with the Audyssey mic.

It seems that the Zoom R24 mixing board will not generate any output with the Audyssey mic, though my computer sound card will. I am going to try to get analog output working on the sound card so I can use it as a preamp.

https://en.wikipedia.org/wiki/Phantom_power

Plug-in-power (PiP), is the low-current 3 V to 5 V supply provided at the microphone jack of some consumer equipment, such as portable recorders and computer sound cards. It is also defined in IEC 61938.[11] It is unlike phantom power since it is an unbalanced interface with a low voltage (around +5 volts) connected to the signal conductor with return through the sleeve; the DC power is in common with the audio signal from the microphone. A Capacitor is used to block the DC from subsequent audio frequency circuits. It is often used for powering electret microphones, which will not function without power. It is suitable only for powering microphones specifically designed for use with this type of power supply. Damage may result if these microphones are connected to true (48 V) phantom power through a 3.5 mm to XLR adapter that connects the XLR shield to the 3.5 mm sleeve.[12] Plug-in-power is covered by Japanese standard CP-1203A:2007[13] A similar line-powering scheme is found in computer sound cards. Both plug-in-power and soundcard power are defined in the second edition of IEC 61938.[14]

These alternative powering schemes are sometimes improperly referred to as "phantom power" and should not be confused with true 48-volt phantom powering described above.

https://suite.io/richard-mudhar/1prr2nh
https://suite.io/richard-mudhar/24hr2nh

My Fluke tells me that the Onkyo is generating 5V at 1mA. I guess that means I need to add a 5V supply and a 5K resistor (4.7K?) to power the Audyssey mic, or use the sound card as my preamp.

Fortunately I was bright enough not to turn on the phantom power on the mixing board. 48V might have damaged the mic.

Will let you know what happens.

Current plan is to use the Zoom as a pre-amp with a 5V/1mA plug-in power adapter (probably DIY) and put 100:1 attenuator on the output of the Behringer.
 

CherylJosie

Member
Apr 18, 2015
51
0
6


I’m going to hazard a guess that you have not equalized your sub? If so, that’s the problem, not the “approach.” If the vocals sound “nasally,” I’m going to hazard a guess that you really mean they sound “thin” and lacking warmth? If so, that’s an indication that your speakers may be too small for the room.

What speakers are you using, and what is the size of your room (cu. ft.)?


The mic with calibration file will cost less than $100. The Yamaha EQ I mentioned can be ebay’d for usually less than $150.


Okay I think I finally get it. You’re ultimately running the signal into the receiver's calibration mic input? I don’t think you mentioned that specifically, so I figured you were using one of the regular inputs. Hence my comments that no post-EQ gain adjustment is necessary.

An easy way to decrease the output after the EQ is to simply use a second pro-audio mixer. Plug the EQ output into one of the channels, then reduce the main output to where you need it.

And the whole idea is fool Audyssey into auto-calibrating a Harman curve? Problem is, it’s never a good idea to use a “pat” house curve. As fully discussed and explored in my house curve article, every room is different. The slope you need will depend on the size of the room and your distance from the speakers.

Honestly, you could easily accomplish the general Harman slope with the receiver's tone controls, after Audyssey has auto-calibrated flat response. A bit up with the bass control, a bit down with the treble control would do it.

Regards,
Wayne A. Pflughaupt

Hey Wayne, your articles stated the obvious that everyone was overlooking. The power response of a speaker changes from the flat on-axis response due to adding in reflected off-axis. Every speaker in every placement goes through this change. Our ears largely hear through it but only in the mid/treble. The rest needs EQ that follows the same basic countour as the rest of the speaker power response with some smoothing, not bending, of the overall curve. That changes from room to room and placement to placement. You nailed it.

That was what I objected to in Harman's study, was the choosing of a curve that sort of fit one speaker in one room and claiming the job was finished because people liked the sound. What else was to be expected? What they needed to do was teach the EQ what the general countour of the un-EQd response is, with smoothing, and use that as the target. Done. Simple, but apparently far simpler for a human that can actually hear the on-axis response through the music and tune it using the DSP in the brain to understand and analyze it.

What I am proposing is to let the auto EQ 'hear' the native power response by measuring it offline and then subtracting a smoothed version of the 'flat' power response target from the feedback via injection into the microphone path, per speaker, as the auto-EQ progresses around the channels. Subtracting EQ from the feedback path amounts to adding EQ to the output so the smoothed version of the native power response is the resulting power response tune. As the power response is tuned, hopefully any aberrations it puts on the perceived on-axis response is tamed. That is all the EQ should be doing IMO.

First measure it in REW over a few positions, average it, curve-fit the average with 2nd order least squares algorithm, then subtract that out from the feedback path through the microphone input. A smoother version of the native speaker response results and the human ear hears that as natural improvement without major disruption to the on-axis response. It works out to a compromise solution area-wise and frequency-wise.

No need to fiddle with room curves and no suboptimal on-axis results. Everything in alignment pretty much where it naturally landed anyway on a speaker-by-speaker basis.

Isn't that the way that humans EQ a system manually? I know I do. The first thing you learn about using manual EQ is that it rarely pays to try to bend something way off where it rests naturally.
 

amirm

Banned
Apr 2, 2010
15,813
37
0
Seattle, WA
Hey veryone thanks for your helpful insights

I am having some difficulty with the Audyssey mic.

