The Official Dirac Live Thread: Start w/1,2,3 🙋‍♂️

[AudioHTIT]

I’d like to point out that this is the Dirac thread. It’s purpose is to help G3P owners (or prospective ones) get better performance out of their processor by using Dirac. There has been a wide latitude of discussion here and I’ve been lax in trying to keep it on track, but when almost every post is off topic or to the detriment of the OP (please read it if you haven’t), I must ask you to stop. If you want to discuss why you don’t like Dirac, what other measurement techniques are available, why stereo is better, or any myriad of non-Dirac subjects, you’re welcome to start you’re own thread for just that purpose. The length of this thread is partly due to OT discussion and it’s not fair to those who come here to ask questions or learn more about Dirac. Thanks to everyone who has contributed to our understanding of Dirac and this complex subject.

[hsamwel]

Aug 1, 2021 20:00:57 GMT -5 leonski said:

hsamwel,
Bass is THE worset thing to measure to get right. I hope you agree. I have no idea about the implimentation of it in DIRAC especially with all this talk of Effects V Music. I personally don't get it.

Either it is right or it is not. With my relatively SIMPLE system, stereo 2-way speakers AND a single sub (located correctly) I get fast, musical bass AND the dinosaurs walking in Jurassic park will

rattle the dishes hung on the wall. My very odd shaped room works in my favor. 8 asymmetrical sides and an off-center peaked ceiling running th 'long' way.

My main objection is that people do NOT generally apply the scientific method TO DIRAC. Few will read and become knowledgable in the various isssues, expecting to set up the mic and take some

readings and have a magically 'corrected' system. I've suggested keeping good records and Measurements now for quite a while.

The guys that get it......spend I don't know how much time with details. And I suspect not a little effort doing it the SAME each time....at least as far as mic locations go, regardelss of the SW settings.


And Yes, I'll agree 100% that the BASS must be right. And I hope you agree that is the toughest TO get right. those huge wavelthghts can be 2x or even 3x the longest dimension of the room.....Gives

rise to standing waves with peaks / valleys and some unpredictable or even un MODELABLE effects.

I've NEVER said DIRAC is Crap. Please show me in a post. What I have said is that it is not like starting a car or mixing some cake batter.

I measured stuff for a good part of my professional career. I even compared peoples measurements of the SAME set of parts. With the SAME tools. And got statistically different results. I'd expect NO LESS

with DIRAC or perhaps even REW.

The people that do BEST are good to share. And their knowledge base should become the 'standard'....

Yes, bass is the absolutely most important thing to get right.
Also the hardest as you say.. Especially with room treatment and manual setup.

However, Dirac makes this easy.
Whatever you think of Dirac and other room calibration programs, Dirac atleast do make a
hugh improvement to the low end. Dirac lets you use curtains to set which frequencies you
like to calibrate. If <500hz is your thing, then easy to setup in Dirac.
Mostly by aligning the phase, impulse response and amplitude. This makes the subs play
alot cleaner and almost feel like integrated large mains bass.
Also almost no rattling in the room.. Even though they feel harder and deeper than ever..
Dirac removes almost all that is bad with subwoofers when setup without room treatment.

Before I got Dirac I measured my subs, used PEQ, REW and my ears to calibrate them.
Took several hours and days.. I was really pleased and my friends where very impressed
with the result. Dirac bested my setup in an hour first try in my room.

But with enough knowledge and time you can get just about the same result with REW.
Room treatment is hard, because you don’t touch the phase and impulse response at all.
Also you need a computer and knowledge where to set those bass traps. Not counting in
the placement of the subs which is, for the most part, the most important thing.

Regarding mic positions.. I have always been meticulous with the MLP and not so with
the other measure points. Why? Because the others are just used as control/info points
for Dirac to the MLP point.
You simply measure a little cube around your head so Dirac can have some more information
on what happens around the MLP. This way it can make more accurate calculations.
The main thing is free line of sight for the extra measuring points.
I used about the same measure cube size though and always have the mic pointing straight up.

