• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Trouble with Dirac Live and lack of bass after correction

spacebar

Active Member
Forum Donor
Joined
Apr 4, 2020
Messages
190
Likes
80
Location
The far side
Hi everyone. First I would like to thank you all for a great forum and really interesting topics. I'm reading/following and learning a lot :) @amirm has some really interesting topics on Room acoustics that made me buy a mic and trying out REW + Dirac. So thank you very much @amirm.

All of this is very new to me so I do make mistakes. You will probably see that in the measurements ;)

So I'm having some trouble with Dirac 2.0 limited version (500Hz). I use a MiniDSP UMIK-1 and going through all the steps. When I'm done and listening to the correction, the sound is worse. The lack of bass and the clarity is reduced. Not so wide and open and very tight sound in a way.

I've tried wide 13 point and wide + 9 point but same results. My (open space) living room/kitchen is mainly concrete walls and the sofa is against the wall. My speakers and slightly toed in to center.

@amirm wrote: "Why the low frequencies? Because that is the region where the room modifies the sound of your speakers the most." https://www.audiosciencereview.com/forum/index.php?threads/subwoofer-low-frequency-optimization.15/
and maybe this is my problem with Dirac. Maybe some corrections and/or bad measurements in this area makes the sound worse.

I like the sound with no Dirac, but some bass correction wouldnt hurt, not much, but some tweaking. I really enjoy the tones and clarity "out of the box", a huge step up from my Yamaha RX-A1070.

I have some screenshots and measurements, and to be honest I feel that I'm doing everything wrong here.

I tested the Harman 8db/10db boost curve compared to Dirac live curve. I like the Harman curve better since it gave more bass in the area I feel is missing out. But none of those are in my ears better than listening to Dirac off.

Maybe my speakers are not the best, or my room, but I dont know.

center-position.jpg


Dirac live room curve.
dirac-live-curve.jpg



This is my speaker measurement.
I dont think I got this right with REW but here you go. This is average from 3 measurements.

MA-Silver-1-average.jpg


If anyone sees something obvious or have some advice please let me know.
Thanks.
 

waynel

Addicted to Fun and Learning
Forum Donor
Joined
Nov 14, 2019
Messages
891
Likes
1,140
1). It looks like the target curve for Dirac that you are using has overall less bass in the 50-150 Hz region which results in a more neutral. But less bass heavy response than what you are used to.
2). Don’t force the Dirac target curve to be above the natural roll off of your speakers below 40hz, this will just result in more distortion as your speakers are not capable of higher output here. to achieve that target curve you will need to add a sub (2 subs are better)
3). Can you clarify what the last plot is?
 
OP
S

spacebar

Active Member
Forum Donor
Joined
Apr 4, 2020
Messages
190
Likes
80
Location
The far side
What’s your equipment?
NAD M10 (I’m testing this unit so I might end up with something else tho).
Monitor Audio Silver 1 (45Hz - 35KHz) sensitivity 87db

Edit: I’m also considering the elac dbr 62 to try something new and different.

1). It looks like the target curve for Dirac that you are using has overall less bass in the 50-150 Hz region which results in a more neutral. But less bass heavy response than what you are used to.
2). Don’t force the Dirac target curve to be above the natural roll off of your speakers below 40hz, this will just result in more distortion as your speakers are not capable of higher output here. to achieve that target curve you will need to add a sub (2 subs are better)
3). Can you clarify what the last plot is?

1. Okey, thanks for the info.
2. Noted, I’ll make the changes and follow the natural roll off. You just confirmed my next question about subs, so thank you for that :)
3. It’s my attempt to measure the speaker with the mic pointing towards the tweeter from a distance using REW to see the curve. I recon this is the SPL in REW.
 
Last edited:

waynel

Addicted to Fun and Learning
Forum Donor
Joined
Nov 14, 2019
Messages
891
Likes
1,140
So the third curve is REW measurements after Dirac? It looks pretty good although it’s showing less bass than the Dirac live average. A pair of subs would really help flatten out the bass down to lower frequencies.
 
