Alternative Dirac Technique

[ansat]

As most of you know, I have been running tons of measurements of my room and have re-run Dirac about 40 times now. After some back and forth with a few users and with some help from flak, I have some guidelines that differ from the what exists already. These have yielded the most repeatability successful results and is now my go to measurement technique.

So here are some guidelines that need to be followed:


Setup and First Measurement

---

• AGC Must be turned OFF in the microphone properties and the gain should be set to 100%.
• Follow Emotiva's instructions for setting the gain and levels in the Dirac application
• Make sure that you hit play on all powered subs to ensure that they are awake. (this includes speakers with built in powered subs)
• Take the First measurement at your main listening position with the microphone tip pointing up and exactly at ear level. Note the height and location as these will become your reference points for all other measurements.

Measurements 2 - 9

---

• All 8 other measurements should be 8-10 in above and below the height of the first measurement. Stay within this 2" range. For those who have microphone stands. I recommend setting the stand up so that the extension all the way down is 9" below the first measurement and then raise the extension 18" and marked the extension pole. Or if your OCD like I am, break out your laser level.
• Take your measurements in "roughly" Dirac's recommended positions. But use the high/low pattern that is suggested.
• Each measurement needs to be at a minimum of 18" away from all previous measurements
• Each measurement needs to be Line of sight to all speakers except the subwoofer
• Measurements do not need to be contained within the seating and it is ok to have measurements behind and in front of the seating.
• If possible keep your measurements 30" away from any wall(unless your listening position is right up against the wall)

Here is an rough estimate where my measurements ended up being located (Blue Dots). Measurements 2-8 were going in the order that Dirac recommended and alternating the heights. For now, I am still only calibrating the first row of seating.



MINIDSP

---

So the minidsp manual provides a treasure trove of knowledge and can be found here


Taken from the minidsp manual

It is important that measurements are performed over a sufficiently broad area. If the measurement area is too small, the result may be over-correction that will sound dry and dull. If using the Chair listening area, spread the microphone positions over a circle with a diameter of at least a meter (three feet). The microphone should be varied in height from the central position by at least 30 cm (one foot) up and down. If using the Sofa or Auditorium listening environment, spread the measurement locations over the full listening area and ensure that microphone height varies by at least 30 cm (one foot) up and down.
While the locations indicated are recommended, you can use a different set of locations if necessary. The important thing is to ensure that the measurement locations are spread over the whole listening area and that the microphone is moved a sufficient distance vertically as well as horizontally.
In some cases, such as when the listening area is very close to the loudspeakers, the size and in particular the height of the measurement area can be reduced, to avoid discrepancies caused by varying output response from the speakers themselves.

Other things that it shares is that if you get clipping at a certain position. to go back a step and reduce the volume and continue with your measurements. It states that there is no need to redo any measurements already taken. This mirrors the information that we received from Emotiva.

MiniDSP also shows the recommended microphone positions and states nothing that the selection of the listening environment selection doing anything different.



If Dirac is not producing the results expected, give this a try and please share your thoughts.

Tony

[linvincible]

thanks for sharing this looks valuable!
can't wait to sell my Marantz to get the XSP!

[villock]

Thanks Tony

[rogersch]

Jan 5, 2015 10:19:01 GMT -5 ansat said:

As most of you know, I have been running tons of measurements of my room and have re-run Dirac about 40 times now.

Wow.


Great work Ansat! You are really helping the XMC-1 community.

[ansat]

Jan 5, 2015 11:31:55 GMT -5 rogersch said:

Jan 5, 2015 10:19:01 GMT -5 ansat said:

As most of you know, I have been running tons of measurements of my room and have re-run Dirac about 40 times now.

Wow.


Great work Ansat! You are really helping the XMC-1 community.

I'm almost there, I almost have an arc (that i prefer) working in my room.

I need emotiva to release the full version of Dirac. And a fix to the bass and treble calibration to seal the deal.

Tony

[hemster]

I have 2 rows of seating with 2 seats in the front row and 3 in the back row. However the back row is on a 1 foot riser. I've been OCD'ing about mic placements for the Dirac readings. Here's what I did:

I've taken 3 across the back row. The middle seat (the money seat!) I took at ear level. The right seat I placed at 4" above and the left seat at 3" below. For the front row I did 2 seats at ear level. The remainder of the readings were with the mic placed between the 2 rows.

How can I improve on mic placements?

Any comments would be welcome.