One thing I like to mention.. I have my setup in the livingroom which the whole family
uses. Because of this I have really limited places I can put subs and also their sizes.
I simply use two 12” placed inbetween my two front speakers.
This is actually good for music.. But I have some hugh bumps in 25hz, 50hz and about
75hz.. Also a couple of big dips around 30-35hz and 60-65hz..
If I had four subs and a dedicated room I maybe could have placed them better for a
smoother response without Dirac.
Dirac fixes almost everything for me without the effort. That’s why I think it’s a great
tool for me.

What Dirac does with the measured data may not be the same as what you used to
measure. I would think their software can handle a bit of different results when making
a filter. The major thing is impulse respone though..   Making each speaker start play at
exactly the same time. This is probably not as hard to measure and probably makes
less difference where people place their mics. However the MLP point is very important.

[hsamwel]

Dirac however redid the whole software suit.. This has had some bugs in some versions.
The good is as long as Dirac is implemented in a processor you can use the same PC/Mac
software for any processor.
They have also worked on some improvements during the last years which has probably
made measurements more sensitive. Can be why some has trouble in their rooms or with
the speakers.

[leonski]

Aug 2, 2021 13:08:18 GMT -5 hsamwel said:

Aug 1, 2021 20:00:57 GMT -5 leonski said:

hsamwel,
Bass is THE worset thing to measure to get right. I hope you agree. I have no idea about the implimentation of it in DIRAC especially with all this talk of Effects V Music. I personally don't get it.

Either it is right or it is not. With my relatively SIMPLE system, stereo 2-way speakers AND a single sub (located correctly) I get fast, musical bass AND the dinosaurs walking in Jurassic park will

rattle the dishes hung on the wall. My very odd shaped room works in my favor. 8 asymmetrical sides and an off-center peaked ceiling running th 'long' way.

My main objection is that people do NOT generally apply the scientific method TO DIRAC. Few will read and become knowledgable in the various isssues, expecting to set up the mic and take some

readings and have a magically 'corrected' system. I've suggested keeping good records and Measurements now for quite a while.

The guys that get it......spend I don't know how much time with details. And I suspect not a little effort doing it the SAME each time....at least as far as mic locations go, regardelss of the SW settings.


And Yes, I'll agree 100% that the BASS must be right. And I hope you agree that is the toughest TO get right. those huge wavelthghts can be 2x or even 3x the longest dimension of the room.....Gives

rise to standing waves with peaks / valleys and some unpredictable or even un MODELABLE effects.

I've NEVER said DIRAC is Crap. Please show me in a post. What I have said is that it is not like starting a car or mixing some cake batter.

I measured stuff for a good part of my professional career. I even compared peoples measurements of the SAME set of parts. With the SAME tools. And got statistically different results. I'd expect NO LESS

with DIRAC or perhaps even REW.

The people that do BEST are good to share. And their knowledge base should become the 'standard'....

Yes, bass is the absolutely most important thing to get right.
Also the hardest as you say.. Especially with room treatment and manual setup.

However, Dirac makes this easy.
Whatever you think of Dirac and other room calibration programs, Dirac atleast do make a
hugh improvement to the low end. Dirac lets you use curtains to set which frequencies you
like to calibrate. If <500hz is your thing, then easy to setup in Dirac.
Mostly by aligning the phase, impulse response and amplitude. This makes the subs play
alot cleaner and almost feel like integrated large mains bass.
Also almost no rattling in the room.. Even though they feel harder and deeper than ever..
Dirac removes almost all that is bad with subwoofers when setup without room treatment.

Before I got Dirac I measured my subs, used PEQ, REW and my ears to calibrate them.
Took several hours and days.. I was really pleased and my friends where very impressed
with the result. Dirac bested my setup in an hour first try in my room.

But with enough knowledge and time you can get just about the same result with REW.
Room treatment is hard, because you don’t touch the phase and impulse response at all.
Also you need a computer and knowledge where to set those bass traps. Not counting in
the placement of the subs which is, for the most part, the most important thing.

