G3P Elevated BM & LFE Issue

[405x5]

My bass is SOOOO SWEET right down to 15HZ !
So great I’m afraid to screw with it.
The only advice I can offer for all this ongoing adventure here is to cut back on the technological overload and try trusting your ears.
Additionally, predetermined manual adjustments between LFE for film and music are mandatory. So easy and fluid….what’s the big deal?? I get more sleep than you for sure (I bets)

[marcl]

Jul 19, 2021 21:36:22 GMT -5 jbrunwa said:

Jul 19, 2021 19:02:50 GMT -5 ttocs said:

Thanks for this info.

Here's some history. I like history, it's how we learn to not repeat it when history exposes failure.
I reported this issue about 10 months ago. I gave them lots of data about it last winter. I've spoken to Damon on the phone and was assured that it would be dealt with in the future. All good so far.

FW2.0 was livable. FW2.3 is not. FW2.3 has improvements I'd like to keep experiencing, but when the sound field switches to one that I'm not setup for, the bass can be disturbingly aggressive. I used to enjoy using All Stereo for unattended listening of Pandora, but not anymore with the exorbitant amount of bass it carries. I setup my Movie Night! events ahead of show time with knowledge of the audio modes that will be in play for the evening so there are no surprises.

Something has changed with 2.3. I don't know enough about it yet to know which direction I should go, but I'm probably going to have to directly compare the FW versions again to prove these play a role in this issue. I've got a play book for doing this as I've done this before, but it's the time needed to do it all in the same session so everything is as consistent as possible. Too many variables to contend with if done on different days.

I appreciate all your dedication, but it really should be up to Emotiva to define how small bass management should work, identify ALL use case scenarios that should be tested, and what fixes should be made. We should all just be enjoying music and video, or building out our systems.

It makes sense to me that bass trim should adjust all bass levels. But if that’s not how it’s supposed to work, it would be really helpful to have document that explains the nuances of bass management and how it is supposed to work. If it’s any consolation they are not alone, as Dirac seems to have been having their own bass control issues.

Thank you!  ttocs considerable efforts to precisely quantify this issue - and my support - are based on our own listening experience, and also in support of many users who have been commenting over the past several months about the sudden change in bass response.  Maybe not everyone hears it, but many do, and now we know why.

A few simple points ...

1 - bass management worked on the XMC-1 and it USED TO work on the XMC-2/RMC un until 2.2/2.3 when it was broken
2 - the only exception to this is that bass management implemented with large front speakers always had a +3db bump due to the mono summing to two speakers (+6 if the center was also large).  That is an incorrect implementation and needs to be fixed to work the way multiple subs work
3 - irrespective of personal preference for a "house curve", any system should be capable of nominally flat response ... with the current firmware it is not - under any configuration with small speakers - capable of flat response
4 - our issue is not with Dirac Live.  Dirac's bass issues are strictly with the Dirac Live Bass Control product, not Dirac Live.  Dirac acknowledged the issue and has a fix in beta, soon to be released.  Of course this does not affect us because we can't get DLBC.
5 - yes, there should be a specification as to how it should work, and the system should be tested to that spec before being delivered


These issues are not about design decisions made by Emotiva.  They are bugs.  I'll take a leap and say there is no system on the market at any price that would design in a minimum +5db bass boost for all small speakers that can't be overridden.

[geebo]

Jul 19, 2021 22:04:54 GMT -5 405x5 said:

My bass is SOOOO SWEET right down to 15HZ !
So great I’m afraid to screw with it.
The only advice I can offer for all this ongoing adventure here is to cut back on the technological overload and try trusting your ears.
Additionally, predetermined manual adjustments between LFE for film and music are mandatory. So easy and fluid….what’s the big deal?? I get more sleep than you for sure (I bets)

Make no mistake, it is immediately obvious to those of us having the issue. It's not subtle and requires no measurements to know something is wrong. The measurements haven't brought this to light so much as the audible bass increase has prompted the measurements to see exactly what is happening.

