AVS Forum banner

3101 - 3120 of 3181 Posts

·
Registered
Joined
·
13,858 Posts
Discussion Starter #3,101
  • Like
Reactions: TheManFromDownUnder

·
** Man of Leisure **
Joined
·
24,058 Posts
Is there a tutorial for this??
A custom target curve can be implemented by loading a text file that contains the attributes of the customization you want. You can download several typical and popular target curves from here: Dirac Live 2/3. Scroll down to the section titled "Target Curves for Dirac Live 2'3".

  • Download the target curve text file to your laptop.
  • Load Dirac Live and open your Dirac project file.
  • On the design page, click the menu icon in the upper left and select "Load target file", and select the text file you downloaded.
  • Select the option to apply the same target to all speaker groups and click OK.

You will now see the target curve applied to your project measurements. Allow the filters to re-calculate, and save the new project to a slot in your processor/
 

·
Registered
Joined
·
429 Posts
How does one create and apply their own target curve?
One way, is at the filter design section of Dirac, you can modify the current curve (either the default it creates or one you loaded). The little points on the curve can be moved up or down or slid along the curve to change frequency. Right clicking the curve allows a new breakpoint to be added. Your resulting new curve can be saved as a file on your computer.

That handles the mechanics, but what you want the target to be is more of a challenge. Usually folks are aiming to do something specific, like increase bass, track their particular speakers, roll off or increase treble. Unfortunately, you can do undesirable things if you don't understand what you are doing.

The things people usually do are things like align how people hear or generally prefer, hence the reference to Harmon curves. The ones on first entry of this discussion, Increase bass and roll off thru mid-bass to flat into the high frequencies. The default Dirac curve raises the bass somewhat and decreases the highest frequencies. As shown on the filter page it is a sloped line. Any of these could be starting point for change. I would recommend becoming familiar with one or more first, to consider what if anything you would change (e.g. add bass to Dirac or roll off treble for Harmon examples).

Doing it on the filter page takes a steady hand! It might be a good idea to hover over it to see its FR and offset to reference. Then you can check you are at the same FR after a change. A computer is easier than a phone or tablet. A big screen computer is easier than a laptop w/o screen.

You can save and load target files. An alternative is to save a file and use a text editor to modify it. This is a little bit tricky as the data is plotted against a horizontal line (e.g. at 0db). When subtracting values to ones below it like adding (e.g. -1 - 3 = -4). If you edit this way you will get to view results when you load it back into the filter page.

I'm assuming you are primarily interested in Dirac. But the idea is a target curve is a frequency response that sounds good (ideally in your room with your speakers in which case you call it a house curve). It can then be used by the system to make EQ changes to achieve the frequency response. Caveat being sufficient headroom and the changes able to be made by EQ (e.g. you can't EQ out of a null).

If you can start with something close then there isn't much do. The Harmon curves linked in first note are great start for the bass response as they track human hearing studies. One of these and treble roll off might be ideal for many.
 

·
Registered
Joined
·
429 Posts
My post was the tutorial ;)

Start with a low-shelf filter at 100Hz, gain 6dB, Q 0.707. Apply to all speakers including the sub.
I am a amateur enthusiast. Your description was confusing to me, mainly because I'm not certain what low-shelf filter means. I do have an idea what the Gain and Q references mean. Thinking your recommending a bass boost. I googled and noticed initial returns associate the term with recording. And a suggested usage is to prevent voices from overlapping. All good but I don't easily see how to get where I want from your quote.

I do get that 1) only change below 100hz, 2) max gain at lowest frequencies is 6dB, and 3) I guess that want to mimic the shape of a filter with a Q= 0.707 (broad). What I think is missing is at what frequency would that filter be applied. From looking at Dirac target curves, I'd be putting it between 15-30hz (and I can understand that it depends on the sub or speakers).