Regarding mic positions.. I have always been meticulous with the MLP and not so with
the other measure points. Why? Because the others are just used as control/info points
for Dirac to the MLP point.
You simply measure a little cube around your head so Dirac can have some more information
on what happens around the MLP. This way it can make more accurate calculations.
The main thing is free line of sight for the extra measuring points.
I used about the same measure cube size though and always have the mic pointing straight up.

One thing I like to mention.. I have my setup in the livingroom which the whole family
uses. Because of this I have really limited places I can put subs and also their sizes.
I simply use two 12” placed inbetween my two front speakers.
This is actually good for music.. But I have some hugh bumps in 25hz, 50hz and about
75hz.. Also a couple of big dips around 30-35hz and 60-65hz..
If I had four subs and a dedicated room I maybe could have placed them better for a
smoother response without Dirac.
Dirac fixes almost everything for me without the effort. That’s why I think it’s a great
tool for me.

What Dirac does with the measured data may not be the same as what you used to
measure. I would think their software can handle a bit of different results when making
a filter. The major thing is impulse respone though..   Making each speaker start play at
exactly the same time. This is probably not as hard to measure and probably makes
less difference where people place their mics. However the MLP point is very important.

So I don't get in 'further trouble', I'll just say I can agree with some of what you say......Disagree with other parts and REALLY disagree with some.

Just to touch one point?    Measurment locations are critical.   Getting them the same for repeat measuremnts more so.  

I would like to propose a TEST?   Let someone else measure your room.   At least do the setup part while you run the software part.  I'd bet that 

a total of 4 'measurers' (you plus 3 volunteers) would get different data.    Some would definately AGREE while others would not.     MIght make a difference

but might not, as well.   I've done MONITOR (visual) calibration to a standard.    Checked my results with an ISF guy.    And you know?   As it turns out, some like

the ISF result and others do NOT.   Even though the ISF guy brings standards and such and can access parts of the software in the display you'd 100% want to leave alone.

I've LOOKED at some of the advanced screens in my OLED set and will NOT TOUCH THEM for love 'nor money.....Now while audio (does it?) doesn't have such a deep menu

sustem, I can see someone preferring one setup over the calibrated setup.   

Problem with my idea?    You've got a LIFE and depending on the logistics (find volunteers.....when and amount of time?) this may not be possible.     But when I was measuring stuff

for a living, I'd get variable (statistically valid) results when having 3 people measure 10 samples EACH, than repeat 2x more for 3 measures each of 10 samples.     ONE person would always

be 'off' relative to the other pair.    And this was a very simple test.   all you had to do was focus on the sample and press 'measure'....then record the readings...

[hsamwel]

Aug 2, 2021 15:22:48 GMT -5 leonski said:

Aug 2, 2021 13:08:18 GMT -5 hsamwel said:

Yes, bass is the absolutely most important thing to get right.
Also the hardest as you say.. Especially with room treatment and manual setup.

However, Dirac makes this easy.
Whatever you think of Dirac and other room calibration programs, Dirac atleast do make a
hugh improvement to the low end. Dirac lets you use curtains to set which frequencies you
like to calibrate. If <500hz is your thing, then easy to setup in Dirac.
Mostly by aligning the phase, impulse response and amplitude. This makes the subs play
alot cleaner and almost feel like integrated large mains bass.
Also almost no rattling in the room.. Even though they feel harder and deeper than ever..
Dirac removes almost all that is bad with subwoofers when setup without room treatment.

Before I got Dirac I measured my subs, used PEQ, REW and my ears to calibrate them.
Took several hours and days.. I was really pleased and my friends where very impressed
with the result. Dirac bested my setup in an hour first try in my room.

But with enough knowledge and time you can get just about the same result with REW.
Room treatment is hard, because you don’t touch the phase and impulse response at all.
Also you need a computer and knowledge where to set those bass traps. Not counting in
the placement of the subs which is, for the most part, the most important thing.

Regarding mic positions.. I have always been meticulous with the MLP and not so with
the other measure points. Why? Because the others are just used as control/info points
for Dirac to the MLP point.
You simply measure a little cube around your head so Dirac can have some more information
on what happens around the MLP. This way it can make more accurate calculations.
The main thing is free line of sight for the extra measuring points.
I used about the same measure cube size though and always have the mic pointing straight up.