[davidl81]

Jul 20, 2021 8:14:00 GMT -5 geebo said:

Jul 19, 2021 22:04:54 GMT -5 405x5 said:

My bass is SOOOO SWEET right down to 15HZ !
So great I’m afraid to screw with it.
The only advice I can offer for all this ongoing adventure here is to cut back on the technological overload and try trusting your ears.
Additionally, predetermined manual adjustments between LFE for film and music are mandatory. So easy and fluid….what’s the big deal?? I get more sleep than you for sure (I bets)

Make no mistake, it is immediately obvious to those of us having the issue. It's not subtle and requires no measurements to know something is wrong. The measurements haven't brought this to light so much as the audible bass increase has prompted the measurements to see exactly what is happening.

Yes, this is not some Amir review that the difference cannot be heard, but maybe seen on a chart. This was instantly noticed once the upgrade to 2.2 was done. And I love some bass, but this was overwhelming bass.

[405x5]

Jul 20, 2021 8:14:00 GMT -5 geebo said:

Jul 19, 2021 22:04:54 GMT -5 405x5 said:

My bass is SOOOO SWEET right down to 15HZ !
So great I’m afraid to screw with it.
The only advice I can offer for all this ongoing adventure here is to cut back on the technological overload and try trusting your ears.
Additionally, predetermined manual adjustments between LFE for film and music are mandatory. So easy and fluid….what’s the big deal?? I get more sleep than you for sure (I bets)

Make no mistake, it is immediately obvious to those of us having the issue. It's not subtle and requires no measurements to know something is wrong. The measurements haven't brought this to light so much as the audible bass increase has prompted the measurements to see exactly what is happening.

Wish I had an answer…I am well aware that quite a few here are having this issue.

[jbrunwa]

Jul 20, 2021 0:48:31 GMT -5 marcl said:

Jul 19, 2021 21:36:22 GMT -5 jbrunwa said:

I appreciate all your dedication, but it really should be up to Emotiva to define how small bass management should work, identify ALL use case scenarios that should be tested, and what fixes should be made. We should all just be enjoying music and video, or building out our systems.

It makes sense to me that bass trim should adjust all bass levels. But if that’s not how it’s supposed to work, it would be really helpful to have document that explains the nuances of bass management and how it is supposed to work. If it’s any consolation they are not alone, as Dirac seems to have been having their own bass control issues.

Thank you!  ttocs considerable efforts to precisely quantify this issue - and my support - are based on our own listening experience, and also in support of many users who have been commenting over the past several months about the sudden change in bass response.  Maybe not everyone hears it, but many do, and now we know why.

A few simple points ...

1 - bass management worked on the XMC-1 and it USED TO work on the XMC-2/RMC un until 2.2/2.3 when it was broken
2 - the only exception to this is that bass management implemented with large front speakers always had a +3db bump due to the mono summing to two speakers (+6 if the center was also large).  That is an incorrect implementation and needs to be fixed to work the way multiple subs work
3 - irrespective of personal preference for a "house curve", any system should be capable of nominally flat response ... with the current firmware it is not - under any configuration with small speakers - capable of flat response
4 - our issue is not with Dirac Live.  Dirac's bass issues are strictly with the Dirac Live Bass Control product, not Dirac Live.  Dirac acknowledged the issue and has a fix in beta, soon to be released.  Of course this does not affect us because we can't get DLBC.
5 - yes, there should be a specification as to how it should work, and the system should be tested to that spec before being delivered


These issues are not about design decisions made by Emotiva.  They are bugs.  I'll take a leap and say there is no system on the market at any price that would design in a minimum +5db bass boost for all small speakers that can't be overridden.

I’m sure you addressed this before, but to clarify for me the way it is supposed to work: EQ including tone presets, tone trims, and (either PEQ or Dirac Live) sum to act on all outputs to the respective speaker, no matter if the signal is from the audio input to the processor or redirected small bass management? Other than reference mode that disables DSP completely there is no signal output to the speakers that bypasses these controls? So that if a particular source has too much or not enough bass, the user can adjust the trim to suit their preference?

I agree that the missing DLBC multi sub bass management is a separate issue, although direct competitors such as Arcam and Monoprice offer it, so in my opinion is quite overdue for Emotiva processors. 