I offer this as a member of the audience that respects what you have done for this community. I realize that effort on my part is to be expected. You might reach more of us if you consider some like myself are not formally trained.
 

·
Registered
Joined
·
93 Posts
I am a amateur enthusiast. Your description was confusing to me, mainly because I'm not certain what low-shelf filter means. I do have an idea what the Gain and Q references mean. Thinking your recommending a bass boost. I googled and noticed initial returns associate the term with recording. And a suggested usage is to prevent voices from overlapping. All good but I don't easily see how to get where I want from your quote.

I do get that 1) only change below 100hz, 2) max gain at lowest frequencies is 6dB, and 3) I guess that want to mimic the shape of a filter with a Q= 0.707 (broad). What I think is missing is at what frequency would that filter be applied. From looking at Dirac target curves, I'd be putting it between 15-30hz (and I can understand that it depends on the sub or speakers).

I offer this as a member of the audience that respects what you have done for this community. I realize that effort on my part is to be expected. You might reach more of us if you consider some like myself are not formally trained.
Well put candor and thanks for your explanation earlier!

I am a rank amatuer with such things and rely heavily on forums such as this to increase my knowledge....I will be watching and learning from the many talented folk who provide their experience and knowledge for dummies like me!!
 

·
Registered
Joined
·
472 Posts
If you are testing new speakers and changing things in your setup, is it ok too measure only one point with dirac so it doesn't take do long everytime and check the results or does this impact the results to much?
 

·
Registered
Joined
·
22 Posts
If you are testing new speakers and changing things in your setup, is it ok too measure only one point with dirac so it doesn't take do long everytime and check the results or does this impact the results to much?
The result will be different from what you will get if you use like 13 positions.
The result will be much more equal with more measurems than with just one.
if you want save time, than at least use the smallest amount of positions, think it was 5 position measurement.

for testing out new speaker or absorber placement you can use REW.
 

·
Registered
Joined
·
472 Posts
The result will be different from what you will get if you use like 13 positions.
The result will be much more equal with more measurems than with just one.
if you want save time, than at least use the smallest amount of positions, think it was 5 position measurement.

for testing out new speaker or absorber placement you can use REW.
Ok thanks!
 

·
Registered
Joined
·
108 Posts
Didn't see it mentioned here, but looks like the partnership between Dirac and Onkyo announced a few months ago is starting to bear fruit with the latest Onkyo receivers announced which supposedly have Dirac Live.

Newly announced Onkyo TX-NR7100 and Onkyo TX-RZ50

Thought it might be of interest to those who might be on the lookout for Dirac in the mid-level receiver category. Seems to be only 7.2.2 though, and no mention of whether DLBC will be supported, but promising news nonetheless. I hope they expand to their higher-channel count offerings as well. Exciting times indeed!
 

·
Registered
Joined
·
155 Posts
3076261

Red without Dirac.
Turquoise Dirac dlbc with roomfeel/Nad curve
Purple normal Dirac dlbc.

What happend here? Some cancelation?
 

·
Registered
Joined
·
155 Posts

·
Registered
Joined
·
429 Posts
View attachment 3076261
Red without Dirac.
Turquoise Dirac dlbc with roomfeel/Nad curve
Purple normal Dirac dlbc.

What happend here? Some cancelation?
Did you see similar in Dirac V3.x app when doing the measurement at that mic position? The other challenge here is you presumably did multiple mic positions during calibration and this chart is likely based on one position (which one would assume was near the first position of the Dirac calibration). Another consideration what speaker/subs are measured in the chart (e.g. are all the ones that provide bass actually playing in the test), though would presume the two measures with DLBC would be same speakers. Are you able to compare the targetfile contents between the two DLBC choices?

It does look like the mic is in a null at 36hz, but if it was in same position as chart for roomfeel chart you would think evidence of it would be there, as Dirac boost would be limited to around 10dB.