One thing I like to mention.. I have my setup in the livingroom which the whole family
uses. Because of this I have really limited places I can put subs and also their sizes.
I simply use two 12” placed inbetween my two front speakers.
This is actually good for music.. But I have some hugh bumps in 25hz, 50hz and about
75hz.. Also a couple of big dips around 30-35hz and 60-65hz..
If I had four subs and a dedicated room I maybe could have placed them better for a
smoother response without Dirac.
Dirac fixes almost everything for me without the effort. That’s why I think it’s a great
tool for me.

What Dirac does with the measured data may not be the same as what you used to
measure. I would think their software can handle a bit of different results when making
a filter. The major thing is impulse respone though..   Making each speaker start play at
exactly the same time. This is probably not as hard to measure and probably makes
less difference where people place their mics. However the MLP point is very important.

So I don't get in 'further trouble', I'll just say I can agree with some of what you say......Disagree with other parts and REALLY disagree with some.

Just to touch one point?    Measurment locations are critical.   Getting them the same for repeat measuremnts more so.  

I would like to propose a TEST?   Let someone else measure your room.   At least do the setup part while you run the software part.  I'd bet that 

a total of 4 'measurers' (you plus 3 volunteers) would get different data.    Some would definately AGREE while others would not.     MIght make a difference

but might not, as well.   I've done MONITOR (visual) calibration to a standard.    Checked my results with an ISF guy.    And you know?   As it turns out, some like

the ISF result and others do NOT.   Even though the ISF guy brings standards and such and can access parts of the software in the display you'd 100% want to leave alone.

I've LOOKED at some of the advanced screens in my OLED set and will NOT TOUCH THEM for love 'nor money.....Now while audio (does it?) doesn't have such a deep menu

sustem, I can see someone preferring one setup over the calibrated setup.   

Problem with my idea?    You've got a LIFE and depending on the logistics (find volunteers.....when and amount of time?) this may not be possible.     But when I was measuring stuff

for a living, I'd get variable (statistically valid) results when having 3 people measure 10 samples EACH, than repeat 2x more for 3 measures each of 10 samples.     ONE person would always

be 'off' relative to the other pair.    And this was a very simple test.   all you had to do was focus on the sample and press 'measure'....then record the readings...

Well, I do agree that the data would differ with different locations for the mic.
How much difference this actually does is another topic.

About 80-90% of the important data is from the MLP.. So as long as everybody places
it the exact same place then each complete measurement would sound basically the
same in my experience. Although not exactly the same if we would go and look at the
numbers in the created Dirac filter..
Also Dirac only defines the MLP with an exact mic location. The other points are up for
interpretation actually.. Maybe they should define their place more detailed?!

But it can be that I have a nice room for measuring and speakers with dispersion that
works great with Dirac and how it measures. So in another setup it might not work out
as it does for me?!

Although I agree for repeated measurements, with comparison in mind, you should
try to have each measuring point as exact as possible. Although without taking notes
and placing the mic with laser I don’t think it’s possible.

I would not like to do a test though.. Doing these measurements is not something
I enjoy doing over and over. It’s actually some months back since I did my last
measure now.. Doing 3-4 in a row.. Well, no! I’ll rather enjoy good music or a movie.

Using Dirac is a free choice. If it doesn’t fit you then don’t use it, simple as that.
Same goes if you don’t like what Dirac does with the sound. Doesn’t mean others
will though.. That’s why there are so many different things in this hobby.

Yeah, I know some people also that prefered some crazy setup rather than the
calibrated one.. Or simply don’t care.
IMO the calibrated is so much better with more natural colors and incredible depth
in the picture.
But generally picture and sound is not comparable this way.
Most people prefer a calibrated picture if they can see the before and after compared.
Also there is a standard setup and something which is correct.
Audio is more emotions and unique to each individual. Yes you can calibrate to a flat
response, But will that be correct in your room? What is correct?
I would guess more than half would not like a ”calibrated” sound if that would even be
possible without extensive changes in each room. Also each speaker brand has its own
signature sound.
Probably why there is no ”isf” standard for audio. Dirac tries though.. But still it is no
standard. Only one way change or improve your sound, depending how you see it.

