[Warden95]

Do you have a screenshot of your volume calibration screen?

No I don’t. I usually match every thing to the lowest driver. Should I boost sub in volume calibration ?

 

[Sounds_Insane]

This can be overcome with pre-Dirac EQ on the sub.

Yeah, PEQ the peaks out, probably cut 10dB, and now I hit my target.

 

[Sounds_Insane]

No I don’t. I usually match every thing to the lowest driver. Should I boost sub in volume calibration ?

it won't matter. If you set the sub volume to +10dB Dirac will just level it back to 0. I think all it uses volume calibration for is so you know all your speakers are playing loud enough to get measured without blowing anything up.

 

[Warden95]

it won't matter. If you set the sub volume to +10dB Dirac will just level it back to 0. I think all it uses volume calibration for is so you know all your speakers are playing loud enough to get measured without blowing anything up.

That what I always thought. You can use the sliders but I like to level match in the output gain

 

[Truthunter]

Two things others have done to remedy:

1) Bring the target level down a few db until it's within the 10db window for Dirac to correct. This results in lower overall output.

or

2) Cut sub 6db in the output tab before Dirac. Use a target that is 6db lower for the sub group. It's the same target as the rest of the groups just 6db lower. Add the 6db back into the sub output tab after Dirac

 

[tgleds]

Two things others have done to remedy:

1) Bring the target level down a few db until it's within the 10db window for Dirac to correct. This results in lower overall output.

or

2) Cut sub 6db in the output tab before Dirac. Use a target that is 6db lower for the sub group. It's the same target as the rest of the groups just 6db lower. Add the 6db back into the sub output tab after Dirac

In 1) you're referring to the target curve?

Also, What do these blue vertical sliders do in Dirac? I dont have a photo of them at the moment, but they are a horizontal dB level lines across the frequency plot.

 

[Sounds_Insane]

In 1) you're referring to the target curve?

Also, What do these blue vertical sliders do in Dirac? I dont have a photo of them at the moment, but they are a horizontal dB level lines across the frequency plot.

Those are the curtains. that sets the range you want Dirac to try to correct to the target curve. there are vertical white dashed lines that is what it thinks the driver will respond best to, but you can go past that with the curtains. Take those as far as you can where Dirac is still correcting to the target. Ideally the curtains should be at least 1 octave past the crossover point you plan to use.

 

[tgleds]

Thanks all, your help is much appreciated. I figured it might be orientation, I will take a look. I forgot I had it forward facing until after I did all the measurements. The system sounds great, but it still needs a bit of ironing /working out a desirable target curve.

Im also crossing my:
XXM325s at 300 LR4
Morel Altos 2500 LR4

Let me know if those Xovers seem out of place. All measurements were taken using PEQ APFs at the Xover points.

Unrelated, but my CC3 HU has a 20k Butterworth that I cannot bypass, so that's why the tweeter measurements drop at 20k. I figured it would be best just to let the crossover do its thing and pull the curtains in to 16k instead of EQing it to 20k.

L MID

R MID

L TWT

R TWT

 

[Sounds_Insane]

Thanks all, your help is much appreciated. I figured it might be orientation, I will take a look. I forgot I had it forward facing until after I did all the measurements. The system sounds great, but it still needs a bit of ironing /working out a desirable target curve.

Im also crossing my:
XXM325s at 300 LR4
Morel Altos 2500 LR4

Let me know if those Xovers seem out of place. All measurements were taken using PEQ APFs at the Xover points.

Unrelated, but my CC3 HU has a 20k Butterworth that I cannot bypass, so that's why the measurements drop at 20k. I figured it would be best just to let the crossover do its thing and pull the curtains in to 16k instead of EQing it to 20k.

L MID
View attachment 505995
R MID
View attachment 505996
L TWT
View attachment 505997
R TWT
View attachment 505998

Drag each type of speaker out of group 1 and put them in their own groups. Tweets, Mids, Midbass, Subs, etc. Then adjust the curtains as needed per group.

Let Dirac focus on correcting the speakers as it needs, where it can do the best job at it.

Follow the quick start guide posted on here.

 

[tgleds]

Those are the curtains. that sets the range you want Dirac to try to correct to the target curve. there are vertical white dashed lines that is what it thinks the driver will respond best to, but you can go past that with the curtains. Take those as far as you can where Dirac is still correcting to the target. Ideally the curtains should be at least 1 octave past the crossover point you plan to use.

I know about the curtains. Im referring to the horizontal lines that slide vertically. They have a dB measurement attached to them. I will grab a picture next time Im on the DSP

 

[Sounds_Insane]

I know about the curtains. Im referring to the horizontal lines that slide vertically. They have a dB measurement attached to them. I will grab a picture next time Im on the DSP

Oh, that is Dirac's target curve. It will go away if you load your own curve or you can go back to it with the icons in the group(s).