[marcl]

I'm compelled to comment on the topic of measurement and listening - and not addressing anyone in particular here. I spend a LOT of time just listening ... sometimes 2hrs a day or more, focused, not looking at my phone or doing something else ... sitting in the MLP and listening. That, in addition to a couple hours of Netflix, Prime or HBO each evening. I'm very familiar with the sound in my room, so if something changes, I notice. Sometimes it's a result of a mistake I made ... and that's good because I hear something, then figure out what it is and fix it. 

I'm a musician and an engineer.  I trust my ears, AND I trust measurement.  It's science.  Sometimes it measures good and sounds bad.  Sometimes it measures bad and sounds good.  There is something to be learned by both.  Sometimes you don't hear a problem until you fix it.

Before Dirac was available I used the User PEQ filters, or at least I tried to. They didn't work on the subs and I could plainly hear - and see in the measurements - that they didn't work. So I went back and forth with Support between March and April 2020 with them saying it DID work ... but it didn't, and it was finally acknowledged in FW 1.10 that it was finally fixed. So did it not work for a year since the processor was released, and nobody knew? As with the Loudness, that did not work at all until 2.2 and is still severely broken if you use it for anything but Stereo?  Bass management with small speaker bass going to large fronts has always had an issue.  It's unreasonable to expect EVERYONE to use a sub for bass management.  This should work properly and it was reported a year ago that it doesn't.

So this issue - the topic of this thread - is not a few guys quibbling over a nuance of a measurement with dubious audibility. We aren't wasting our lives (and in my case my retirement) constantly measuring and looking for things to complain about. We measure as a check on our hearing, and to verify and quantify the problem. I make changes to my system that require rerunning Dirac, and I sometimes see issues.

Finally, may I ask ... please don't patronize or ridicule those of us who dig into the technical aspects of being an audiophile as well as LISTEN intently and for the pleasure of hearing the music the best that it can be.  We're enjoying the sound of our systems, I assure you.  But we are also contributing our time to help Emotiva at least make these products perform in a manner that is stable and meets nominal specifications both published as well as consistent with commonly accepted expectations.

[marcl]

Jul 20, 2021 11:14:01 GMT -5 jbrunwa said:

Jul 20, 2021 0:48:31 GMT -5 marcl said:

Thank you!  ttocs considerable efforts to precisely quantify this issue - and my support - are based on our own listening experience, and also in support of many users who have been commenting over the past several months about the sudden change in bass response.  Maybe not everyone hears it, but many do, and now we know why.

A few simple points ...

1 - bass management worked on the XMC-1 and it USED TO work on the XMC-2/RMC un until 2.2/2.3 when it was broken
2 - the only exception to this is that bass management implemented with large front speakers always had a +3db bump due to the mono summing to two speakers (+6 if the center was also large).  That is an incorrect implementation and needs to be fixed to work the way multiple subs work
3 - irrespective of personal preference for a "house curve", any system should be capable of nominally flat response ... with the current firmware it is not - under any configuration with small speakers - capable of flat response
4 - our issue is not with Dirac Live.  Dirac's bass issues are strictly with the Dirac Live Bass Control product, not Dirac Live.  Dirac acknowledged the issue and has a fix in beta, soon to be released.  Of course this does not affect us because we can't get DLBC.
5 - yes, there should be a specification as to how it should work, and the system should be tested to that spec before being delivered


These issues are not about design decisions made by Emotiva.  They are bugs.  I'll take a leap and say there is no system on the market at any price that would design in a minimum +5db bass boost for all small speakers that can't be overridden.

I’m sure you addressed this before, but to clarify for me the way it is supposed to work: EQ including tone presets, tone trims, and (either PEQ or Dirac Live) sum to act on all outputs to the respective speaker, no matter if the signal is from the audio input to the processor or redirected small bass management? Other than reference mode that disables DSP completely there is no signal output to the speakers that bypasses these controls? So that if a particular source has too much or not enough bass, the user can adjust the trim to suit their preference?

I agree that the missing DLBC multi sub bass management is a separate issue, although direct competitors such as Arcam and Monoprice offer it, so in my opinion is quite overdue for Emotiva processors. 

Okay there are a few things here and I'll go through step by step.