[leonski]

I hear and understand about 'defined mic positions'.....other than the MLP.

However, this should NOT prevent the individual user from defining his OWN positions, per DIRAC instructions and guidance.     

This would ensure consistency during REPEAT measures, where you may have changed some settings.....

To this end?   I URGE careful record keeping and use of a tape measure and even a 'map' of the room......

IF I were very familiar with DIRAC, usage and running it....I should be able to come to YOUR house, and useing YOUR directions / setup actually be

able to DUPLICATE your results.....within a narrow margin of error.   

You could actually determine margin of error IN YOUR SYSTEM....by running DIRAC 1x Daily for a week, using the EXACT setup parameters each time...

You should, by such means, be able to determine SYSTEM ERROR as opposted to OPERATOR ERROR......Both of which play into the final results.

I do NOT know....Does the microphone come with a 'generic' correction file OR is does each mic come with its OWN dedicated correction file generated 

thru measureing the output OF THAT MIC?     I suspect a fully NIST traceable and calibrated mic would be a LOT OF $$$ and should be recalibrated annually in regular use.

[marcl]

Aug 3, 2021 14:54:50 GMT -5 leonski said:

I hear and understand about 'defined mic positions'.....other than the MLP.

However, this should NOT prevent the individual user from defining his OWN positions, per DIRAC instructions and guidance.     

This would ensure consistency during REPEAT measures, where you may have changed some settings.....

To this end?   I URGE careful record keeping and use of a tape measure and even a 'map' of the room......

IF I were very familiar with DIRAC, usage and running it....I should be able to come to YOUR house, and useing YOUR directions / setup actually be

able to DUPLICATE your results.....within a narrow margin of error.   

You could actually determine margin of error IN YOUR SYSTEM....by running DIRAC 1x Daily for a week, using the EXACT setup parameters each time...

You should, by such means, be able to determine SYSTEM ERROR as opposted to OPERATOR ERROR......Both of which play into the final results.

I do NOT know....Does the microphone come with a 'generic' correction file OR is does each mic come with its OWN dedicated correction file generated 

thru measureing the output OF THAT MIC?     I suspect a fully NIST traceable and calibrated mic would be a LOT OF $$$ and should be recalibrated annually in regular use.

Excuse me leonski, but after these months of lively discussion I've just lost track .... have you used Dirac, and do you use it now?  Just trying to refresh my sense of context as I read your posts.  Thanks!

[marcl]

I'd like to comment further on the post I did a week or so ago with measurements taken with Dirac and REW with no filter correction.  

To refresh your memory:




My point was simple:   Given that the measurements taken with Dirac and REW with no correction applied are essentially indistinguishable from each other despite the much-touted (but never quantified) differences that are presumed to exist between how Dirac measures and how REW measures .... I postulate that using REW to do a post-calibration measurement will also produce a result that would be indistinguishable from a post-calibration Dirac measurement ... IF Dirac were to make such a measurement ... which they choose NOT do ever do.  

Now, that is for frequency/amplitude response and impulse response which are the only measurements that Dirac shows us.  But since both applications do sweep measurements which result in phase and amplitude (vs frequency) data and derive the impulse response using an inverse Fourier Transform (i.e. neither slams a true impulse through our speakers!), then we can conclude that any other properties in the time/amplitude/frequency domains would be similarly indistinguishable from each other using the two methods.

[hsamwel]

Aug 3, 2021 16:52:11 GMT -5 marcl said:

I'd like to comment further on the post I did a week or so ago with measurements taken with Dirac and REW with no filter correction.  

To refresh your memory:

View AttachmentView Attachment


My point was simple:   Given that the measurements taken with Dirac and REW with no correction applied are essentially indistinguishable from each other despite the much-touted (but never quantified) differences that are presumed to exist between how Dirac measures and how REW measures .... I postulate that using REW to do a post-calibration measurement will also produce a result that would be indistinguishable from a post-calibration Dirac measurement ... IF Dirac were to make such a measurement ... which they choose NOT do ever do.  