[ansat]

Jan 5, 2015 13:13:40 GMT -5 hemster said:

I have 2 rows of seating with 2 seats in the front row and 3 in the back row. However the back row is on a 1 foot riser. I've been OCD'ing about mic placements for the Dirac readings. Here's what I did:

I've taken 3 across the back row. The middle seat (the money seat!) I took at ear level. The right seat I placed at 4" above and the left seat at 3" below. For the front row I did 2 seats at ear level. The remainder of the readings were with the mic placed between the 2 rows.

How can I improve on mic placements?

Any comments would be welcome.

hemster,
I am in the same boat and that is why I have not tackled the rear 4 seats in my room (just the front 3). (Still holding out to see if we can get a 2nd dirac speaker set)

But what my testing showed me is that the other 8 microphone position locations really don't matter. What matters is 4 high/low measurements in front of the listeners and 4 high/low measurements behind the listeners. I would follow the same guidelines (although I have not started my testing with the rear seating) using the 18" difference between heights and do 4 measurements 18" in front of the listening position and 4 behind ear level of the rear seating (as long as it does not get to close to a wall) (In my room both measurements will be about 18" in front of each rows seating due to a wall 1" behind the rear seating). So for the guidelines - The most important is line of sight to the speakers, followed by not measuring to close to a wall (too close to a wall renders the measurement almost useless as your decay will usually be terrible). Followed by the 18" between the high and low positions.


Tony


Edit - if you look at my measurement positions you will notice that the blue dot to the left had to be brought in as it would lose line of sight to the front 3 and be too close to a wall if I were to measure it behind the chair. So I brought the mic on the inside to keep it line of site.

[cawgijoe]

Tony....the key statement you make is that the other 8 mic position locations don't matter.

I read a very good review by Ken Rockwell on Dirac Live. If you get a chance Google it. He also mentions that those positions really don't matter. The key is the first sweet spot position.

Mic position was the most confusing part for me when I ran my first Dirac. I kept looking at that couch and trying to figure out how to do it with an L shaped seating layout that didn't look anything like the simple couch. Thanks for guiding me there.

[fbczar]

Tony, Does the impulse response measurement have any value relative to subwoofers?  If so how do you read an impulse response graph for a subwoofer?

[socketman]

What cal file are you using now with dirac.

[ansat]

Jan 5, 2015 14:54:32 GMT -5 fbczar said:

Tony, Does the impulse response measurement have any value relative to subwoofers?  If so how do you read an impulse response graph for a subwoofer?

In the dirac application. not really, especially if you are running multiple subwoofers as timing changes to phase align can be read as reflections.

I find it useful for phase alignment via REW. But that is out of the realm for a beginner tutorial and will require different equipment and needs loopback connection. There are also those who don't like to use that method. For treatment placement I don't bother. With sub Frequencies hanging out in the room for 300+ms (in a treated room), It is just too much work.
Tony

[dubs]

Tony, I want to make sure I understand this. I have rear speakers that come just about to the top of the back of my couch. I have a open kitchen with a half wall to my living area so they have to be at that height or my wife would be a unhappy camper. To keep in line of sight all my settings would have to be above the couch back wether in front or back.  The half wall is about 4 1/2 ft away from the back of the couch. Your thoughts?

[ansat]

Jan 5, 2015 15:38:58 GMT -5 dubs said:

Tony, I want to make sure I understand this. I have rear speakers that come just about to the top of the back of my couch. I have a open kitchen with a half wall to my living area so they have to be at that height or my wife would be a unhappy camper. To keep in line of sight all my settings would have to be above the couch back wether in front or back.  The half wall is about 4 1/2 ft away from the back of the couch. Your thoughts?

go as low as you can on the couch to maintain line of sight to the tweeter. You don't want to take any measurements out side of the bubble you created. I would still only go 9" higher then ear level for the upper rear measurements. Something kinda like this
Red is the high measurements - Blue is the low measurements.


Tony

[deewan]

What confuses me the most is the requirement for line of sight for all speakers. I understand why line of sight would be important, but how attainable is it for most surround setups? As I understand after the first mic position at the main listening point (location and height), how would someone go about making sure a mic position that is lower and in front of the main mic position have line of sight of the rear speakers?

Here is a pic of my room. If I were to place the mic 8-10 inches below the main listening positions height and slightly in front of the main listening position's location, I would think I would lose line of sight for both rear speakers.


I would think the same situation would happen if I were to do anything behind the main listening position and below, I would lose line of sight to the front three and possible side surround speakers.

Am I missing the simple answer to this problem?

Edit: Sorry, I was slow to reply and it appears a similar questions was asked. But I still worry about line of sight for anything lower than the main listening position's height.