When you set a speaker size to small, and set a crossover frequency, bass below that crossover for that speaker is sent to whatever speakers are configured to receive bass management.  Level in the Preset should affect that channel such that both the bass and the sound above the crossover are affected by the Level control.  If you apply User EQ or Dirac filters, they would apply to the specific speaker channel above the crossover point.  Below the crossover point, User EQ or Dirac filters would apply to the speaker that is playing the bass portion of the channel.  Trims should also affect all the sound for a channel.  If you increase the Surround trim by 3db, then both the sound of the Surround speakers as well as the bass from the Surround channel being sent to bass management speakers should increase 3db.

The fundamental problem here, is that as of FW2.2 for ANY configuration using small speakers - one, two or three subs; or large speakers - the small speaker bass is boosted several db.  This is irrespective of Trims, Levels, User PEQ, Dirac ..... it is always boosted and can't be made flat.  The XMC-1 did not behave this way, nor did the XMC-2/RMC before FW 2.2.  All of this is true for any audio mode that includes Bass Management, and there's nothing we as users can do to fix it.

There are other aspects of this that make it worse for specific to configurations and audio modes.  But this is the fundamental problem that was introduced in 2.2, communicated to Emotiva, and which they specifically said was NOT addressed in 2.3.

[jbrunwa]

Jul 20, 2021 11:33:36 GMT -5 marcl said:

Okay there are a few things here and I'll go through step by step.

When you set a speaker size to small, and set a crossover frequency, bass below that crossover for that speaker is sent to whatever speakers are configured to receive bass management.  Level in the Preset should affect that channel such that both the bass and the sound above the crossover are affected by the Level control.  If you apply User EQ or Dirac filters, they would apply to the specific speaker channel above the crossover point.  Below the crossover point, User EQ or Dirac filters would apply to the speaker that is playing the bass portion of the channel.  Trims should also affect all the sound for a channel.  If you increase the Surround trim by 3db, then both the sound of the Surround speakers as well as the bass from the Surround channel being sent to bass management speakers should increase 3db.

The fundamental problem here, is that as of FW2.2 for ANY configuration using small speakers - one, two or three subs; or large speakers - the small speaker bass is boosted several db.  This is irrespective of Trims, Levels, User PEQ, Dirac ..... it is always boosted and can't be made flat.  The XMC-1 did not behave this way, nor did the XMC-2/RMC before FW 2.2.  All of this is true for any audio mode that includes Bass Management, and there's nothing we as users can do to fix it.

There are other aspects of this that make it worse for specific to configurations and audio modes.  But this is the fundamental problem that was introduced in 2.2, communicated to Emotiva, and which they specifically said was NOT addressed in 2.3.

Thanks. Maybe I am splitting hairs, so let me know if this question is irrelevant. Do the distance/delay parameters work in a similar manner? Is the redirected bass played using the delay/distance parameter value of the small speaker? I ask because my surrounds will be farther away from listening position than the ideal distance.

[muslhead]

Jul 20, 2021 8:14:00 GMT -5 geebo said:

Jul 19, 2021 22:04:54 GMT -5 405x5 said:

My bass is SOOOO SWEET right down to 15HZ !
So great I’m afraid to screw with it.
The only advice I can offer for all this ongoing adventure here is to cut back on the technological overload and try trusting your ears.
Additionally, predetermined manual adjustments between LFE for film and music are mandatory. So easy and fluid….what’s the big deal?? I get more sleep than you for sure (I bets)

Make no mistake, it is immediately obvious to those of us having the issue. It's not subtle and requires no measurements to know something is wrong. The measurements haven't brought this to light so much as the audible bass increase has prompted the measurements to see exactly what is happening.

^^^^^^^^
This

[marcl]

Jul 20, 2021 13:53:57 GMT -5 jbrunwa said:

Jul 20, 2021 11:33:36 GMT -5 marcl said:

Okay there are a few things here and I'll go through step by step.

When you set a speaker size to small, and set a crossover frequency, bass below that crossover for that speaker is sent to whatever speakers are configured to receive bass management.  Level in the Preset should affect that channel such that both the bass and the sound above the crossover are affected by the Level control.  If you apply User EQ or Dirac filters, they would apply to the specific speaker channel above the crossover point.  Below the crossover point, User EQ or Dirac filters would apply to the speaker that is playing the bass portion of the channel.  Trims should also affect all the sound for a channel.  If you increase the Surround trim by 3db, then both the sound of the Surround speakers as well as the bass from the Surround channel being sent to bass management speakers should increase 3db.