OP
S

spacebar

Active Member
Forum Donor
Joined
Apr 4, 2020
Messages
190
Likes
80
Location
The far side
So the third curve is REW measurements after Dirac? It looks pretty good although it’s showing less bass than the Dirac live average. A pair of subs would really help flatten out the bass down to lower frequencies.
Hello. No this is before. I did a measurement before I started doing Dirac. I have not done a measurement after Dirac to compare. Sorry, my bad and I’ll do one tomorrow to see any changes.
 
OP
S

spacebar

Active Member
Forum Donor
Joined
Apr 4, 2020
Messages
190
Likes
80
Location
The far side
I found a review with measurements https://hometheaterhifi.com/reviews...-audio-silver-1-bookshelf-loudspeaker-review/
I didn’t know about the effect of the grill
Yet its grille barely affected the sound or measured performance. I suspect the difference in how much the grille affects the upper midrange performance depends on tweeter dispersion. A flat faceplate tweeter fires more upper midrange energy at the grille frame than a waveguide-loaded tweeter such as the Mini A’s. More upper midrange energy directed at the grille frame means more upper midrange energy to diffract off the grille frame: therefore, subjectively and objectively greater sonic degradation.
 

AudioJester

Addicted to Fun and Learning
Forum Donor
Joined
Jan 7, 2020
Messages
645
Likes
833
I always wonder about dirac <100Hz. It gives the impression to users that their little bookshelf speakers can have boosted output down to <20Hz! There needs to be some input about the speakers and what they are capable of. Could this type of boost not damage amps or speakers??
 

waynel

Addicted to Fun and Learning
Forum Donor
Joined
Nov 14, 2019
Messages
891
Likes
1,140
I always wonder about dirac <100Hz. It gives the impression to users that their little bookshelf speakers can have boosted output down to <20Hz! There needs to be some input about the speakers and what they are capable of. Could this type of boost not damage amps or speakers??
I believe that the auto generated Dirac targets do not try to boost below the speaker roll off
 

tecnogadget

Addicted to Fun and Learning
Joined
May 21, 2018
Messages
555
Likes
965
Location
Madrid, Spain
Try to move the hole curve 5dB below (just the same curve) and listen. That way Dirac will work lowering the peaks instead of lifting the nulls. If you still feel you are missing something, make it 3dB hotter (more output) from 200hz or 100hz to 20hz.
 

AudioJester

Addicted to Fun and Learning
Forum Donor
Joined
Jan 7, 2020
Messages
645
Likes
833
I believe that the auto generated Dirac targets do not try to boost below the speaker roll off
Ok, that makes sense. So based on initial measurements showing natural roll-off?
 
OP
S

spacebar

Active Member
Forum Donor
Joined
Apr 4, 2020
Messages
190
Likes
80
Location
The far side
Thanks for all the feedback people. I was a little frustrated and started from scratch again and took some new measurements with more points and selected the wide sofa option. I was interested to see if I could improve the sound and compare it to the previous measurement. The answer is yes, the sound was improved. So I think I was not the best in terms of microphone placement etc... but hey, I'm learning. I followed the 40-60cm recommendation from Dirac.

Remember I have only Dirac limited version so it stops at 500Hz. Detected range is from 28.4Hz so all my adjustments/corrections is from 28.4 to 500.

1. This image is the measurements from center position.
DL-Mes.jpg


2. And here is the default room curve in Dirac that I used in my first Filter, very flat and no bass.
DL-default.jpg



3. This is the Dirac Live Room Curve you can download and is my starting point for later adjustments.
DL-Curve.jpg


4. To replicate the sound I like with Dirac disabled, I ended up with this. I'm getting closer and I'm able to improve the bass (bring it back). However the voices and details is not "there yet". I'm confused about what causes the lack of details and clarity when Dirac is enabled. It seems it tries to tighten the sound. I like more space and details. With Dirac enabled the voices is further away from me and background details etc is more compressed. I have a hard time putting words on this since English is not my first language. Sorry about that!
DL-Custom-Match.jpg


So my question is:
  1. Does Dirac change the whole Hz spectre even if its only limited to 500Hz? Like the other areas above 500Hz, it makes corrections but I'm not allowed to make any adjustments
  2. Is my filter (image 4) very bad or okey? Something I should consider? I did listen to many custom filters and the lack of clarity and voice came from around 100Hz area, it did improve when I started to increase the db, and I tried to stay in the peaks like you told me to.
An observation with Dirac enabled is that when I mute the sound, its not totally muted and I can still hear the sound from the speaker when I put my ear against it. This does not happen when Dirac deactivated. Maybe a feature, maybe a bug, I dunno...

