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

[ttocs]

The Dirac sweeps play in the upper 80dB range, so well above the noise floor, except when the kid across the street starts his car.

[hsamwel]

Feb 7, 2021 17:18:19 GMT -5 ttocs said:

The Dirac sweeps play in the upper 80dB range, so well above the noise floor, except when the kid across the street starts his car.

Yes, but lowering the mic gain makes sure you won’t pick up any unwanted noise from the room
or your house and around.
Dead silent rooms with noise floor lower than 40db of course have no issues.
But those in the high 40’s and above 50 can have issues with ambient noise if main volume is set too low.

This is actually one of the things you should check before doing the calibration. The volume settings
lets you ”see” how you are doing. Then set the mic gain and volume accordingly.

[ttocs]

Feb 7, 2021 17:34:55 GMT -5 hsamwel said:

Feb 7, 2021 17:18:19 GMT -5 ttocs said:

The Dirac sweeps play in the upper 80dB range, so well above the noise floor, except when the kid across the street starts his car.

Yes, but lowering the mic gain makes sure you won’t pick up any unwanted noise from the room
or your house and around.
Dead silent rooms with noise floor lower than 40db of course have no issues.
But those in the high 40’s and above 50 can have issues with ambient noise if main volume is set too low.

This is actually one of the things you should check before doing the calibration. The volume settings
lets you ”see” how you are doing. Then set the mic gain and volume accordingly.

Absolutely agree !!

Back around Dirac 3.0.2 I was vocal about lowering the Mic Gain and raising the Master Volume, but stopped when Dirac started messing with the whole arrangement and things changed, plus some reported speaker damage so I don't want any part in that. Everyone should be aware of how their system operates and adjust accordingly.

I still lower the Mic Gain and raise the Master which is why the sweeps are a tic under 90dB for speakers, I just measured to check this. But when I demonstrate an issue for someone I Always use the Mic Gain at 100% so it's universal and easy for others to see and compare. Everyone's system has a different gain structure, so I no longer recommend what I do, that's for me and my system.

I think 89dB SPL is plenty for the speakers because the subs play a few dB higher, and yes, that's how I want it, and again I don't recommend it for everyone. I started doing that around 3.0.7 or 3.0.11 when they lowered the sweep levels and that didn't suit me and my subs. Right now I'm using 3.0.11 until the crashes are fixed.

An important note about this: It also depends on which mic is used and what its gain level is set to. A UMIK-1 set at 18dB Gain is limited to 96dB SPL. I've done testing to find the limits of my UMIK-1 mics and when both of them are set at 18dB they max out at 94dB before the bass end starts sagging. So there is a limit. I set one of my UMIK-1 mics to a lower Gain so I can use it for higher SPL for some tests of the subwoofers, specifically for LFE. You just unscrew the base end of the mic and pull out the circuit board and change the DIP switches. It's not something one should do all the time, so that's one reason I have two.

Kinda like when Dirty Harry said "A man's got to know his limitations", a Dirac user's got to know the mic's limitations!

[marcl]

I had an exchange with Flavio about this and he basically said as long as Dirac doesn't throw an error the levels are okay. So if you don't get the low level signal to noise error, and it doesn't shut down due to a clipping error, it's happy. So he says.  Nominally, yes, okay, it will sound "good".

But it's so system dependent. And how you have the levels set during calibration DOES affect the levels you end up with in your processor, sometimes requiring you to listen several db higher or lower on the Volume scale. Some of us will say this matters, some will say it doesn't.

In my system, if I set all levels the same in Volume Calibration (within 1db) it measures my sub at a level about 8db quieter than everything else, and my listening level has to be increased several db. If I set the sub level at -20db and all the other speakers at -28db in Volume Calibration, then it measures the sub at the same level as the other speakers, and my listening level does not have to be increased. This is repeatable and predictable for my system. And all of this with mic level left at 100% and my room ambient around 43dbC.

Then you have the option of setting any given speaker group's Target Curve with a + or - bias.  You can shape all the curves, but then say "my sub has a 3db dip at 60Hz and I want to fix that by lowering the sub Target Curve by 3db in Filter Design, and then raise my sub level in the G3P by 3db to compensate" .... (I say these things  ) ... and sure enough it all works out.  Dirac corrects the dip without affecting other speaker levels, the sub comes out a little quiet, and I raise the sub level back up.  You can do the same with any speaker group that has a big dip.



To get it all right and understand in a repeatable way what's happening, you need to at least use a level meter to verify levels after calibration, and preferably check frequency response and distortion with REW.