The fundamental problem here, is that as of FW2.2 for ANY configuration using small speakers - one, two or three subs; or large speakers - the small speaker bass is boosted several db.  This is irrespective of Trims, Levels, User PEQ, Dirac ..... it is always boosted and can't be made flat.  The XMC-1 did not behave this way, nor did the XMC-2/RMC before FW 2.2.  All of this is true for any audio mode that includes Bass Management, and there's nothing we as users can do to fix it.

There are other aspects of this that make it worse for specific to configurations and audio modes.  But this is the fundamental problem that was introduced in 2.2, communicated to Emotiva, and which they specifically said was NOT addressed in 2.3.

Thanks. Maybe I am splitting hairs, so let me know if this question is irrelevant. Do the distance/delay parameters work in a similar manner? Is the redirected bass played using the delay/distance parameter value of the small speaker? I ask because my surrounds will be farther away from listening position than the ideal distance.

That's a good question and it gets down to how Dirac works vs User EQ.  But the simple answer is that the delay setting is not a problem because both Dirac and User set the distances for the speaker playing the sound.

First Dirac ... it knows nothing of bass management and knows nothing of small vs large.  It corrects each speaker individually and sets their distance and level.  So if bass management goes to the subs, then the bass gets the sub delay, correction and level from Dirac, and the small speakers get their own delay, correction and level from Dirac which you only hear above the crossover point.  If you use large fronts for bass management, then the bass gets the delay, correction and level of the large fronts from Dirac.

If you're using User EQ, you manually set all the distances for the actual speaker distances.  Then when you measure each speaker you are measuring the small speaker with bass management included and you can make a filter to correct the full response of that speaker including bass management.  In that case (when I used User EQ) I corrected the subs first and applied the filters for them.  Then I measured the small speakers including the subs playing their bass.  That way the small speaker EQ could possibly smooth across the crossover point.  Doing that was how I found out a year ago that if bass management goes to the large fronts, User PEQ isn't applied properly.  I really don't know if that was fixed.

[ttocs]

210720-01  5.1 User EQ, All Small 150Hz-XO, Lsub Mono, Surround sound field, Levels were done after selecting the 5.1 speaker arrangement.

Just one subwoofer.


Even as a thumbnail you can see that the cursor is placed in a generous location. LFE is at +11dB.

[ttocs]

210720-02 5.1 User EQ, All Small 150Hz-XO, Lsub Mono, Surround sound field, Levels were done after selecting the 5.1 speaker arrangement.

Just one subwoofer, but this time the subwoofer level was adjusted down.


This is a good example of why having multiple subs and using multiple sub channels so one channel is Mono and the other is LFE will help when you can adjust each channel for proper level, even when the processor levels are wrong. With Left Sub Mono, the elevated bass was lowered, but at the expense of losing LFE level which needs to be higher.

If the Center Sub is setup for LFE, then its level can be left where it should be and the Left Sub Mono can be reduced to be the same as above the XO for Small channels. This won't fix what happens when using All Stereo which is really whacky.

If you only have one sub, like in these last two examples, you can connect the Left Sub Mono to the subwoofer through one connection like the Left/Right input, and then connect the Center Sub LFE output to the LFE input on the subwoofer, assuming your sub has both which can be used concurrently. If you have miniDSP then connect both sub outputs from processor to each input on miniDSP, then just one connection to the single subwoofer, and do the routing within the miniDSP.

These are just a couple of workarounds, but again, they don't solve all the problems, they just help.

[ttocs]

210720-03 5.1 User EQ, All Small 150Hz-XO, Lsub Mono, All Stereo sound field, Levels were done after selecting the 5.1 speaker arrangement.

Just one subwoofer. The subwoofer was not adjusted back up to where it should be when using Levels in the processor. Instead it was left where it was from 210720-02 which is at -12dB in Levels, it won't go any lower.

[ttocs]

210721-01 5.1.2 (to force DD Surround to be available) User EQ, DD Surround, 150Hz-XO, Lsub Mono, all Levels equal.