Now, that is for frequency/amplitude response and impulse response which are the only measurements that Dirac shows us.  But since both applications do sweep measurements which result in phase and amplitude (vs frequency) data and derive the impulse response using an inverse Fourier Transform (i.e. neither slams a true impulse through our speakers!), then we can conclude that any other properties in the time/amplitude/frequency domains would be similarly indistinguishable from each other using the two methods.

Btw Marcl you have quite high amount of early reflections/reverberation in your impulse measurements..
Is this because of the room or your speakers? Or both?

[marcl]

Aug 3, 2021 18:02:54 GMT -5 hsamwel said:

Aug 3, 2021 16:52:11 GMT -5 marcl said:

I'd like to comment further on the post I did a week or so ago with measurements taken with Dirac and REW with no filter correction.  

To refresh your memory:

View AttachmentView Attachment


My point was simple:   Given that the measurements taken with Dirac and REW with no correction applied are essentially indistinguishable from each other despite the much-touted (but never quantified) differences that are presumed to exist between how Dirac measures and how REW measures .... I postulate that using REW to do a post-calibration measurement will also produce a result that would be indistinguishable from a post-calibration Dirac measurement ... IF Dirac were to make such a measurement ... which they choose NOT do ever do.  

Now, that is for frequency/amplitude response and impulse response which are the only measurements that Dirac shows us.  But since both applications do sweep measurements which result in phase and amplitude (vs frequency) data and derive the impulse response using an inverse Fourier Transform (i.e. neither slams a true impulse through our speakers!), then we can conclude that any other properties in the time/amplitude/frequency domains would be similarly indistinguishable from each other using the two methods.

Btw Marcl you have quite high amount of early reflections/reverberation in your impulse measurements..
Is this because of the room or your speakers? Or both?

I don't think you can really evaluate early reflections directly from the impulse that I posted, but here's the Energy Time Curve (ETC) for this, the left front, as measured by REW.



Rule of thumb is that you'd like reflections to very quickly drop below -20db, which they do.  I have had some issues in the past with reflections that showed up strongly in the ETC at 8ms, 18ms and even further out.  Using the ETC I was able to place absorption to intercept these reflections.  In one notable case (the 18ms) it was causing an imaging issue which went away when I placed an absorber on the wall to the right of MLP.  The ETC is quite useful for tracking down reflections using ~1ms=1ft.

Here's the Spectrogram for the left channel showing the relatively fast decay, except in the region of the 40Hz resonance due to the length mode.

[hsamwel]

Aug 3, 2021 14:54:50 GMT -5 leonski said:

I hear and understand about 'defined mic positions'.....other than the MLP.

However, this should NOT prevent the individual user from defining his OWN positions, per DIRAC instructions and guidance.     

This would ensure consistency during REPEAT measures, where you may have changed some settings.....

To this end?   I URGE careful record keeping and use of a tape measure and even a 'map' of the room......

IF I were very familiar with DIRAC, usage and running it....I should be able to come to YOUR house, and useing YOUR directions / setup actually be

able to DUPLICATE your results.....within a narrow margin of error.   

You could actually determine margin of error IN YOUR SYSTEM....by running DIRAC 1x Daily for a week, using the EXACT setup parameters each time...

You should, by such means, be able to determine SYSTEM ERROR as opposted to OPERATOR ERROR......Both of which play into the final results.

I do NOT know....Does the microphone come with a 'generic' correction file OR is does each mic come with its OWN dedicated correction file generated 

thru measureing the output OF THAT MIC?     I suspect a fully NIST traceable and calibrated mic would be a LOT OF $$$ and should be recalibrated annually in regular use.

Well, most people have defined their own cube size for the mic placements.
As recommended by Dirac.
Myself I use of about 60x60cm (2x2ft) with the MLP in the absolute middle.
About 1ft high, but the lower measurements I can’t measure as low as I want
because of my sofa.
Again, Dirac makes use of these extra points but does not change the overall
sound if I change positions an inch or two either way. Most of the changes come
from the MLP.