[kempf]

Feb 8, 2021 6:29:51 GMT -5 marcl said:

View Attachment

To get it all right and understand in a repeatable way what's happening, you need to at least use a level meter to verify levels after calibration.

This ist what I do, using the the testones on the 2016 Dolby Atmos demo disc. I have to incresae all back and height channels about 2.5-3 db. Sounds amazing.

[marcl]

Feb 8, 2021 7:11:45 GMT -5 kempf said:

Feb 8, 2021 6:29:51 GMT -5 marcl said:

View Attachment

To get it all right and understand in a repeatable way what's happening, you need to at least use a level meter to verify levels after calibration.

This ist what I do, using the the testones on the 2016 Dolby Atmos demo disc. I have to incresae all back and height channels about 2.5-3 db. Sounds amazing.

Ahhh you bring up an interesting issue.  If you use the internal Levels on your G3P, or use test tones from REW, or use a demo disc as you describe, or use an ATV4K with Surround Signal Generator from Studio Six Digital .... do you get the  same results?

I don't have the demo disc, but with the others ... results can vary.

[KeithL]

Bear in mind that, when use use a steady tone and a meter, you are taking a long term continuous measurement.

This means that you will get different results than you would with a sweep tone or time gated measurement window.

The reason is that the RT60 at various frequencies will come into play....

Basically, depending on how "live" your room is at various frequencies, some frequencies will build up more than others over a fraction of a second.

(When you use test tones and a meter you are taking a measurement that has settled to steady state over several seconds.)

So, for example, if your room is heavily padded, it will have a lower RT60 at higher frequencies, which will have a stronger effect on meter measurements...

So you should expect a meter measurement to show decreasing levels at high frequencies compared to what you would get with a fast sweep...

Feb 8, 2021 7:18:06 GMT -5 marcl said:

Feb 8, 2021 7:11:45 GMT -5 kempf said:

This ist what I do, using the the testones on the 2016 Dolby Atmos demo disc. I have to incresae all back and height channels about 2.5-3 db. Sounds amazing.

Ahhh you bring up an interesting issue.  If you use the internal Levels on your G3P, or use test tones from REW, or use a demo disc as you describe, or use an ATV4K with Surround Signal Generator from Studio Six Digital .... do you get the  same results?

I don't have the demo disc, but with the others ... results can vary.

[kempf]

Feb 8, 2021 10:07:29 GMT -5 KeithL said:

Bear in mind that, when use use a steady tone and a meter, you are taking a long term continuous measurement.

This means that you will get different results than you would with a sweep tone or time gated measurement window.

The reason is that the RT60 at various frequencies will come into play....

Basically, depending on how "live" your room is at various frequencies, some frequencies will build up more than others over a fraction of a second.

(When you use test tones and a meter you are taking a measurement that has settled to steady state over several seconds.)

So, for example, if your room is heavily padded, it will have a lower RT60 at higher frequencies, which will have a stronger effect on meter measurements...

So you should expect a meter measurement to show decreasing levels at high frequencies compared to what you would get with a fast sweep...

Feb 8, 2021 7:18:06 GMT -5 marcl said:

Ahhh you bring up an interesting issue.  If you use the internal Levels on your G3P, or use test tones from REW, or use a demo disc as you describe, or use an ATV4K with Surround Signal Generator from Studio Six Digital .... do you get the  same results?

I don't have the demo disc, but with the others ... results can vary.

The testones on the Atmos disc are the same as the one's implemented in the RMC 1 I think is called pink noise.


Thanks

[marcl]

Feb 8, 2021 10:07:29 GMT -5 KeithL said:

Bear in mind that, when use use a steady tone and a meter, you are taking a long term continuous measurement.

This means that you will get different results than you would with a sweep tone or time gated measurement window.

The reason is that the RT60 at various frequencies will come into play....

Basically, depending on how "live" your room is at various frequencies, some frequencies will build up more than others over a fraction of a second.

(When you use test tones and a meter you are taking a measurement that has settled to steady state over several seconds.)

So, for example, if your room is heavily padded, it will have a lower RT60 at higher frequencies, which will have a stronger effect on meter measurements...

So you should expect a meter measurement to show decreasing levels at high frequencies compared to what you would get with a fast sweep...

Feb 8, 2021 7:18:06 GMT -5 marcl said:

Ahhh you bring up an interesting issue.  If you use the internal Levels on your G3P, or use test tones from REW, or use a demo disc as you describe, or use an ATV4K with Surround Signal Generator from Studio Six Digital .... do you get the  same results?

I don't have the demo disc, but with the others ... results can vary.