[dubs]

Thank you Tony. I think what many, including me are finding what a precise tool Emo gave us for room correction in Dirac. To be honest, after my first run I've been quite happy with the sound of Dirac. I did have to bump the bass for my taste and adjust the volume to my liking but even my wife was impressed. She's like Mikey, (showing my age) she hates everything when it come to my music hobby but having played music and been in bands most of my youth she understands.  I hear many of the problems and complaints and not being that literate in the hardware,  computer and software components of the music industry I can understand completely. There are so many variations in rooms, wall treatments, equipment, amps, etc. that I can see that one size does not fit all with Dirac.  These problems were bound to happen with this precise hardware. Least we forget that most of the time that Dirac is in equipment costing many many times more than we paid so I bet most who have it run it with very high end equipment. (Not that Emotiva Isn't) With your help I'm sure most of us will achieve the sound we are looking for.  I feel like the guy in "Days of Thunder". I could play the drums so I just played but I couldn't read music. We depend on good people as yourself to help us. Sometimes as hard as you try, you speak over our heads. You are under no obligation to talk any different but for us little guys any plain English is a help. That's why I like following You and Keith from Emo so much. I'm being educated. Take care!

[hemster]

deewan, you are right about the challenge of low positions and yet having line of sight to the rear speakers. I have this issue for my front row because the back row is on a riser and for low position measurements in the front row I cannot get line of sight for rear speakers. 

[Mike Ronesia]

Just tried running Dirac. It fails on about half of the positions, but changing nothing and rerunning it seems to work. The room is almost dead quiet. At the end of the 9 positions it failed with a out of phase sub message. I tried starting over and it made it through after again redoing about half of the positions. I didn't bother saving it or even listening because it just seems to be throwing up too many errors. I'll start from scratch when I have more time, and I'll turn off the AGC thing. Question, are there more detailed instructions in the manual and does it explain all of these error messages and what they mean? Where are people placing their tripod? Behind the couch? I sure wish the mic had a standard tripod thread on it as there is no way to use the little one in the positions they suggest. The graphs showed the surrounds almost perfectly flat above 80hz, the fronts not so much even though they are the same speakers but running through their matched subs. I might look at those connections.

Cheers
Mark

[hemster]

mark, AGC is likely the culprit for sporadic failures. I had to get creative with the method for mic placement. I ended up placing a pillow across the top of the seat back and then placing a 2x4 on it to steady the mic stand on.

[ansat]

Jan 5, 2015 15:59:43 GMT -5 deewan said:

What confuses me the most is the requirement for line of sight for all speakers. I understand why line of sight would be important, but how attainable is it for most surround setups? As I understand after the first mic position at the main listening point (location and height), how would someone go about making sure a mic position that is lower and in front of the main mic position have line of sight of the rear speakers?

Here is a pic of my room. If I were to place the mic 8-10 inches below the main listening positions height and slightly in front of the main listening position's location, I would think I would lose line of sight for both rear speakers.


I would think the same situation would happen if I were to do anything behind the main listening position and below, I would lose line of sight to the front three and possible side surround speakers.

Am I missing the simple answer to this problem?

Edit: Sorry, I was slow to reply and it appears a similar questions was asked. But I still worry about line of sight for anything lower than the main listening position's height.

deewan if you have to go higher for the measurement - do so. Line of sight is more important then the height requirements. And remember, these are just guidelines and so are the locations.


Something like attached could work as well.
Red is high / Blue is low


Tony

Attachments:

[simpleman68]

Thanks for your continued help/research on this one Tony.
I had a hard time getting Dirac to see the XMC when signed onto it. I reset to factory defaults on the XMC and reinstalled Dirac software and all was good.

I'm not sure about my mic settings and got miscellaneous clipping alerts (failed run) throughout testing although dropping sub volume helped once or twice.
I dropped that damn Seaton to near rock bottom before starting like I had to with the old Audyssey rig on the Denon 4311.

The Center sounds unnaturally loud, the sub is almost dead and the room feels pretty dead. I lost about 10dB average but understand it may take some getting used to.
It just doesn't hit my ear like the Audyssey set up did. I'm sure I did plenty wrong on the setup here and felt a bit lost at times with mig gain etc.

I have 3 rows of seating. 4 reclining fronts, 4 reclining middle row (8" riser)  and a half dozen bar stools behind the bar behind the 2nd row.

I'll post a few graphs for guidance and input. If nothing else perhaps folks will figure out what NOT to do. lol
Scott

Facing rear



Facing front



L-R before




Sub before





L-R after



Sub after



All after