Unless there is something horribly wrong with a room, yes I can take my laptop
and mic to another room and processor and pretty much duplicate what I got
in my room. With limitation however. Sound does NOT behave the same in every
room. But the same size room with the same layout and same speakers placed
in the same positions, yes. There are many variables, this is why Dirac gets it wrong
sometimes. It’s after all an automated process. Good but not perfect.

Emotiva mic has one calibration file for all the mics.
UMIK-1 has as well, I think it’s a batch calibration file.
UMIK-1 can be had with a unique per mic calibration done by Cross Spectrum.

[leonski]

Aug 3, 2021 16:20:07 GMT -5 marcl said:

Aug 3, 2021 14:54:50 GMT -5 leonski said:

I hear and understand about 'defined mic positions'.....other than the MLP.

However, this should NOT prevent the individual user from defining his OWN positions, per DIRAC instructions and guidance.     

This would ensure consistency during REPEAT measures, where you may have changed some settings.....

To this end?   I URGE careful record keeping and use of a tape measure and even a 'map' of the room......

IF I were very familiar with DIRAC, usage and running it....I should be able to come to YOUR house, and useing YOUR directions / setup actually be

able to DUPLICATE your results.....within a narrow margin of error.   

You could actually determine margin of error IN YOUR SYSTEM....by running DIRAC 1x Daily for a week, using the EXACT setup parameters each time...

You should, by such means, be able to determine SYSTEM ERROR as opposted to OPERATOR ERROR......Both of which play into the final results.

I do NOT know....Does the microphone come with a 'generic' correction file OR is does each mic come with its OWN dedicated correction file generated 

thru measureing the output OF THAT MIC?     I suspect a fully NIST traceable and calibrated mic would be a LOT OF $$$ and should be recalibrated annually in regular use.

Excuse me leonski, but after these months of lively discussion I've just lost track .... have you used Dirac, and do you use it now?  Just trying to refresh my sense of context as I read your posts.  Thanks!

I did metrology for a living.   That is what qualifies me to comment.

However?   I have offered to travel to someplace where someone is using this system so I could see it in action.     I'd love to see what 'common practice' is and the results....

[ttocs]

Aug 3, 2021 18:42:08 GMT -5 hsamwel said:

UMIK-1 has as well, I think it’s a batch calibration file.
UMIK-1 can be had with a unique per mic calibration done by Cross Spectrum.

All UMIK-1 mics, whether purchased direct from miniDSP or Cross Spectrum Labs come with its own "Unique Calibration File". I own both versions, miniDSP and CSL, and both are very similar in results. I don't know when miniDSP began doing individual calibrations vs the batch cals, but I bought mine 2nd quarter 2020 and it was advertised as a unique cal at that time.

There's a $19 price difference once shipping is considered, so it's not enough to stop one from buying the "more expensive" version since there's an extra bit of quality assurance you might say.

They're both great.

[marcl]

Aug 3, 2021 18:53:46 GMT -5 leonski said:

Aug 3, 2021 16:20:07 GMT -5 marcl said:

Excuse me leonski, but after these months of lively discussion I've just lost track .... have you used Dirac, and do you use it now?  Just trying to refresh my sense of context as I read your posts.  Thanks!

I did metrology for a living.   That is what qualifies me to comment.

However?   I have offered to travel to someplace where someone is using this system so I could see it in action.     I'd love to see what 'common practice' is and the results....

Sorry I was not disputing your right to comment.  I literally could not remember your Dirac experience.

[Lsc]

Quick question. Do you guys equalize the volume for all speakers prior to running the measurements?

Also, if you are not running Dirac and don’t plan to…thank you for that information.

[marcl]

Aug 4, 2021 10:16:09 GMT -5 Lsc said:

Quick question. Do you guys equalize the volume for all speakers prior to running the measurements?

Also, if you are not running Dirac and don’t plan to…thank you for that information.