Just one subwoofer. The cursor is generously placed at 69dB, which makes the bass to be elevated at +4dB, and LFE is a proper +10dB.

[ttocs]

I am comparing current measurements to some I did months ago and want to be confident of all the data, so, I will be retweaking my Main speakers back to stock and using a analog interface to verify that the speakers play as flat as possible for when they are used full range, so, no added woofers, just the stock speaker as purchased brand new.

Here is an image posted on page 2 of this thread which used a Dirac EQ, and a comparison to using User EQ. Both are with earlier firmwares. Some sections are blank due to not having done as many combos compared to current tests.
Top row is: FW2.2, Surround, Dirac EQ.
Bottom row is: FW2.0, Surround, Dirac EQ.


FW2.2, Surround, User EQ.
Attachment Deleted

Once I reset my speakers I'll run these again with FW2.0, FW2.1, FW2.2, and FW2.3, all just using Surround and User EQ for simplicity. It'll make for a fun Sunday.

[jbrunwa]

Jul 21, 2021 12:39:24 GMT -5 ttocs said:

I am comparing current measurements to some I did months ago and want to be confident of all the data, so, I will be retweaking my Main speakers back to stock and using a analog interface to verify that the speakers play as flat as possible for when they are used full range, so, no added woofers, just the stock speaker as purchased brand new.

Here is an image posted on page 2 of this thread which used a Dirac EQ, and a comparison to using User EQ. Both are with earlier firmwares. Some sections are blank due to not having done as many combos compared to current tests.
View Attachment
View Attachment

Once I reset my speakers I'll run these again with FW2.0, FW2.1, FW2.2, and FW2.3, all just using Surround and User EQ for simplicity. It'll make for a fun Sunday.

Thanks again for this. It might be a pain and I’m not even sure if it could be done with REW but wouldn’t the best way to measure be directly from the balanced line out to eliminate the speakers and mic.

[KeithL]

Indeed.

Trying to accurately measure signal levels using speakers and a microphone rarely delivers accurate results and is almost never repeatable.

There are simply too many variables - and they tend to vary far too much.

Using a microphone is obviously unavoidable when measuring room and speaker response.

However the proper and accurate way to measure electronic signals is simply to measure them directly.

While some digital multimeters have a reasonably wide and flat frequency response many are only really accurate for measuring AC at frequencies around 60 Hz.

(A meter that can be used for the full range of audio frequencies will have a specified accuracy over a specified frequency range.)

However most digital gear, including even low end sound cards, has a reasonably wide flat frequency response.

(Most programs intended for that sort of measurements also offer some way to loop the output through the input to calibrate for the variations between sound cards anyway.)

There are also many "little black box measurement devices" that are specifically designed to perform audio analyses and display measurement results on your computer.

I would also absolutely recommend REW for this...

And REW can be used to measure and analyze electronic signal inputs directly as well as to take measurements with a microphone.

REW not only supports a wide variety of tests but also includes documentation that explains what they do and how to interpret the results.

Test discs and signal generating software, which provide a digital test signal, at known levels and frequencies, are also usually very accurate.
There are plenty of sets of test tones, and software signal generators, available at very low cost - many of them actually free.

You can also do many tests with standard audio programs...
For example, you can play test tones from a test CD, then record and examine the results with something like Audacity.

You can simply play pink noise, or a sweep tone, from a test CD, then record the results for later analysis, using your computer's sound card...

Or look at the results directly using something like REW, or Audacity, or Adobe Audition.

However you should definitely AVOID so-called test tones and test tone videos on YouTube.

Many of those have been proven NOT to contain what they claim they do... probably due to the way YouTube processes their videos.

(In contrast, a digital audio signal played directly, or direct digital playback of a test CD, will usually be extremely accurate.)

And, of course, if you have the budget, devices like the Audio Precision analyzers have all the standard tests built in and ready to go for you.

Note that, if you use an external signal source, like a test CD...

It becomes trivial to compare the results with and without Dirac enabled...

Or to compare the results with Dirac to those you achieve with manual adjustments...

(Just remember that different types of measurements will often yield different results... so you need to know how to interpret the results of the specific type of test you're using.)