I use band-limited pink noise.  Looking at the RTA in REW it looks like the noise signals in the XMC are very similar to the speaker and sub bandwidth used by REW.  So pretty apples:apples.   In the Studio Six Digital app I use full bandwidth pink noise.  Good thing with the Studio Six Digital is you can measure the Atmos speakers which you can't with REW.  You can also sweep the Atmos speakers and get an idea of response with the REW RTA.

[okjazz]

Feb 8, 2021 10:07:29 GMT -5 KeithL said:

Bear in mind that, when use use a steady tone and a meter, you are taking a long term continuous measurement.

This means that you will get different results than you would with a sweep tone or time gated measurement window.

The reason is that the RT60 at various frequencies will come into play....

Basically, depending on how "live" your room is at various frequencies, some frequencies will build up more than others over a fraction of a second.

(When you use test tones and a meter you are taking a measurement that has settled to steady state over several seconds.)

So, for example, if your room is heavily padded, it will have a lower RT60 at higher frequencies, which will have a stronger effect on meter measurements...

So you should expect a meter measurement to show decreasing levels at high frequencies compared to what you would get with a fast sweep...

Feb 8, 2021 7:18:06 GMT -5 marcl said:

Ahhh you bring up an interesting issue.  If you use the internal Levels on your G3P, or use test tones from REW, or use a demo disc as you describe, or use an ATV4K with Surround Signal Generator from Studio Six Digital .... do you get the  same results?

I don't have the demo disc, but with the others ... results can vary.

Keith, can Emotiva make available an audio driver (e.g.. an Emotiva customized Asio driver type) that can be loaded into REW to allow addressing directly every single speakers offered by an Emotiva processor, including the height speakers, for measurement purposes? The latest general Asio driver do only allow addressing up to 8 channels in a Windows environment.

[foggy1956]

Feb 8, 2021 19:43:05 GMT -5 okjazz said:

Feb 8, 2021 10:07:29 GMT -5 KeithL said:

Bear in mind that, when use use a steady tone and a meter, you are taking a long term continuous measurement.

This means that you will get different results than you would with a sweep tone or time gated measurement window.

The reason is that the RT60 at various frequencies will come into play....

Basically, depending on how "live" your room is at various frequencies, some frequencies will build up more than others over a fraction of a second.

(When you use test tones and a meter you are taking a measurement that has settled to steady state over several seconds.)

So, for example, if your room is heavily padded, it will have a lower RT60 at higher frequencies, which will have a stronger effect on meter measurements...

So you should expect a meter measurement to show decreasing levels at high frequencies compared to what you would get with a fast sweep...

Keith, can Emotiva make available an audio driver (e.g.. an Emotiva customized Asio driver type) that can be loaded into REW to allow addressing directly every single speakers offered by an Emotiva processor, including the height speakers, for measurement purposes? The latest general Asio driver do only allow addressing up to 8 channels in a Windows environment.

Sure, right after they finish up DLBM😂😂😂

[jeffrey40sw]

I solved my Dirac sound quality issues by getting the UMIK-1. In my opinion Emotiva should offer that mic as an upcharge or just add 50-75$ and only send out that mic. Every issue I had is gone after getting an accurate mic. The default Dirac curve even sounds good. I switched back and forth between the Dirac curve and the Harmon +4 curve and there was very little difference. So little in fact I will keep both in the presets depending on what I'm listening too or my mood. The only thing I could tell was the Harmon curve was slightly more vocal forward. looking at the curves side by side and all else being equal that's about what I would expect too. Maybe a tad brighter on the high end with the Harmon but that would depend on the tweeter in your speakers as well. Prior to that the Dirac curve sounded like garbage and looking at the curve there is no reason it should. That's my two cents and 45 years of audio knowledge for what its worth. I would like to think years of building my own speakers, working in a speaker shop, and writing reviews for a well known but unmentionable audio web site should count for something lol. Of couse should Emotiva want to start a new department dedicated to calibrating and testing their mics, they know where to reach me. This PACS stuff I do now would start to seem pretty boring then.

[klinemj]

Feb 9, 2021 15:14:44 GMT -5 jeffrey40sw said:

I solved my Dirac sound quality issues by getting the UMIK-1. In my opinion Emotiva should offer that mic as an upcharge or just add 50-75$ and only send out that mic. Every issue I had is gone after getting an accurate mic. The default Dirac curve even sounds good. I switched back and forth between the Dirac curve and the Harmon +4 curve and there was very little difference. So little in fact I will keep both in the presets depending on what I'm listening too or my mood. The only thing I could tell was the Harmon curve was slightly more vocal forward. looking at the curves side by side and all else being equal that's about what I would expect too. Maybe a tad brighter on the high end with the Harmon but that would depend on the tweeter in your speakers as well. Prior to that the Dirac curve sounded like garbage and looking at the curve there is no reason it should. That's my two cents and 45 years of audio knowledge for what its worth. I would like to think years of building my own speakers, working in a speaker shop, and writing reviews for a well known but unmentionable audio web site should count for something lol. Of couse should Emotiva want to start a new department dedicated to calibrating and testing their mics, they know where to reach me. This PACS stuff I do now would start to seem pretty boring then.