If you have level adjustments made in a Preset prior to doing Dirac calibration, they won't affect the calibration itself but they will be applied on top of Dirac level adjustments after calibration.

So, if you don't intend to measure and verify levels with a meter after doing a Dirac calibration, then definitely zero all the levels and rely on Dirac's level adjustment.

If you're not using Dirac and are going to do measurements and apply PEQ (User) filters, then yes, you should zero the levels first ... and then verify and adjust after.

[okjazz]

Aug 4, 2021 10:34:48 GMT -5 marcl said:

Aug 4, 2021 10:16:09 GMT -5 Lsc said:

Quick question. Do you guys equalize the volume for all speakers prior to running the measurements?

Also, if you are not running Dirac and don’t plan to…thank you for that information.

If you have level adjustments made in a Preset prior to doing Dirac calibration, they won't affect the calibration itself but they will be applied on top of Dirac level adjustments after calibration.

So, if you don't intend to measure and verify levels with a meter after doing a Dirac calibration, then definitely zero all the levels and rely on Dirac's level adjustment.

If you're not using Dirac and are going to do measurements and apply PEQ (User) filters, then yes, you should zero the levels first ... and then verify and adjust after.

Marcl, in your last sentence, you advise "you should zero the levels first ... ". Did you intend to say "you should equalize the levels first ... "? The two bold words don't have the same meaning.

[ttocs]

marcl when you were saying that "something" was different with Dirac, I think it was last week or just before, were you using Saved and Restored settings? I posted in the BMbug thread about a weird corruption that occurred only when using Dirac, but I can only come up with that it is related to Restored settings. I did a Factory Reset, manually entered all settings, then ran Dirac and it's back to normal, well, except for the BMbug. But at least it seems that Levels are back to being more useful.

One thing I hadn't considered until writing this is that I've only run Dirac once since the Factory Reset, and everything is still fresh. Might change after some reboots and more Dirac stuff, who knows?

[marcl]

Aug 4, 2021 12:49:33 GMT -5 okjazz said:

Aug 4, 2021 10:34:48 GMT -5 marcl said:

If you have level adjustments made in a Preset prior to doing Dirac calibration, they won't affect the calibration itself but they will be applied on top of Dirac level adjustments after calibration.

So, if you don't intend to measure and verify levels with a meter after doing a Dirac calibration, then definitely zero all the levels and rely on Dirac's level adjustment.

If you're not using Dirac and are going to do measurements and apply PEQ (User) filters, then yes, you should zero the levels first ... and then verify and adjust after.

Marcl, in your last sentence, you advise "you should zero the levels first ... ". Did you intend to say "you should equalize the levels first ... "? The two bold words don't have the same meaning.

I think you should set all the levels to 0.0db, measure with REW, create filters, then with the filters enabled adjust the levels so all channels are equal.   If you try to set the channels to equal levels before doing the EQ the levels will be affected by peaks that you will be fixing with the filters.  

[marcl]

Aug 4, 2021 14:17:38 GMT -5 ttocs said:

marcl when you were saying that "something" was different with Dirac, I think it was last week or just before, were you using Saved and Restored settings? I posted in the BMbug thread about a weird corruption that occurred only when using Dirac, but I can only come up with that it is related to Restored settings. I did a Factory Reset, manually entered all settings, then ran Dirac and it's back to normal, well, except for the BMbug. But at least it seems that Levels are back to being more useful.

One thing I hadn't considered until writing this is that I've only run Dirac once since the Factory Reset, and everything is still fresh. Might change after some reboots and more Dirac stuff, who knows?

Interesting.   Since getting the XMC-2 a year and a half ago I have always restored settings after doing a factory reset.

What I noticed with Dirac is specific to setting levels in Dirac and then the actual measurement levels of each channel during calibration.  I assumed it had to do with how loud Dirac was telling the processor to play the calibration sweeps.  I know that I had done a Dirac calibration after installing 2.3 and it behaved as it had before.  Then a week or so later the behavior changed and I had not changed anything ... same Dirac 3.0.14 and FW2.3, and no factory reset or restore.