It seems that the Zoom R24 mixing board will not generate any output with the Audyssey mic, though my computer sound card will. I am going to try to get analog output working on the sound card so I can use it as a preamp.

https://en.wikipedia.org/wiki/Phantom_power



https://suite.io/richard-mudhar/1prr2nh
https://suite.io/richard-mudhar/24hr2nh

My Fluke tells me that the Onkyo is generating 5V at 1mA. I guess that means I need to add a 5V supply and a 5K resistor (4.7K?) to power the Audyssey mic, or use the sound card as my preamp.

Fortunately I was bright enough not to turn on the phantom power on the mixing board. 48V might have damaged the mic.

Will let you know what happens.

Current plan is to use the Zoom as a pre-amp with a 5V/1mA plug-in power adapter (probably DIY) and put 100:1 attenuator on the output of the Behringer.
I have never seen an AVR mic that needs phantom power. I would not send it juice.
 

amirm

Banned
Apr 2, 2010
15,813
37
0
Seattle, WA
That was what I objected to in Harman's study, was the choosing of a curve that sort of fit one speaker in one room and claiming the job was finished because people liked the sound. What else was to be expected? What they needed to do was teach the EQ what the general countour of the un-EQd response is, with smoothing, and use that as the target. Done. Simple, but apparently far simpler for a human that can actually hear the on-axis response through the music and tune it using the DSP in the brain to understand and analyze it.
How does a human only hear the on-axis response?
 

NorthStar

Member
Feb 8, 2011
24,305
1,323
435
Vancouver Island, B.C. Canada
True, all AVRs use small plastic cheap mics. ...Two dollars @ most to make by the manufacturers, with a thin extension cord (about 20 feet long and not interchangeable, both the mic and the thin cable).

Anthem AVRs use a better mic though, and it comes with a small tripod too. ...All better quality, not as cheap looking.

And no AVR comes with a phantom power supply, none. ...No need for with a cheap tiny plastic mic of a simple AV receiver.
Even one that costs $5,000 @ retail.

That should give us a good hint about what they think the people who sell hi-end AV receivers.

* In reference to the last post from the last page (post #40).
 
Last edited:

Fitzcaraldo215

New Member
Nov 3, 2014
394
2
0
Hey veryone thanks for your helpful insights

I am having some difficulty with the Audyssey mic.



It seems that the Zoom R24 mixing board will not generate any output with the Audyssey mic, though my computer sound card will. I am going to try to get analog output working on the sound card so I can use it as a preamp.

https://en.wikipedia.org/wiki/Phantom_power



https://suite.io/richard-mudhar/1prr2nh
https://suite.io/richard-mudhar/24hr2nh

My Fluke tells me that the Onkyo is generating 5V at 1mA. I guess that means I need to add a 5V supply and a 5K resistor (4.7K?) to power the Audyssey mic, or use the sound card as my preamp.

Fortunately I was bright enough not to turn on the phantom power on the mixing board. 48V might have damaged the mic.

Will let you know what happens.

Current plan is to use the Zoom as a pre-amp with a 5V/1mA plug-in power adapter (probably DIY) and put 100:1 attenuator on the output of the Behringer.

I think Amir is right about the Audyssey mike. I do not think it needs phantom power.

Pardon me, Cheryl if I shake my head a little bit. You seem in many ways well prepared technically, and it is clear you are determined to solve this puzzle you have laid out for yourself. But, from a distance, it seems a mismatch to try to adapt crappy, imprecise tools like the Audyssey mike on one hand, then be concerned about fine tuning the target curve on the other hand. I am not suggesting you replace the mike and continue, because you will be unable to overcome the mike calibration curve built into the AVR for the Audyssey mike. We have no idea what that looks like.

Again, I suggest you get the right tools for the job, and I think Dirac running on a PC as your "player" with HDMI output to your AVR bypassing Audyssey altogether would be an excellent way to go. True, that would require the better part of $1,000 for the Mch PC version of Dirac including a good, individually calibrated USB mike. (I use a UMIK-1 calibrated by Cross Spectrum myself - under $100.) if you do not like Dirac, there are other PC Room EQ packages like Acourate or Audiolense. They might be even better, in some ways, but they do not appear to be as easy to use.

Another concern that dawned on me about your approach is latency and how that might affect your Audyssey calibration. If you run the Audyssey mike through your PC, its sound card and some EQ software, there might be a latency delay that could sabotage the resulting calibration, because Audyssey is not expecting it and cannot deal with it.

One aspect of that, which is the inaccurate channel delays and level trims determined by Audyssey in your workaround configuration, can be discarded via manual input of those parameters into your AVR from a straight Audyssey calibration. But, the filter calculations might still be messed up because of the latency delay in measuring response.

The problem is Audyssey is too much of a black box. We just do not know enough about its inner workings, and there is no way to find out. It is too proprietary.
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu

Steve Williams
Site Founder | Site Owner | Administrator
Ron Resnick
Site Co-Owner | Administrator
Julian (The Fixer)
Website Build | Marketing Managersing