Good to hear your issue is resolved. And, it could be you actually just had a bad mic from Emotiva. Most don't find night and day differences with Dirac as a function of the mic. I previously had the XMC-1 and compared Dirac with the Emotiva Mic to that with a Cross Spectrum UMIK-1. I didn't have any noticable difference other than I just felt more confident in the UMIK-1 because I had an individual calibration file.

Mark

[marcl]

Can someone help me out? I've been looking and looking for the source of the "Harman Curves". I'm looking for a source that originates from Harman, presumably from Olive, Toole or another of their colleagues. I'm looking for a reference to where the curves came from, and how the folks at Harman believe the curves apply. Specifically how Harman recommends they be applied to equalization of speakers in listening rooms.

[ttocs]

Feb 9, 2021 16:34:17 GMT -5 marcl said:

Can someone help me out? I've been looking and looking for the source of the "Harman Curves". I'm looking for a source that originates from Harman, presumably from Olive, Toole or another of their colleagues. I'm looking for a reference to where the curves came from, and how the folks at Harman believe the curves apply. Specifically how Harman recommends they be applied to equalization of speakers in listening rooms.

I find a lot of info for headphones.
DEVELOPMENT OF HARMAN HEADPHONE TARGET CURVE

A link from Dirac Live 2 page
Target Curves for Dirac Live 2/3

And a link from that linked page above used as a reference
AES.ORG

edit: Here's an important note about using the target curves for all speakers in Dirac:
"Note: Use the very same target curve for all your speakers including the subwoofer. Do NOT use different target curves amongst speakers in the crossover region to the subwoofer(s) or you'll end up with a combined response that no longer follows the intended curve. Adjust high frequency fall-off according to your speaker's in-room response."

edit2: marcl this might be it.
A Study of Listener Bass and Loudness Preferences over Loudspeakers and Headphones

[marcl]

Feb 9, 2021 17:07:38 GMT -5 ttocs said:

Feb 9, 2021 16:34:17 GMT -5 marcl said:

Can someone help me out? I've been looking and looking for the source of the "Harman Curves". I'm looking for a source that originates from Harman, presumably from Olive, Toole or another of their colleagues. I'm looking for a reference to where the curves came from, and how the folks at Harman believe the curves apply. Specifically how Harman recommends they be applied to equalization of speakers in listening rooms.

I find a lot of info for headphones.
DEVELOPMENT OF HARMAN HEADPHONE TARGET CURVE

A link from Dirac Live 2 page
Target Curves for Dirac Live 2/3

And a link from that linked page above used as a reference
AES.ORG

edit: Here's an important note about using the target curves for all speakers in Dirac:
"Note: Use the very same target curve for all your speakers including the subwoofer. Do NOT use different target curves amongst speakers in the crossover region to the subwoofer(s) or you'll end up with a combined response that no longer follows the intended curve. Adjust high frequency fall-off according to your speaker's in-room response."

edit2: marcl this might be it.
A Study of Listener Bass and Loudness Preferences over Loudspeakers and Headphones

And ....


www.audiosciencereview.com/forum/index.php?threads/some-comments-from-floyd-toole-about-room-curve-targets-room-eq-and-more.10950/

[ttocs]

. . . . and here's a page of links to papers by the cast of characters involved, Toole, Olive, McMullen.
Link to many papers published on AES

[ttocs]

Feb 9, 2021 17:21:21 GMT -5 marcl said:

And ....

www.audiosciencereview.com/forum/index.php?threads/some-comments-from-floyd-toole-about-room-curve-targets-room-eq-and-more.10950/

The second post on that page says a lot. I've read these posts before when searching for Toole related info. Thanks.

[leonski]

I see in the FAQs that Spectrum has their equipment cal'd by a 3rd party who is 'Traceable' and 'NIST' listed. Good.

[klinemj]

Feb 9, 2021 19:45:01 GMT -5 leonski said:

I see in the FAQs that Spectrum has their equipment cal'd by a 3rd party who is 'Traceable' and 'NIST' listed. Good.

I like that also...I'm an engineer and did work on quality systems and improving methods to ensure they met our needs (way back in the late 80's/early 90's). For any method, starting with solid, calibrated equipment is key.

Mark