Whichever is on the screen is what Dirac will use, it won't use both targets.

 

[tgleds]

1) @Sounds_Insane So I grouped each set and adjusted the curtains as per your rec, and now I have a wonderful stage. That obviously had a huge impact on the Dirac algo. It may be time to update the guide to suggest that from the start?

2) Here's the horizontal lines I was referring to. Got a picture of them before I opened the curtains, to my best knowledge it appears that they apply a linear target within the curtains instead of the target curve, which I believe you referred to this as the 'Dirac target curve'.

3) Pointing the sub towards the rear had a huge impact on my 50-90Hz null.
REAR FACING:

FORWARD FACING:

 

[Sounds_Insane]

1) @Sounds_Insane So I grouped each set and adjusted the curtains as per your rec, and now I have a wonderful stage. That obviously had a huge impact on the Dirac algo. It may be time to update the guide to suggest that from the start?

2) Here's the horizontal lines I was referring to. Got a picture of them before I opened the curtains, to my best knowledge it appears that they apply a linear target within the curtains instead of the target curve, which I believe you referred to this as the 'Dirac target curve'.

Awesome, glad the stage is coming together. Don't make my mistake when I got into Dirac and be satisfied with the first tune. The more you do and experiment the better it will get.

Yes, those blue lines are the Dirac built-in target curves. You can move them up or down to adjust it, but once you load a target it defaults to that instead. Those lines do not give you as much control, but you can convert it to points if you want to make a custom curve inside Dirac.

 

[tgleds]

Just realized that only the editors of the guide have access to view the comments. I copied and pasted it here:


Finalize language for:

1) Once you've finalized your XO points, save the config, then load to the next plugin preset
2) bypass all XOs and add the following APFs (Q=.7 if using LR4):
-Sub @ Sub-to-MB XO
-Midbass @ Sub-to-MB XO and MB-to-Mid XO
-Mid @ Sub-to-MB XO, MB-to-Mid XO, and Mid to Tweet XO
-Tweet @ Sub-to-MB XO and Mid to Tweet XO
3) Run Dirac measurements
4) Bypass APFs and un-bypass XOs
Note: no more cascading the mid-bass LPF to the sub

The reason the Mids have 3 APFs and the Tweets only have 2 APFs is because of phase alignment? e.g. Tweets don't need all 4 APFs since the 2 lower APFs cancel eachother in the phase summation...

Is my understanding in the ball park?

 

[squiers007]

The reason the Mids have 3 APFs and the Tweets only have 2 APFs is because of phase alignment? e.g. Tweets don't need all 4 APFs since the 2 lower APFs cancel eachother in the phase summation...

Is my understanding in the ball park?

The purpose of the APFs is to mimic the phase shift induced by the xover. Since we are measuring with xovers off you need an APF for each xover that particular driver is going to be effected by. The mids only need 3 APFs if you are using BM. If you do not use BM (it's not a requirement btw, some just prefer it), then the mids would only have 2 APFs. The tweeter would only get 1 APF if not using BM as well.

Hopefully that makes sense.

 

[tgleds]

I brought the SUB target curve down 15 dB then re-boosted in PEQ (12) and Routing (3).

The SUB response is now blending well.

 

[squiers007]

I brought the SUB target curve down 15 dB then re-boosted in PEQ (12) and Routing (3).

The SUB response is now blending well.

View attachment 506124

Do some pre-eq and knock that big peak at 40hz down and it should work even better.

 

[tgleds]

Do some pre-eq and knock that big peak at 40hz down and it should work even better.

Youre suggesting the Dirac algo works better with 'proper' pre-Dirac PEQ?

Is there a problem with using as much as 15 dB for the post-Dirac correction?

 

[squiers007]

Do some pre-eq and knock that big peak at 40hz down and it should work even better.

Youre suggesting the Dirac algo works better with 'proper' pre-Dirac PEQ?

Is there a problem with using as much as 15 dB for the post-Dirac correction?

So long as the post Dirac correction is to the output gains, not PEQ them your good. What I was suggesting was to do pre-Dirac eq on that peak so Dirac doesn't have to waste processing on it, plus this will make the level of the sub appear lower to Dirac so it won't try to cut as much level out

 

[tgleds]

So long as the post Dirac correction is to the output gains, not PEQ them your good. What I was suggesting was to do pre-Dirac eq on that peak so Dirac doesn't have to waste processing on it, plus this will make the level of the sub appear lower to Dirac so it won't try to cut as much level out

Should I be performing a band-pass pre-Dirac PEQ cutting on all of the drivers? Or only the ones with sharp peaks? Looks like most of my drivers are all being corrected by Diracs EQ in the +/- 10 dB range in the band pass. Not sure if this is ideal.