You probably should redo making the filter used and see if it measures the same. If you redo the Dirac measurements, you might save a snip of the measurement at the mic position you do the above charts at for comparison.
 

·
Registered
Joined
·
919 Posts
JBL SDP 55.
If you have any crossovers set in the SDP 55, I would try setting the speakers to large and reloading the Dirac filter. They could be interfering with the Dirac filters.

I have an Arcam AV40. Dirac support told me that the internal crossovers were not being disabled correctly with firmware 1.42 and advised I set the speakers to large in the AV40. I don't know if it was fixed in 1.44. I still have everything set to large since Dirac handles all the crossovers anyway.
 

·
** Man of Leisure **
Joined
·
24,058 Posts
Do you guys check volume levels after dirac?
Yes. If pre-calibration levels are not set properly, the output level of all speakers can be impacted, I.e. the output can be lower than desired. Even without checking, I can immediately tell when the output level is low. On my HTP-1, I can also observe the Dirac trims which will immediately tell me whether to expect low output levels. And finally, I typically check output levels with REW during post-calibration measurement sweeps.

Are you experiencing low speaker output levels? My speakers measure between 72 and 73dB, which is pretty close to normal, IMO.
 

·
Registered
Joined
·
13,858 Posts
Discussion Starter #3,117 (Edited)
Do you guys check volume levels after dirac?
Yes. If pre-calibration levels are not set properly, the output level of all speakers can be impacted, I.e. the output can be lower than desired.
To be more specific, it's the subwoofer level in relation to the other speakers you need to have an eye on. The level relationship between other speakers is usually a given. Make sure the subwoofer level ends up roughly in the range of your other speakers.
If there's a large difference amongst other speakers you need to harmonize amp input sensitivity or consider getting new speakers with roughly the same efficiency.
 

·
Registered
Joined
·
13,858 Posts
Discussion Starter #3,118
View attachment 3076261
Red without Dirac.
Turquoise Dirac dlbc with roomfeel/Nad curve
Purple normal Dirac dlbc.

What happend here? Some cancelation?
Define "normal Dirac dlbc" – which box in DL did you select?
How many subs? How are they connected?
How was this measured? What test signal was sent to which channel?
 

·
Registered
Joined
·
285 Posts
(...)
 

·
Registered
Joined
·
155 Posts
Did you see similar in Dirac V3.x app when doing the measurement at that mic position? The other challenge here is you presumably did multiple mic positions during calibration and this chart is likely based on one position (which one would assume was near the first position of the Dirac calibration). Another consideration what speaker/subs are measured in the chart (e.g. are all the ones that provide bass actually playing in the test), though would presume the two measures with DLBC would be same speakers. Are you able to compare the targetfile contents between the two DLBC choices?

It does look like the mic is in a null at 36hz, but if it was in same position as chart for roomfeel chart you would think evidence of it would be there, as Dirac boost would be limited to around 10dB.

You probably should redo making the filter used and see if it measures the same. If you redo the Dirac measurements, you might save a snip of the measurement at the mic position you do the above charts at for comparison.
No.
L&R and subs. It Are the same postion and 1 minute after.
Yes, I can try later to do that.
If you have any crossovers set in the SDP 55, I would try setting the speakers to large and reloading the Dirac filter. They could be interfering with the Dirac filters.

I have an Arcam AV40. Dirac support told me that the internal crossovers were not being disabled correctly with firmware 1.42 and advised I set the speakers to large in the AV40. I don't know if it was fixed in 1.44. I still have everything set to large since Dirac handles all the crossovers anyway.[/QUOTE
I thought the problem eas solved.

Define "normal Dirac dlbc" – which box in DL did you select?
How many subs? How are they connected?
How was this measured? What test signal was sent to which channel?
The curve with JBL/latest Dirac. The third. Dlbc.
2. xlr in 15 and 16.
rew with umik-1. l&r with subs Connect.
 
3101 - 3120 of 3181 Posts
Top