Jul 21, 2021 14:09:33 GMT -5 jbrunwa said:

Jul 21, 2021 12:39:24 GMT -5 ttocs said:

I am comparing current measurements to some I did months ago and want to be confident of all the data, so, I will be retweaking my Main speakers back to stock and using a analog interface to verify that the speakers play as flat as possible for when they are used full range, so, no added woofers, just the stock speaker as purchased brand new.

Here is an image posted on page 2 of this thread which used a Dirac EQ, and a comparison to using User EQ. Both are with earlier firmwares. Some sections are blank due to not having done as many combos compared to current tests.
View Attachment
View Attachment

Once I reset my speakers I'll run these again with FW2.0, FW2.1, FW2.2, and FW2.3, all just using Surround and User EQ for simplicity. It'll make for a fun Sunday.

Thanks again for this. It might be a pain and I’m not even sure if it could be done with REW but wouldn’t the best way to measure be directly from the balanced line out to eliminate the speakers and mic.

[marcl]

Jul 21, 2021 14:49:12 GMT -5 KeithL said:

Indeed.

Trying to accurately measure signal levels using speakers and a microphone rarely delivers accurate results and is almost never repeatable.

There are simply too many variables - and they tend to vary far too much.

Using a microphone is obviously unavoidable when measuring room and speaker response.

However the proper and accurate way to measure electronic signals is simply to measure them directly.

While some digital multimeters have a reasonably wide and flat frequency response many are only really accurate for measuring AC at frequencies around 60 Hz.

(A meter that can be used for the full range of audio frequencies will have a specified accuracy over a specified frequency range.)

However most digital gear, including even low end sound cards, has a reasonably wide flat frequency response.

(Most programs intended for that sort of measurements also offer some way to loop the output through the input to calibrate for the variations between sound cards anyway.)

There are also many "little black box measurement devices" that are specifically designed to perform audio analyses and display measurement results on your computer.

I would also absolutely recommend REW for this...

And REW can be used to measure and analyze electronic signal inputs directly as well as to take measurements with a microphone.

REW not only supports a wide variety of tests but also includes documentation that explains what they do and how to interpret the results.

Test discs and signal generating software, which provide a digital test signal, at known levels and frequencies, are also usually very accurate.
There are plenty of sets of test tones, and software signal generators, available at very low cost - many of them actually free.

You can also do many tests with standard audio programs...
For example, you can play test tones from a test CD, then record and examine the results with something like Audacity.

You can simply play pink noise, or a sweep tone, from a test CD, then record the results for later analysis, using your computer's sound card...

Or look at the results directly using something like REW, or Audacity, or Adobe Audition.

However you should definitely AVOID so-called test tones and test tone videos on YouTube.

Many of those have been proven NOT to contain what they claim they do... probably due to the way YouTube processes their videos.

(In contrast, a digital audio signal played directly, or direct digital playback of a test CD, will usually be extremely accurate.)

And, of course, if you have the budget, devices like the Audio Precision analyzers have all the standard tests built in and ready to go for you.

Note that, if you use an external signal source, like a test CD...

It becomes trivial to compare the results with and without Dirac enabled...

Or to compare the results with Dirac to those you achieve with manual adjustments...

(Just remember that different types of measurements will often yield different results... so you need to know how to interpret the results of the specific type of test you're using.)

Jul 21, 2021 14:09:33 GMT -5 jbrunwa said:

Thanks again for this. It might be a pain and I’m not even sure if it could be done with REW but wouldn’t the best way to measure be directly from the balanced line out to eliminate the speakers and mic.

All this is true, but not really relevant to the issues in this thread, right?

Like, simply put .....  everything ttocs (and I) have observed and measured has been with the exact same systems, microphone, software, etc. that we've used for months ... years, even.   This is a real bug, because we know exactly when it changed.  

[JKCashin]

I have a spectrum analyzer, well, actually a digital oscilloscope with ability to do FFT spectrum calculation. I could test the outputs as long as the XMC-2 uses a common ground referenced to mains input, i.e., as long as the chassis ground is electrically equivalent to output ground pin. Does anyone know if this is the case? If not I will need to get an isolator. Sigh.

Also does anyone know what is the peak output voltage of the XMC-2 on the XLRs WRT ground? I even have the ability to do up to 4 measurements at once!