Again, thanks for all the help.
 

Soniclife

Major Contributor
Forum Donor
Joined
Apr 13, 2017
Messages
4,213
Likes
4,951
Location
UK
Hello. No this is before. I did a measurement before I started doing Dirac. I have not done a measurement after Dirac to compare. Sorry, my bad and I’ll do one tomorrow to see any changes.
What distance was that measurement taken at, and what smoothing was used? If this was from the listening position your starting position is very good, but it does not look like the Dirac measurement, so I'm confused.
 

QMuse

Major Contributor
Joined
Feb 20, 2020
Messages
3,124
Likes
2,713
Try to move the hole curve 5dB below (just the same curve) and listen. That way Dirac will work lowering the peaks instead of lifting the nulls. If you still feel you are missing something, make it 3dB hotter (more output) from 200hz or 100hz to 20hz.

I also think the height of the target curve was the problem as Dirac refused to boost LF that much. i would try lowering the entire target curve by 5-7dB, regenerating the filters and measure again with REW the final result.

P.S when measuring with REW use RTA MMM or if measuring with sweeps and making average (as you did) do it for each channel separately, not with both channels playing at the samme time.
 
OP
S

spacebar

Active Member
Forum Donor
Joined
Apr 4, 2020
Messages
190
Likes
80
Location
The far side
What distance was that measurement taken at, and what smoothing was used? If this was from the listening position your starting position is very good, but it does not look like the Dirac measurement, so I'm confused.

Hi. This was taken 1 meter from the speaker and not listening position. I used 1/12 smoothing with REW. I just wanted to test REW and my speaker before I started with Dirac.
 

Soniclife

Major Contributor
Forum Donor
Joined
Apr 13, 2017
Messages
4,213
Likes
4,951
Location
UK
Hi. This was taken 1 meter from the speaker and not listening position. I used 1/12 smoothing with REW. I just wanted to test REW and my speaker before I started with Dirac.
That's what I guessed as there was no roll off at the top end.
I second the idea to try using MMM at the listening position with no EQ to see how that looks.
 
OP
S

spacebar

Active Member
Forum Donor
Joined
Apr 4, 2020
Messages
190
Likes
80
Location
The far side
I also think the height of the target curve was the problem as Dirac refused to boost LF that much. i would try lowering the entire target curve by 5-7dB, regenerating the filters and measure again with REW the final result.

P.S when measuring with REW use RTA MMM or if measuring with sweeps and making average (as you did) do it for each channel separately, not with both channels playing at the samme time.
Okey, if I understand you correctly this will flatten out my curve from 28 to around 100 to match the dip after 100hz? Or do you mean lowering the whole curve from 28 to 500 with 5 to 7db?

Wouldn’t I loose the bass and the clarity like before, but sound wise it’s more correct?
 

Soniclife

Major Contributor
Forum Donor
Joined
Apr 13, 2017
Messages
4,213
Likes
4,951
Location
UK
I also think the height of the target curve was the problem as Dirac refused to boost LF that much. i would try lowering the entire target curve by 5-7dB, regenerating the filters and measure again with REW the final result
How does Dirac handle this when it only does up to 500hz? Would that not result in very quiet bass with a jump up from 500?
 
OP
S

spacebar

Active Member
Forum Donor
Joined
Apr 4, 2020
Messages
190
Likes
80
Location
The far side
That's what I guessed as there was no roll off at the top end.
I second the idea to try using MMM at the listening position with no EQ to see how that looks.
Noted. I’ll have to do that tomorrow. Thanks for the tip qmuse.
 
Top Bottom